4. Hides JA, Richardson C, Jull G (2001) Long term effects of Specific Stabilizing Exercises for First Episode Low Back Pain. Spine 26 ( )

Transcription

1

2 CONTENTS 1. Boren K, Conrey C, Le Coguic J, Paprocki L, Voight M, Robinson K, (2011) Electromyographic analysis of gluteus medius and gluteus maximus during rehabilitation exercises. The International Journal of Sports Physical Therapy 6 (3) Cruz-Díaz D, Romeu M, Velasco C, (2018) The effectiveness of 12 weeks of Pilates intervention on disability, pain and kinesiophobia in patients with chronic low back pain: a randomized controlled trial Clinical Rehabilitation Hartvigsen J, Hancock MJ, Kongsted A, Louw Q, Ferreira ML, Genevay S, Hoy D, Karppinen J, Pransky G, Sieper J, Smeets RJ, Underwood (2018) What low back pain is and why we need to pay attention. The Lancet Hides JA, Richardson C, Jull G (2001) Long term effects of Specific Stabilizing Exercises for First Episode Low Back Pain. Spine 26 ( ) 5. Hodges P (2008) Transversus abdominis: a different view of the Elephant. British Journal of Sports Medicine 42(2008) Hodges P, Moseley G.L. (2003) Pain and motor control of the lumbopelvic region: effects and possible mechanisms. J. Electromyography and Kinesiology 13 ( ) 7. Junginger B, Baessler K, Sapsford R, Hodges PW (2009) Effect of abdominal and pelvic floor tasks on muscle activity, abdominal pressure and bladder neck Int Urogynecol J (2010) 21: Laws A, Williams S, Wilson C (2017) The Effect of Clinical Pilates on Functional Movement in Recreational Runners Int J Sports Med 2017; 38: Mallin G, Murphy S The effectiveness of a 6-week Pilates Programme on outcome measures in a population of chronic neck pain patients: a pilot study J Bodyw Mov Ther 17 (3):

3 10. Pool-Goudzward A, Vleeming A, Stoeckart R, Snijders C, Mens J (1998) Insufficient lumbopelvic stability: A clinical, anatomical and biomechanical approach to a specific low back pain. Manual Therapy 3 (1) Stolze LR, Allison SC, Childs JD (2012) Derivation of a Preliminary Clinical Prediction Rule for Identifying a Subgroup of Patients With Low Back Pain Likely to Benefit From Pilates-Based Exercise. Journal of orthopaedic & sports physical therapy 42 (5) Tsao H, Hodges P W, Galea M P (2008) Reorganization of the motor cortex is associated with postural control deficits in recurrent low back pain. Brain 131(2008) Urquart D M, Hodges P W, Allen T J, Story I H (2005) Abdominal muscle recruitment during a range of voluntary exercises. Manual Therapy 10(2005) Vleeming A, Schuenke MD,Danneels L, Willard FH (2014) The functional coupling of the deep abdominal and paraspinal muscles: the effects of simulated paraspinal muscle contraction on force transfer to the middle and posterior layer of the thoracolumbar fascia J. Anat. 225, Wells C, Kolt GS, Marshall P, Bialocerkowski A (2013) The Definition and Application of Pilates Exercise to Treat People With Chronic Low Back Pain: A Delphi Survey of Australian Physical Therapists J. Of Phys. Therapy. 94 (6) Withers GA (2017) Pilates as a therapeutic exercise In touch Journal of Physiotherapy Withers GA (2010) The role of Pilates in lumbar spine instability training Fitpro Journal 20-21

4 Int Urogynecol J (2010) 21:69 77 DOI /s z ORIGINAL ARTICLE Effect of abdominal and pelvic floor tasks on muscle activity, abdominal pressure and bladder neck Baerbel Junginger & Kaven Baessler & Ruth Sapsford & Paul W. Hodges Received: 31 December 2008 / Accepted: 8 August 2009 / Published online: 3 September 2009 # The International Urogynecological Association 2009 Abstract Introduction and hypothesis Although the bladder neck is elevated during a pelvic floor muscle (PFM) contraction, it descends during straining. This study aimed to investigate the relationship between bladder neck displacement, electromyography (EMG) activity of the pelvic floor and abdominal muscles and intra-abdominal pressure (IAP) during different pelvic floor and abdominal contractions. Methods Nine women without PFM dysfunction performed maximal, gentle and moderate PFM contractions, maximal and gentle transversus abdominis (TrA) contractions, bracing, Valsalva and head lift. Bladder neck position was assessed with perineal ultrasound. PFM and abdominal muscle activities were recorded with a vaginal probe and fine-wire electrodes, respectively. IAP was recorded with a rectal balloon. Results Bladder neck elevation only occurred during PFM and TrA contractions. PFM EMG and IAP increased during all tasks from 0.5 (gentle TrA) to 45.7 cmh 2 O (maximal Valsalva). Conclusion Bladder neck elevation was only observed when the activity of PFM EMG was high relative to the IAP increase. Keywords Bladder neck movement. Intra-abdominal pressure. Muscle EMG activity. Pelvic floor re-education. Perineal ultrasound Abbreviations ASIS Anterior superior iliac spine APFQ Australian Pelvic Floor Questionnaire EMG Electromyography Hz Hertz IAP Intra-abdominal pressure khz Kilohertz lowtra Lower TrA MHz Megahertz midtra Middle TrA OE Obliquus externus abdominis muscle OI Obliquus internus abdominis muscle PUS Perineal ultrasound RA Rectus abdominis muscle RMS Root mean square TrA Transversus abdominis muscle VAL Valsalva Introduction B. Junginger : R. Sapsford : P. W. Hodges (*) NHMRC Centre of Clinical Research Excellence in Spinal Pain, Injury and Health, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane, Queensland, Australia p.hodges@uq.edu.au B. Junginger : K. Baessler Charité University Hospital, Berlin, Germany Voluntary contraction of pelvic floor muscles (PFM) elevates the bladder neck [1 4] and compresses the urethra [5]. The contraction also provides a firm base against which the urethra is closed by the increased intra-abdominal pressure (IAP) [6, 7]. These factors contribute to the maintenance of continence, but the extent of elevation of the bladder neck is probably not determined by PFM activity alone; increased IAP may prevent elevation or induce caudal displacement of the bladder neck [1, 3, 7].

5 70 Int Urogynecol J (2010) 21:69 77 Consistent with this proposal, opposite directions of pelvic floor motion have been observed but not quantified during voluntary contraction of the PFM (elevation) and a Valsalva (depression), but contraction of other muscles rather than IAP has been argued to explain the results [8]. Abdominal and PFM are co-activated during voluntary [9, 10] and involuntary [11, 12] tasks. The abdominal activation may displace the bladder neck but this effect may be different between the structurally and functionally different muscles of the abdominal wall (obliquus internus abdominis (OI), obliquus externus abdominis (OE), rectus abdominis (RA) and transversus abdominis (TrA)). Furthermore, the timing of the abdominal and PFM contraction in relation to the IAP can be important. Although continence may be dependent on the net effect of each of these factors [13], no study has comprehensively investigated multiple elements simultaneously or the potential of subtle differences in coordinative mechanisms. Variable patterns of abdominal muscle co-activity occur during PFM contraction. PFM activity has been reported with activity of all abdominal muscles [9], selective activation of TrA [14] or co-activation of TrA and OI [10, 15]. The variation in results may be due to differences in contraction intensity (e.g. gentle vs. maximal contraction), the instruction used and the recording methods (e.g. intramuscular or surface electromyography (EMG)). Conservative rehabilitation of PFM dysfunction requires optimisation of the muscle s function. Although PFM strength must be restored, PFM co-ordination must be established to stabilise and elevate the bladder neck. For PFM rehabilitation, it may be important to understand the relationship between PFM and abdominal muscle recruitment patterns, IAP and consequent bladder neck elevation to plan optimal and individual treatment. This study aimed to investigate normal muscle recruitment patterns and the relationship between displacement of the bladder neck, PFM, OI, OE, RA and TrA EMG and IAP during abdominal and pelvic floor tasks with submaximal efforts. Although a co-contraction of the TrA and PFM has been demonstrated [9], it is unclear whether this pattern is also present during abdominal tasks such as Valsalva and brace which would involve OI and OE contractions. Furthermore, tasks and contractions with submaximal effort which would appear to be predominant in daily life have not been studied much. The specific aims of our study were: (1) to compare displacement of the bladder neck, IAP, PFM and abdominal EMG between a range of different abdominal and PFM contractions that aimed to induce different pressures and muscle activation patterns including submaximal efforts and (2) to compare activity of the PFM and abdominal muscles and IAP between maximal contractions of each muscle. Methods Subjects A convenience sample of nine female volunteers without PFM disorders with a mean (range) age, height and weight of 42 (32 59) years, 165 ( ) cm and 66 (57-72) kg, respectively, participated in the study. Four women were nulliparous and five women had one or two normal vaginal deliveries. Subjects were excluded if they had a history of back or pelvic pain in the last 6 months, hip or abdominal surgery or a history of PFM dysfunction as motor control deficits have been shown in these diseases. The Australian Pelvic Floor Questionnaire (APQF) [16] was used to screen for PFM disorders. We did not perform a vaginal examination but checked for pelvic organ prolapse on perineal ultrasound [17]. Subjects were also excluded if they had a history of laparotomy. This study was approved by the Institutional Ethics Committee and conformed to the Declaration of Helsinki. Bladder neck displacement Displacement of the bladder neck was assessed with perineal ultrasound (PUS) using a Logiq9 ultrasound (GE Medical, USA) with a curved transducer (3.5 6 MHz). Subjects were positioned supine with the hips and knees slightly flexed and abducted. The transducer was positioned on the perineum in the sagittal plane to view the pubic symphysis, bladder and urethra. Images were made at rest and during the experimental tasks. The position of the bladder neck was estimated using a coordinate system through the pubic symphysis [3] (Fig. 1). Displacement of the bladder neck in the anterior posterior and caudal cranial directions and a net displacement vector were calculated (Fig. 1). EMG Vaginal PFM EMG was recorded with a Periform probe (Neen Mobilus Healthcare group, UK) that has previously been shown to record PFM EMG without recording crosstalk from adjacent muscles such as hip muscles [11, 18]. Vaginal EMG and perineal ultrasound were recorded in separate trials as placement of the intravaginal EMG probe interfered with the perineal ultrasound image. Abdominal muscle EMG was recorded on the right side with a combination of fine-wire and surface electrodes. Fine-wire recordings were made with bipolar electrodes fabricated from Teflon-coated stainless-steel wire (75 μm, 1-mm Teflon removed, tips bent back 1 and 2 mm, threaded into a hypodermic needle mm/ mm). Electrodes were inserted with ultrasound

6 Int Urogynecol J (2010) 21: Fig. 1 a Measurement of bladder neck movement. The x-axis of the coordinate system was aligned to bisect the pubic symphysis; the y- axis is perpendicular to this line. The positions of the bladder neck at rest (start) and at contraction (end) are identified and the vectors are calculated as shown. b Bladder neck movements during the experimental tasks. Mean and standard deviation are shown. PFM pelvic floor muscle, TrA transversus abdominis guidance (7-MHz linear vector transducer) [19] into the lower TrA (lowtra) and the OI, 2 cm medial and inferior to the anterior superior iliac spine (ASIS), the middle TrA (midtra) and OE, halfway between the ASIS and the rib cage, and RA, 2 cm lateral and inferior to the umbilicus [20]. Surface EMG electrodes were placed over OI medial to the ASIS ( 15 from horizontal) and over OE inferior to the rib cage in parallel with the muscle [21]. EMG data were amplified 2,000 times, band-pass-filtered between 10 Hz and 1 khz and sampled at 2 khz using Spike2 software and a Power1401 Data acquisition system (CED, UK). Root mean square (RMS) EMG amplitude was calculated for 1 s at rest and during the tasks. EMG data were normalised to the maximum RMS EMG amplitude recorded for each muscle across a series of maximal contractions. IAP IAP was recorded with a custom-made air-filled pressure catheter inserted into the rectum. The catheter was attached to a pressure transducer (Valydine, USA), covered with a condom and lubricated. The catheter was positioned cranial to the external anal sphincter. Correct placement was confirmed when IAP increased during a sniff but not to direct compression by contractions of the PFM and anal sphincter muscles. Pressure data were amplified and sampled at 1,000 Hz. Experimental tasks Subjects performed a series of six abdominal and PF manoeuvres and standardised maximal contractions of each muscle. Subjects were positioned supine with the knees and hips flexed over a pillow. Gentle PFM contraction Subjects gently contracted the PFM with an effort of 2 on a 15-point modified Borg scale [22] indicating a very light effort. The Borg scale was developed for rating an individual s perceived exertion during exercise and was used here to standardise the effort between tasks. One investigator checked the EMG and provided feedback. Subjects were instructed to relax completely and then gently lift and tighten the PFM with a very light effort and maintain the contraction before breathing in. Moderate PFM contraction Using a similar procedure to that described above, subjects performed a moderate PFM contraction with an effort of 8 on the Borg scale, indicating a moderate or a somewhat hard effort. Isolated contraction of TrA Subjects were instructed to contract TrA with an effort of 2 on the Borg scale without activity of the other abdominal muscles. Subjects were instructed to very gently draw in or flatten the lower abdominal wall below the umbilicus whilst ensuring that the other abdominal muscles remained relaxed, which was confirmed by observation of the EMG recordings. Subjects were also provided with feedback from palpation of the muscle and real-time ultrasound imaging of the abdominal muscles in the lateral abdominal wall [23]. Brace contraction Subjects were instructed to brace the abdominal muscles by tightening the abdominal muscles whilst widening the waist with an effort equivalent to 2 on the modified Borg scale. Valsalva Subjects performed a forced expiration against a closed glottis with an effort equivalent to 2 on the modified Borg scale.

7 72 Int Urogynecol J (2010) 21:69 77 Head lift Subjects performed a very gentle sit-up with an effort of 2 on the modified Borg scale. For all tasks, subjects were instructed to take a relaxed breath in and out, completely relax the muscles and then to perform the manoeuvre without breathing. During the abdominal tasks, they were not given any instruction related to the PFM. Subjects indicated with a finger when they were relaxed and when they had performed the task to indicate the time to freeze the ultrasound and place time markers on the EMG and IAP recording. Tasks were repeated three times with a break of 1 2 min between repetitions. All procedures were completed twice: once with the intravaginal EMG in situ and once with perineal ultrasound measurement. The correct abdominal muscle manoeuvres were confirmed by EMG changes that were assessed on the computer screen by the investigator. Standardised maximal voluntary contractions were performed at the start of the trial in the supine position. Maximal contractions were maintained for 5 s and separated by s. The tasks included: PFM maximal elevation and tightening of the PFM; maximal TrA maximal forced expiration ( huff ) to residual lung volume; OI ipsilateral rotation of the trunk in crook lying position with the arms at 90 in front of the body and resistance applied to the knee and forearms; OE contralateral rotation of the trunk in crook lying position with the arms at 90 in front of the body and resistance applied to the knee and forearms; RA trunk flexion with resistance applied to the upper trunk and thighs; IAP maximal Valsalva manoeuvre against a closed glottis. Statistical analysis Displacement of the bladder neck and changes in IAP between tasks were compared with a repeated-measures analysis of variance (ANOVA). To determine whether the displacement and change in IAP were different to no change, data for each task were compared to zero with a t test for single samples. To adjust for multiple comparisons, the significance was adjusted to P< (0.05 divided by number of tasks) for this comparison. Normalised EMG data were compared between rest and contraction and between tasks and between muscles using a repeated-measures ANOVA with two repeated measures (contraction, task) and one independent factor (muscle). Post hoc analysis was undertaken with Duncan s multiplerange test. IAP and EMG were compared between maximal contractions of the PFM and abdominal muscles with repeatedmeasures ANOVA and post hoc Duncan s multiple-range test. Normalised EMG amplitude during the maximal contractions of the PF and abdominal muscles was compared between tasks (repeated measure) and muscles (EMG analysis only: independent factor) with a repeatedmeasures ANOVA and post hoc testing with Duncan s multiple-range test. The alpha level was set at P<0.05. Results The bladder neck was elevated between 1.0 (0.3) and 3.3 (1.5) mm during the contraction of the pelvic floor muscle (gentle and moderate) and TrA (interaction: task measure P<0.0001, post hoc: all P<0.0002) but not with the brace, the Valsalva and the head lift (all: P>0.56; Fig. 1). The bladder neck elevation during the moderate pelvic floor contraction was greater than during all other tasks (all other tasks: P<0.002). Bladder neck elevation during the gentle pelvic floor contraction was greater than during all abdominal manoeuvres (P<0.04). Contraction of the TrA resulted in a greater bladder neck elevation than a head lift or brace (P<0.03). There was no difference in bladder neck elevation between the head lift, the brace and the Valsalva tasks (P>0.33). According to the coordinate system used in the present study, the motion was primarily in the ventral direction and there was no significant cranial motion (y-axis change) of the bladder neck in any task (P>0.21). IAP increased in all tasks from 0.46 cmh 2 O (TrA) to 1.59 cmh 2 O (VAL; all: P<0.05; Fig. 3). Brace and Valsalva increased the IAP more than TrA contraction (main effect: task: P<0.02, post hoc: P<0.02; P<0.002, respectively). The IAP increase during the TrA contraction was less than during the moderate pelvic floor muscle task (P<0.046). PFM EMG increased with all pelvic floor and abdominal tasks (interaction: task contraction muscle P<0.0002, post hoc: all P<0.01; Fig. 3). In the gentle pelvic floor task, the PFM EMG activity was less than in the moderate pelvic floor task (P<0.01) but greater than during brace, head lift and Valsalva (P=0.01). PFM activity was similar during a gentle PFM and TrA contraction (P=0.179). PFM EMG activity was greater during the moderate pelvic floor task (P=0.014) than all other tasks (all: P<0.01). Low TrA EMG activity increased with all manoeuvres except the gentle pelvic floor task (all: P<0.007; gentle pelvic floor: P=0.54; Fig. 2). Although the increase in lowtra EMG was not statistically significant during the gentle pelvic floor task, there was a small but variable increase in activity of this muscle for all subjects. MidTrA EMG increased with the brace (P<0.02) and Valsalva (P<0.09) tasks (Fig. 2). There was no change in midtra EMG with the TrA, gentle and moderate pelvic floor contractions (all: P>0.18). OI EMG increased with the moderate pelvic floor, the brace and Valsalva tasks (all: P<0.02), but not with the

8 Int Urogynecol J (2010) 21: Fig. 2 Mean normalised EMG amplitude at rest and during contraction for all tasks. Mean and standard deviation are shown. PFM pelvic floor muscle, lowtra low region of transversus abdominis, midtra mid-region of TrA, OI obliquus internus abdominis, OE obliquus externus abdominis, RA rectus abdominis, GP gentle pelvic floor, MP moderate pelvic floor, TrA isolated transversus abdominis, BR brace, VAL Valsalva, HL head lift gentle tasks (head lift, TrA, PFM; all: P>0.19; Fig. 2). EMG of the OI was greater during moderate pelvic floor contraction, the brace and the Valsalva than during gentle pelvic floor contraction (all: P<0.02) and greater during brace than TrA and head lift tasks (all: P<0.04). There was no increase in the OE EMG (P>0.22) or in the RA EMG (P>0.10) in any task. Figure 3 shows the relationship between bladder neck elevation, IAP and abdominal muscle activity. Normalised EMG values for all abdominal muscles were summed to provide a gross estimation of total EMG of the abdominal muscles, which all contribute to the increase in IAP. The EMG, IAP and kinematic data suggest that the bladder neck is elevated during contractions of the PFM although the IAP is increased. PFM activity is sufficient to overcome the pressure and leads to elevation. During the abdominal manoeuvres (except the TrA task), the bladder neck does not elevate. In these cases, IAP is increased (to the same level that was recorded during the PFM contractions) in association with abdominal muscle activity but without sufficient PFM EMG (less than during the PFM contractions) to overcome the downward pressure of the IAP on the PFM. During the TrA contraction, despite the increase in abdominal muscle activity, the IAP was associated with PFM elevation. In this case, IAP was lower than the other abdominal tasks, but PFM EMG was not different to that recorded during the PFM contractions. There was no difference between the IAP increase during the maximal PFM contraction compared to the maximal RA, OE and OI contractions (all: P>0.21). The increase in IAP was greater during the maximal Valsalva (P<0.025) than all tasks except the maximal TrA muscle contraction (P=0.40). During the maximal PFM contraction, activity of all of the abdominal muscles increased between 8.2% (7.1%; midtra) and 32.8% (25.3%; lowtra) of the maximal voluntary contraction (Fig. 4). As expected, the PFM were active to a greater percentage of maximum than the abdominal muscles during the PFM task (interaction: task muscle P<0.0001, post hoc: all P<0.02). There was a trend for lowtra to be more active than RA (P=0.07), and there was no difference between the other muscles (P>0.09). The PFM were strongly active (between 28.4% (23.6%) MVC (maximal Valsalva) and 64.8% (24.3%) MVC (maximal OE contraction)) during all of the abdominal maximal voluntary contraction tasks (OE, OI, RA, TrA, IAP; Fig. 4). The PFM were more active during the maximal PFM and the OE contractions than during the maximal Valsalva. There was no significant difference between other tasks (P>0.05). Discussion The results of this study show that bladder neck elevation during a PFM contraction is influenced by the relationship between PFM activity and IAP. Bladder neck elevation occurred consistently only during PFM and gentle TrA contractions. The abdominal tasks gentle brace, gentle Valsalva and gentle head lift increased the IAP and prevented significant bladder neck elevation. When other muscles, such as the OI, the OE and the RA muscle contracted, the PFM activity was not sufficient to overcome the greater increase in IAP and the bladder neck was not elevated. Petros et al. [8] have presented an alternative view from X-ray observations that depression of the pelvic floor during straining may be due to contraction of a pelvic muscle other than puborectalis rather than pressure. Although our data cannot show whether IAP or some other muscle force causes the reduced depression, the tight relationship between IAP, pelvic floor activity and position change imply that bladder neck elevation occurs when PFM contraction is sufficient to counteract the bladder neck descent caused by downward force of the IAP [24]. Our study also showed that a co-contraction of PFM and TrA is already present with submaximal efforts. Methodological considerations A number of aspects require consideration for interpretation of the results of our study. First, superficial EMG recordings have the potential for crosstalk. For the vaginal probe, crosstalk from the hip rotator and abdominal muscles has

9 74 Int Urogynecol J (2010) 21:69 77 Fig. 3 Relationship between intra-abdominal pressure (IAP: black dots), pelvic floor muscle (PFM) EMG (white dots), abdominal muscle (Abdo) EMG (grey dots) and bladder neck displacement (grey been described [25, 26]. However, PFM recordings have been shown to be minimally affected by crosstalk during low and moderate levels of contraction similar to those used in this study [11]. Strong PFM contraction during the bars) during the experimental tasks. Note that elevation of the bladder neck is related to adequate PFM activity relative to IAP. Mean and standard deviations are shown maximal trunk rotation tasks (OE and OI) with resistance to thigh and arms was unexpected as the IAP increase, and the associated demand on PFM activity was relatively small in these tasks. This could suggest that the high PFM activity Fig. 4 Amplitude of electromyographic (EMG) activity and intraabdominal pressure (IAP) during the standardised maximal voluntary contractions. Note the maximal activity of the target muscles and IAP during the appropriate task and the high pelvic floor muscle (PFM) activity during most tasks. Mean and standard deviation are shown. TrA transversus abdominis, OI obliquus internus abdominis, OE obliquus externus abdominis, RA rectus abdominis

10 Int Urogynecol J (2010) 21: in these tasks may be due to crosstalk from the hip rotator muscles. Second, it is both an advantage and a disadvantage that fine-wire EMG electrodes record the activity from a specific area of a muscle. Fine-wire EMG is specific for a muscle but results do not necessarily apply to the whole muscle. For the purpose of this study, it was a requirement to selectively record activity of TrA, OI, OE and RA. As surface electrodes cannot differentiate between OI and TrA, fine-wire electrodes were the method of choice [19, 27]. Third, it was necessary to standardise the effort of contraction. A modified 15-point Borg scale was used for this purpose. A maximal contraction (perceived Borg scale 15) was performed at the beginning of the study which helped provide an anchor for the contractions at very gentle (2 out of 15) and moderate (8 out of 15) intensities. Fourth, although a consistent bladder base elevation during a correct PFM contraction has been recorded using transabdominal ultrasound [28], this technique is affected by abdominal wall movement. Perineal ultrasound provides a more sensitive assessment of bladder neck displacement [29]. Fifth, we recruited a convenience sample of four nulliparous and five multiparous women. As they did not complain of any pelvic floor disorders and did not demonstrate pelvic organ prolapse or incontinence on ultrasound, we assumed that their motor control would be intact. For the same reason, however, we excluded women with back pain [30] and women who had undergone caesarean section. Sixth, we chose to study tasks with submaximal effort. For comparisons and standardisation, maximal contractions were performed for each appropriate muscle and the IAP. We purposely omitted coughing, laughing, etc. because these are complex and very rapid actions requiring quick inspiration and then forced expiration involving all muscles of the abdominal capsule. As we were interested in the reaction of the pelvic floor and bladder neck when TrA, OI and OE are predominately contracted, effort and tasks with minimal interference of other muscles were selected. Co-ordination between the pelvic floor and abdominal muscles Although several studies have investigated activity of the PFM during abdominal muscle activation, the results have been variable. The findings of the present study may explain the discrepancies in some of the previous literature. Unlike the present study, Bø et al. [31] reported descent of the bladder measured using transabdominal ultrasound during activation of TrA and the PFM in some women. However, the present data show that when TrA muscle contractions are performed accurately with EMG confirmation, the bladder neck is elevated. Without precise concurrent measurement of abdominal muscle activity, differences in recruitment strategies may induce differences in PFM EMG, IAP and bladder neck position. Consistent with our data, Thompson et al. [32] hypothesised that different strategies for recruitment of the abdominal muscles could explain the bladder neck descent during PFM contraction. They showed that symptomatic women who depressed the bladder base during attempted PFM contraction had greater activity of abdominal muscles (recorded with surface electrodes) and less activity of PFM compared to asymptomatic women [33]. Asymptomatic women had more PFM activation and less abdominal activation. The data from our study confirm that bladder neck descent may be due to poor pattern and more exertion of abdominal muscles and therefore higher IAP with insufficient PFM contraction to prevent bladder neck descent. Nearly all previous studies have used maximal contraction of the PFM. Our data show that this involves co-activation of all abdominal muscles and increased IAP. With our intention to examine normal patterns of muscle recruitment, our participants performed standardised tasks with submaximal effort which are not as complex and rapid as coughing and laughing. With regard to pelvic floor muscle training, some programmes focus on strength and randomised controlled trials have shown good outcomes [34]. However, these programmes have not examined whether the bladder neck is elevated or not. It was an aim of this study to examine which contractions and tasks lead to bladder neck elevation as we deem this essential for rehabilitation. Several authors have suggested that PFM is accompanied by contraction of TrA [9, 10, 14]. All of the earlier studies involved higher efforts and therefore more co-contraction of other abdominal muscles. This is likely to explain the variation in patterns of activity that have been observed (e.g. activation of all of the abdominal muscles) [9, 33]. In some studies [10], it is unclear how much effort was used as no standardisation of a maximum contraction was reported. Our study shows differences in the activation of the middle and lower regions of TrA when contractions of the PFM and TrA were performed. Previous work has identified morphological [35] and functional differences [20, 36 38] between regions of the abdominal wall. These differences in activity of the parts of TrA have clinical importance as TrA is usually palpated at the level of the anterior superior iliac spine. Our findings suggest that this is appropriate as the lower region of TrA was more active than the fibres in the more cranial middle region during PFM and gentle TrA contractions. The present study confirms previous data on activation of the PFM during the Valsalva manoeuvre even with submaximal effort [3, 39]. Using intraurethral pressure measurements, it has indirectly been demonstrated that there is an active component during increased abdominal pressure during coughing, e.g. [40, 41]. Although relaxa-

11 76 Int Urogynecol J (2010) 21:69 77 tion of the PFM is required to observe maximum bladder neck descent during clinical evaluation of the pelvic organ support including the condition of the connective tissue, this does not seem to be the natural tendency and is not achieved by all women [39, 42]. As the objective of the present study was to evaluate the automatic activation of PFM, no instruction for PFM relaxation was provided during the Valsalva task. It can be argued that a Valsalva manoeuvre is different from straining per definition. A Valsalva method is used to test the patency of the Eustachian tubes or to equalise pressure. Simultaneous contraction of the pelvic floor appears necessary to avoid any involuntary leakage of urine or flatus and stool. During straining, e.g. when evaluating the Valsalva leak point pressure or pelvic organ support, a contracted PFM acts as a confounder [42]. Implications for pelvic floor rehabilitation Our findings have important clinical implications for PFM rehabilitation. First, the outcomes of our study suggest that maximal contraction of the PFM is associated with activity of all abdominal muscles which increases the IAP. As women with stress urinary incontinence have been shown to have greater activation of OE during voluntary PFM contractions [33] and during postural perturbations [18], the associated IAP increase during maximal PFM contractions should be considered in the rehabilitation. Attempts to normalise the co-ordination between abdominal and PFM may be an important step in rehabilitation. This requires multifaceted assessment of the pelvic floor with consideration of bladder neck position, IAP and activity of the PFM and abdominal muscles. The second clinical implication is that the initial activation of the PFM could be achieved by gentle contraction of TrA in some women. This might be helpful in women who have problems with the perception of their PF and who are not able to contract their PFM because of decreased PF awareness. However, it would be necessary to confirm that the PF contracts during this task as it cannot be assumed and work is required with women with PFM dysfunction to ensure that the same relationship exists in that population. An early step in training such as this could then be followed by appropriate training of PFM contraction for strength and co-ordination which are likely to be important for control of continence. Conclusion In women without pelvic floor disorders, we demonstrated that, although there was co-contraction of the lower part of the TrA muscle and the pelvic floor muscle during all tasks, bladder neck elevation occurred only consistently during PFM and TrA contractions. When the OI was also recruited with Valsalva, head lift and brace, the PFM co-contraction was not sufficient to overcome the greater increase in IAP to elevate the bladder neck. Because maximal PFM contraction is associated with activity of all abdominal muscles and considerable increase of the IAP, we chose submaximal efforts and confirmed that PFM and TrA are recruited concomitantly. Acknowledgement Paul Hodges is an NHMRC Principal Research Fellow (Australian National Health and Medical Research Council). Conflicts of interest References None. 1. Peschers U, Schaer G, Anthuber C, Delancey JO, Schuessler B (1996) Changes in vesical neck mobility following vaginal delivery. Obstet Gynecol 88: Dietz HP, Steensma AB, Vancaillie TG (2003) Levator function in nulliparous women. Int Urogynecol J Pelvic Floor Dysfunct 14: Schaer GN, Koechli OR, Schuessler B, Haller U (1995) Perineal ultrasound for evaluating the bladder neck in urinary stress incontinence. Obstet Gynecol 85: King JK, Freeman RM (1998) Is antenatal bladder neck mobility a risk factor for postpartum stress incontinence? Br J Obstet Gynecol 105: Baessler K, Miska K, Draths R, Schuessler B (2005) Effects of voluntary pelvic floor contraction and relaxation on the urethral closure pressure. Int Urogynecol J Pelvic Floor Dysfunct 16: DeLancey JO (1994) Structural support of the urethra as it relates to stress urinary incontinence: the hammock hypothesis. Am J Obstet Gynecol 170: Howard D, Miller JM, Delancey JO, Ashton-Miller JA (2000) Differential effects of cough, Valsalva and continence status on vesical neck movement. Obstet Gynecol 95: Petros PE, Ulmsten U (1997) Role of the pelvic floor in bladder neck opening and closure I: muscle forces. Int Urogynecol J Pelvic Floor Dysfunct 8: Sapsford RR, Hodges PW, Richardson CA, Cooper DH, Markwell SJ, Jull GA (2001) Co-activation of the abdominal and pelvic floor muscles during voluntary exercises. Neurourol Urodyn 20: Neumann P, Gill V (2002) Pelvic floor and abdominal muscle interaction: EMG activity and intra-abdominal pressure. Int Urogynecol J Pelvic Floor Dysfunct 13: Hodges PW, Sapsford R, Pengel LH (2007) Postural and respiratory functions of the pelvic floor muscles. Neurourol Urodyn 26: Hemborg B, Moritz U, Hamberg J, Holmstrom E, Lowing H, Akesson I (1985) Intra-abdominal pressure and trunk muscle activity during lifting. III. Effect of abdominal muscle training in chronic low-back patients. Scand J Rehabil Med 17: Smith MD, Coppieters MW, Hodges PW (2007) Postural activity of the pelvic floor muscles is delayed during rapid arm movements in women with stress urinary incontinence. Int Urogynecol J Pelvic Floor Dysfunct 18: Sapsford RR, Hodges PW (2001) Contraction of the pelvic floor muscles during abdominal maneuvers. Arch Phys Med Rehabil 82:1081 8

12 Int Urogynecol J (2010) 21: Thompson JA, O Sullivan PB, Briffa K, Neumann P, Court S (2005) Assessment of pelvic floor movement using transabdominal and transperineal ultrasound. Int Urogynecol J Pelvic Floor Dysfunct 16: Baessler K, O Neill S, Maher CF, Battistutta D (2009) Australian pelvic floor questionnaire: a validated interviewer-administered pelvic floor questionnaire for routine clinic and research. Int Urogynecol J Pelvic Floor Dysfunct 20: Dietz HP, Haylen BT, Broome J (2001) Ultrasound in the quantification of female pelvic organ prolapse. Ultrasound Obstet Gynecol 18: Smith MD, Coppieters MW, Hodges PW (2007) Postural response of the pelvic floor and abdominal muscles in women with and without incontinence. Neurourol Urodyn 26: Hodges PW, Richardson CA (1997) Feedforward contraction of transversus abdominis is not influenced by the direction of arm movement. Exp Brain Res 114: Urquhart DM, Hodges PW (2005) Differential activity of regions of transversus abdominis during trunk rotation. Eur Spine J 14: Ng JK-F, Kippers V, Richardson CA (1998) Muscle fibre orientation of abdominal muscles and suggested surface EMG electrode positions. Electromyogr Clin Neurophysiol 38: Borg G, Dahlstrom H (1962) The reliability and validity of a physical work test. Acta Physiol Scand 55: Hides JA, Richardson CA, Jull GA (1998) Use of real-time ultrasound imaging for feedback in rehabilitation. Man Ther 3: Wise B, Cutner A, Cardozo L, Abbott D, Burton G (1992) The assessment of bladder neck movement in postpartum women using perineal ultrasonography. Ultrasound Obstet Gynecol 2: Peschers UM, Gingelmaier A, Jundt K, Leib B, Dimpfl T (2001) Evaluation of pelvic floor muscle strength using four different techniques. Int Urogynecol J Pelvic Floor Dysfunct 12: Workman DE, Cassisi JE, Dougherty MC (1993) Validation of surface EMG as a measure of intravaginal and intra-abdominal activity: implications for biofeedback-assisted Kegel exercises. Psychophysiology 30: Hodges PW, Richardson CA (1996) Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 21: Sherburn M, Murphy CA, Carroll S, Allen TJ, Galea MP (2005) Investigation of transabdominal real-time ultrasound to visualise the muscles of the pelvic floor. Aust J Physiother 51: Thompson JA, O Sullivan PB, Briffa NK, Neumann P (2007) Comparison of transperineal and transabdominal ultrasound in the assessment of voluntary pelvic floor muscle contractions and functional manoeuvres in continent and incontinent women. Int Urogynecol J Pelvic Floor Dysfunct 18: Hides JA, Richardson CA, Jull GA (1996) Multifidus recovery is not automatic after resolution of acute, first-episode low back pain. Spine 21: Bo K, Sherburn M, Allen T (2003) Transabdominal ultrasound measurement of pelvic floor muscle activity when activated directly or via a transversus abdominis muscle contraction. Neurourol Urodyn 22: Thompson JA, O Sullivan PB, Briffa NK, Neumann P (2006) Differences in muscle activation patterns during pelvic floor muscle contraction and Valsalva maneuver. Neurourol Urodyn 25: Thompson JA, O Sullivan PB, Briffa NK, Neumann P (2006) Altered muscle activation patterns in symptomatic women during pelvic floor muscle contraction and Valsalva manoeuvre. Neurourol Urodyn 25: Bo K, Talseth T, Holme I (1999) Single blind, randomised controlled trial of pelvic floor exercises, electrical stimulation, vaginal cones and no treatment in management of genuine stress incontinence in women. BMJ 318: Urquhart DM, Barker PJ, Hodges PW, Story IH, Briggs CA (2005) Regional morphology of the transversus abdominis and obliquus internus and externus abdominis muscles. Clin Biomech 20: Hodges P, Cresswell A, Thorstensson A (1999) Preparatory trunk motion accompanies rapid upper limb movement. Exp Brain Res 124: Urquhart DM, Hodges PW, Allen TJ, Story IH (2005) Abdominal muscle recruitment during a range of voluntary exercises. Man Ther 10: Urquhart DM, Hodges PW, Story IH (2005) Postural activity of the abdominal muscles varies between regions of these muscles and between body positions. Gait Posture 22: Peschers UM, Fanger G, Schaer GN, Vodusek DB, Delancey JO, Schuessler B (2002) Bladder neck mobility in continent and incontinent nulliparous women. BJOG 108: Constantinou CE (1985) Resting and stress urethral pressures as a clinical guide to the mechanisms of continence in the female patient. Urol Clin North Am 12: Petros PE, Ulmsten U (1995) Urethral pressure increase on effort originates from the urethra, and continence from musculovaginal closure. Neurourol Urodyn 14: Ornö AK, Dietz HP (2007) Levator co-activation is a significant confounder of pelvic organ descent on Valsalva maneuver. Ultrasound Obstet Gynecol 30:346 50

13 IJSPT ORIGINAL RESEARCH ELECTROMYOGRAPHIC ANALYSIS OF GLUTEUS MEDIUS AND GLUTEUS MAXIMUS DURING REHABILITATION EXERCISES Kristen Boren, DPT 1 Cara Conrey, DPT 1 Jennifer Le Coguic, DPT 1 Lindsey Paprocki, DPT 1 Michael Voight, PT, DHSc, SCS, OCS, ATC, CSCS 1 T. Kevin Robinson, PT, DSc, OCS 1 ABSTRACT Purpose/Background: Previous research studies by Bolga, Ayotte, and Distefano have examined the level of muscle recruitment of the gluteal muscles for various clinical exercises; however, there has been no cross comparison among the top exercises from each study. The purpose of this study is to compare top exercises from these studies as well as several other commonly performed clinical exercises to determine which exercises recruit the gluteal muscles, specifically the gluteus medius and maximus, most effectively. Methods: Twenty-six healthy subjects participated in this study. Surface EMG electrodes were placed on gluteus medius and maximus to measure muscle activity during 18 exercises. Maximal voluntary muscle contraction (MVIC) was established for each muscle group in order to express each exercise as a percentage of MVIC and allow standardized comparison across subjects. EMG data were analyzed using a root-mean-square algorithm and smoothed with a 50 millisecond time reference. Rank ordering of the exercises was performed utilizing the average percent MVIC peak activity for each exercise. Results: Twenty-four subjects satisfied all eligibility criteria and consented to participate in the research study. Five of the exercises produced greater than 70%MVIC of the gluteus medius muscle. In rank order from highest EMG value to lowest, these exercises were: side plank abduction with dominant leg on bottom (103%MVIC), side plank abduction with dominant leg on top (89%MVIC), single limb squat (82%MVIC), clamshell (hip clam) progression 4 (77%MVIC), and font plank with hip extension (75%MVIC). Five of the exercises recruited gluteus maximus with values greater than 70%MVIC. In rank order from highest EMG value to lowest, these exercises were: front plank with hip extension (106%MVIC), gluteal squeeze (81%MVIC), side plank abduction with dominant leg on top (73%MVIC), side plank abduction with dominant leg on bottom (71%MVIC), and single limb squat (71%MVIC). Four of the exercises produced greater than 70%MVIC for both gluteus maximus and medius muscles. Conclusions: Higher %MVIC values achieved during performance of exercises correlate to muscle hypertrophy. 20,22 By knowing the %MVIC of the gluteal musculature that occurs during various exercises, potential for strengthening of the gluteal muscles can be inferred. Additionally, exercises may be rank ordered to appropriately challenge the gluteal musculature during rehabilitation. Keywords: gluteus medius, gluteus maximus, muscle recruitment, rehabilitation exercise 1 Belmont University, Nashville, TN, USA This study was approved by the Institutional Review Board at Belmont University and informed consent was obtained from all subjects. This project was completed for partial fulfillment of a degree. The authors would like to thank Dr. John Halle and Dr. Patrick Sells for their advice and assistance with statistical analysis, as well as Mr. Anthony Carey for the donation of Core-Tex TM equipment. CORRESPONDING AUTHOR Lindsey Paprocki 1351 Emir Street, Green Bay, WI paprockil@sbcglobal.net The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 206

14 INTRODUCTION The lower extremity functions in a kinematic chain, leading many researchers in recent years to examine the mechanical effect of weak proximal musculature on more distal segments. 1,2 Previous research by Distefano, 3 Bolgla, 4 and Ayotte 5 has sought to determine the most appropriate exercises to strengthen the gluteal muscles due to their role in maintaining a level pelvis and preventing hip adduction and internal rotation during single limb support. 1,6 Measurement of such femoral torsion and pelvic rotation in the transverse plane, along with measurement of pelvic tilt in the sagittal plane can indicate abnormal alignment of the hip joint. 7 Numerous pathologies have been described which are related to the inability to maintain proper alignment of the pelvis and the femur, including: tibial stress fracture, 8 low back pain, 9,10 iliotibial band friction syndrome, 1,11 anterior cruciate ligament injury, 1,12 and patellofemoral pathology. 2,13,14,15,16,17 While Distefano, 3 Bolgla, 4 and Ayotte 5 have examined a wide range of exercises used to strengthen the hip musculature, to the knowledge of the authors, no cross comparison amongst the top exercises from each study has been performed. Similar to Distefano, 3 Ayotte, 4 and Bolgla, 5 exercises examined in the current study were rank ordered according to their recruitment of specific gluteal musculature and expressed as a percent of the subject s maximum volitional isometric contraction (MVIC). By knowing the approximate percentage of MVIC (%MVIC) recruitment of each of the gluteal muscles in a wide variety of exercises, the exercises may be ranked to appropriately challenge the gluteal musculature. MVIC was established in the standard manual muscle testing positions for gluteus medius and maximus, as described by Daniels and Worthingham. 18 The use of the sidelying abduction position is supported by the results of Widler, 19 where similarity in EMG activity for weight bearing and sidelying abduction (ICC s and for the respective positions) demonstrated that it is acceptable to use the MVIC value obtained during the standard manual muscle test position in order to establish a percentage MVIC for a weight bearing exercise. Several previously published research articles helped to establish the parameters for determining a sufficient level of muscle activation for strength gains referenced in the current study. Anderson found that in order for strengthening adaptation to occur, muscle stimuli of at least 40-60% of a subject s MVIC must occur. 20 When quantifying muscular strength, work by Visser correlates the use of a MVIC and a one-repetition maximum. 21 In order to gain maximal muscular hypertrophy, Fry s work suggests an 80-95% of a subject s one repetition maximum must be achieved. 22 Based on the work by Anderson, 20 Visser, 21 and Fry, 22 for the purposes of this study, exercises producing greater than 70%MVIC were deemed acceptable for enhancement of strength. Distefano examined electromyography (EMG) signal amplitude normalized values of gluteus medius and gluteus maximus muscles during exercises of varying difficulty in order to determine which exercises most effectively recruit these muscles. 3 Rank order of exercises and %MVIC of Distefano s study can be viewed in Table 1. Of the top five exercises for the gluteus medius described by Distefano, the authors of the current study chose to reexamine sidelying hip abduction, single limb squat, and the single limb deadlift. Lateral band walk was not included in the current study as the researchers wished to only examine exercises that required no external resistance. Research by Bolgla and Uhl also examined the magnitude of hip abductor muscle activation during rehabilitative exercises. 4 Their results may be viewed in Table 2. Of the exercises studied by Bolgla et al, the authors of the current study chose only to look at the pelvic drop and sidelying hip abduction. These two exercises were chosen since the primary intention of the current study was to compare an exercise s recruitment of the gluteal musculature, and not the activation effects of weight bearing versus non-weight bearing on the musculature. Finally, Ayotte et al. used EMG to analyze lower extremity muscle activation of the pelvic stabilizers as well as the quadriceps complex during five unilateral weight bearing exercises, 5 displayed in Table 3. The authors of the current study elected to forgo analyzing a single-limb wall squat and a single-limb minisquat due to their similarity to the single-limb squat. Forward step-up and lateral step-up were included in the current analysis. The current study serves to compare top exercises from these previously published studies, as well as several other commonly performed The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 207

15 Table 1. Findings of Distefano et al. 3 Values are described as %MVIC, followed by rank in parentheses. Table 2. Findings by Bolgla and Uhl, 4 represented as %MVIC. Table 3. Findings of Ayotte et al. 5 Values are described as %MVIC, followed by rank in parentheses. clinical exercises in order to determine the exercises that are most effective at recruiting the gluteus maximus and medius. METHODS Subjects This study was approved by the Institutional Review Board of Belmont University. A total of 26 subjects were recruited from within the university and surrounding community through flyers and word of mouth. Healthy subjects who were able to perform exercise for approximately one hour were included in the study and reported to the laboratory for a single testing session. At this time they completed an informed consent form as well as a health history form and comprehensive lower quarter screen to identify exclusionary criteria. Pain when performing exercises, current symptoms of injury, history of ACL injury or any lower extremity surgery within past two years, and age of less than 21 years were criteriariteria for exclusion. Testing Procedures EMG data were collected and analyzed on the dominant leg, identified by which leg the subject used to kick a ball. 3,5,23 Alcohol wipes were used to clean the skin over the gluteal region prior to electrode placement. Schiller Blue Surface electrodes (Schiller America Inc.; Doral, FL) were placed over the gluteus medius and gluteus maximus muscles of the subject s dominant The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 208

16 Figure 1. Maximum voluntary isometric contraction testing example set up. side, 4 per standard EMG protocol. 24 In order to ensure consistent electrode placement throughout testing, electrodes were secured with surgical tape. Placement was confirmed by viewing EMG signals while separately activating each muscle. Subjects then performed a sub-maximal warm-up for five minutes on a stationary bicycle while watching a brief video of the exercises to be performed in order to familiarize subjects with exercise technique. A five-second MVIC was performed three times in the standard manual muscle testing protocol positions for each gluteal muscle 18,19 with one minute of rest between each contraction. A strap was secured around the distal femur during muscle testing for both muscles to ensure standardization of resistance (Figure 1). Verbal encouragement was given with each trial. Exercise order was randomized using a random pattern generator 25 in order to avoid any order bias due to fatigue. Subjects were barefoot while performing exercises to prevent any potential variations that may have occurred due to footwear. Two minutes of rest was given between the performance of each exercise. Subjects performed eight repetitions of each exercise, three practice repetitions and five repetitions that were used for data collection. Exercises were performed to a metronome set at 60 beats per minute to standardize the rate of movement across subjects. To replicate a clinical setting, researchers chose to use visual analysis of movement to ensure proper exercise technique rather than an electrogoniometer or movement analysis software since both of these Figure 2. CorTex TM equipment (Performance Dynamics, San Diego, CA). procedures are unlikely to be available in a clinic. To ensure proper exercise technique, each subject was allowed three practice repetitions prior to data collection and any necessary verbal and tactile cues by the instructing researcher. A description of each exercise may be found in Appendix A. After completing all exercises, the subject s MVIC was reassessed to ensure electrodes had not been displaced during testing. The equipment used for the conditions which required an unstable surface is the Core-Tex Balance Trainer (Performance Dynamics; San Diego, CA), a new piece of exercise equipment which is a platform mounted on a half-sphere atop a circular basin lined with ball bearings, creating an unstable and rapidly accelerating surface (Figure 2). The Core-Tex was developed to train a healthy fitness population; however, it may also be used to train individuals during rehabilitation in a clinical setting. Data Analysis All data were rectified and smoothed using a rootmean-square algorithm, and smoothed with a 50 millisecond (msec) time reference. Peak amplitudes were averaged over a 100 msec window of time, 50 msec prior to peak and 50 msec after the peak. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 209

17 Table 4. Results for Gluteus Medius recruitment, %MVIC and rank for all exercises. To determine MVIC, the middle 3/5 ths time for each manual muscle test trial was isolated and the peak value determined. The highest peak value out of the three trials was recorded and determined to be the MVIC. In order to establish %MVIC for each exercise performed by an individual subject, data were collected for the last five repetitions of each exercise. If the EMG data were clearly cyclic, the middle three repetitions were analyzed. If it was difficult to determine when a repetition started and stopped on visual analysis of EMG data, then the middle 3/5 ths of the total time to perform the five repetitions was analyzed. The highest peak out of the three repetitions was then divided by MVIC to yield %MVIC for that individual. To determine %MVIC values for rank ordering of exercises, the %MVIC for each muscle was averaged between all subjects for each exercise. RESULTS Twenty-four subjects satisfied all eligibility criteria and consented to participate in the research study. Data from one subject were excluded due to faulty data from the EMG leads for both muscles, and data from another subject were excluded due to faulty data from the EMG lead for gluteus maximus only. There were a few other isolated instances of faulty data from EMG leads, in which case the subject s data were excluded from analysis for that specific exercise. The number of subjects included in data analysis for each exercise can be referenced in Tables 4 and 5. Due to the advanced level of some of the exercises included in the current study, such as single limb bridge on unstable surface and side plank, some subjects were unable to successfully complete all exercises. In these instances, subject data were not included in data analysis for that specific exercise. Peak amplitudes, The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 210

18 Table 5. Results for Gluteus Maximus recruitment, %MVIC and rank for all exercises. Table 6. Top exercises for muscle activation of both gluteus medius and gluteus maximus (>70% MVIC). expressed as %MVIC for gluteus medius and gluteus maximus, are rank ordered in Tables 4 and 5. Five of the exercises produced greater than 70%MVIC of the gluteus medius muscle. In rank order from highest EMG value to lowest, these exercises were: side plank abduction with dominant leg on bottom (103%MVIC), side plank abduction with dominant leg on top (89%MVIC), single limb squat (82%MVIC), clamshell (hip clam) progression 4 (77%MVIC), and font plank with hip extension (75%MVIC). Five of the exercises recruited gluteus maximus with values greater than 70%MVIC. In rank order from highest EMG value to lowest, these exercises were: front plank with hip extension (106%MVIC), gluteal squeeze (81%MVIC), side plank abduction with dominant leg on top (73%MVIC), side plank abduction with dominant leg on bottom (71%MVIC), and single limb squat (71%MVIC). Table 6 displays the exercises that The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 211

19 produced greater than 70%MVIC for both gluteus medius and maximus muscles. These exercises included front plank with hip extension (75%MVIC, 106%MVIC), side plank abduction with dominant leg on top (89%MVIC, 73%MVIC), side plank abduction with dominant leg on bottom (103%MVIC, 71%MVIC), and single limb squat (82%MVIC, 71%MVIC) for gluteus medius and maximus respectively. DISCUSSION The main objective of this study was to examine muscle activity during common clinical exercises used to strengthen the gluteus medius and gluteus maximus muscles. This study sought to analyze and compare information reported in previous studies by Distefano, Bolga, and Ayotte regarding ranking of various therapeutic exercises using %MVIC. The secondary objective was to describe %MVIC for other commonly used therapeutic exercises not previously reported upon. The authors of this study chose to examine peak amplitude averaged over a 100 ms window, 50 ms prior to peak and 50 ms after the peak, during repetitions five, six and seven, the highest of which was converted to %MVIC. This methodology is similar to studies by both Distefano 3 and Bolgla. 4 Ayotte et al. averaged EMG activity over a 1.5 sec window during the concentric phase of each exercise. 5 Due to slight differences in data collection and data analysis between the current study, and studies conducted by Distefano, Bolgla and Ayotte, interpretation of results and similarities across studies will predominantly address the sequence of rank order as opposed to absolute values for the %MVIC. 3,4,5 There were two exercises where %MVIC was found to be higher than MVIC, side plank abduction with dominant leg down (103%MVIC) for gluteus medius and front plank with hip extension (106%MVIC) for gluteus maximus. There are several possibilities as to why these findings may have occurred. One possibility is that subjects lacked sufficient motivation to perform a true maximal contraction during MVIC testing, despite the fact that verbal encouragement was given to all subjects during max testing of both muscles. Another possibility is that subjects were not able to truly give a maximum effort during the manual muscle test. Authors of previous research have reported that in order to obtain a true maximum contraction, it is necessary to superimpose an interpolated twitch, which is an electrically stimulated contraction, on top of the maximum voluntary contraction. 26 Current research in electrophysiology is further examining this phenomenon with mixed results regarding sensitivity of various interpolated twitch techniques, differences in methodology, and interpretation of their results. 27,28,29 Future researchers using MVIC for standardization across subjects should follow this research closely in order to ensure the most accurate methodology is used for establishing maximal voluntary muscle contractions. A final possibility is that with these exercises there was substantial co-contraction of the core musculature, which may have led to higher values than could be obtained during isolated volitional contraction. In the MMT positions used to establish MVIC the pelvis is stabilized against the surface of the table with relatively isolated muscle recruitment. In both of the above mentioned exercises, the pelvis does not have external support and higher EMG values could reflect increased activity due to an increased need for stabilization resulting in synergistic co-contraction. Future research may need to examine differences in muscle recruitment and activation patterns in exercises that test isolated muscle function versus ones that require core stabilization resulting in co-contraction. Gluteus Medius Table 7 depicts the top gluteus medius exercises determined by the authors of the current study as referenced to the exercises examined in studies performed by Distefano, 3 Bolgla, 4 and Ayotte. 5 The authors of the current study found highest %MVIC peak values for side plank abduction with dominant leg on bottom (103%MVIC), side plank abduction with dominant leg on top (89%MVIC), single limb squat (82%MVIC), clamshell progression 4 (77%MVIC), and front plank (75%MVIC) as outlined in Table 7. Four of the top five exercises were not previously examined by Distefano, 3 Bolga, 4 or Ayotte. 5 All of these exercises exhibited greater than 70%MVIC, the peak amplitude necessary for enhancement of strength, suggesting they may have benefits for gluteus medius strengthening. However, these exercises are all very challenging and would not be appropriate for initial strengthening in patients with weak core musculature due to their high degree of difficulty and the amount of core stabilization required. The possible exception may be clamshell progression 4, due to the stabilization provided to the subject when lying on the floor to perform the The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 212

20 Table 7. Comparison of rank order of exercises for recruitment of gluteus medius between the current study and Distefano, 3 Bolgla, 4 and Ayotte, 5 using %MVIC *Single-limb wall squat exercise. While the top exercises in this study produced the greatest peak amplitude EMG values, it is also important to consider functional demands and dosage when selecting an exercise for muscle training and strengthening, especially in early stages of rehabilitation of a weak or under-recruited muscle. The top gluteus medius exercises from Distefano s study were sidelying hip abduction (81%MVIC), single-limb squat (64%MVIC), and single limb dead lift (58%MVIC). 3 With the exception of single limb squat, the current study found similar rank order with values of 63%MVIC, 82%MVIC, and 56%MVIC respectively. Of note, Distefano s subjects performed the single limb squat to a predetermined knee flexion angle of approximately 30 degrees, 3 while the current study had the subjects perform the exercise to a predetermined chair height of 47 cm. This difference in methodology may account for the difference in findings across the two studies. The methodology used by Distefano may allow for greater normalization, as squatting to a predetermined knee flexion angle allows for equal challenge to all subjects, where as squatting to a predetermined height creates a greater challenge for taller subjects. Bolga s top exercise for gluteus medius was the pelvic drop (57%MVIC). 4 The current study found a similar value at 58%MVIC, although this exercise was ranked 11 th out of the 22 exercises evaluated. This exercise should not be discounted; however, as it is a functional training exercise for pelvic stabilization in single limb stance, and many gait abnormalities and lower extremity pathologies are the result of the gluteus medius muscle s inability to properly and effectively stabilize the pelvis during single limb stance. Bolga found sidelying abduction to have a value of 42%MVIC, 4 which is significantly lower than the findings in either the Distefano 3 or the current study. In general, qualitative movement analysis during performance of sidelying abduction reveals poor technique with frequent substitution using the tensor fascia lata muscle demonstrated through increased hip flexion during abduction, which may have accounted for the low value found in the Bolga study. 6 Furthermore, subjects in both the Distefano and the Bolgla study maintained the bottom leg in neutral hip extension and knee extension, 3,4 while subjects in the current study were allowed to flex the bottom hip and knee in order to provide greater support and stabilization during abduction of the top leg. Ayotte s top exercise was the unilateral wall squat (52%MVIC), 5 which is comparable to the single limb squat, ranking in the top three exercises in both the current study and in Distefano s study, 3 although the external stabilization provided in the unilateral wall squat should be considered. Ayotte ranked forward step-up (44%MVIC) higher than lateral step-up (38%MVIC), 5 whereas the authors of the current study ranked lateral step-up (60%MVIC) higher than forward step-up (55%MVIC). It should be noted that subjects were allowed upper extremity external support during the exercise in Ayotte s study which may The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 213

21 Table 8. Comparison of rank order of exercises for recruitment of gluteus maximus between the current study and Distefano, 3 and Ayotte, 5 using %MVIC. *Single-limb squat account for these differences, 5 along with differences in data analysis described previously. Gluteus Maximus Table 8 depicts the top exercises for gluteus maximus of the current study. These include front plank with hip extension (106%MVIC), gluteal squeeze (81%MVIC), side plank abduction with dominant leg on top (73%MVIC), side plank abduction with dominant leg on bottom (71%MVIC), and single limb squat (71%MVIC). The top four exercises from the current study were not performed in other studies. Bolgla s study did not include assessment of performance of the gluteus maximus so will not be included in the discussion below. 4 Distefano s top exercises were single limb squat (59%MVIC), single limb dead lift (59%MVIC), and sidelying hip abduction (39%MVIC). 3 Subjects performing these same exercises in the current study produced results of 71%MVIC, 59%MVIC, and 51%MVIC, respectively, demonstrating the same rank order of muscle activity as these exercises in the Distefano study. 3 The only differences in rank ordering between the current study and Distefano s for gluteus maximus were between clamshell progression 1 and sidelying abduction; 3 however, within each study there was less than 5%MVIC difference for each exercise when determining rank order (Table 8). As previously noted, differences in technique and substitution are common occurrences during the performance of sidelying abduction which may account for the differences found between the two studies. Ayotte ranked forward step-up (74%MVIC) higher than lateral step-up (56%MVIC), 5 whereas the current study ranked lateral step-up (64%MVIC) higher than forward step-up (55%MVIC). Again, differences could be attributed to variances in technique or the ability of subjects in Ayotte s study to use external upper extremity support 5 as well as differences in data analysis. The low ranking for stable single limb bridge (11 th ) and unstable single limb bridge (14 th ) was somewhat surprising as both are common exercises used clinically for gluteus maximus strengthening. There were several instances of subjects reporting hamstring cramping during bridging on the unstable surface, which led the researchers to suspect substitution with the hamstrings during this exercise. The same may hold true for bridging on the stable surface, however there were fewer complaints. Future studies should examine muscle recruitment and activation patterns of gluteus maximus and the hamstrings during various bridging activities. The effect of a subject s attention to volitional contraction of a muscle during an exercise should also be considered. The gluteal squeeze was the only exercise where verbal cues were explicitly given to maximally contract the gluteal muscles while performing the exercise, which could possibly have contributed to its high ranking for performance by the gluteus maximus. Future research should examine the difference in amount, if any, noted in muscle recruitment when verbal instructions are given to concentrate on the muscle contraction while The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 214

22 performing the exercise versus no verbal instructions during performance. The effects of tone of voice, volume of cues, and frequency of verbal cueing are unknown. CONCLUSION Anderson and Fry have previously reported that higher %MVIC values with exercises correlate to muscle hypertrophy. 20,22 By knowing the %MVIC of the gluteus maximus and medius that occurs during various exercises, the potential for strengthening these muscles can be inferred. Subsequently, exercises may be ranked to appropriately challenge the gluteus maximus and medius during rehabilitation. The authors of the current study found patterns within their results consistent with previous research published by Distefano and Bolgla. 3,4 The authors conclude that differences in data collection and analysis as well as the use of external upper extremity support may have accounted for the differences noted between the current study and the study by Ayotte. 5 One of the purposes of the current study was to provide a rank ordered list of exercises for the recruitment of the gluteus maximus and medius. These rank ordered lists may help form the basis for a graded rehabilitation program. For patients early in the rehabilitation process, the clinician should systematically determine which muscle they are wishing to strengthen and use less difficult (lower %MVIC) exercises. In order to maximally challenge a patient s gluteus maximus and medius, the authors recommend using a front plank with hip extension, a single limb squat, and a side plank on either extremity with hip abduction. REFERENCES 1. Leetun D, Ireland M, Wilison J, et al. Core Stability Measures as Risk Factors for Lower Extremity Injury in Athletes. Med Sci Sports Exercise. 2004; 36: Souza R, Powers C. Differences in Hip Kinematics, Muscle Strength, and Muscle Activation Between Subjects With and Without Patellofemoral Pain. J Orthop Sports Phys Ther. 2009; 39: Distefano L, Blackburn J, Marshall S, et al. Gluteal Activation During Common Therapeutic Exercises. J Orthop Sports Phys Ther. 2009; 39: Bolgla L, Uhl T. Electromyographic Analysis of Hip Rehabilitation Exercises in a Group of Healthy Subjects. J Orthop Sports Phyl Ther. 2005; 35: Ayotte N, Stetts D, Keenan G, et al. Electromyographical Analysis of Selected Lower Extremity Muscles During 5 Unilateral Weight- Bearing Exercises. J Orthop Sports Phys Ther. 2007; 37: Grimaldi A. Assessing Lateral Stability of the Hip and Pelvis. Manual Therapy. 2010, Pages Levangie P. The Hip complex. In: Levangie P, Norkin C. Joint Structure and Function: A Comprehensive Analysis, 4 th ed. Philadelphia, PA: F.A. Davis Company; 2005: Milner C, Hamill J, Davis I. Distinct Hip and Rearfoot Kinematics in Female Runners With a History of Tibial Stress Fracture. J Orthop Sports Phys Ther. 2010; 40: Nelson-Wong E, Flynn T, Callaghan J. Development of Active Hip Abduction as a Screening Test for Identifying Occupational Low Back Pain. J Orthop Sports Phys Ther. 2009; 39: Nelson-Wong E, Gregory D, Winter D, et al. Gluteus Medius Muscle Activation Patterns as a Predictor of Low Back Pain During Standing. Clin Bio. 2008; 23: Ferber R, Noehren B, Hamill J, et al. Competitive Female Runners With a History of Iliotibial Band Syndrome Demonstrate Atypical Hip and Knee Kinematics. J Orthop Phys Ther. 2010; 40: Hewett, T, Myer G, Ford K, Anterior Cruciate Ligament Injuries in Female Athletes: Part 1, Mechanisms and Risk Factors. Am J Sports Med. 2006; 34: Magalhaes E, Fukuda T, Sacramento S, et al. A Comparison of Hip Strength Between Sedentary Females With and Without Patellofemoral Pain Syndrome. J Orthop Sports Phys Ther. 2010; 40: Souza R, Draper C, Fredericson M, et al. Femur Rotation and Patellofemoral Joint Kinematics: A Weight- Bearing Magnetic Resonance Imaging Analysis. J Orthop Sports Phys Ther. 2010; 40: McKenzie K, Galea V, Wessel J, et al. Lower Extremity Kinematics of Females With Patellofemoral Pain Syndrome While Stair Stepping. J OrthopSports Phys Ther 2010; 40: Fukuda T, Rossetto F, Magalhaes E, et al. Short-Term Effects of Hip Abductors and Lateral Rotators Strengthening in Females With Patellofemoral Pain Syndrome: A Randomized Controlled Clinical Trial. J Orthop Sports Phys Ther. 2010; 40: Ireland M, Wilson J, Bellantyne B, et al. Hip Strength in Females With and Without Patellofemoral Pain. J Orthop Sports Phys Ther. 2003; 33: The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 215

23 18. Hislop H, Montgomery J. Daniels and Worthingham s Muscle Testing: Techniques of Manual Examination. St. Louis, MO; Elsevier Saunders; Widler K, Glatthorn J, Bizzini M, et al. Assessment of Hip Abductor Muscle Strength. A Validity and Reliability Study. J Bone Joint Surg Am. 2009; 91: Anderson L, Magnusson S, Nielsen M, et al. Neuromuscular Activation in Conventional Therapeutic Exercises and Heavy Resistance Exercises: Implications for Rehabilitation. Phys Ther. 2006; 86: Visser J, Mans E, van den Berg-Vos RM, et al. Comparison of Maximal Voluntary Isometric Contraction and Hand-held Dynamometry in Measuring Muscle Strength of Patients with Progressive Lower Motor Neuron Syndrome. Neuromuscul Disord. 2003; 13: Fry, A. The Role of Resistance Exercise Intensity on Muscle Fibre Adaptations. Sports Med. 2004; 34: Pages Reimer R, Wikstrom E. Functional Fatigue of the Hip and Ankle Musculature Cause Similar Alterations In Single Leg Stance. J SMS. 2010; 13: bluesensor.php Dowling J, Konert E, Ljucovic P, et al. Are Humans Able to Voluntarily Elicit Maximum Muscle Force. Neurosci Lett. 1994; 179: Berger M, Watson B, Doherty T. Effect of maximal voluntary contraction on the amplitude of the compound muscle action potential: implications for the interpolated twitch technique. Muscle Nerve. 2010; 42: Folland J, Williams A. Methodological Issues with the Interpolated Twitch Technique. J Electro Kinesiology. 2007; 17: Shield A, Zhou S. Assessing Voluntary Muscle Activation with The Twitch Interpolation Technique. Sports Med. 2004; 34: The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 216

24 APPENDIX A 1. Clamshell (hip clam) Progression: Each exercise is performed with the subject sidelying on the non-dominant side. (Figure 3) a. Progression 1 (upper left): Start position is sidelying with hips flexed to approximately 45 degrees, knees flexed, and feet together. Subject externally rotates the top hip to bring the knees apart for one metronome beat and returns to start position during the next beat. b. Progression 2 (upper right): Start position identical to progression 1; however, in this progression subject keeps the knees together while internally rotating the top hip to lift the top foot away from the bottom foot for one metronome beat, returning to the start position during the next beat c. Progression 3 (lower left): The subject is positioned identical to progressions 1 and 2, but with the top leg raised parallel to the ground. The subject maintains the height of the knee while internally rotating at the hip by bringing the foot toward the ceiling for one beat and then returns to the start position during the next beat. d. Progression 4 (lower right): The subject is positioned the same as progression 3, but with the hip fully extended. As in progression 3, the subject maintains the height of the knee and internally rotates at the hip by bringing the foot toward the ceiling for one beat and returns to the start position with knee and ankle in line during the next beat. 2. Pelvic drop: Subject stands with dominant leg on the edge of a 5 cm box (right), and then lowers the heel of the non-dominant leg to touch the ground without bearing weight, for one beat (left). Subject returns foot to the height of the box while keeping the hips and knees extended for one beat. (Figure 4) Figure 3. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 217

25 Figure Sidelying abduction: Start with subject sidelying on non-dominant side. Subject flexes the hip and knee of the support side and then abducts the dominant leg to approximately 30 degrees while maintaining neutral or slight hip extension and knee extension with the toes pointed forward for a count of two beats up and two beats down. (Figure 5) 4. Side Plank with Abduction, dominant leg up: (Start with subject in a side plank position with dominant leg up. Subject is instructed to keep shoulders, hips, knees, and ankles in line bilaterally, and then to rise to plank position with hips lifted off ground to achieve neutral alignment of trunk, hips, and knees. The subject is allowed upper extremity support as seen Figure 5. Figure 6. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 218

26 Figure 7. on left. While balancing on elbows and feet, the subject raises the top leg into abduction (right) for one beat and then lowers leg for one beat. Subject maintains plank position throughout all repetitions (Figure 6). 5. Side Plank with Abduction, dominant leg down: Exercise position is identical to Exercise 4 except on the opposite side. Subject is instructed to abduct the nondominant uppermost leg for two beats and lowers leg for two beats. Subject maintains plank position throughout all repetitions. 6. Front Plank with Hip Extension: Start with subject prone on elbows in plank with trunk, hips, and knees in neutral alignment (left). Subject lifts the dominant leg off of the ground, flexes the knee of the dominant leg, and extends the hip past neutral hip alignment by bringing the heel toward the ceiling (right) for one beat and then returns to parallel for one beat. (Figure 7) 7. Single Limb Bridging on Stable Surface: Start with subject in hook-lying position (left). The subject is instructed to bridge on both legs by keeping the feet on the floor and raising hips off the ground to achieve neutral trunk, hip, and knee alignment for one beat. From this position, the subject extends the knee of the non-dominant leg to full knee extension while keeping the femurs parallel (right) for one beat, returns the non-dominant leg to the bridge position for one beat, and then lowers the body back to the ground for one beat (Figure 8). 8. Single Limb Bridge on Unstable Surface: Subject is positioned as in Exercise 7 and places the dominant foot in the center of the Core-Tex (left). Figure 8. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 219

27 Figure 9. The subject performs the same sequence as above (right) while maintaining the disc of the Core-Tex in the center. (Figure 9) for three beats. Subjects were allowed two-finger unilateral upper extremity support on the frame of the Core-Tex for balance assist. (Figure 10). 9. Hip Circumduction on Stable Surface: The subject places the non-dominant leg on the outside of the base of the Core-Tex and stands to the side of the Core-Tex on the dominant leg (left). The subject performs a single limb squat while tracing the toe of the non-dominant leg on the outside of the Core-Tex base (right) in an arc for three beats, then traces the toe back to the start, while returning to a standing position 10. Hip Circumduction on Unstable Surface: In standing, the subject places the non-dominant foot on the outer edge of the Core-Tex and stands to the side of the Core-Tex on the dominant leg (left). The subject then performs a single limb squat on the dominant leg while drawing an arc with the non-dominant foot, extending the arc away from the subject for three beats (right). The subject then returns the foot to the Figure 10. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 220

28 Figure 11. starting position by drawing the foot in, while returning to a standing position for three beats. Subjects were allowed upper extremity support as in Exercise 9. (Figure 11) Figure Single Limb Squat: Subject stands on the dominant leg, slowly lowering the buttocks to touch a chair 47cm in height for two beats and then extends back to standing for two beats. (Figure 12) Figure 13. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 221

29 Figure Single Limb Deadlift: Subject stands on the dominant leg and slowly flexes at the hip, keeping the back straight, to touch the floor with the opposite hand for two beats. Subject then extends at the hip to standing for two beats. Subjects were permitted to have knees either straight or slightly bent in the case that hamstring tightness limited subject s ability to touch the floor. (Figure 13) 13. Dynamic Leg Swing: Subject is positioned in standing on the dominant leg, and then begins to swing the non-dominant leg (with the knee flexed) into hip flexion (left) and extension (right) at a rate of one beat forward and one beat backward. Subjects were instructed to move through a smooth range of hip motion and to not allow their trunk to move out of the upright position. (Figure 14) Figure Forward Step-up: Beginning with both feet on the ground, subject steps forward onto a 20cm step with the dominant leg for one beat. Subject then steps up with the non-dominant leg during the next beat. Subject then lowers the non-dominant leg back to the ground for one beat followed by the dominant leg during the next beat. (Figure 15) The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 222

30 were recorded in order to measure activity as both the stabilizing and moving leg. (Figure 17) 17. Skater Squat: Subject stands on the dominant leg and performs a squat to a comfortable knee flexion angle for two beats down and two beats up with non-dominant leg extended at the hip and flexed at the knee. The torso twists during the squat. The toe of the non-dominant leg was permitted to touch the ground between repetitions. (Figure 18) 18. Gluteal Squeeze: In standing with feet shoulder-width apart, subject squeezes gluteal muscles for two beats and then relaxes for two beats. Subjects were instructed to maximally contract the gluteal musculature during the exercise. Figure Lateral Step-up: Subject stands on the edge of a 15cm box on the dominant leg and squats slowly to lower the heel of the non-dominant leg toward floor for one beat and then returns to start position during the next beat. (Figure 16) 16. Quadruped Hip Extension: In quadruped (left) the subject extends the dominant leg at the hip, while keeping the knee flexed 90 degrees, to lift the foot toward the ceiling (right) to achieve neutral hip extension for two beats and then returns the dominant leg to the start position for two beats. This exercise was repeated with the non-dominant leg and EMG values Figure 18. Figure 17. The International Journal of Sports Physical Therapy Volume 6, Number 3 September 2011 Page 223

31 768393CRE / Clinical RehabilitationCruz-Díaz et al. research-article2018 Original Article CLINICAL REHABILITATION The effectiveness of 12 weeks of Pilates intervention on disability, pain and kinesiophobia in patients with chronic low back pain: a randomized controlled trial Clinical Rehabilitation 1 9 The Author(s) 2018 Reprints and permissions: sagepub.co.uk/journalspermissions.nav DOI: journals.sagepub.com/home/cre David Cruz-Díaz 1, Marta Romeu 2, Carmen Velasco-González 1, Antonio Martínez-Amat 1 and Fidel Hita-Contreras 1 Abstract Objective: To assess the effectiveness of 12 weeks of Pilates practice on disability, pain and kinesiophobia in patients with chronic non-specific low back pain. Design: This is a randomized controlled trial. Setting: This study was conducted in the university laboratory. Subjects: A total of 64 participants with chronic non-specific low back pain were included. Interventions: Participants were randomly allocated to intervention group consisted in Pilates intervention during 12 weeks (n = 32) or control group who received no treatment (n = 32). Main measures: Disability, pain and kinesiophobia were assessed by Roland Morris Disability Questionnaire, visual analogue scale and Tampa Scale of Kinesiophobia, respectively. Measurements were performed at baseline, at 6 and 12 weeks after study completion. Results: There were significant differences between groups with observed improvement in Pilates intervention group in all variables after treatment (P < 0.001). Major changes on disability and kinesiophobia were observed at six weeks of intervention with no significant difference after 12 weeks (P < 0.001). Mean changes of the intervention group compared with the control group were 4.00 (0.45) on the Roland Morris Disability Questionnaire and 5.50 (0.67) in the Tampa Scale of Kinesiophobia. Pain showed better results at six weeks with a slightly but statistically significant improvement at 12 weeks with Visual Analogue Scale scores of 2.40 (0.26) (P < 0.001). Conclusion: Pilates intervention in patients with chronic non-specific low back pain is effective in the management of disability, pain and kinesiophobia. Keywords Pilates, chronic low back pain, kinesiophobia, therapeutic exercise Received: 30 August 2017; accepted: 10 March Department of Health Sciences, Faculty of Health Sciences, University of Jaén, Jaén, Spain 2 Unit of Pharmacology, Department of Basic Medical Sciences, Faculty of Medicine and Health Sciences, NFOC Group, Universitat Rovira i Virgili, Reus, Spain Corresponding author: David Cruz-Díaz, Department of Health Sciences, Faculty of Health Sciences, University of Jaén, E Jaén, Spain. dcruz@ujaen.es

32 2 Clinical Rehabilitation 00(0) Introduction Therapeutic exercise is considered one of the most effective treatment options in the improvement of pain and disability, associated with chronic nonspecific low back pain. 1,2 Among these exercise modalities, the Pilates method has been reported to be effective in the management of chronic low back pain and has been widely recommended by healthcare providers. 3 The combination of Pilates training with physical therapy intervention in patients with chronic low back pain has proved to be superior to physical therapy alone in the long term. 4 Pilates principles include motor control, deep trunk muscle activation and pelvic floor muscles activation, 5 which may play an important role in the improvement of pain and disability in this population group. A proper monitoring of the muscular pattern activation and the evaluation of deep trunk muscle thickness would provide additional data to elucidate the mechanism of action due to Pilates intervention. Although Pilates method has been deemed to be effective in previous research, some studies present limitations such as small sample size, the absence of control group, high rate of drop-out or inaccurate description of the intervention. 6 A recent systematic review found that there was low to moderate quality evidence that Pilates is more effective than minimal intervention for those with chronic low back pain with contradictory findings. 6 The role of Pilates principles application and its influence in the management of pain and disability remains unclear. Therefore, the aim of this study was to assess the effectiveness of 12 weeks of Pilates intervention on pain, function, kinesiophobia and deep trunk muscle thickness in patients with chronic non-specific low back pain. Methods This is a single-blind randomized controlled trial conducted at the physiotherapy laboratories of the University of Jaén. Recruitment process was based on informative panels located in the University campus and Medical Centers of Jaén (Spain). Patients from the university and general population who responded to the announcement were screened by an expert clinician and invited to be enrolled in the study if the following inclusion criteria were meet: age between 18 and 50 years; suffering from low back pain for at least three months; absence of radiculopathy or other damages to the spine such as fractures, stenosis or tumors; not habitual Pilates practitioners; not receiving physical therapy during the trial or immediately prior thereto; and enough physical autonomy to participate in the physical activities required by the study. This study was registered (NCT: NCT ) and was approved by the Human Ethics Committee of the University of Jaén and meets the CONSORT (Consolidated Standards of Reporting Trials) statement and guidelines. 7 Patients who met the inclusion criteria and accepted to be enrolled in the study were randomly allocated into experimental and control groups. Participants were randomized into Pilates or control group using sealed opaque envelopes that were created at each institution prior to the initiation of the investigation by an independent researcher not involved with the intervention in a 1:1 ratio. Participants were evaluated at three different times during the intervention in the physiotherapy laboratory by an independent assessor blinded to the allocation and intervention. Outcomes measured were disability, pain, kinesiophobia and muscular thickness and were assessed at baseline prior to the beginning of the intervention, after 6 and 12 weeks of treatment. Disability was assessed using the Roland Morris Disability Questionnaire, a short and simple measure with contrasted validity, reliability and responsiveness. The questionnaire is a 24-item scale, the scores of which range from 0 (no disability) to 24 (high disability) Pain was measured using a visual analogue scale. The visual analogue scale consists of a 10-cm line, with the left extremity representing (absence of pain) and the right extremity indicating (great pain). Participants were asked to indicate in the scale their current level of pain, higher values being related to more intense pain. 11 Fear of movement/injury or reinjury was assessed using the Spanish version of the Tampa Scale of

33 Cruz-Díaz et al. 3 Kinesiophobia, a 17-item with scores ranged from 17 (absence of fear) to 68 (highest fear). 12, 13 Tampa Scale of Kinesiophobia has been reported to correlate with the Roland Morris Disability Questionnaire, 14 the primary outcome measure of this study, and has presented good reliability in patients with chronic non-specific low back pain. 15 Transversus abdominis activation was assessed to evaluate the possible change in the deep trunk muscle function using a real-time ultrasound scanning, MyLab 25 Gold (Esaote, Inc., Paris, France) with a 60-mm, 5-MHz curvilinear array in brightness mode at rest and during abdominal drawing-in maneuver. The transversus abdominis was tested in the supine hook-lying position (subject lying supine, with feet placed on the table, hips flexed to visually approximated 45 and knees to 90 ). The thickness of transversus abdominis was defined as the distance between the upper and lower borders of the fascia of the transversus abdominis and the percentage change in thickness was calculated. 16 Patients assigned to experimental group were included in a Pilates intervention which consisted of two sessions per week of 50 minutes during 12 weeks. The Pilates sessions were conducted by an expert Pilates physiotherapist instructor with 10 years of experience. The intervention was divided in three different parts. Each session start with a warm-up with breathing exercises, pelvis tilt centering, deep trunk and pelvic floor muscles activation and joint mobility. The principal part of the session consisted in strength and flexibility exercises involving the trunk, upper and lower limbs. Finally, a cool down section with some stretching exercises was conducted. All the exercises proposed by the instructor could be performed at different difficulty levels (basic, intermediate and advanced) in order to be adapted to patients physical condition. A more detailed description of the protocol can be found in Table 1. In order to collect any adverse event both during and after the Pilates session, patients were instructed to record any discomfort in the given booklet at the beginning of the study. Patients who were allocated to control group received a booklet with chronic non-specific low back pain information, to minimize potential dropout and disappointment with not receiving any Table 1. Intervention program. Pilates Mat 1. Warm-ups 2. Single leg stretch 3. Double leg stretch 4. Criss-cross 5. Single straight leg 6. Roll up 7. Rolling 8. Side kick: front/back 9. Side kick: small circles 10. Spine twist 11. Rowing Rowing Pull straps Pull straps Swimming 16. Teaser Leg pull back 18. Leg pull front 19. Mermaid 20. Rolling down 21. Cool down treatment. Patients of the control group who attended to the assessment session were offered to be incorporated to the same Pilates protocol performed by the intervention group after study completion. Sample size estimation was designed to have at least 80% power to detect a 2.5-point betweengroup difference in the scores of the primary outcome measure, the Roland Morris Questionnaire. Sample size calculation was performed with ENE 3.0 (GlaxoSmithKline, Brentford, UK) for a common standard deviation of 3.7 points in the Roland Morris Disability Questionnaire taking as a reference the data reported by Morton; 17 using a twogroup one-tailed t test with 80% power at the 0.05 level required 28 subjects per group. Considering a drop-out rate of 15%, the final sample population was 32 patients per group. Data were analyzed using the SPSS version 23.0 (SPSS Inc., Chicago, IL, USA) statistical package. Distributions were checked using Kolmogorov Smirnov test to ensure that parametric assumptions were met. In order to

34 4 Clinical Rehabilitation 00(0) compare the variables between groups, Student t test or non-parametric equivalent, Mann Whitney U test, was used. To assess differences between evaluation times within groups, the one-way analysis of variance (ANOVA) with repeated measures or the non-parametric alternative, the Friedman test, was used. Chi-square was used to compare descriptive data of the participants. Results A total of 64 patients (32 Pilates group and 32 control group) were enrolled in the study (Figure 1). All participants completed the study in the experimental group and two patients included in the control group were excluded to loss the assessment session. The sociodemographic data of Pilates and control group participants are shown in Table 2. No significant difference was found among pre-intervention characteristics of the control and the experimental groups. The weight was similar between the groups in the three moments where sampled; however, the body mass index was significantly different between the groups in the pre-intervention (P = 0.020), but not 6 or 12 weeks post-intervention. Table 3 shows the pre- and post-intervention (6 and 12 weeks) outcome measures of transversus abdominis thickness, disability, pain and kinesiophobia. None of the variables showed differences between groups before the intervention. However, all variables, except the transversus abdominis, improved significantly in the group of Pilates with respect to control group, both at 6 to 12 weeks postintervention. In the group of Pilates, all measures at six weeks improve their values respect to pre-intervention and the improvement is maintained or is higher at 12 weeks post-intervention. Discussion The main finding of this study was that 12 weeks of Pilates intervention was effective in reducing pain intensity and improving disability, fear of movement and deep trunk muscle thickness in patients with chronic non-specific low back pain. There were reported no adverse events during the intervention in the Pilates group whose participants showed high adherence to treatment with no drop-outs. Pilates group showed an improvement on disability and function with a significant change in the Roland Morris Disability Questionnaire score from baseline to 6 and 12 weeks, with no changes observed in the control group with an R 2 = 0.179; P < The effectiveness of Pilates in the management of patients with chronic non-specific low back pain has been addressed by several studies Our results agree with previous reported data where Pilates has been deemed to be superior to no treatment or minimal intervention in this population group. 4,19,20 It was observed a change of five points in the Roland Morris Disability Questionnaire after 12 weeks of Pilates intervention. These results are consistent with the existing research of previous studies using the same outcome measure but showing greater improvement. 3,21,22 The longer duration of the present Pilates intervention may be one factor that explains the better results obtained in disability and function. This hypothesis has been advocated by Natour et al., 3 who suggested that better scores could be related to longer intervention time. However, in our study, major change on disability was obtained after six weeks of Pilates intervention, with no observed within-group change between 6 and 12 weeks. Thus, motor control learning skills and Pilates methodology may play an important role in the Pilates intervention effectiveness. Regarding pain perception, it was observed a significant improvement on pain in the Pilates group with no change in the control group from baseline to 6 and 12 weeks, respectively, P < Our results in the intervention group, ranged from 4.70 ( ) at baseline to 1.95 ( ) at the end of the intervention. In contrast to the obtained results on disability, it was observed a significant and progressive improvement from 6 to 12 weeks of intervention. This may indicate that longer intervention could be related to pain perception improvement, although function did not follow this pattern. As with disability results, self-reported pain improvement was greater than those reported by similar

35 Cruz-Díaz et al. 5 Figure 1. Flowchart of the study. studies A possible explanation may be that the management of the deep trunk muscle activation due to Pilates practice could improve the perception of pain. These results agree with the reported data of Ferreira et al., 25 who suggested that pain may be responsible of the onset of deep trunk muscle dysfunction. The improvement observed on pain and transversus abdominis in this study may contribute to support this hypothesis. With regard to this statement, it is widely extended among Pilates literature that deep trunk muscle activation is related to the improvement on pain and disability. 26 Some authors have concluded that the improvement of deep trunk

36 6 Clinical Rehabilitation 00(0) Table 2. Baseline characteristics of participants at all evaluation times. Pilates (PG) n = 32 Control (CG) n = 30 Mean SD pt Mean SD pt P-value PG versus CG Gender (female/male) (%) 21/11 20/ Age (years) Height (m) Weight a (kg) Pre w a ns w a, b ns BMI (kg/m 2 ) Pre w ns ns w ns ns Occupational status (%) Primary Secondary University Marital status (%) Single Married Divorced BMI: body mass index; PG: experimental Pilates group; CG: control group, control; SD: standard deviation; CI: confidence interval; Pre: pre-intervention; 6w: six weeks post-intervention; 12w: 12 weeks post-intervention; pt: P-values between pre-6w-12w evaluation times. a: P < 0.05 with respect to Pre; b: P < 0.05 with respect to 6w; ns: not significant. a Non-normal distributed data, values are expressed as median and 95% CI. muscle activation due to Pilates practice could be an important component in achieving positive results in patients with chronic non-specific low back pain. 3,4,26 Nevertheless, until recently, there is no evidence about the improvement on deep trunk muscle activity after Pilates training. In our study, transversus abdominis thickness was assessed and was observed major change after six weeks in the Pilates group with a more slightly but constant improvement during the rest of the intervention until the completion of the study after 12 weeks. Control group did not present any significance difference from baseline to 6 and 12 weeks; P < Kinesiophobia is an important variable in patients with chronic non-specific low back pain because of its relationship with disability and symptom perpetuation. 27 The lack of activity due to fear of movement, may induce muscle atrophy and therefore worsening of symptoms. 27 Some studies have reported the benefit of Pilates intervention in the improvement of kinesiophobia. 20,24,26 However, to our knowledge, only Da Luz Jr et al. 20 and Miyamoto et al. 24 have studied the influence of Pilates on kinesiophobia in patients with chronic non-specific low back pain with contradictory findings. Our results support the positive findings reported by Da Luz Jr et al.20 in contrast with Miyamoto et al., 24 whose results showed no change after the intervention. The improvement on pain and disability could be related to an increased physical activity which may have a positive influence on kinesiophobia. Fear avoidance beliefs about physical activity may lead in decreased neuromuscular control of the deep trunk muscle activation which has been reported to be related to chronic low back pain. Thus, the improvement of patient s confidence and their involvement in physically demanding task could contribute to a better neuromuscular function. Following the recommendation of a recent systematic review about the effectiveness of

37 Cruz-Díaz et al. 7 Table 3. Pre- and post-intervention measures of transversus abdominis thickness, disability, pain and kinesiophobia at all evaluation times. Pilates (PG) n = 32 Control (CG) n = 30 Difference (CG-PG) P-value Median 95% CI Median 95% CI Mean SD PG versus CG RM a Pre ( ) 9.00 ( ) w 5.00 ( ) 9.00 ( ) < w 5.00 ( ) 9.00 ( ) <0.001 VAS a Pre 4.70 ( ) 5.15 ( ) w 2.05 ( ) 4.85 ( ) < w 1.95 ( ) 4.35 ( ) <0.001 Tampa a Pre ( ) ( ) w ( ) ( ) < w ( ) ( ) <0.001 TrAR a Pre 5.75 ( ) 6.25 ( ) w 6.00 ( ) 6.15 ( ) w 6.00 ( ) 6.20 ( ) TrAC a Pre 6.95 ( ) 7.00 ( ) w 8.25 ( ) 7.10 ( ) < w 9.00 ( ) 6.90 ( ) <0.001 TrA% a Pre ( ) ( ) w ( ) ( ) < w ( ) ( ) <0.001 TrAR: transversus abdominis thickness in relaxation; TrAC: transversus abdominis thickness in activation; TrA%: TrA activation; RM: Roland Morris disability test; VAS: visual analogue scale of pain; Tampa: Tampa Scale of Kinesiophobia; PG: experimental Pilates group; CG: control group, control; CI: confidence interval; Pre: pre-intervention; 6w: six weeks post-intervention; 12w: 12 weeks post-intervention. a Non-normal distributed data, values are expressed as median and 95% CI. Pilates intervention in patients with chronic nonspecific low back pain, 6 the authors have sought to avoid some methodological issues observed in previous research such as the presence of a control group, concealed allocation or assessors blinding. This study yielded new results regarding the importance of deep trunk muscle training in the improvement of chronic low back pain. However, additional research is needed to confirm the present findings and for a better understanding of the influence of muscle training approach in this population group. Moreover, although the use of ultrasound measurement is well-documented and reported good reliability and validity, the inter-examiner reliability could be considered as a risk of bias. 25 Conflicting results showed in previous research regarding Pilates effectiveness could benefit from a better understanding of muscle activity changes due to Pilates training. Pilates group participants experienced a noticeable improvement in all outcome measures after 12 weeks of intervention. The effectiveness of Pilates in the management of patients with chronic low back pain together with the absence of adverse events and the rapid improvement observed in the intervention group suggests that Pilates is a valuable treatment option that could be incorporated during the rehabilitation process in this population group. Nevertheless, a followup period to observe the long-term effects of the Pilates intervention is required. The evaluation of the achieved results over time will enable researchers to know more about how these

38 8 Clinical Rehabilitation 00(0) improvements are maintained to prevent relapses and to develop a therapeutic protocol for patients with chronic low back pain. Clinical Messages The Pilates method was effective in improving disability, pain and kinesiophobia in patients with chronic non-specific low back pain. Transversus abdominis thickness increased after 12 weeks of Pilates intervention. No adverse events or symptoms aggravation were observed during the intervention. Acknowledgements The authors would like to express their gratitude to Ángela Molina and Juanjo Molina of Fremap Hospital, for their cooperation in the recruitment process, and especially to the patients who participated in the study. Declaration of Conflicting Interests The author(s) declared the following potential conflicts of interest with respect to the research, authorship and/or publication of this article: We certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated, and, if applicable, we certify that all financial and material supports for this research (e.g. NIH or NHS grants) and work are clearly identified in the title page of the manuscript. Funding The author(s) received no financial support for the research, authorship and/or publication of this article. ORCID id David Cruz-Díaz References 1. Carter IR and Lord JL. Clinical inquiries. How effective are exercise and physical therapy for chronic low back pain? J Fam Pract 2002; 51: Hayden JA, Van Tulder MW, Malmivaara A, et al. Exercise therapy for treatment of non-specific low back pain. Cochrane Database Syst Rev 2005; 20: CD Natour J, Cazotti Lde A, Ribeiro LH, et al. Pilates improves pain, function and quality of life in patients with chronic low back pain: a randomized controlled trial. Clin Rehabil 2015; 29: Cruz-Díaz D, Martínez-Amat A, Osuna-Pérez MC, et al. Short-and long-term effects of a six-week clinical Pilates program in addition to physical therapy on postmenopausal women with chronic low back pain: a randomized controlled trial. Disabil Rehabil 2016; 38: Wells C, Kolt GS and Bialocerkowski A. Defining Pilates exercise: a systematic review. Complement Ther Med 2012; 20: Yamato TP, Maher CG, Saragiotto BT, et al. Pilates for low back pain. Cochrane Database Syst Rev 2015; 2: CD Chan L, Heinemann AW and Roberts J. Elevating the quality of disability and rehabilitation research: mandatory use of the reporting guidelines. Ann Phys Rehabil Med 2014; 57: Roland M and Morris R. A study of the natural history of back pain, part I: the development of a reliable and sensitive measure of disability in low back pain. Spine 1983; 8: Stratford PW and Binkley JM. Applying the results of self-report measures to individual patients: an example using the Roland-Morris questionnaire. J Orthop Sports Phys Ther 1999; 29: Beurskens A, De Vet H, Van der Heijden G, et al. Measuring the functional status of patients with low back pain: assessment of the quality of four disease-specific questionnaires. Spine 1995; 20: Boonstra AM, Preuper HRS, Reneman MF, et al. Reliability and validity of the visual analogue scale for disability in patients with chronic musculoskeletal pain. Int J Rehabil Res 2008; 31: Gómez-Pérez L, López-Martínez AE and Ruiz-Párraga GT. Psychometric properties of the Spanish version of the Tampa Scale for Kinesiophobia (TSK). J Pain 2011; 12(4): French DJ, France CR, Vigneau F, et al. Fear of movement/(re)injury in chronic pain: a psychometric assessment of the original English version of the Tampa Scale for Kinesiophobia (TSK). Pain 2007; 127: Vlaeyen JWS, Kole-Snijders AMJ, Rotteveel AM, et al. The role of fear of movement/(re)injury in pain disability. J Occup Rehabil 1995; 5: Swinkels-Meewisse E, Swinkels R, Verbeek A, et al. Psychometric properties of the Tampa Scale for Kinesiophobia and the fear-avoidance beliefs questionnaire in acute low back pain. Man Ther 2003; 8: Hides JA, Miokovic T, Belavy DL, et al. Ultrasound imaging assessment of abdominal muscle function during drawing-in of the abdominal wall: an intrarater reliability study. J Orthop Sports Phys Ther 2007; 37: Morton J. Manipulation in the treatment of acute low back pain. J Man Manip Ther 1999; 7: Cruz-Díaz D, Martínez-Amat A, De la Torre-Cruz MJ, et al. Effects of a six-week Pilates intervention on balance and fear of falling in women aged over 65 with chronic

39 Cruz-Díaz et al. 9 low-back pain: a randomized controlled trial. Maturitas 2015; 8: Wajswelner H, Metcalf B and Bennell K. Clinical Pilates versus general exercise for chronic low back pain: randomized trial. Med Sci Sports Exerc 2012; 44: Da Luz Jr MA, Costa LO, Fuhro FF, et al. Effectiveness of mat Pilates or equipment-based Pilates exercises in patients with chronic nonspecific low back pain: a randomized controlled trial. Phys Ther 2014; 94: O Brien N, Hanlon M and Meldrum D. Randomised, controlled trial comparing physiotherapy and Pilates in the treatment of ordinary low back pain. Phys Ther Rev 2006; 11: Rydeard R, Leger A and Smith D. Pilates-based therapeutic exercise: effect on subjects with nonspecific chronic low back pain and functional disability: a randomized controlled trial. J Orthop Sports Phys Ther 2006; 36: Lauridsen KH, Hartvigsen J, Manniche C, et al. Responsiveness and minimal clinically important difference for pain and disability instruments in low back pain patients. BMC Musculoskelet Disord 2006; 7: Miyamoto GC, Costa LO, Galvanin T, et al. Efficacy of the addition of modified Pilates exercises to a minimal intervention in patients with chronic low back pain: a randomized controlled trial. Phys Ther 2013; 93: Ferreira PH, Ferreira ML and Hodges PW. Changes in recruitment of the abdominal muscles in people with low back pain: ultrasound measurement of muscle activity. Spine 2004; 29: Cruz-Díaz D, Bergamin M, Gobbo S, et al. Comparative effects of 12 weeks of equipment based and mat Pilates in patients with chronic low back pain on pain, function and transversus abdominis activation. A randomized controlled trial. Complement Ther Med 2017; 33: Karayannis NV, Smeets RJ, Van den Hoorn W, et al. Fear of movement is related to trunk stiffness in low back pain. PLoS ONE 2013; 8(6): e67779.

40 Long-Term Effects of Specific Stabilizing Exercises for First-Episode Low Back Pain SPINE Volume 26, Number 11, pp E243 E , Lippincott Williams & Wilkins, Inc. Julie A. Hides, PhD, MPhtySt, BPhty,* Gwendolen A. Jull, MPhty, FACP* and Carolyn A. Richardson, PhD, BPhty(Hons)* Study Design. A randomized clinical trial with 1-year and 3-year telephone questionnaire follow-ups. Objective. To report a specific exercise intervention s long-term effects on recurrence rates in acute, first-episode low back pain patients. Summary of Background Data. The pain and disability associated with an initial episode of acute low back pain (LBP) is known to resolve spontaneously in the short-term in the majority of cases. However, the recurrence rate is high, and recurrent disabling episodes remain one of the most costly problems in LBP. A deficit in the multifidus muscle has been identified in acute LBP patients, and does not resolve spontaneously on resolution of painful symptoms and resumption of normal activity. Any relation between this deficit and recurrence rate was investigated in the long-term. Methods. Thirty-nine patients with acute, first-episode LBP were medically managed and randomly allocated to either a control group or specific exercise group. Medical management included advice and use of medications. Intervention consisted of exercises aimed at rehabilitating the multifidus in cocontraction with the transversus abdominis muscle. One year and three years after treatment, telephone questionnaires were conducted with patients. Results. Questionnaire results revealed that patients from the specific exercise group experienced fewer recurrences of LBP than patients from the control group. One year after treatment, specific exercise group recurrence was 30%, and control group recurrence was 84% (P 0.001). Two to three years after treatment, specific exercise group recurrence was 35%, and control group recurrence was 75% (P 0.01). Conclusion. Long-term results suggest that specific exercise therapy in addition to medical management and resumption of normal activity may be more effective in reducing low back pain recurrences than medical management and normal activity alone. [Key Words: multifidus, low back pain, rehabilitation] Spine 2001;26:E243 E248 The major costs of low back pain (LBP) have been identified with two groups: those who develop chronic LBP and those who have recurrent disabling episodes of LBP. 12 These two groups incur 85% of the total costs. 13,23,33 Efforts have been made to identify the 2 3% of patients who go on to develop chronic symptoms, 22 but little is known about the factors that lead to recurrence. From the *Department of Physiotherapy, The University of Queensland, Brisbane, Australia, and the Department of Physiotherapy, Mater Misericordiae Public Hospitals, South Brisbane, Queensland, Australia Acknowledgment date: October 27, Acceptance date: February 13, Device status category: 1. Conflict of interest category: 12. It is documented and generally accepted that a single episode of acute LBP has a favorable natural history with respect to symptom reduction and restoration of function and work capacity in the short term. 1 In the majority of cases, the pain associated with an initial acute episode resolves within 2 4 weeks. 5,7 11,21 It is estimated that 2 3% of patients go on to develop disabling chronic LBP after an acute episode. 5,18,22 However, the course of LBP for most primary care patients is recurrent rather than acute or chronic in the usual sense of these terms. 39 When the frequency of low back pain recurrences following an acute episode is examined, the recurrence rate is found to be staggeringly high. Recurrence rates range from 60% to 86% for patients suffering recurrences, particularly in the first year after the acute episode. 3,35 37 Bergquist-Ullman and Larsson 3 conducted a detailed study of 217 workers in an industrial setting in Sweden. The median duration of pain for the initial episode was 35 days and short-term resolution of painful symptoms occurred in the majority of cases (70% within 2 months, 86% within 3 months). However, during the 1-year follow-up, 62% of the patients experienced at least one recurrence of LBP and a further 36% experienced two or more recurrences. The median time from resolution of the initial episode to the first recurrence of LBP was only 2 months. These high figures would suggest that it is important to identify the factors that may relate to this vulnerability to recurrence. Although several processes are likely to be involved, the model provided by Panjabi 26,27 could provide an explanation for recurrences after painful symptoms have subsided. This model of spinal stability encompasses the passive, active, and neural control subsystems. It has been proposed that instability at the spinal segmental level is a loss of control or excessive motion in the spinal segment s neutral zone, which is associated with injury, degenerative disc disease, and muscle weakness. 26,27 It has been shown in in vitro biomechanical studies that muscles can provide segmental stabilization by controlling motion in the neutral zone, and the neutral zone can be returned to within physiologic limits by effective muscle control. 14,28,41 While various muscles may be able to control and protect the spinal segments, one muscle that has been investigated in relation to this role is the lumbar multifidus. The multifidus provides segmental stiffness and controls motion in the neutral zone. 14,28,34,41 Further evidence of this stabilizing role has been provided by in vivo animal research. 20 Investigations have also demonstrated a relation between multifidus muscle dysfunc- E243

41 E244 Spine Volume 26 Number tion and poor functional outcome and recurrence of LBP following disc surgery. 29,32 Optimal functioning of the muscle system is desirable to control and protect the spinal segments following injury. Despite initial resolution of painful symptoms, failure to protect spinal segments could increase the likelihood of a recurrence of symptoms. Specific exercises targeting the multifidus and transversus abdominis muscles have been shown to decrease pain and disability in chronic low back pain patients. 25 Our research has shown the occurrence of localized segmental dysfunction of the multifidus muscle after an initial episode of acute unilateral LBP. 15,16 To establish the low back pain recurrence rates in the two groups, the present study presents the follow-ups of the patients from the study at 1 year and 3 years after treatment. Methods During a 6-month period, patients were recruited from a hospital accident and emergency department. 16 Men and women were eligible for the initial study if they were aged 18 to 45 years, were experiencing their first episode of unilateral mechanical LBP for less than 3 weeks, and presented to the accident and emergency department because of this condition. Inclusion and exclusion criteria are provided in detail elsewhere. 16 Thirty-nine patients were accepted into the study. All patients gave their consent and the Medical Ethical Review Committees of the University of Queensland and the Mater Adult Hospital, Brisbane, Australia approved the study. Assessment Procedures. Assessments for the short-term phase of the trial were performed by two independent examiners, who were blinded to group allocation and patient presentation. The following assessments were conducted to establish baseline levels and to monitor improvement over time: pain (McGill Pain Questionnaire and Visual Analogue Scales), disability (Roland Morris Disability Index), range of motion (using inclinometers), habitual activity levels, 2 and muscle crosssectional area (using ultrasound imaging). 16 The aim of the long-term follow-ups was to determine the incidence of recurrence of LBP. To meet this aim, a telephone questionnaire was selected as the most appropriate assessment tool. Methodologic research has indicated that well-designed telephone interviews provide results comparable to face-to-face interviews 6 and investigations of pain data obtained in this way also support the validity of telephone interviews. 38,39 The questionnaires were administered by a research assistant who was not involved in the first stage of the study and who was blind to group allocation. The questionnaires used to determine the recurrence rate of LBP episodes during the 1-year and 3-year year follow-up periods were devised especially for the patients in this study, as the information sought was specific to the design and methods implemented. The questionnaire consisted of three groups of questions, and took approximately 5 minutes to complete. Questions related to episodes of LBP experienced in the year after the study (1-year follow-up) and then in years two to three (3-year follow-up). A general opening question was used to determine whether patients had experienced any episodes of LBP in the time period in question. Subsequent questions determined the number of episodes experienced in that time frame, their length, severity, precipitating factors, and treatment sought. Ideally, it would have been useful to reimage the patients multifidus muscles. This was not possible because many of the patients had relocated interstate or overseas. Intervention and Patient Management. Patients in Group 1 (control group) received medical management, including advice on bedrest, absence from work, prescription of medication, and advice to resume normal activity as tolerated, whereas those in Group 2 (specific exercise group) additionally performed specific localized exercises aimed at restoring the stabilizing protective function of the multifidus. The exercises were designed specifically to activate and train the isometric holding function of the multifidus muscle at the affected vertebral segment (in cocontraction with the transversus abdominis muscle). Contraction of the multifidus was confirmed by realtime ultrasound imaging. This rehabilitation approach is described in detail elsewhere. 17,19,30,31 The intervention period was 4 weeks, and patients from the specific exercise group were seen twice per week in this period. Statistical Analysis. Data analysis was performed using the SPSS statistics program. Comparability of baseline measurements between the two groups was assessed using a one-way analysis of variance (ANOVA) to examine differences in all baseline measurements. ANOVA also was used to examine differences between groups over time for all outcome measures used. For ultrasound imaging data, the percentage difference between the painful and nonpainful side was calculated for each vertebral level measured. Analysis of muscle recovery was conducted using the data from the most affected vertebral level (i.e., the vertebral level with the largest percentage difference between sides). For the 1-year and 2 3-year follow-up analysis, the data were expressed as the likelihood of recurring LBP in the control group relative to that in the intervention group. A relative risk ratio of 1.00 indicates that patients in both groups were equally likely to report recurring LBP. A large risk ratio indicates that the treatment was effective, while a ratio less than one would indicate that the treatment increased the likelihood of recurrence. The significance of the treatment was determined with a 2 test. Because the three patients who were lost for 2 3-year follow-up were all from the control group, the analysis was repeated using the best case analysis, assuming that the three patients had all completely recovered, and did not suffer recurrences in this period. Results Study Sample Patients were randomly allocated to Group 1 (control, n 19) or Group 2 (specific exercise, n 20). The demographics for the groups are shown in Table 1. Baseline Characteristics Comparability between groups was found to be satisfactory at baseline for age, height, weight, duration of symptoms, premorbid activity, and outcome measures used. 16 Primary Outcomes for Weeks 1 4 Results of the short-term study have been presented in detail in an earlier report, 16 but in summary, ultrasound imaging revealed that asymmetry of the multifidus muscle was present with diminished muscle size evident on

42 Stabilizing Exercises for Low Back Pain Hides et al E245 Table 1. Demographic Data for Groups 1 (Control) and 2 (Specific Exercise) Group 1 (Control) n 19 Group 2 (Specific Exercise) n 20 Mean SD Range Mean SD Range Age (years) Gender 9 male, 10 female 7 male, 13 female Height (cm) Weight (kg) Duration of Symptoms (Days) Smokers 7 4 Worker s Compensation 5 8 Pre-morbid Activity Levels Work Sport Leisure the patient s nominated painful side in all cases. The difference between the sides at the most affected vertebral level was expressed as a percentage of the CSA for the unaffected side at that level. The mean of these percentages was 22% 8.7% for the control group and 26% 8.7% for the specific exercise group (range, 12 46%). Results at follow-up immediately after the intervention period and at a 10-week follow-up examination revealed that multifidus muscle recovery was not spontaneous on remission of painful symptoms in control group patients. In the control group, multifidus CSA at the most affected vertebral level remained 16.8% 9.3% less at 4 weeks and 14% 6.3% less at ten weeks. Muscle recovery was more rapid and more complete in patients in Group 2 who received specific and localized exercises (P ). Multifidus CSA at the most affected vertebral level was only 0.7% 2.5% less at 4 weeks and 0.2% 3.3% less at ten weeks. The other outcome measurements of disability and physical function were similar for the two groups at the 4-week examination (pain and disability had completely resolved in 90% of the patients). Although they resumed normal levels of activity, patients in Group 1 still exhibited significantly decreased multifidus muscle size at the 10-week follow-up examination, and the difference between groups was still significant (P ). Long-Term Follow-Up Study Sample The response rate to the questionnaire at 1 year was 100%, with all 39 patients interviewed. Three patients could not be contacted for the 3-year interview, despite records of work, residence, mobile phone, and stable relative contact. All three were from the control group. For the 3-year follow-up interview, questions related to recurrence of symptoms in the previous 2 years. Overall Recurrence Rate and Risk of Recurrences Results of the contingency 2 analysis revealed that, in the year after the initial episode, patients in the control group were 12.4 times more likely to experience recurrences of LBP than patients in the specific exercise group (x 2 (1) 12.41, P 0.001). Additionally, these patients were 9 times more likely to experience LBP recurrences in years 2 3 (x 2 (1) 9.31, P 0.01). The risk of pain for each group is presented in Table 2, along with confidence intervals. In year 1, approximately 1 patient in the specific exercise group reported pain for every 3 patients who did not, whereas approximately 4 patients in the control group reported recurrences for every 1 that did not. In years 2 3, the likelihood of reporting recurrences of LBP in the exercise group increased slightly to around 2:5, while the likelihood of recurrences in the control group reduced to 10:3. A repeat analysis of the data using the best case analysis revealed that patients in the control group were still 5.9 times more likely to suffer recurrences of LBP than patients in the specific exercise group in years 2 3 (x 2 (1) 5.92, P 0.015). Figure 1 shows the pattern of recurrence over time for each patient of the two groups. Figure 1(a) shows the control group patients recurrence patterns and 1(b) shows the specific exercise group patients recurrence patterns. Number and Severity of Recurrent Episodes For the first year, the mean number of episodes reported by those in the control group was compared with episodes on average for the specific exercise group. Recurrent episodes of LBP were rated as as se- Table 2. Risk of Recurrent Episodes of LBP and Confidence Limits for Each Group in Year 1 and Years 2 3 Year 1 Year 2 3 Risk 95% Confidence Limits Risk 95% Confidence Limits Exercise Control* * Less than 5 subjects in the control group reported no recurrences of LBP in both years.

43 E246 Spine Volume 26 Number exercise group reported persistent low level LBP that was subsequently aggravated by activities such as lifting. The number of specific episodes reported by the remaining patients in the two groups were similar (control group, mean episodes, specific exercise group, mean ). Recurrent episodes of LBP were rated as as severe as the original episode by 2 of 12 (17%) of the control group and 1 of 7 (14.2%) of the specific exercise group. Precipitating Factors At 1 year, a traumatic incident initialed the recurrences in 3 of 16 (19%) of the control group. These included bending and lifting (2 patients) and a trampoline accident. In contrast, 4 of 6 (67%) of the specific exercise group could relate traumatic incidences to recurrences. These included carrying a patient and slipping, heavy lifting (2 patients), and an incident that involved pulling a heavy sail on a boat. For years 2 3, a traumatic incident was related to recurrences in the preceding 2 years by 5 of 12 (42%) of the control group, and all (7 of 7) of the patients from the specific exercise group who experienced recurrences. The three patients in the specific exercise group who only reported episodes in years 2 3 related them to high-trauma incidents including a motor vehicle accident, a work-related heavy lifting incident, and an injury in representative level football. Apart from these cases, patients of both groups most commonly reported precipitating incidents related to lifting. Figure 1. Pattern of LBP recurrence over time for each patient of the two groups. A, Control group. B, specific exercise group. 16 of 19 (84%) of the control group reported recurrences in the first year after the acute episode compared with 6 of 20 (30%) of the specific exercise group. In years 2 3, of the 16 patients from the control group who had experienced recurrences in the first year, 2 were lost to follow-up. 12 of 14 (86%) of the control group patients who experienced recurrences in the first year reported continuing recurrences. For years 2 3 there were recurrences reported in 12 of 16 (75%). For the specific exercise group, of the 6 who experienced recurrences in the first year, 4 continued to have recurrences during years 2 and 3. Three subjects who had not experienced recurrences in the first year reported acute injuries during years 2 through 3, with 7 of 20 reporting recurrences in years 2 through 3. vere as the original episode by 9 of 16 (56%) of the control group and 2 of 6 (33%) of the specific exercise group. For years 2 3, of those who experienced recurrences, 5 of 12 of the control and 4 of 7 of the specific Treatment Sought In the first year, treatment was sought by 8 of 19 (42%) of the patients from the control group and 3 of 20 (15%) of the specific exercise group. In all cases, this treatment consisted of medical management (time off from work, advice, medications) and physiotherapy treatment. A variety of physiotherapy treatments were reported. However, the patients did not report that the treating physiotherapists had prescribed specific multifidus exercises. Contamination of the exercise outcome from the 1-year follow-up can therefore be considered minimal. For years 2 3, 4 of 16 (25%) of the control group sought treatment in the 2-year period in comparison with 4 of 20 (20%) of the specific exercise group. Control group patients accessed physiotherapy, medical management, and one had received an orthopedic consult, whereas patients from the specific exercise group received physiotherapy only. Patients Lost to Follow-Up The three patients who were lost to follow-up for years 2 3 reported quite different patterns of recurrence over the 1-year follow-up period. Patient 12, at 1 year, reported that recurrent episodes started within 2 4 weeks of the 10-week initial trial period. She had experienced several episodes. The aggravating factor was prolonged sitting (studying), after which she reported experiencing pain at night. She did not suffer any traumatic predisposing injuries to precipitate these recurrences, but reported

44 Stabilizing Exercises for Low Back Pain Hides et al E247 that they were milder than the original incident, for which she sought treatment. In contrast, patient 13 was one of the 16% from the control group who had not experienced any recurrences at the 1-year follow-up. After bending over to make a bed during the 1 year follow-up period, patient 19 experienced one 2-week episode of LBP as severe as the original incident, for which treatment was sought. This had resulted in time off work (2 days) but medications were not used. Discussion The results from the control group, who were managed medically and advised to resume normal activity, reflect the reported high recurrence rate of LBP that occurs after the initial episode. 3,35 37 Their recurrence rate at 1 year (84%) is similar to the rates previously reported and furthermore, for 56% of these subjects, the recurrences were reported as being as severe and disabling as the original episode. In contrast, the group to whom specific exercise was given to the multifidus reported only 30% recurrence at 1 year and these were reported as being as severe in only 33% of cases. Results from the control group lead us to agree with the report of Von Korff and Saunders 40 in that the course of LBP for most primary care patients is recurrent rather than acute or chronic in the usual sense of these terms. Furthermore, as expressed by Von Korff and Saunders, 40 it is necessary to assess not only the short-term outcome of the index episode but also the long-term outcomes over a sufficient period of time. The positive natural history of acute LBP in the short-term, without provision of long-term follow-up, may have led to an underestimation of the importance of early intervention, which aims to prevent recurrences. Results from the control group highlight that the greatest number of recurrences (especially severe disabling ones) occur predominantly in the first year after the original episode. Few detailed reports of long-term follow-up of acute LBP are available. The most detail for the year following the initial episode is provided by Bergquist-Ullman and Larsson. 3 However, little information is available for longer-term outcomes. Von Korff and Saunders 39 report that LBP recurrence rates were similarly high at follow-up at 2 years. In this study, recurrence rates remained high for the control group for years two to three (75%), but episodes reported as equally severe as the original episode decreased from 56% to 17%. This investigation therefore demonstrated some moderation in LBP over time, and it has been previously reported that the risk of recurrence lessens 2 years after an acute episode. 24 As the highest rate of severe disabling recurrences occurred in the first year after the initial episode and one of the major costs of LBP is in association with those who have recurrent disabling episodes of LBP, 12 it would appear that intervention may have its maximal benefits in this period. However, long-term positive effects of the intervention used were demonstrated in this study (30% recurrence at 1 year to 35% recurrence rate for years two to three). This long-term benefit was achieved with a short intervention period (4 weeks). There is now biomechanical evidence to explain the role of the multifidus in stabilization of the lumbar segments. 14,28,34,41 The rehabilitation approach aimed at retraining the multifidus for its functional role of protection and control of movements of the vertebral segments. 14,28,34,41 It is now possible to hypothesize how this approach may be effective to account for the longterm differences between the control and specific exercise groups. Following an acute injury to the low back, a deficit in the multifidus may leave the injured segment susceptible to further injury. Specific exercise therapy may be required to restore normal muscle function, with the long-term sequelae of a deficient multifidus in control subjects being a susceptibility to further injury and recurrence of LBP. Furthermore, the biomechanical model provided by Cholewicki and McGill 4 may help to explain why recurrences occurred with seemingly little provocation, especially in the control group subjects. The model highlighted the importance of muscles that provide spinal segmental support, not only during high demand activities such as heavy lifting, but during low load activity requiring only low muscle forces. Deficient stabilization of lumbar segments caused by a deficient multifidus may explain LBP recurrence with minimal or no predisposing incidents. This study provides one step forward in the knowledge concerning the long-term effects of conservative management for LBP patients. The results are promising in that they suggest that specific exercises help to reduce the high recurrence rate of LBP after the initial acute episode, and this pilot study may be used to determine a design model for further research. The limitations of this study include the small sample size and limited outcome measures (telephone questionnaire) for long-term follow-up. More evidence in a larger study population is required to further substantiate the findings of this study. Conclusion The results from this study showed that subjects with acute, first-episode LBP who received specific exercise therapy in addition to medical management and resumption of normal activity experienced fewer recurrences of LBP in the long-term than subjects who received only medical management and resumed normal activity. Biomechanical research may explain why it is important to focus on particular muscles for their stabilizing functions in rehabilitation. Additional research on larger subject populations is required, and other factors will obviously be involved in low back pain recurrence. However, in terms of prevention of recurrences, this study might represent one step forward in the optimal management of the acute low back pain patient.

45 E248 Spine Volume 26 Number Key Points Following an initial episode of acute low back pain (LBP), the recurrence rate is high. This randomized clinical trial followed acute low back pain patients who undertook specific stabilization exercises and control subjects for 3 years. Results showed decreased recurrence of low back pain episodes in the specific exercise group compared with the control group. Acknowledgments The authors thank the patients studied, Helen King for interviewing patients, the staff of the Mater Misericordiae Adult Hospital, Craig Shaw for statistical advice, Dr D.H. Cooper for helpful suggestions with this manuscript, The Menzies Foundation, The JP Kelly Mater Research Foundation, The Wenkart Foundation, The Physiotherapy Research Foundation, and the Manual Therapy Special Group (Australia) for financial support. References 1. Andersson GBJ, Svensson HO, Oden A. The intensity of work recovery in low back pain. Spine 1983;8: Baecke J, Burema J, Frejters J. A short questionnaire for the measurement of habitual activity in epidemiological studies. Am J Clin Nutr 1982;36: Bergquist-Ullman M, Larsson U. Acute low back pain in industry: a controlled prospective study with special reference to therapy and confounding factors. Acta Orthop Scand 1970;170(suppl):S Cholewicki J, McGill SM. Mechanical stability of the in vivo lumbar spine: implications for injury and low back pain. Clin Biomech 1996;11: Coste J, Delecoeuillerie G, et al. Clinical course and prognostic factors in acute low back pain: an inception cohort study in primary care practice. BMJ 1994;308(6928): De Leeuw ED, van der Zouwen J. Data quality in telephone and face to face surveys: a comparative meta-analysis. In Groves RM, ed. Telephone Survey Methodology. New York: John Wiley and Sons, 1988: Deyo RA. The role of the primary care physician in reducing work absenteeism and costs due to back pain. In: Deyo RA, ed. Occupational Back Pain. Philadephia: Hanley and Belfus, 1987: Dillane JB, Fry J, Kalton G. Acute back syndrome a study from general practice. BMJ 1966;2: Dixon AJ. Problems of progress on back pain research. Rheumatol Rehabil 1973;12: Evans C, Gilbert JR, Taylor DW, et al. A randomized controlled trial of flexion exercises, education and bed rest for patients with acute low back pain. Physiother Can 1987;39: Farrell JP, Twomey LT. Acute low back pain: comparison of two conservative treatment approaches. Med J Aust 1982;1: Frymoyer JW. Epidemiology. In: Frymoyer JW, Gordon SL, ed. New Perspectives in Low Back Pain. Illinois: American Academy of Othopaedic Surgeons, 1988: Frymoyer JW, Pope MH, Clements JH, et al. Risk factors in low back pain: an epidemiological study. J Bone Joint Surg [Am] 1983;65: Goel VK, Kong W, Han JS, et al. A combined finite element and optimization investigation of lumbar spine mechanics with and without muscles. Spine 1993;18: Hides JA, Stokes MJ, Saide M, et al. Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 1994;19: Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after resolution of acute first episode low back pain. Spine 1996;21: Hides JA, Richardson CA, Jull GA. Use of real-time ultrasound imaging for feedback in rehabilitation. Manual Therapy 1998;3: Jenkins EM, Borenstein DG. Exercise for the low back pain patient. Clin Rheumatol 1994;8: Jull GA, Richardson CA. Rehabilitation of active stabilization of the lumbar spine. In: Twomey LT, Taylor JR, eds. Physical Therapy of the Low Back, 2nd ed. New York: Churchill Livingstone, 1994: Kaigle AM, Holm SH, Hansson TH. Experimental instability in the lumbar spine. Spine 1995;20: Kelsey JL. Epidemiology of Musculoskeletal Disorders. New York: Oxford University Press, Lehmann TR, Spratt KF, et al. Predicting long-term disability in low back injured workers presenting to a spine consultant. Spine 1993;18: Morris A. Identifying workers at risk to back injury is not guesswork. Occup Health Saf 1985;55: Nachemson A. Advances in low back pain. Clin Orth 1985;200: O Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 1997;22: Panjabi M. The stabilizing system of the spine: Part I: Function, dysfunction, adaptation and enhancement. J Spinal Disord 1992a;5: Panjabi M. The stabilizing system of the spine. Part II: Neutral zone and instability hypothesis. J Spinal Disord 1992b;5: Panjabi M, Abumi K, Duranceau J, et al. Spinal stability and intersegmental muscle forces: a biomechanical model. Spine 1989;14: Rantanen J, Hurme M, Ralck B, et al. The lumbar multifidus muscle five years after surgery for a lumbar intervertebral disc herniation. Spine 1993; 18: Richardson CA, Jull GA. Muscle control pain control. What exercises would you prescribe? Manual Therapy 1995;1: Richardson CA, Jull GA, Hodges PW, et al. Therapeutic Exercise for Spinal Segmental Stabilization in Low Back Pain Churchill Livingstone: Edinburgh. 32. Sihvonen T, Herno A, Paljarvi L, et al. Local denervation atrophy of paraspinal muscles in postoperative failed back syndrome. Spine 1993;18: Spengler DM, Bigos SJ, Martin NA, et al. Back injuries in industry: a retrospective study. I. Overview and cost analysis. Spine 1986;11: Steffen R, Nolte LP, Pingel TH. Rehabilitation of post-operative segmental lumbar instability. A biomechanical analysis of the rank of the back muscles. Rehabilitation 1994;33: Troup JD, Martin JW, Lloyd DC. Back pain in industry: a prospective survey. Spine 1981;6: Valkenberg HA, Haanen HCM. The epidemiology of low back pain. In White AA III, Gordon SL, eds. American Academy of Orthopaedic Surgeons Symposium on Idiopathic Low Back Pain. St. Louis: Mosby, 1982: Von Korff M, Deyo RA, Cherkin D, et al. Back pain in primary care: outcomes at one year. Spine 1993;18: Von Korff M, Dworkin SF, LeResche L. Graded classification of chronic pain: an epidemiologic evaluation. Pain 1990;40; Von Korff M, Ormel J, Keefe F, et al. Grading the severity of chronic pain. Pain 1992;50: Von Korff M, Saunders K. The course of back pain in primary care. Spine 1996;21: Wilke HJ, Wolf S, Claes LE, et al. Stability increase of the lumbar spine with different muscle groups: a biomechanical in vitro study. Spine 1995;20: Address reprint requests to Julie A. Hides Department of Physiotherapy Mater Misericordiae Adult Hospital Raymond Terrace South Brisbane, Queensland 4101 AUSTRALIA backclin@mater.org.au

46 Journal of Electromyography and Kinesiology 13 (2003) Pain and motor control of the lumbopelvic region: effect and possible mechanisms Paul W. Hodges a,, G. Lorimer Moseley a,b a Department of Physiotherapy, The University of Queensland, Brisbane, Qld 4072, Australia b Royal Brisbane Hospital, Brisbane Qld, Australia Abstract Many authors report changes in the control of the trunk muscles in people with low back pain (LBP). Although there is considerable disagreement regarding the nature of these changes, we have consistently found differential effects on the deep intrinsic and superficial muscles of the lumbopelvic region. Two issues require consideration; first, the potential mechanisms for these changes in control, and secondly, the effect or outcome of changes in control for lumbopelvic function. Recent data indicate that experimentally induced pain may replicate some of the changes identified in people with LBP. While this does not exclude the possibility that changes in control of the trunk muscles may lead to pain, it does argue that, at least in some cases, pain may cause the changes in control. There are many possible mechanisms, including changes in excitability in the motor pathway, changes in the sensory system, and factors associated with the attention demanding, stressful and fearful aspects of pain. A new hypothesis is presented regarding the outcome from differential effects of pain on the elements of the motor system. Taken together these data argue for strategies of prevention and rehabilitation of LBP 2003 Elsevier Science Ltd. All rights reserved. Contents 1. Introduction Changes in motor control of the lumbopelvic region Possible mechanisms for pain to affect motor control of the trunk muscles Possible outcomes of motor control changes Introduction Changes in motor control and function of the trunk muscles have been reported frequently in the literature. These changes range from changes in recruitment to reduced strength and endurance of the trunk muscles. Notably, patterns of hyperactivity and hypoactivity have been reported and a variety of hypotheses have been developed to explain the effects and mechanisms of the Corresponding author. Tel.: ; fax: address: p.hodges@shrs.uq.edu.au (P.W. Hodges). changes. The majority of available hypotheses are broadly consistent with two main theories that propose; (i) that changes in muscle activity cause spinal pain (muscle tension or pain spasm pain model), or (ii) changes in muscle activity serve to restrict spinal motion (pain adaptation model). Experimental (e.g. [11,70,95,113]), and clinical [7,50] data suggest that the muscle tension model is too simplistic, and offer support for the pain adaptation model [63], however considerable debate exists [114]. The purpose of this paper is to review the evidence for changes in motor control, discuss possible mechanisms for these changes and the effect of these changes on function of the lumbopelvic region /03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved. doi: /s (03)

47 362 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) Changes in motor control of the lumbopelvic region Although early studies of trunk muscle function focused on the strength and endurance of the trunk muscles in patients with LBP (e.g. [94,101]), more recently the focus has shifted to issues of motor control. The challenge of motor control of the lumbopelvic region is immense and must serve to move and control the spine in a range of environments and with a complex interaction between internal and external forces. The challenge is further complicated by the fact that without muscle the spine and pelvis are inherently unstable [62,77]. Trunk muscles must have sufficient strength and endurance to satisfy the demands of control, but the efficacy of the muscle system is dependent on its controller, the central nervous system (CNS) [77]. The CNS must continually interpret the status of stability and movement, plan mechanisms to overcome predictable challenges and rapidly initiate activity in response to unexpected challenges. It must interpret the afferent input from peripheral mechanoreceptors and other sensory systems, consider this input and the impending requirements against an internal model of body dynamics and then generate a coordinated response of the trunk muscles so that the muscle activity occurs at the correct time, with the correct amplitude, and so on. Further, muscle activity must be coordinated to maintain control of the spine within a hierarchy of interdependent levels; control of intervertebral translation and rotation, control of spinal posture/orientation, control of body with respect to the environment [44,45,77]. Finally, unlike the muscles of the limb, trunk muscles perform a variety of homeostatic functions in addition to movement and control of the trunk (e.g. respiration and continence) [39]. In view of the complex requirements of trunk muscle control, it is not surprising that aspects of control are altered in people with LBP. Many studies report changes in motor control in people with acute and chronic LBP (e.g. [45,81,89]). While there is considerable variability in results, we have consistently found differential changes in activity between the deep and superficial trunk muscles. In terms of deep intrinsic trunk muscle activity, there is evidence of delayed activity of transversus abdominis (TrA), the deepest of the abdominal muscles (recorded with intramuscular EMG electrodes), in association with rapid limb movements in people with chronic LBP [47]. It is well accepted that the CNS initiates a sequence of muscle activity involving the limb and trunk muscles in advance of limb movements to prepare the body for the predictable disturbance to stability from the reactive forces caused by movement [3,4,8,46]. This sequence of responses is feedforward, that is, it is preplanned by the CNS and occurs in advance of the movement. Therefore, these responses precede any afferent input from the movement [4]. While changes have been identified in these tasks in feedforward activity of both deep and superficial muscles, the most consistent change (between subjects and movements) occurs in TrA [45]. The changes in TrA have been replicated when pain is induced by intramuscular injection of hypertonic saline into the longissimus at L4 [41]. Notably, the changes observed in patients were identified in people who had a history of LBP but were in remission from their symptoms. Although these studies have reported a delay in activation, it is likely that the change is not confined to this parameter, but instead may be reflective of a change in control. For example tonic activity of TrA, which is normally observed during repetitive trunk [17] and limb movements [40], is reduced during experimentallyinduced pain [41]; relative EMG activity of rectus abdominis and EMG activity recorded with electrodes over the inferiolateral abdominal wall is altered in people with chronic LBP during a novel task to move the abdominal wall inwards [75]. There is preliminary evidence that the deep paraspinal muscles show similar changes in activity. During functional tasks, there is reduced amplitude of activity of multifidus, the deepest back muscle in the lumbar region, in people with LBP [60,89] and altered responses have been observed during loading of the trunk. For example, when a load is unexpectedly dropped into the hands, there is normally a short-latency response of the paraspinal muscles [59,73]. In healthy control subjects, studies have reported earlier activity of the deep [73] and superficial [59] fibres of multifidus when the loading can be anticipated compared to trials when the load cannot be anticipated [73]. However, when people with sciatica catch a load that is predictable, the response of the paraspinal muscles (recorded with surface electrodes) does not occur earlier than the unpredictable trials [59]. Others, using an unexpected loading paradigm, report both delayed [67,109] and no change [113] in activity of the paraspinal muscles. The apparent greater specificity of the delay to predictable tasks suggests that the change is dependent on input from higher centres of the CNS. The reported changes in activity of multifidus are consistent with changes in its morphology and fatigability, which in turn could be explained by altered use of the muscle. For example, studies report changes in muscle fibre composition [83] and increased fatigability [6,87], and reduced cross-sectional area of multifidus has been identified as little as 24 hours after the onset of acute, unilateral LBP [37], although it is not clear as to whether this is a premorbid phenomenon. In summary, the evidence seems to suggest that, with LBP, there is an alteration in control of the deep intrinsic spinal muscles that consistently manifests as hypoactivity. The possible implications of these changes are discussed below. Although others argue that paraspinal muscles react to

48 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) pain and injury with hyperactivity (e.g. [49,90,116]) this may vary between components of the paraspinal group. Due to the ease of accessibility of the superficial trunk muscles to surface EMG recordings, there is a large literature that investigates changes in these muscles in LBP. Despite the popularity of the muscle tension model, there is considerable debate about the presence of augmented activity of the paraspinal muscles. Studies have had variable results, some reporting increased activity [1,111], others reporting decreased [89] or asymmetrical activity [16] and others reporting no change in activity [15], for a review see [114]. One finding that has been consistently observed in people with LBP is sustained activity of the erector spinae muscles at the end of range of spinal flexion, a point at which the erector spinae muscles are normally inactive (the so-called flexion relaxation response) [88]. Importantly, this normal response is only lost in a subset of patients, suggesting that factors other than the presence of pain influence this change in activity (e.g. fear of pain, see below). Nonetheless, this finding has been replicated by experimental pain [113] and has been shown to limit intervertebral motion [56]. In this regard, changes in paraspinal muscle activity during gait may serve a similar purpose. The normal periods of silence in erector spinae activity between heel contacts are reduced in LBP patients and in otherwise symptomatic participants given experimentally induced LBP [2], which may serve to splint the area during this period. Variability in superficial trunk muscle activity associated with pain has been observed in other tasks. In a study by Radebold and colleagues [81] in which a load was removed from the trunk, augmentation of superficial trunk muscles was observed, but only in a subset of patients. Experimentally elicited pain caused variable responses of the superficial trunk muscles in association with rapid limb movements [41]. However, importantly, although there was considerable inter-subject variability in the pattern of superficial trunk muscle activity, at least one superficial muscle was augmented during pain in every subject. Notably the hypoactivity of the intrinsic spinal muscle, TrA, was a consistent finding across the group. In addition to changes in muscle recruitment, impairment of other elements of motor control has been identified in people with LBP. For example changes in balance control and sensory aspects. Balance has been shown to be impaired in people with LBP when standing on one [64] or two legs [10] or sitting [82], and people with poor performance in a test of standing balance have an increased risk of LBP [100]. Because both feedforward and feedback-mediated components of motor control are dependent on sensory input, any change in sensory input is likely to be important. Numerous studies have reported reduced acuity [32] and impaired ability to perform repositioning tasks [9] in people with LBP. Other more complex elements of control have also been found to be altered in LBP. For instance people with LBP have a slower reaction time [65], and slow reaction time has been associated with musculoskeletal injuries (including LBP) in a variety of sports [99]. Although there is marked variability between individuals and studies, the relationship between pain and motor control of the spine appears complex. Importantly, there is increasing evidence for differential changes in activity of the deep and superficial trunk muscles with pain. In this regard, two issues require consideration. What are the possible mechanisms for this change and what are the potential outcomes in terms of spinal function? 3. Possible mechanisms for pain to affect motor control of the trunk muscles It is not certain whether pain causes changes in motor control or whether motor control changes lead to pain, or both. Farfan [25] and Panjabi [77], amongst others, have presented models that suggest that deficits in motor control lead to poor control of joint movement, repeated microtrauma and pain. Consistent with this model, Janda [53] has argued that people who have mild neurological signs (e.g. minor coordination difficulties) are more likely to have pain as adults. Furthermore slow reaction times have been linked to increased risk of musculoskeletal injury [99]. However, the converse may also be true. Perhaps pain leads to changes in motor control. Numerous studies using experimental models of pain have provided support for this hypothesis by reproduction of changes in control that have been identified in clinical populations [2,41,113]. Consequently, a number of mechanisms have been proposed to explain the effect of pain on motor control (Fig. 1). These include changes in excitability at the spinal or cortical level, changes in proprioception or afferent mediated control, or specific cortical effects imparted by aspects of pain, such as its demand on CNS resources, stress or fear. The following sections will review each of these possible mechanisms. Widespread changes in excitability have been identified at many levels of the motor system during pain. Acute experimental pain has been shown to cause changes in spinal motoneuron activity [70,96,97]. For instance, increased stretch reflex amplitude of the soleus muscle has been reported after intramuscular injection of hypertonic saline [70]. Others report reduced amplitude of motor potential evoked by transcranial magnetic stimulation over the motor cortex in response to experimental pain [102]. However, these responses may be task or muscle specific as other studies have reported no changes in excitability of the motoneuron or motor cortex [29,112]. Those authors argued that changes in motor drive may occur upstream of the motor cortex, for

49 364 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) Fig. 1. Possible mechanisms for pain to affect motor control. Multiple mechanisms have been proposed for pain to affect motor control. It is unlikely that the simple inhibitory pathways (left) can mediate the complex changes in motor control of the trunk muscles. The most likely candidates are changes in motor planning via a direct influence of pain on the motor centres, fear-avoidance, or due to changes in the sensory system. instance, involving areas associated with motor planning. Reflex inhibition of motoneuron excitability has also been suggested to occur in association with swelling [92] and injury to joint structures [24], which has been argued to indicate polysynaptic inhibition at a spinal level [93]. While this may be a factor in clinical populations, it cannot explain the findings of studies of experimental pain that are not associated with oedema and injury, and similar effects cannot be produced by injection of similar volumes of isotonic saline [33]. Evidence from several groups argues that changes in trunk muscle activity in LBP may not be mediated by simple changes in excitability. Zedka et al [113] were unable to identify changes in the short latency response of the paraspinal muscles to a mechanical tap to the muscle following pain induced by injection of hypertonic saline into the muscle (changes in this component would be consistent with changes in motoneuron excitability). These authors did find changes in later components of the response that can be influenced by input from higher centres. We have shown several changes in coordination of the trunk muscles in association with pain that are inconsistent with a change in excitability or delayed transmission of the motor command. For example, when people move an arm rapidly, normally the response of TrA is independent of the direction of arm movement [46]. If the delay in response observed during pain was due to a change in excitability it may be predicted that the response would remain consistent between movement directions, although delayed. However, this is not the case. The response of TrA in people with LBP is earlier with shoulder extension than the other movements, which is similar to the response of the superficial trunk muscles, normally under differential control [45,73]. Also, in a healthy population, when the preparation for movement is reduced, despite slowing of the response of the prime mover of the arm and the oblique abdominal muscles, the response of TrA is not affected [48]. However, in people with LBP, the response is delayed along with the increased reaction time of the movement in the reduced preparation trials [42]. Taken together, these findings are likely to represent a change in motor planning. Consistent with the identification of changes in motor planning there is compelling evidence that pain has strong effects at the supraspinal level [20,38,57,61,64,66,104]. Both short- and long-term changes are thought to occur in activity of the supraspinal structures including the cortex with pain. Many studies have reported changes during experimental pain in activity of regions of the brain involved in movement planning and performance (see [20]). One area that has been consistently found to be affected is the anterior cingulate cortex (ACC) [79]. The ACC has also been reported to be chronically active in people with chronic LBP [51]. The ACC has long been thought to be important in motor responses and directly projects to motor and supplementary motor areas [80]. Hypothetically at least, activation of these cortical regions during pain may influence movement control directly and mediate the changes reported above. However, confirmation of this hypothesis is difficult because movement is not permitted in many imaging studies. Other authors have identified increased activity in areas of somatosensory cortex activated by noxious cutaneous stimulation of the finger and back in people with LBP [27]. Furthermore, the area activated increased as a function of the duration of their pain. These changes may contribute to the perpetuation of pain in the absence of peripheral nociception, but may also contribute to the motor changes. Further work is required to clarify these findings as they relate to motor control.

50 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) Although nociceptive stimulation and pain may disrupt motor output directly, it is also possible that an effect is caused by aspects of pain, such as its attention demanding requirements, stress or fear. In terms of attention demand, it is widely considered that pain utilises attentional resources, probably by virtue of its direct relevance for survival (see [80] for review). Several studies support this hypothesis. For example, recordings of event-related potentials in the cortex [86], brain imaging studies [20], cognitive performance tasks [18,19,22] and a combination of these methods [61] indicate increased latencies and/or error rates in the presence of pain. Thus pain may lead to changes in movement coordination as a result of the increased demand placed on information processing resources. While several authors have identified slower reaction times in people with LBP, which may be attributable to this mechanism [66], we have recently shown that performance of an attentiondemanding task does not replicate the changes in trunk muscle activity seen in people with LBP [72]. In this study subjects rapidly moved an arm in response to a visual stimulus while performing an attentiondemanding task. Although the reaction time of the arm movement was delayed, the response of the deep trunk muscles (TrA and deep MF) occurred earlier relative to the deltoid response (i.e. opposite to the changes seen in LBP). There was no change in the activity of the superficial abdominal or paraspinal muscles. A further possibility is that the stress associated with pain produces the change in control of the trunk muscles. Numerous studies have shown that stress (i.e. perception of threat) may affect motor control [54,103,108]. Notably, trunk muscle activity during a lifting task is altered when the task is performed in the presence of psychosocial stressors [69] and shoulder muscle activity during a keyboard task is altered by work-related stress [23]. Furthermore, changes in paraspinal muscle activity in chronic pain patients have been linked to subjective measures of distress and anxiety, rather than just the intensity of pain [26,28,106]. We have tested the effect of stress on the postural response of the trunk muscles during rapid arm movements by repeating the attentiondemanding task described above, but with negative feedback of performance and other negative psychosocial cues [72]. Although the addition of stress did not replicate the changes that we had identified with experimentally induced pain, there was a delay in the response of the deep trunk muscles relative to tasks when the attention demand was non-stressful, indicating some effect of stress. Another alternative argues that the changes in control may relate to the fear associated with pain. The notion that fear is important in behavioural and motor output associated with pain is not new, with the fear-avoidance model gaining considerable support in the literature (see [105] for review). In brief, the fear avoidance model argues that fear of pain and (re)injury prevents normal return to activity, which leads to deconditioning and disability [105]. Although the primary application of the fear avoidance model has been in consideration of behavioural response to pain and injury, corresponding findings have been reported in the pattern of motor control [68]. Several studies have reported differences in trunk muscle activity between fearful and non-fearful back pain patients. For instance fearful patients have a greater reduction in endurance of the paraspinal muscles [6] and less relaxation of the paraspinal muscles at the end of trunk flexion [107] than non-fearful patients and controls. Furthermore, it has been suggested that chronic LBP patients have increased paraspinal muscle activity when they are exposed to personally relevant stressors but not when they are exposed to general stressors [26]. Finally, when pain-free subjects rapidly move an arm, but are subjected to moderately painful electrical shocks to the back that are unpredictable in time and amplitude, the response of TrA and deep MF is delayed in a manner that is similar to that seen with experimentally induced LBP [72]. While the latter finding does not confirm that fear of pain causes the changes seen in people with LBP it does suggest that fear may at least replicate the changes. Moreover, it is possible that both pain and fear of pain act directly on the motor centres through a common mechanism. It is important to consider that fear of pain may explain why people who have a history of pain have delayed activity of TrA. Furthermore, if fear of pain can disrupt the normal control of the trunk muscles, this may provide a link between psychosocial factors and physiological changes that lead to recurrence of pain. It could also be interpreted that these changes in motor control are an adaptation to limit loading and prevent recurrence. However, we propose that these adaptive strategies may provide a short term solution with long term sequelae (see below). If pain or other supraspinal mechanisms such as fear can disrupt motor control, why does this lead to the relatively consistent finding of reduced activity of deep intrinsic spinal muscles and increased activity of the large superficial muscles? The explanation may lie in the pain-adaptation model of Lund and colleagues [63]. This model stipulates that in the event of pain, the alteration in motor control serves to limit movement. During movement, this involves a decrease in agonist muscle activity and an increase in antagonist activity so as to limit the velocity, force and range of movement [95]. This pattern of response has been observed in clinical and experimental pain studies for many regions of the body including the jaw [95] and trunk [113]. In terms of control of a segment such as the trunk, the response may also involve general stiffening of the body segment(s) by muscle co-activation. Panjabi [77] and Cholewicki [13] predicted that such a response would

51 366 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) increase vertebral control and is consistent with the augmented activity of the large, superficial trunk muscles. Consistent with this, there is evidence of relative stiffening of the spine in pain. Moe-Nilssen et al. [71] reported reduced trunk movement during gait during experimentally induced pain, and [35] showed that trunk movement following a support surface translation is reduced during pain. Hypothetically, if the general stiffness of the spine is increased, the CNS may perceive the demand for fine-tuning to be diminished, leading to reduced activity of the deep intrinsic spinal muscles despite the potential long-term sequelae of this strategy (see below). After resolution of the pain, this adapted strategy may also resolve, or, in the presence of ongoing fear of pain or other reinforcement, persist to chronicity. This hypothesis requires investigation. An additional factor to consider is that accurate control of movement is dependent on the sensory element of the motor system. Inaccurate afferent input would affect all aspects of motor control from simple reflex responses (e.g. those arising from stimulation of mechanoreceptors in the muscles [113] or other elements of the spine [52,91,116]) to complex movements that are dependent on an accurate internal model of body dynamics (see [34]), which allows the CNS to predict the interaction between internal and external forces. Several studies have reported decreased acuity to spinal motion in LBP [98] and impaired ability to accurately reposition with LBP [9,32]. In addition, muscle spindle sensitivity is altered by pain (e.g. [78] and muscle activity [30], thus any change in activation may adversely affect perception of movement. Finally, several studies have argued that sensory acuity may be reduced by fatigue [12], thus decreased muscle endurance with injury or pain may lead to impaired sensory acuity via increased fatigability. 4. Possible outcomes of motor control changes In view of the differential changes in the deep intrinsic muscles and the superficial muscles in the presence of pain, it is critical to consider possible sequelae of these changes. All trunk muscles are required for control and stability of the spine [77] and it is clear that stability is dependent on the interplay between an array of muscles, both intrinsic and superficial [5,14,115]. Yet, there is considerable redundancy in the motor system with many muscles potentially able to perform similar functions. A change in strategy of trunk muscle control, toward increased stiffening of the spine via increased activity of large superficial muscles, which has been predicted [13,77] and shown to occur (see above), would seem to satisfy the demands for spinal function. However, we propose several side effects of this strategy that may compromise lumbopelvic health and potentially lead to long-term sequelae. The basis for this hypothesis is that the contribution of the deep intrinsic spinal muscles to trunk control is that of fine-tuning of intervertebral motion. Although it is unlikely that differentiation in muscle function can be described in a dichotomous manner, in general it has been suggested that, in contrast to intrinsic muscles, the large and superficial trunk muscles that transcend the lumbar spine and pelvis have a more significant contribution to prevention of buckling of the spine [5,14,115] and to balance external loads [5]. These are also the muscles that have the greatest potential to generate torque to move the trunk. In contrast, in vivo [43,55,85], in vitro [110] and modelling studies [110] argue that the deep intrinsic muscles, such as TrA and the deep fibres of multifidus, are critical for the control of intervertebral motion. Thus, data suggests that the deep muscles might provide the fine-tuning as a component of the complex interdependent activity of the trunk muscles to stabilise the spine. We suggest that in the pain adaptation model, the response to pain of stiffening the spine with increased activity of the large muscles may be at the cost of a loss of this fine tuning. Other factors require consideration. First, movement is an important element of spinal function. It is known that in healthy subjects the CNS uses movement rather than simple stiffening of the spine to overcome challenges to stability [44] and reduce energy expenditure [58]. A strategy of trunk stiffening, although requiring less complex neural control, may compromise optimal spinal function. Second, co-activation of the superficial muscles may have a loading cost. The superficial trunk muscles generate torque at the trunk. This torque must be overcome by antagonist activation in order to keep the spine upright, and this co-activation results in a compressive load on the spine [31]. Excessive compression, which results in increased intradiscal pressure and loading through the posterior elements of the spine has long been considered to be a risk factor for spinal degeneration and pain [74]. If greater demand is placed on the superficial muscle system, the loading may be increased. Third, trunk muscles are involved in functions other than spinal control and movement. As the superficial abdominal muscles depress the rib cage and are involved in forced expiration [21], increased activity of these muscles in people with pain may lead to compromised respiratory function, for example restricted movement of the chest wall. In contrast, TrA has a limited effect on rib cage motion due to its horizontal fibre orientation and contributes to expiratory airflow via rostral displacement of the abdominal contents [21]. In a recent study we have shown that of the abdominal muscles only TrA can coordinate respiratory and postural functions [40]. Thus, changes in trunk muscle activity may be problematic from a systemic point of view. While each of these hypotheses requires further investigation, additional support comes from the litera-

52 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) ture which suggests that reorganisation of the control of the deep and superficial trunk muscles through motor learning strategies [84] leads to reduced pain and disability associated with LBP [76] and reduced recurrence of pain [36]. In summary, relatively consistent patterns of change in activity have been identified in the literature, although there are considerable inter-individual differences. While the mechanisms of these changes are not completely understood, there is compelling evidence to suggest that pain may be responsible for the change, at least in some individuals. The consequence of these changes may potentially be a factor in the recurrence of LBP. Taken together these data are consistent with contemporary strategies for rehabilitation of patients with LBP. References [1] J.G. Arena, R.A. Sherman, G.M. Bruno, T.R. Young, Electromyographic recordings of 5 types of low back pain subjects and nonpain controls in different positions, Pain 37 (1) (1989) [2] L. Arendt-Nielsen, T. Graven-Nielsen, H. Svarrer, P. Svensson, The influence of low back pain on muscle activity and coordination during gait: a clinical and experimental study, Pain 64 (2) (1996) [3] A.S. Aruin, M.L. Latash, Directional specificity of postural muscles in feed-forward postural reactions during fast voluntary arm movements, Experimental Brain Research 103 (1995) [4] V. Belenkii, V.S. Gurfinkel, Y. Paltsev, Elements of control of voluntary movements, Biofizika 12 (1) (1967) [5] A. Bergmark, Stability of the lumbar spine. A study in mechanical engineering, Acta Orthopedica Scandinavica 60 (Suppl 230) (1989) [6] H.J. Biederman, G.L. Shanks, W.J. Forrest, J. Inglis, Power spectrum analysis of electromyographic activity: discriminators in the differential assessment of patients with chronic low back pain, Spine 16 (10) (1991) [7] E.B. Blanchard, F. Andrasik, J.G. Arena, S.J. Teders, Variation in meaning of pain descriptors for different headache types as revealed by psychophysical scaling, Headache 22 (3) (1982) [8] S. Bouisset, M. Zattara, A sequence of postural adjustments precedes voluntary movement, Neuroscience Letters 22 (1981) [9] S. Brumagne, P. Cordo, R. Lysens, S. Verschueren, S. Swinnen, The role of paraspinal muscle spindles in lumbosacral position sense in individuals with and without low back pain, Spine 25 (8) (2000) [10] N.N. Byl, P.L. Sinnott, Variations in balance and body sway in middle-aged adults: subjects with healthy backs compared with subjects with low back dysfunction, Spine 16 (1991) [11] N.F. Capra, J.Y. Ro, Experimental muscle pain produces central modulation of proprioceptive signals arising from jaw muscle spindles, Pain 86 (1 2) (2000) [12] J.E. Carpenter, R.B. Blasier, G.G. Pellizzon, The effects of muscle fatigue on shoulder joint position sense, American Journal of Sports Medicine 26 (2) (1998) [13] J. Cholewicki, M.M. Panjabi, A. Khachatryan, Stabilizing function of trunk flexor extensor muscles around a neutral spine posture, Spine 22 (19) (1997) [14] J. Cholewicki, I.J. Van Vliet, Relative contribution of trunk muscles to the stability of the lumbar spine during isometric exertions, Clinical Biomechanics 17 (2) (2002) [15] G.A. Collins, M.J. Cohen, B.D. Naliboff, S.L. Schandler, Comparative analysis of paraspinal and frontalis emg, heart rate and skin conductance in chronic low back pain patients and normals to various postures and stresses, Scandinavian Journal of Rehabilitation Medicine 14 (1982) [16] J.R. Cram, J.C. Steger, Emg scanning in the diagnosis of chronic pain, Biofeedback and Self Regulation 8 (1983) [17] A.G. Cresswell, H. Grundstrom, A. Thorstensson, Observations on intra-abdominal pressure and patterns of abdominal intra-muscular activity in man, Acta Physiologica Scandinavica 144 (1992) [18] G. Crombez, C. Eccleston, F. Baeyens, P. Eelen, Attentional disruption is enhanced by the threat of pain, Behavioural Research and Therapy 36 (2) (1998) [19] G. Crombez, C. Eccleston, F. Baeyens, B. van Houdenhove, A. van den Broeck, Attention to chronic pain is dependent upon pain-related fear, Journal of Psychosomatic Research 47 (5) (1999) [20] S.W. Derbyshire, A.K. Jones, F. Gyulai, S. Clark, D. Townsend, L.L. Firestone, Pain processing during three levels of noxious stimulation produces differential patterns of central activity, Pain 73 (3) (1997) [21] A. DeTroyer, M. Estenne, Functional anatomy of the respiratory muscles, in: M.J. Belman (Ed.), Respiratory Muscles: Function in Health and disease, W.B. Saunders Co, Philadelphia, 1988, pp [22] C. Eccleston, G. Crombez, Pain demands attention: a cognitive affective model of the interruptive function of pain, Psychological Bulletin 125 (3) (1999) [23] K. Ekberg, J. Eklund, Psychological stress and muscle activity during data entry at visual display units, Work Stress 9 (1995) [24] J. Ekholm, G. Eklund, S. Skoglund, On reflex effects from knee joint of cats, Acta Physiologica Scandinavica 50 (1960) [25] H.F. Farfan, Mechanical disorders of the low back, Lea and Febiger, Philadelphia, [26] H. Flor, N. Birbaumer, Symptom-specific psychophysiological responses in chronic pain patients, Psychophysiology 29 (1992) [27] H. Flor, C. Braun, T. Elbert, N. Birbaumer, Extensive reorganization of primary somatosensory cortex in chronic back pain patients, Neuroscience Letters 224 (1) (1997) 5 8. [28] H. Flor, D.C. Turk, Psychophysiology of chronic pain: do chronic pain patients exhibit symptom-specific psychophysiological responses?, Psychological Bulletin 105 (2) (1989) [29] S.C. Gandevia, G.M. Allen, J.E. Butler, J.L. Taylor, Supraspinal factors in human muscle fatigue: evidence for suboptimal output from the motor cortex, Journal of Physiology (London) 490 (Pt 2) (1996) [30] S.C. Gandevia, D.I. McCloskey, D. Burke, Kinaesthetic signals and muscle contraction, Trends in Neurosciences 15 (2) (1992) [31] M.G. Gardner-Morse, I.A. Stokes, The effects of abdominal muscle coactivation on lumbar spine stability, Spine 23 (1) (1998) [32] K.P. Gill, M.J. Callaghan, The measurement of lumbar proprioception in individuals with and without low back pain, Spine 23 (3) (1998) [33] T. Graven-Nielsen, P. Svensson, L. Arendt-Nielsen, Effects of experimental muscle pain on muscle activity and coordination during static and dynamic motor function, Electroencephalography and Clinical Neurophysiology 105 (2) (1997) [34] V.S. Gurfinkel, The mechanisms of postural regulation in man, Soviet Scientific Reviews. Section F. Physiology and general biology 7 (1994)

53 368 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) [35] S. Henry, Postural responses in persons with low back pain, in: J. Duysens, B.C.M. Smits-Engelsman, H. Kingma (Eds.), Control of posture and gait, International Society for Posture and Gait Research, Maastricht, 2001, pp [36] J.A. Hides, G.A. Jull, C.A. Richardson, Long term effects of specific stabilizing exercises for first episode low back pain, Spine 26 (2001) [37] J.A. Hides, M.J. Stokes, M. Saide, G.A. Jull, D.H. Cooper, Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain, Spine 19 (2) (1994) [38] P. Hodges, Changes in motor planning of feedforward postural responses of the trunk muscles in low back pain, Experimental Brain Research 141 (2001) [39] P. Hodges, S. Gandevia, Activation of the human diaphragm during a repetitive postural task, Journal of Physiology (London) 522 (2000) [40] P. Hodges, S. Gandevia, Changes in intra-abdominal pressure during postural and respiratory activation of the human diaphragm, Journal of Applied Physiology 89 (3) (2000) [41] P. Hodges, G. Moseley, A. Gabrielsson, S. Gandevia, Acute experimental pain changes postural recruitment of the trunk muscles in pain-free humans. Society for Neuroscience Abstracts, [42] P.W. Hodges, Changes in motor planning of feedforward postural responses of the trunk muscles in low back pain, Experimental Brain Research 141 (2) (2001) [43] P. Hodges, A. Kaigle-Holm, S. Holm, L. Ekstrom, A.G. Cresswell, T. Hansson, A. Thorstensson, In vivo evidence that postural activity of the respiratory muscles increases intervertebal stiffness of the spine: porcine studies. Society for Neuroscience Abstracts 2002; [44] P.W. Hodges, A.G. Cresswell, A. Thorstensson, Preparatory trunk motion accompanies rapid upper limb movement, Experimental Brain Research 124 (1999) [45] P.W. Hodges, C.A. Richardson, Inefficient muscular stabilisation of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdominis, Spine 21 (1996) [46] P.W. Hodges, C.A. Richardson, Feedforward contraction of transversus abdominis in not influenced by the direction of arm movement, Experimental Brain Research 114 (1997) [47] P.W. Hodges, C.A. Richardson, Delayed postural contraction of of transversus abdominis associated with movement of the lower limb in people with low back pain, Journal of Spinal Disorders 11 (1) (1998) [48] P.W. Hodges, C.A. Richardson, Transversus abdominis and the superficial abdominal muscles are controlled independently in a postural task, Neuroscience Letters 265 (2) (1999) [49] S. Holm, A. Indahl, M. Solomonow, Sensorimotor control of the spine, J Electromyography and Kinesiology 12 (3) (2002) [50] K.A. Holroyd, D.B. Penzien, K.G. Hursey, D.L. Tobin, L. Rogers, J.E. Holm, P.J. Marcille, J.R. Hall, A.G. Chila, Change mechanisms in EMG biofeedback training: cognitive changes underlying improvements in tension headache, Journal of Consultation and Clinical Psychology 52 (6) (1984) [51] J.C. Hsieh, M. Belfrage, S. Stone-Elander, P. Hansson, M. Ingvar, Central representation of chronic ongoing neuropathic pain studied by positron emission tomography, Pain 63 (2) (1995) [52] A. Indahl, A. Kaigle, O. Reikeras, S. Holm, Electromyographic response of the porcine multifidus musculature after nerve stimulation, Spine 20 (24) (1995) [53] V. Janda, Muscles, central nervous motor regulation and back problems, in: I.M. Korr (Ed.), The Neurobiologic Mechanisms in Manipulative Therapy, New York, Plenum Press, 1978, pp [54] G. Jones, A. Cale, Goal difficulty, anxiety and performance, Ergonomics 40 (3) (1997) [55] A.M. Kaigle, S.H. Holm, T.H. Hansson, 1997 Volvo award winner in biomechanical studies. Kinematic behavior of the porcine lumbar spine: a chronic lesion model, Spine 22 (24) (1997) [56] A.M. Kaigle, P. Wessberg, T.H. Hansson, Muscular and kinematic behavior of the lumbar spine during flexion extension, Journal of Spinal Disorders 11 (2) (1998) [57] T. Kuukkanen, E. Malkia, Effects of a three-month active rehabilitation program on psychomotor performance of lower limbs in subjects with low back pain: a controlled study with a nine-month follow-up, Perceptual Motor Skills 87 (3, Pt 1) (1998) [58] C.J. Lamoth, O.G. Meijer, P.I. Wuisman, J.H. van Dieen, M.F. Levin, P.J. Beek, Pelvis thorax coordination in the transverse plane during walking in persons with nonspecific low back pain, Spine 27 (4) (2002) E92 E99. [59] V. Leinonen, M. Kankaanpaa, M. Luukkonen, O. Hanninen, O. Airaksinen, S. Taimela, Disc herniation-related back pain impairs feed-forward control of paraspinal muscles, Spine 26 (16) (2001) E367 E372. [60] K.-A. Lindgren, T. Sihvonen, E. Leino, M. Pitkänen, H. Manninen, Exercise therapy effects on fuctional radiographic findings and segmental electromyographic activity in lumbar spine instability, Archives of Physical Medicine and Rehabilitation 74 (1993) [61] J. Lorenz, B. Bromm, Event-related potential correlates of interference between cognitive performance and tonic experimental pain, Psychophysiology 34 (4) (1997) [62] D.B. Lucas, B. Bresler, Stability of the ligamentous spine, Technical Report esr. 11 no. 40. Biomechanics Laboratory, University of California, Berkeley and San Francisco, [63] J.P. Lund, R. Donga, C.G. Widmer, C.S. Stohler, The pain-adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity, Canadian Journal of Physiology and Pharmacology 69 (5) (1991) [64] S. Luoto, H. Aalto, S. Taimela, H. Hurri, I. Pyykko, H. Alaranta, One-footed and externally disturbed two-footed postural control in patients with chronic low back pain and healthy control subjects. A controlled study with follow-up, Spine 23 (19) (1998) [65] S. Luoto, M. Heliövaara, H. Hurri, H. Alaranta, Static back endurance and the risk of low-back pain, Clinical Biomechanics 10 (6) (1995) [66] S. Luoto, S. Taimela, H. Hurri, H. Alaranta, Mechanisms explaining the association between low back trouble and deficits in information processing. A controlled study with follow-up, Spine 24 (3) (1999) [67] M. Magnusson, A. Aleksiev, D. Wilder, M. Pope, K. Spratt, S. Lee, Unexpected load and asymmetric posture as etiologic factors in low back pain, European Spine Journal 5 (1) (1996) [68] C.J. Main, P.J. Watson, What harm pain behavior? Psychological and physical factors in the development of chronicity, Bulletin of Hospital Joint Disorders 55 (4) (1996) [69] W.S. Marras, K.G. Davis, C.A. Heaney, A.B. Maronitis, W.G. Allread, The influence of psychosocial stress, gender, and personality on mechanical loading of the lumbar spine, Spine 25 (23) (2000) [70] D.A. Matre, T. Sinkjaer, P. Svensson, L. Arendt-Nielsen, Experimental muscle pain increases the human stretch reflex, Pain 75 (2 3) (1998) [71] R. Moe-Nilssen, A.E. Ljunggren, E. Torebjork, Dynamic adjustments of walking behavior dependent on noxious input in experimental low back pain, Pain 83 (3) (1999)

54 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) [72] G.L. Moseley, P.W. Hodges, S.C. Gandevia, Attention demand, anxiety and acute pain cause differential effects on postural activation of the abdominal muscles in humans. Society for Neuroscience Abstracts, [73] G.L. Moseley, P.W. Hodges, S.C. Gandevia, External perturbation to the trunk is associated with differential activity of the deep and superficial fibres of lumbar multifidus, in: 4th Interdisciplinary World Congress on Low Back and Pelvic Pain, Montreal, Canada, [74] A. Nachemson, J.M. Morris, In vivo measurement of intradiscal pressure: Discometry, a method for the determination of presure in the lower lumbar discs, Journal of Bone and Joint Surgery 46A (1964) [75] P.B. O Sullivan, L. Twomey, G.T. Allison, Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention, Journal of Orthopedic and Sports Physical Therapy 27 (2) (1998) [76] P.B. O Sullivan, L.T. Twomey, G.T. Allison, Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis, Spine 22 (24) (1997) [77] M.M. Panjabi, The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, and enhancement, Journal of Spinal Disorders 5 (4) (1992) [78] J. Pedersen, P. Sjolander, B.I. Wenngren, H. Johansson, Increased intramuscular concentration of bradykinin increases the static fusimotor drive to muscle spindles in neck muscles of the cat, Pain 70 (1) (1997) [79] R. Peyron, B. Laurent, L. Garcia-Larrea, Functional imaging of brain responses to pain. A review and meta-analysis (2000), Neurophysiology Clinics 30 (5) (2000) [80] D.D. Price, Psychological Mechanisms of Pain and Analgesia, IASP Press, Seattle, [81] A. Radebold, J. Cholewicki, M.M. Panjabi, T.C. Patel, Muscle response pattern to sudden trunk loading in healthy individuals and in patients with chronic low back pain, Spine 25 (8) (2000) [82] A. Radebold, J. Cholewicki, G.K. Polzhofer, H.S. Greene, Impaired postural control of the lumbar spine is associated with delayed muscle response times in patients with chronic idiopathic low back pain, Spine 26 (7) (2001) [83] J. Rantanen, M. Hurme, B. Falck, H. Alaranta, F. Nykvist, M. Lehto, S. Einola, H. Kalimo, The lumbar multifidus muscle five years after surgery for a lumbar intervertebral disc herniation, Spine 18 (5) (1993) [84] C.A. Richardson, G.A. Jull, P.W. Hodges, J.A. Hides, Therapeutic Exercise for Spinal Segmental Stabilisation in Low Back Pain: Scientific Basis and Clinical Approach, Churchill Livingstone, Edinburgh, [85] C.A. Richardson, C.J. Snijders, J.A. Hides, L. Damen, M.S. Pas, J. Storm, The relation between the transversus abdominis muscles, sacroiliac joint mechanics, and low back pain, Spine 27 (4) (2002) [86] J.P. Rosenfeld, K. Bhat, A. Miltenberger, M. Johnson, Eventrelated potentials in the dual task paradigm: P300 discriminates engaging and non-engaging films when film-viewing is the primary task, International Journal of Psychophysiology 12 (3) (1992) [87] S.H. Roy, C.J. DeLuca, D.A. Casavant, Lumbar muscle fatigue and chronic low back pain, Spine 14 (9) (1989) [88] O. Shirado, T. Ito, K. Kaneda, T.E. Strax, Flexion relaxation phenomenon in the back muscles. A comparative study between healthy subjects and patients with chronic low back pain, American Journal of Physical Medicine and Rehabilitation 74 (2) (1995) [89] T. Sihvonen, K.A. Lindgren, O. Airaksinen, H. Manninen, Movement disturbances of the lumbar spine and abnormal back muscle electromyographic findings in recurrent low back pain, Spine 22 (3) (1997) [90] M. Solomonow, B. Zhou, R.V. Baratta, M. Zhu, Y. Lu, Neuromuscular disorders associated with static lumbar flexion: a feline model, Journal of Electromyography and Kinesiology 12 (2) (2002) [91] M. Solomonow, B.H. Zhou, M. Harris, Y. Lu, R.V. Baratta, The ligamento muscular stabilizing system of the spine, Spine 23 (23) (1998) [92] J.D. Spencer, K.C. Hayes, I.J. Alexander, Knee joint effusion and quadriceps reflex inhibition in man, Archives of Physical Medicine and Rehabilitation 65 (1984) [93] M. Stokes, A. Young, The contribution of reflex inhibition to arthrogenous muscle weakness, Clinical Science 67 (1984) [94] N. Suzuki, S. Endo, A quantitative study of trunk muscle strength and fatigability in the low-back-pain syndrome, Spine 8 (1) (1983) [95] P. Svensson, L. Arendt-Nielsen, L. Houe, Sensory-motor interactions of human experimental unilateral jaw muscle pain: a quantitative analysis, Pain 64 (1995) [96] P. Svensson, A. De Laat, T. Graven-Nielsen, L. Arendt-Nielsen, Experimental jaw-muscle pain does not change heteronymous h- reflexes in the human temporalis muscle, Experimental Brain Research 121 (3) (1998) [97] P. Svensson, T.S. Miles, T. Graven-Nielsen, L. Arendt-Nielsen, Modulation of stretch-evoked reflexes in single motor units in human masseter muscle by experimental pain, Experimental Brain Research 132 (1) (2000) [98] S. Taimela, M. Kankaanpaa, S. Luoto, The effect of lumbar fatigue on the ability to sense a change in lumbar position. A controlled study, Spine 24 (13) (1999) [99] S. Taimela, U.M. Kujala, Reaction times with reference to musculoskeletal complaints in adolescence, Perceptual and Motor Skills 75 (1992) [100] E. Takala, E. Viikari-Juntura, Do functional tests predict low back pain?, Spine 25 (16) (2000) [101] A. Thorstensson, Å. Arvidson, Trunk muscle strength and low back pain, Scandinavian Journal of Rehabilitation Medicine 14 (1982) [102] M. Valeriani, D. Restuccia, V. Di Lazzaro, A. Oliviero, P. Profice, D. Le Pera, E. Saturno, P. Tonali, Inhibition of the human primary motor area by painful heat stimulation of the skin, Clinical Neurophysiology 110 (8) (1999) [103] G.P. van Galen, M. van Huygevoort, Error, stress and the role of neuromotor noise in space oriented behaviour, Biological Psychology 51 (2 3) (2000) [104] S. Venna, H. Hurri, H. Alaranta, Correlation between neurological leg deficits and reaction time of upper limbs among lowback pain patients, Scandinavian Journal of Rehabilitation Medicine 26 (1994) [105] J.W. Vlaeyen, S.J. Linton, Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art, Pain 85 (3) (2000) [106] J.W. Vlaeyen, H.A. Seelen, M. Peters, P. de Jong, E. Aretz, E. Beisiegel, W.E. Weber, Fear of movement/(re)injury and muscular reactivity in chronic low back pain patients: an experimental investigation, Pain 82 (3) (1999) [107] P.J. Watson, C.J. Booker, Evidence for the role of psychological factors in abnormal paraspinal activity in patients with chronic low back pain, Journal of Musculoskeletal Pain 5 (1997) [108] R. Weinberg, V. Hunt, The interrelationships between anxiety, motor performance and electromyography, Journal of Motor Behaviour 8 (1976) [109] D.G. Wilder, A.R. Aleksiev, M.L. Magnusson, M.H. Pope, K.F. Spratt, V.K. Goel, Muscular response to sudden load. A tool to evaluate fatigue and rehabilitation, Spine 21 (22) (1996)

55 370 P.W. Hodges, G.L. Moseley / Journal of Electromyography and Kinesiology 13 (2003) [110] H.J. Wilke, S. Wolf, L.E. Claes, M. Arand, A. Wiesend, Stability increase of the lumbar spine with different muscle groups: a biomechanical in vitro study, Spine 20 (2) (1995) [111] S.L. Wolf, J.V. Basmajian, Assessment of paraspinal electromyographic activity in normal subjects and chronic low back pain patients using a muscle biofeedback device, in: E. Asmussen, K. Jorgensen (Eds.), Biomechanics iv b, University Park Press, Baltimore, 1977, pp [112] M. Zedka, M. Chan, A. Prochazka, Voluntary control of painful muscles in humans. Society for Neuroscience Abstracts, [113] M. Zedka, A. Prochazka, B. Knight, D. Gillard, M. Gauthier, Voluntary and reflex control of human back muscles during induced pain, Journal of Physiology (London) 520 (1999) [114] J.H. van Dieën, L. Selen, J. Cholewicki, Trunk muscle activation in low-back pain patients, an analysis of the literature. J Electromyogr Kinesiol 2003;13 doi: /S (03) [115] S.M. McGill, S. Grenier, N. Kavcic, J. Cholewicki, Coordination of muscle activity to assure stability of the lumbar spine. J Electromyogr Kinesiol 2003;13 doi: /S (03) [116] M. Solomonow, R.V. Baratta, B.-H. Zhou, E. Burger, A. Zieske, A. Gedalia, Muscular dysfunction elicited by creep of lumbar viscoelastic tissues. J Electromyogr Kinesiol 2003;13 doi: /S (03) Paul Hodges is an NHMRC Senior Research Fellow and Associate Professor at the University of Queensland where he heads the Human Neuroscience Unit. Paul s research involves integration of neuroscience and biomechanics to investigate the nervous system control of joint stability and movement. Key areas of research include investigation of: task conflict of the trunk muscles, in vivo and in vitro studies of the biomechanical effect of contraction of intrinsic spinal muscles on the spine, mechanisms for pain to affect motor control, strategies used by the central nervous system to control joint stability, and the mechanism of efficacy of therapeutic exercise for musculoskeletal pain. In addition to his research in Brisbane, Paul has ongoing collaborations with laboratories in Sydney, Stockholm (Sweden), and Portland (USA). He has published more than 50 research papers in international physiology and medical journals, has presented more than 15 keynote lectures at International back pain and motor control conferences in Australia, Europe and North America and has co-authored a clinical text. Paul s research has resulted in several awards from the Australian Society for Medical Research and the International Society for Biomechanics and led to him being awarded the Young Australian of the Year Award for Science and Technology for Lorimer Moseley completed his Ph.D. at the Pain Management and Research Centre, University of Sydney, where he investigated the psychophysiology of pain and spinal control. He has also conducted several clinical trials into the use of high level pain physiology education as a therapeutic strategy. Lorimer is an Australian NHMRC Clinical Research Fellow currently researching perceptual and motor mechanisms of pain at Royal Brisbane Hospital and The University of Queensland. His particular interests are the interface between physiotherapy, psychology and physiology; conceptual paradigms of pain; and therapeutic strategies for the integration of pain sciences and clinical practice, particularly for those with chronic and recurrent pain.

56 Downloaded from bjsm.bmj.com on 28 January 2009 Editorials Transversus abdominis: a different view of the elephant Paul Hodges It is good to see that clinical and research hypotheses are debated in the literature. The purpose of science is to challenge ideas and to consider alternative interpretations of observations. Within this, the place for neurophysiological/biomechanical studies in clinical research is not to predict the potential efficacy of a clinical approach, but to try to understand the mechanisms that underlie it. This is helpful as it provides a means to refine, improve, and direct intervention and provides a platform to develop rationales for intervention, particularly when we are faced with complex patients who do not fit the clinical prediction rule or the narrow criteria adopted for inclusion in clinical trials. If we understand the mechanisms we have a powerful tool to rationalise and test interventions. The developing debate about the role of transversus abdominis is healthy for rational consideration of motor control interventions for back pain. I welcome this opportunity to comment on the opinions and interpretations of Allison et al 1 and Cook. 2 As indicated by Allison et al in their paper published in JOSPT, 3 it is not the data that are questioned; it is the interpretation. It seems that we have a recurrence of the issue of the six blind men and the elephant, where we see the same animal, but from different perspectives, and draw different conclusions. There are a number of assumptions that require consideration to challenge the interpretation of Allison et al 1 and the opinion of Cook. 2 A key issue is that to conclude that a single observation from a single task refutes the conclusion of a whole range of different methodologies/tasks seems unfounded. CAN PHYSIOLOGICAL DATA INFLUENCE CLINICAL EFFECTIVENESS? In response to the editorial by Cook, 2 the first thing to consider is that the results of physiological/biomechanical studies cannot be used to challenge the outcomes of Correspondence to: Paul Hodges, NHMRC Centre of Clinical Research Excellence in Spinal Pain, Injury and Health, School of Health and Rehabilitation Sciences The University of Queensland Brisbane Qld 4072 Australia; p.hodges@uq.edu.au clinical trials and systematic reviews. The fact that the control of transversus abdominis may not be as simple as once thought does not challenge the positive clinical outcomes from interventions that include strategies to train this muscle. It suggests that we need to take a look at the potential mechanisms for efficacy of the approach. Although Cook argues that the efficacy of the approach is disappointing, it is worth taking stock of the current status of the literature. Most systematic reviews 4 5 suggest that motor control training that includes training of the deep trunk muscles has a large effect size when applied to specific populations, and a reduced effect size when applied to a generic non-specific low back pain group. This is not surprising. Our challenge is to identify those who benefit most from the interventions. This is the goal of several research groups around the world and applies equally to most therapeutic interventions. DIFFERENT VIEWS OF THE ELEPHANT Allison et al 1 make a number of assumptions that require further consideration. These authors appear to assume that: (1) a muscle can only do one thing at a time, and in unilateral arm movements transversus abdominis can only contribute to rotation; (2) if the muscle does not turn on at the same time on both sides it can do nothing (despite the fact that activity is present on both sides, albeit a little later on the ipsilateral side, at the time the arm starts to move); (3) observation of asymmetrical activity in an arm movement task refutes all other data of unique activation of transversus abdominis during tasks such as walking, trunk movements and trunk perturbations (this data was not mentioned by Allison et al 1 ); and (4) clinical approaches are restricted to teaching patients to activate transversus abdominis bilaterally. TRANSVERSUS ABDOMINIS CAN ACT TO ACHIEVE TWO GOALS AT ONCE In response to the first assumption that a muscle can only do one thing at a time, there is an abundant literature that shows that muscles can be activated to achieve two or more goals concurrently. It has been well documented that transversus abdominis is active asymmetrically during trunk rotation. 6 How transversus abdominis contributes to axial rotation is unclear for a number of reasons. First, transversus abdominis is active with both directions of rotation (but greater when rotating the thorax towards the side of the muscle), and second, the muscle has a trivial moment arm to generate rotation torque. 7 The contribution of the muscle to rotation may relate to control of the linea alba, while the contralateral obliquus externus (OE) and ipsilateral obliquus internus (OI) contribute most to the torque. 6 Thus, the finding by Allison et al that transversus abdominis is an axial rotator is not new. Furthermore, in an earlier study that combined modelling of reactive trunk moments and measurement of trunk muscle EMG it was argued that the timing of transversus abdominis may be linked to the control of axial rotation. 8 The problem is that Allison et al 1 have assumed that if transversus abdominis is active to control rotation then the muscle cannot do anything else. We have documented in a number of experiments that this is not the case and the nervous system can coordinate the activity of muscle to achieve multiple goals concurrently. In some simple examples we have shown that activation of transversus abdominis involves multiple components. For instance, transversus abdominis activity is tonic during gait (perhaps to contribute to some aspects of spinal control), but this activity is phasically modulated in association with breathing (to assist expiration), and has peaks of activity associated with heel strike events (which are coordinated with periods of peak reactive force from the foot contact and the time of change in direction of rotation of the trunk). 9 Similarly, activity of the diaphragm 10 and pelvic floor muscles, 11 during repetitive arm movements, includes tonic activation as well as phasic activation with movement and breathing. Recent data show that the diaphragm, like transversus abdominis, is involved in axial rotation of the trunk. 12 The fact that transversus abdominis is active earlier on the contralateral side (but bilaterally by the time the movement starts) may simply reflect the mechanical demand to both influence axial rotation (including the control of rotation, which is an aspect of stability) and contract bilaterally to contribute to Br J Sports Med December 2008 Vol 42 No

57 Downloaded from bjsm.bmj.com on 28 January 2009 Editorials control of multiple mechanical demands on the trunk. ASYMMETRICAL ACTIVITY OF TRANSVERSUS ABDOMINIS IS STILL MECHANICALLY USEFUL The second assumption is that if the contraction is not symmetrical it cannot do anything for the spine. But, as mentioned above, Allison et al s 1 data show that, while the onset was not simultaneous, the muscle was active on both sides in a feedforward manner (i.e. active before any feedback could be available to induce activation of the muscle (,50 ms after the onset of deltoid EMG)) 13 and both sides were active at the time of onset of deltoid EMG (well before the movement started and well before any reactive forces would affect the spine). Although Allison et al 1 argue that the lack of activity of transversus abdominis on the other side or activity of the rectus abdominis (RA) would limit the potential for transversus abdominis to generate force, this is not what their data show, as transversus abdominis on the other side is active by the time the movement starts and may well contribute to spinal control in spite of a later onset. Furthermore, direct measurement of intra-abdominal pressure (IAP) clearly demonstrates that there is a mechanical output of the muscle contraction before the arm moves In-vivo and modelling 17 studies show that IAP contributes to spinal control. Other studies show that transversus abdominis is the abdominal muscle most closely correlated with IAP changes. 18 Modelling studies that suggest that transversus abdominis does very little for spinal stability 19 consider only the role of transversus abdominis as a flexor (for which it has a trivial moment arm). This ignores the biomechanical data that show that transversus abdominis can contribute to spinal control via IAP or fascial tension Both human and animal studies show that activation of transversus abdominis has a mechanical effect on the spine and the pattern of activation of the muscle in Allison s data is not inconsistent with that assumption. What that data shows is that, in addition to the bilateral activation at the time the movement starts, it also has activity that could be consistent with an additional role in control of axial rotation. An issue that is accurately indicated by Allison et al 1 is that there are no data to show that the spine is less optimally controlled when activation of transversus abdominis is changed. This is challenging to test in humans (because it is difficult to take the muscle away) and cannot be tested in current biomechanical models (as few include the contributions of IAP and fascial tension). Animal studies are underway to test the effect of reduced deep muscle activation on spine biomechanics and it is hoped that this will shed light on this issue. IT S NOT JUST ABOUT ARM MOVEMENTS; DATA FROM OTHER METHODS SUPPORT THE ROLE OF TRANSVERSUS ABDOMINIS IN SPINAL CONTROL The third issue relates to Allison et al s 1 and Cook s 2 failure to consider the wealth of data from numerous groups using other experimental designs. Arm movement tasks provide a window of opportunity to study the system, but there are many other models which have provided insight into the function of transversus abdominis. In trunk movements, 22 isometric trunk tasks, 22 trunk perturbations in sitting 23 and lying, 24 transversus abdominis is active in a manner that is unique amongst the trunk muscles, that is, it is active with forces and movements in opposite directions in the sagittal plane. These tasks do not include an axial rotation component and that may make the interpretation easier (as it does not involve the functional requirement to combine activity for rotation with other aspects). These observations and the interpretation of unilateral flexion and extension movements raise the question: why does the nervous system use a muscle in a similar manner with two opposite directions of movement? I concur with Allison et al 1 that this does not necessarily mean that it contributes to spinal control. That is simply a hypothesis that we have gone on to test in a number of biomechanical studies, with results that show the activation can control spinal motion. To conclude that a single observation from a single task refutes the conclusion of a whole range of different methodologies seems unfounded. IS THE DEBATE ABOUT TRANSVERSUS ABDOMINIS MISSING A CRITICAL ISSUE? Having highlighted some of the assumptions made by Allison et al 1 and Cook 2 it is also worth considering some issues in the clinical literature as a whole that have spawned some of this current debate. In my view the whole debate around transversus abdominis is missing a critical issue. Back pain is not an issue of a single muscle, it is associated with complex changes across a whole system. Although the early studies focussed on this muscle, an abundant literature has evolved that shows that the changes in back pain are complex and involve many muscles and many control properties. One of the factors that have perpetuated the confusion is that, although the changes in transversus abdominis appear to be relatively consistent, the changes in the other muscles are variable, and therefore harder to find in a non-specific pain population. Recent work even suggests that many people with back pain may have increased stability rather than decreased stability, 29 potentially as a result of increased activity of the more superficial trunk muscles, and this puts a whole new perspective on the meaning of optimal spinal control; not simply to increase stability, but to find a balance between too much and too little. 30 It is increasingly clear that rehabilitation should not target a single muscle, but instead should involve careful evaluation of a whole system. While changes in transversus abdominis (and other muscles such as multifidus) can be a useful marker of dysfunction in the system (and recent data show that patients with delayed transversus abdominis do better with a motor control training approach than people without a delay) (unpublished data), to limit treatment to this muscle is unlikely to be beneficial. The days of contracting transversus abdominis as the primary exercise and then sending the patient away are over. Instead, training of transversus abdominis should be part of the intervention, when appropriate for the patient and the changes in their control system. Cook 2 argues that training transversus abdominis bilaterally may be redundant as that may not be the way the muscle functions. Although this may not always be the case, it is likely to be so in some tasks. But evidence that the muscle is not symmetrical (although bilateral) cannot be used to say that bilateral training is not effective or appropriate. There is a developing literature that shows that training muscles in this way changes the control of the muscle in other tasks. Not only does it change the timing of activation of transversus abdominis in arm movement and gait tasks ; it also changes the organisation of the motor cortex 33 and this change is related to the change in timing during an arm movement task. Notably, this was only achieved by cognitive bilateral activation, and not by simple activation as part of a sit-up 31 or other abdominal bracing manoeuvre. 34 Furthermore, focussed attention on activation of the deeper muscles 942 Br J Sports Med December 2008 Vol 42 No 12

58 Downloaded from bjsm.bmj.com on 28 January 2009 Editorials can also change the activation of many muscles of the trunk. 35 These data suggest that training of bilateral activation is an effective training stimulus to change the way the muscle is activated in function, despite the fact that this may not be the only way it is active in function. This principle of a training stimulus that does not reflect every function is true for many exercise approaches. For instance, eccentric loading is effective in management of tendinopathy, but this is not the only way those muscles function. WHY ARE MOTOR CONTROL INTERVENTIONS SUCCESSFUL IN TREATING BACK PAIN? Finally, I agree with Cook 2 and Allison et al 1 when they argue that we do not know why motor control interventions are effective. We don t know that the effect is explained by increased stability of the spine due to activation of transversus abdominis and other deep muscles. Core work by our group has focussed on this very issue over recent years. In a series of studies we aimed to evaluate the potential mechanisms for efficacy of a motor control approach to the management of neck pain. These studies showed that the motor intervention not only changed the control of the deep neck flexor muscles (Jull et al, unpublished data), but was also associated with improvements in posture 36 and neck proprioception. 37 In terms of the back, recent data in a small clinical trial suggest that the intervention can reduce the muscular stabilisation of the trunk by reducing activity of more superficial muscles. 35 This could suggest that the approach leads to more optimal control. This is being followed up in a large randomised controlled clinical trial. There are many candidate mechanisms and we need to keep an open mind to make sure that we do not miss the wood for the trees. While we have investigated the potential role of optimisation of control of the spine by changing the activation of the system of trunk muscles, including transversus abdominis, the truth is likely to be more complex. That is the joy of science, to hypothesise and then challenge new ideas. The ultimate goal is to understand so that we can identify better treatments. THE CHALLENGE FOR THE FUTURE In summary, the data provided by Allison et al 1 add richness to our understanding of the control of the deep muscles and ultimately the control of the trunk. As highlighted above, the data do not refute the original hypotheses of the role of transversus abdominis in trunk control; in fact, they are very congruent with the evolution of our understanding of the function of the deep muscles. The basic observations from the early studies that were conducted 15 years ago provided a starting point. 25 The subsequent data have shaped and evolved the interpretation. We all agree that clinical practice often adopts research findings in a simplified and hardline approach. Allison et al s 1 data do not refute the viability and potential efficacy of the approach. Current literature suggests that the clinical application of the findings is beneficial. In an ideal world the experimental testing of an idea would be completed and all issues resolved and understood before implementation into practice, but this is not practical as nothing would ever be implemented. And, after all, research that has a clinical application must be done in an iterative manner with communication back and forth between clinicians and researchers. In that way clinical practice can inform research and research can be accurately implemented into practice. The challenge for us all is to keep our blinkers off and keep an open mind when looking at our data and looking at patients so that we have a chance to move forward. Competing interests: None declared. Accepted 11 November 2008 Br J Sports Med 2008;42: doi: /bjsm REFERENCES 1. Allison GT. Transversus abdominis and core stability: has the pendulum swung? Br J Sports Med 2008;42: Cook J. Jumping on bandwagons: taking the right clinical message from research. Br J Sports Med 2008;42: Allison GT, Morris SL, Lay B. Feedforward responses of transversus abdominis are directionally specific and act asymmetrically: implications for core stability theories. J Orthop Sports Phys Ther 2008;38: Ferreira PH, Ferreira ML, Maher CG, et al. Specific stabilisation exercise for spinal and pelvic pain: a systematic review. Aust J Physiother 2006;52: Hauggaard A, Persson AL. Specific spinal stabilisation exercises in patients with low back pain - a systematic review. Phys Ther Rev 2007;12: Urquhart DM, Hodges PW. Differential activity of regions of transversus abdominis during trunk rotation. Eur Spine J 2005;14: Urquhart DM, Barker PJ, Hodges PW, et al. Regional morphology of the transversus abdominis, obliquus internus and obliquus externus abdominis muscles. Clin Biomech 2005;20: Hodges PW, Cresswell AG, Daggfeldt K, et al. Three dimensional preparatory trunk motion precedes asymmetrical upper limb movement. Gait Posture 2000;11: Saunders SW, Rath D, Hodges PW. Postural and respiratory activation of the trunk muscles changes with mode and speed of locomotion. Gait Posture 2004;20: Hodges P, Gandevia S. Activation of the human diaphragm during a repetitive postural task. J Physiol 2000;522: Hodges PW, Pengel HM, Sapsford R. Postural and respiratory functions of the pelvic floor muscles. Neurourol Urodyn In press. 12. Hodges P. Differential activity of the right and left costal diaphragm during non-respiratory tasks in humans. In: Proceedings Australian Neuroscience Society; Sydney. 13. Aruin AS, Latash ML. Directional specificity of postural muscles in feed-forward postural reactions during fast voluntary arm movements. Exp Brain Res 1995;103: Hodges PW, Butler JE, McKenzie D, et al. Contraction of the human diaphragm during postural adjustments. J Physiol 1997;505: Hodges P, Kaigle Holm A, Holm S, et al. Intervertebral stiffness of the spine is increased by evoked contraction of transversus abdominis and the diaphragm: in vivo porcine studies. Spine 2003;28: Hodges P, Shirley D, Eriksson AEM, et al. Intraabdominal pressure can directly increase stiffness of the lumbar spine. J Biomech 2005;38: Cholewicki J, Juluru K, McGill SM. Intra-abdominal pressure mechanism for stabilizing the lumbar spine. J Biomech 1999;32: Cresswell AG, Thorstensson A. Changes in intraabdominal pressure, trunk muscle activation and force during isokinetic lifting and lowering. Eur J Appl Physiol Occup Physiol 1994;68: Kavcic N, Grenier S, McGill SM. Determining the stabilizing role of individual torso muscles during rehabilitation exercises. Spine 2004;29: Tesh KM, ShawDunn J, Evans JH. The abdominal muscles and vertebral stability. Spine 1987;12: Barker P, Guggenheimer K, Grkovic I, et al. Effects of tensioning the lumbar fasciae on segmental stiffness during flexion and extension. Spine 2005;31: Cresswell AG, Grundstrom H, Thorstensson A. Observations on intra-abdominal pressure and patterns of abdominal intra-muscular activity in man. Acta Physiol Scand 1992;144: McCook D, Vicenzino B, Hodges P. Activity of deep abdominal muscles increases during submaximal flexion and extension efforts but antagonist cocontraction remains unchanged. J Electromyogr Kinesiol. Published Online First 19 December doi: /j.jelekin Eriksson Crommert AE, Thorstensson A. Trunk muscle coordination in reaction to load-release in a position without vertical postural demand. Exp Brain Res 2008;185: Hodges PW, Richardson CA. Inefficient muscular stabilisation of the lumbar spine associated with low back pain: A motor control evaluation of transversus abdominis. Spine 1996;21: Hodges PW, Richardson CA. Delayed postural contraction of of transversus abdominis associated with movement of the lower limb in people with low back pain. J Spinal Disord 1998;11: van Dieen JH, Selen LP, Cholewicki J. Trunk muscle activation in low-back pain patients, an analysis of the literature. J Electromyogr Kinesiol 2003;13: Hodges P, Cholewicki J, Coppieters M, et al. Trunk muscle activity is increased during experimental back pain, but the pattern varies between individuals. In: Proceedings International Society for Electrophysiology and Kinesiology, 2006, Turin, Italy. 29. Hodges P, Van den Hoorn W, Dawson A, et al. Changes in the mechanical properties of the trunk in low back pain may be associated with recurrence. J Biomech. In press. 30. Hodges P, Cholewicki J. Functional control of the spine. In: Vleeming A, Mooney V, Stoeckart R, eds. Movement, Stability and Lumbopelvic Pain. Edinburgh: Elsevier, In press. Br J Sports Med December 2008 Vol 42 No

59 Downloaded from bjsm.bmj.com on 28 January 2009 Editorials 31. Tsao H, Hodges PW. Immediate changes in feedforward postural adjustments following voluntary motor training. Exp Brain Res. In press. 32. Tsao H, Hodges PW. Persistence of changes in postural control following training of isolated voluntary contractions in people with recurrent low back pain. J Electromyogr Kinesiol. In press. 33. Tsao H, Galea M, Hodges P. Skilled motor training induces reorganisation of the motor cortex and is associated with improved postural control in chronic low back pain. In: Proceedings World Congress on Pain, 2008, Glasgow, UK. 34. Hall L, Tsao H, MacDonald D, et al. Immediate effects of co-contraction training on motor control of the trunk muscles in people with recurrent low back pain. J Electromyogr Kinesiol. In press. doi: / j.jelekin Hodges P, Druit T, Tsao H. Practice of skilled trunk muscle activation changes coordination in an untrained task. In: Proceedings International Society for Electrophysiology and Kinesiology, Niagara Falls. 36. Falla D, Jull G, Russell T, et al. Effect of neck exercise on sitting posture in patients with chronic neck pain. Phys Ther 2007;87: Jull G, Falla D, Treleaven J, et al. Retraining cervical joint position sense: the effect of two exercise regimes. J Orthop Res 2007;25: Why glucocorticoids should be removed from the World Antidoping Agency s list of banned products John W Orchard Sports medicine clinicians and researchers should all be familiar with the concepts of false positives and false negatives. Research to test a hypothesis about a link between, say, a risk factor and a disease can potentially be wrong in either of two ways. The findings might falsely show a link when in reality one does not exist (a type I or a error), or they might fail to show a link when there really is one (a type II or b error). 12 Hopefully most of the time, if studies are well conducted, the likelihood of both of these errors is reduced. Similar errors in both directions can potentially occur in drug testing in sport, although the nature of false positives and false negatives is somewhat different from that in other clinical testing. The rigorous methods of collection and the use of A and B samples mean that many sources of potential laboratory error are minimised. The false positive and false negative phenomena in doping may be better referenced to the athlete s intent to cheat using performance-enhancing drugs. An athlete who takes a so-called undetectable anabolic steroid is a true drug cheat, but one who might produce a false-negative drug test because the structure of the undetectable drug is not yet known by the testing authorities. By comparison, the athlete who inadvertently takes a banned drug (particularly one with minimal performance-enhancing potential) not to cheat, but to treat a legitimate medical condition, may test Correspondence to: John W Orchard, Sports Medicine at Sydney University, Cnr. Western Ave. & Physics Rd, University of Sydney NSW 2006, Australia; jorchard@med.usyd.edu.au positive in a doping test. Such a result may be considered a false positive with respect to intent to cheat using a performance-enhancing drug, even though the testing process was accurate in finding the drug in the athlete s system. Intent to cheat is so difficult to prove or disprove that WADA takes a pragmatic approach and enforces strict liability for all positive tests. 3 Strict liability means that denying intention to cheat is not relevant if the results of a drug test are positive. If the excuse of lack of intention was generally accepted, most true cheats would deny intent and many would escape prosecution as a result. History suggests, though, that there are some drug suspensions which were probably false positives with respect to intent to cheat. Perhaps the first such case was Rick DeMont of the USA, who lost a swimming gold medal after apparently being prescribed ephedrine by his team doctor to treat asthma at the Munich Olympics in A similar case saw Andrea Raducan of Romania stripped of her Olympic gold medal in gymnastics in 2000 after a positive test for pseudoephedrine, apparently taken as a medication to treat a cold. Babette Pluim has recently highlighted cases of suspected false positives from the sport of tennis. 4 Alarmingly, it was calculated that as many as 68% of the doping charges in tennis over the previous 5 years were false positives. 4 Glucocorticoids were one of the major classes responsible for the suspected false positive cases. Because glucocorticoids are extremely commonly used in general medicine and have not been shown to enhance performance in humans, 4 it can be strongly argued, using Bayes theorem, that this drug class is particularly likely to produce false positive results. Bayes theorem, developed centuries ago by an English clergyman, is a formula to calculate positive predictive value (the likelihood that a positive test actually represents a true positive). 1 The numerator is the number of true positive cases with the denominator being the sum of both true positive and false positive cases. One of the principles of Bayes theorem is that the likelihood of a positive result being a true positive is proportional to the prevalence of the condition being tested for in the sample population. The principles of Bayes theorem are used for screening in other areas of medicine, like cancer detection. 5 Mammograms, for example, are recommended for postmenopausal women, but not for younger women. This is because breast lesions detected by mammogram in older women are somewhat likely to be malignant (because the prevalence of breast cancer is relatively high). By comparison, breast lesions in younger women are extremely likely to be benign (because the prevalence of breast cancer is very low). It is generally calculated by screening experts that a mammogram performed in a young woman is much more likely to cause harm (by falsely identifying a suspicious lesion which is, in fact, benign) than it is to lead to benefit (by identifying a suspicious lesion which is a true malignancy). 5 As women get older and breast cancer becomes more likely, then the value of screening tests, such as mammograms, increases. In sports medicine, Bayes theorem has been used to argue against routine ECG screening of asymptomatic young athletes. 6 Bayes theorem as it applies to drug testing can be demonstrated by considering a theoretical population of elite athletes and two theoretical drugs which we can call G and A. Drug G is a glucocorticoid and is commonly used to treat medical conditions such as asthma and sinus congestion. Therapeutic use exemption (TUE) is available for athletes 944 Br J Sports Med December 2008 Vol 42 No 12

60 Series Low back pain 1 What low back pain is and why we need to pay attention Jan Hartvigsen*, Mark J Hancock*, Alice Kongsted, Quinette Louw, Manuela L Ferreira, Stéphane Genevay, Damian Hoy, Jaro Karppinen, Glenn Pransky, Joachim Sieper, Rob J Smeets, Martin Underwood, on behalf of the Lancet Low Back Pain Series Working Group Low back pain is a very common symptom. It occurs in high-income, middle-income, and low-income countries and all age groups from children to the elderly population. Globally, years lived with disability caused by low back pain increased by 54% between 1990 and 2015, mainly because of population increase and ageing, with the biggest increase seen in low-income and middle-income countries. Low back pain is now the leading cause of disability worldwide. For nearly all people with low back pain, it is not possible to identify a specific nociceptive cause. Only a small proportion of people have a well understood pathological cause eg, a vertebral fracture, malignancy, or infection. People with physically demanding jobs, physical and mental comorbidities, smokers, and obese individuals are at greatest risk of reporting low back pain. Disabling low back pain is over-represented among people with low socioeconomic status. Most people with new episodes of low back pain recover quickly; however, recurrence is common and in a small proportion of people, low back pain becomes persistent and disabling. Initial high pain intensity, psychological distress, and accompanying pain at multiple body sites increases the risk of persistent disabling low back pain. Increasing evidence shows that central pain-modulating mechanisms and pain cognitions have important roles in the development of persistent disabling low back pain. Cost, health-care use, and disability from low back pain vary substantially between countries and are influenced by local culture and social systems, as well as by beliefs about cause and effect. Disability and costs attributed to low back pain are projected to increase in coming decades, in particular in low-income and middle-income countries, where health and other systems are often fragile and not equipped to cope with this growing burden. Intensified research efforts and global initiatives are clearly needed to address the burden of low back pain as a public health problem. Introduction Low back pain is an extremely common symptom experienced by people of all ages. 1 3 In 2015, the global point prevalence of activity-limiting low back pain was 7 3%, implying that 540 million people were affected at any one time. Low back pain is now the number one cause of disability globally. 4 The largest increases in disability caused by low back pain in the past few decades have occurred in low-income and middle-income countries, including in Asia, Africa, and the Middle East, 5 where health and social systems are poorly equipped to deal with this growing burden in addition to other priorities such as infectious diseases. Rarely can a specific cause of low back pain be identified; thus, most low back pain is termed non-specific. Low back pain is characterised by a range of biophysical, psychological, and social dimensions that impair function, societal participation, and personal financial prosperity. The financial impact of low back pain is cross-sectoral because it increases costs in both health-care and social supports systems. 6 Disability attributed to low back pain varies substantially among countries, and is influenced by social norms, local health-care approaches, and legislation. 7 In low-income and middle-income countries, formal and informal social-support systems are negatively affected. While in high-income countries, the concern is that the prevalent health-care approaches for low back pain contribute to the overall burden and cost rather than reducing it. 8 Spreading high-cost health-care models to Key messages Published Online March 21, S (18)30480-X See Online/Comment S (18) See Online/Viewpoint S (18) This is the first in a Series of two papers about low back pain *Joint first authors Members listed at the end of the report Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark (Prof J Hartvigsen PhD, A Kongsted PhD); Nordic Institute of Chiropractic and Clinical Biomechanics, Odense, Low back pain is an extremely common symptom in populations worldwide and occurs in all age groups, from children to the elderly population Low back pain was responsible for 60 1 million disability-adjusted life-years in 2015, an increase of 54% since 1990, with the biggest increase seen in low-income and middle-income countries Disability from low back pain is highest in working age groups worldwide, which is especially concerning in low-income and middle-income countries where informal employment is common and possibilities for job modification are limited Most episodes of low back pain are short-lasting with little or no consequence, but recurrent episodes are common and low back pain is increasingly understood as a long-lasting condition with a variable course rather than episodes of unrelated occurrences Low back pain is a complex condition with multiple contributors to both the pain and associated disability, including psychological factors, social factors, biophysical factors, comorbidities, and pain-processing mechanisms For the vast majority of people with low back pain, it is currently not possible to accurately identify the specific nociceptive source Lifestyle factors, such as smoking, obesity, and low levels of physical activity, that relate to poorer general health, are also associated with occurrence of low back pain episodes Costs associated with health care and work disability attributed to low back pain vary considerably between countries, and are influenced by social norms, health-care approaches, and legislation The global burden of low back pain is projected to increase even further in coming decades, particularly in low-income and middle-income countries Published online March 21,

61 Series Denmark (Prof J Hartvigsen, A Kongsted); Department of Health Professions, Faculty of Medicine and Health Sciences, Macquarie University, Sydney, Australia (M J Hancock PhD); Faculty of Medicine and Health Sciences, Physiotherapy Division and Department of Health and Rehabilitation Sciences, Stellenbosch University, Tygerberg, South Africa (Prof Q Louw PhD); Institute of Bone and Joint Research, Sydney Medical School, The University of Sydney, Sydney, Australia (M L Ferreira PhD); Division of Rheumatology, University Hospitals of Geneva, Geneva, Switzerland (S Genevay MD); University of Sydney, Sydney, Australia (D Hoy PhD); Medical Research Centre Oulu, University of Oulu and University Hospital, Oulu, Finland (Prof J Karppinen PhD); Department of Family Medicine and Community Health, University of Massachusetts Medical School, Worcester, MA USA (G Pransky MD); Department of low-income and middle-income countries will compound rather than alleviate the burden. Low back pain is therefore an urgent global public health concern. Genetic factors Biophysical factors Pain experience Nociceptive input* Central pain processing Disability Psychological factors Comorbidities Social factors Figure 1: Contributors to low back pain and disability The model includes key contributors to low back pain and disability but does not attempt to represent the complex interactions between different contributors. *Nociceptive input includes non-identifiable sources in non-specific low back pain, neurological sources (eg, radicular pain) and specific pathology (eg, fractures). Against this backdrop, we present a series of two papers and a Viewpoint. The aim of this paper is to present a current understanding of what low back pain is, its burden and global impact, as well as an overview of causes and the course of low back pain. The evidence for the effectiveness of current treatments and promising new directions for managing low back pain is presented in paper two, 9 and the Viewpoint is a worldwide call to action. 10 The approach for this Series involved the constitution of a team of leading international experts on back pain from different professional backgrounds and from countries around the globe who convened for a workshop in Buxton, UK, in June, 2016, to outline the structure of each paper. For this paper, we identified scientific studies using broad search terms in MEDLINE (PubMed) and Scopus. To identify potentially relevant papers from lowincome and middle-income countries, we also searched Google Scholar and the African Index Medicus Database. To minimise selection bias and to ensure high-quality evidence was selected, systematic reviews were preferred and sought when possible. However, we also used information from large population-based cohorts, international clinical guidelines, and the Global Burden of Disease (GBD) 2015 study. Primary research from low-income and middle-income regions excluded from systematic reviews was also referenced where appropriate. Panel 1: Potential nociceptive contributors to low back pain that have undergone investigation Intervertebral disc Although some imaging and clinical findings increase the likelihood that pain is arising from the intervertebral disc (with the reference standard of discography), no investigation has accurately identified a disc problem as contributing to an individual s pain; 14 there is no widely accepted reference standard for discogenic pain Facet joint Injecting facet joints with local anaesthetic can cause temporary relief of pain; 15 however, the Framingham Heart Study (3529 participants) did not find an association between radiological osteoarthritis of facet joints and presence of low back pain; 16 clinical identification of individuals whose facet joints are contributing to their pain is not possible. 17 Vertebral endplates (Modic changes) Modic changes are vertebral endplate abnormalities seen on MRI with specific subchondral and vertebral bone marrow features that can be classified according to different signal intensities into type 1, type 2, and type 3; endplate defects and disc herniation might predispose to the development of Modic changes; one theory is that the pro-inflammatory response, caused by structural damage to the disc or endplate, could allow microbial infiltration, autoimmune reactions, or both, that intensify and extend nociceptor stimulation by chemical or mechanical stimuli; 18 a low-grade infection by Propionibacterium acnes might promote the development of Modic changes; 19 the relevance of these findings to clinical practice is, however, unclear; a systematic review concluded that Modic type 1 changes are associated with low back pain; 20 a subsequent study, including 1142 people, found that Modic type 2 changes were associated with disability (odds ratio 1 56, 95% CI ), but not pain (1 36, ); 21 identification of individuals in whom Modic changes are contributing to their pain is not possible. What is low back pain? Low back pain is a symptom not a disease, and can result from several different known or unknown abnormalities or diseases. It is defined by the location of pain, typically between the lower rib margins and the buttock creases. 11 It is commonly accompanied by pain in one or both legs and some people with low back pain have associated neurological symptoms in the lower limbs. For nearly all people presenting with low back pain, the specific nociceptive source cannot be identified and those affected are then classified as having so-called non-specific low back pain. 12 There are some serious causes of persistent low back pain (malignancy, vertebral fracture, infection, or inflammatory disorders such as axial spondyloarthritis) that require identification and specific management targeting the cause, but these account for a very small proportion of cases. People with low back pain often have concurrent pain in other body sites, and more general physical and mental health problems, when compared with people not reporting low back pain. 13 The combined effect on individuals of low back pain and comorbidity is often more than the effect of the low back pain or the comorbidity alone and results in more care, yet typically a poorer response to a range of treatments. 13 Thus, many people living with low back pain have diverse problems in which psychological, social, and biophysical factors as well as comorbidities and pain-processing mechanisms impact 2 Published online March 21,

62 Series on both the pain experience and the associated disability (figure 1). Causes of low back pain Although clinical tests are unable to accurately identify the tissue source of most low back pain, several structures are innervated and have been shown to produce pain when stimulated. In some cases local anaesthetic relieves the pain (panel 1). 14,15 Many imaging (radiography, CT scan, and MRI) findings identified in people with low back pain are also common in people without such pain, and their importance in diagnosis is a source of much debate. 22 Nevertheless, at least in people younger than 50 years, some MRI abnormalities are more common in those with low back pain than in those without. A systematic review (14 case-control studies; 3097 participants) found several MRI findings had a reasonably strong association with low back pain, including Modic type 1 change (odds ratio [OR] 4 0, 95% CI ), disc bulge (7 5, ), disc extrusion (4 4, ), and spondylolysis (5 1, ; table 1). 20 However, evidence is insufficient to know whether MRI findings can be of use to predict the future onset, or the course, of low back pain. 23 Importantly, no evidence exists that imaging improves patient outcomes 24 and guidelines consistently recommend against the routine use of imaging for people with low back pain Neurological symptoms associated with low back pain Radicular pain and radiculopathy Radicular pain occurs when there is nerve-root involvement; commonly termed sciatica. The term sciatica is used inconsistently by clinicians and patients for different types of leg or back pain and should be avoided. 29 The diagnosis of radicular pain relies on clinical findings, including a history of dermatomal leg pain, leg pain worse than back pain, worsening of leg pain during coughing, sneezing or straining, 30 and straight leg raise test. Radiculopathy is characterised by the presence of weakness, loss of sensation, or loss of reflexes associated with a particular nerve root, or a combination of these, and can coexist with radicular pain. People with low back pain and radicular pain or radiculopathy are reported to be more severely affected and have poorer outcomes compared with those with low back pain only. 31 Disc herniation in conjunction with local inflammation is the most common cause of radicular pain and radiculopathy. Disc herniations are, however, a frequent finding on imaging in the asymptomatic population, 22 and they often resolve or disappear over time independent of resolution of pain. 32 Lumbar spinal stenosis Lumbar spinal stenosis is clinically characterised by pain or other discomfort with walking or extended standing that radiates into one or both lower limbs and is typically relieved by rest or lumbar flexion (neurogenic Number of studies OR (95% CI) Prevalence asymptomatic (95% CI) claudication). 33 It is usually caused by narrowing of the spinal canal or foramina due to a combination of degenerative changes such as facet osteoarthritis, ligamentum flavum hypertrophy, and bulging discs. Expert consensus is that the diagnosis of the clinical syndrome of lumbar spinal stenosis requires both the presence of characteristic symptoms and signs as well as imaging confirmation of narrowing of the lumbar spinal canal or foramina. 34 Symptoms of lumbar spinal stenosis are thought to result from venous congestion or ischaemia of the nerve roots in the cauda equina due to compression. 33 Specific pathological causes of low back pain Potential causes of low back pain that might require specific treatment include vertebral fractures, inflammatory disorders (eg, axial spondyloarthritis), malignancy, infections, and intra-abdominal causes (panel 2). A study of 1172 new presentations of acute (<2 weeks) episodes of low back pain in primary care in Australia found specific causes of back pain in 0 9% of participants, with fracture being by far the most common (eight of 11 cases), followed by inflammatory disorders (two of 11 cases). 37 A review from Uganda of 204 patients referred to a hospital orthopaedic clinic with a primary complaint of low back pain, showed that 4% of patients had serious spinal abnormalities due to tuberculosis, 3 5% had vertebral compression fractures, 1% brucellosis, and 1% had malignancy. 52 These differences in the patterns of specific pathological causes could reflect the ongoing burden of infectious diseases and their manifestations as low back pain in low-income countries. So-called red flags are case Prevalence symptomatic (95% CI) p value Heterogeneity Intervertebral disc degeneration-related outcomes Disc degeneration ( ) 34% (32 38) 57% (55 60) 0 01 High Modic change ( ) 12% (10 15) 23% (22 27) 0 43 High Modic type 1 change ( ) 3% (0 7 9) 7% (5 9) 0 04 Low Internal disc rupture-related outcomes Annular fissure ( ) 11% (9 14) 20% (18 23) 0 06 High High Intensity Zone ( ) 10% (7 13) 10% (8 13) 0 17 High Disc displacement-related outcomes Disc bulge ( ) 6% (4 9) 43% (38 48) 0 03 High Disc protrusion ( ) 19% (17 22) 42% (39 45) 0 00 High Disc extrusion ( ) 2% (0 1 4) 7% (5 9) <0 01 Low Other outcomes Spondylolysis ( ) 2% (0 5) 9% (7 12) <0 01 Low Spondylolisthesis ( ) 3% (2 6) 6% (4 9) 0 20 Low Central spinal canal stenosis ( ) 14% (10 19) 60% (55 64) 0 17 High Data are modified from Brinjikji et al (2015). 20 Heterogeneity (I²) was graded "low" only for "0" values since no CI for I² was presented. Prevalence data presented for reference only. OR=odds ratio. Table 1: Strength of association between MRI findings and low back pain in younger adults Rheumatology, Charité, Campus Benjamin Franklin, Berlin, Germany (Prof J Sieper MD); Department of Rehabilitation Medicine, Maastricht University, Maastricht, Netherlands (Prof R J Smeets PhD); Libra Rehabilitation and Audiology, Eindhoven, Netherlands (Prof R J Smeets); and Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick, Coventry, UK (Prof M Underwood MD) Correspondence to: Prof Martin Underwood, Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick, Coventry, CV4 7AL, UK m.underwood@warwick.ac.uk Published online March 21,

63 Series Panel 2: Specific pathological causes of low back pain Vertebral fracture Symptomatic minimal trauma vertebral fractures due to osteoporosis are rare under the age of 50 years but the incidence increases rapidly with age. 35 Although age-specific incidence is not changing, with an ageing population, the population burden is increasing. A systematic review (14 studies) found post-test probability for having a symptomatic vertebral fracture was 9% (95% CI 3 25) for those who were older (men aged >65 years, women aged >75 years), 33% (10 67) for those with a history of long-term corticosteroid use, and 62% (49 74) when a contusion or abrasion was present. The probability of a minimal trauma vertebral fracture being present when multiple risk factors (at least three of female, age >70, severe trauma, and long-term use of glucocorticoids) were present was 90% (34 99). 36 The predictive value of such a decision rule is, however, not greatly different from clinical assessment. 37 Symptomatic minimal trauma vertebral fractures have been shown in some studies to have a major health impact with a mean of 158 days of restricted activity and a third of those affected still have significant back pain after 2 years. 35 In some studies, minimal trauma vertebral fractures are also associated with a two-to-eight times increased risk of mortality. 35 Axial spondyloarthritis Axial spondyloarthritis is a chronic inflammatory disease that mainly affects the axial skeleton in young people (peak of onset years). Although traditionally thought to be a disease of young men, there is only a slight male predominance in population studies. 38 The term axial spondyloarthritis covers both people who have already developed structural damage in the sacroiliac joints or spine visible, or both, on radiographs (radiographic axial spondyloarthritis; also termed ankylosing spondylitis) and those who have not yet developed such structural damage (non-radiographic spondyloarthritis). 39 Non-radiographic spondyloarthritis is a prodrome of axial spondyloarthritis that might subsequently produce structural bony damage in the axial skeleton. 40 The prevalence of radiological disease is between 0 3 and 0 8% in western countries and is dependent on the HLA-B27 prevalence in a given population. 38 The typical presentation of axial spondyloarthritis includes morning stiffness, mostly in the lower back, with improvement seen with exercise but not with rest. In a Danish cohort of 759 people aged years with chronic low back pain, the discriminative value of inflammatory back pain symptoms for axial spondyloarthritis was low with sensitivity and specificity ranging between 50% and 80% depending on the criteria being used. 41 However, around 30% of those referred to secondary care with symptoms of inflammatory back pain receive a final diagnosis of axial spondyloarthritis. 42 Around 5% of European people presenting with chronic low back pain in primary care could have axial spondyloarthritis. 43 There is often a delay between the onset of (back pain) symptoms and making a diagnosis of axial spondyloarthritis of 5 years or longer. People with axial spondyloarthritis are commonly misdiagnosed with non-specific low back pain. Since effective treatments are now available for axial spondyloarthritis, a specialist rheumatology referral is advised for people who are suspected of having an axial spondyloarthritis. Malignancy Back pain is a common symptom in people with metastatic cancer; vertebral metastases occur in 3 5% of people with cancer, and 97% of spinal tumours are metastatic disease. 44 Nevertheless, malignancy is an uncommon cause of low back pain. Past history of malignancy is the most useful indicator for identifying such disease in people presenting with low back pain; however, it only increases the post-test probability to 7% (95% CI 3 16) in primary care, and to 33% (22 46) in the emergency setting. 36 The common solid tumours metastasising to the spine are adenocarcinomas ie, breast, lung, prostate, thyroid, and gastrointestinal. A past history of other tumours is less important. Myeloma typically presents as persistent bone pain in people aged 60 years and older. Infections Spinal infections include spondylodiscitis, vertebral osteomyelitis, epidural abscess, and rarely facet joint infection. Bacterial infections are divided into pyogenic (eg, Staphylococcus aureus and S epidermidis) and granulomatous diseases (eg, tuberculosis, brucellosis). Although rare, these disorders are associated with a substantial mortality; up to 3% for epidural abscesses, 6% for spinal osteomyelitis, and possibly as high as 11% for pyogenic spondylodiscitis In high-income countries, granulomatous diseases are mainly encountered in immigrant populations; pyogenic infections are seen largely in older patients (mean age years). 48 In low-income countries, tuberculosis affects a broader span of ages (mean age years), and could represent up to a third of spinal infections. 48 People with chronic comorbidities, particularly immunosuppressive disorders, and intravenous drug users, are at higher risk of spinal infections. Recent increases in the incidence of spinal infection are attributed to an ageing population with inherent comorbidities plus improved case ascertainment related to the availability of modern imaging techniques. 47,49 Cauda equina syndrome Although not strictly a cause of low back pain, cauda equina compression, which mainly arises from disc herniation, can have catastrophic consequences. It is rare and most primary care clinicians will not see a true case in a working lifetime. 50 Early diagnosis and surgical treatment are probably helpful; therefore, there needs to be a low threshold for further assessment when there has been a new onset of perianal sensory change or bladder symptoms, or bilateral severe radicular pain with low back pain of any duration. 50 The cardinal clinical features are urinary retention and overflow incontinence (sensitivity 90%, specificity 95%) Published online March 21,

64 Series history or clinical findings believed to increase the risk of a serious disease; however, 80% of people with acute low back pain have at least one red flag despite less than 1% having a serious disorder. 37 Nearly all recommended individual red flags are uninformative and do not substantially change post-test probabilities of a serious abnormality. 36 The very low specificity of most red flags contributes to unnecessary specialist referrals and imaging. 53 Clinicians do, however, need to consider if the overall clinical picture might indicate a serious cause for the pain, remembering that the picture can develop over time. 53 The US guideline for imaging advises deferral of imaging pending a trial of therapy when there are weak risk factors for cancer or axial spondyloarthritis. 54 How common is low back pain? Low back pain is uncommon in the first decade of life, but prevalence increases steeply during the teenage years; around 40% of 9 18-year olds in high-income, medium-income, and low-income countries report having had low back pain. 55,56 Most adults will have low back pain at some point. 57 The median 1-year period prevalence globally in the adult population is around 37%, it peaks in mid-life, and is more common in women than in men (figure 2). 1 Low back pain that is accompanied by activity limitation increases with age. 58 The mean prevalence in high-income countries is higher than in middle-income and low-income countries (32 9% [SD 19 0] vs 25 4% [25 4] vs 16 7% [16 7]), but globally there is no difference between rural and urban areas. 1 Jackson pooled results from 40 publications dealing with prevalence of persistent low back pain in 28 countries from Africa, Asia, the Middle East, and South America (n=80 076) and found that chronic low back pain was 2 5 (95% CI ) times more prevalent in working population than in non-working populations for reasons that are not clear. 59 The gender pattern in low-income and middle-income regions might also differ from that of high-income countries and even differ between low-income regions. For example, men seem to report low back pain more often than women in Africa. 56 This was not the case in Latin America, 60 which might reflect African culture, in which men often do hard physical labour, as well as gender inequalities, which might result in women underreporting their low back pain. Burden and impact of low back pain Overall disability The GBD 2015 study calculated disease burden for 315 causes in 195 countries and territories from 1990 to 2015 and provides a comprehensive assessment of the patterns and levels of acute and chronic diseases and burden and disability of those worldwide. 61 Low back pain was responsible for around 60 1 million years lived with disability (YLD) in 2015, an increase of 54% since It is the number one cause of disability Prevalence (%) Women Men Age group (years) Figure 2: Median prevalence of low back pain, with IQR, according to sex and midpoint of age group, reproduced from Hoy et al 1 with permission from John Wiley and Sons globally, as well as in 14 of the 21 GBD world regions. 4 Less than 28% of prevalent cases (n=151 million) fell in the severe and most severe categories; however, these cases accounted for 77% of all disability caused by low back pain (46 5 million YLDs). 62 Thus, most people with low back pain have low levels of disability, but the additive effect of those, combined with high disability in a substantial minority, result in the very high societal burden. In high-income countries, disabling back pain is linked to socioeconomic status, job satisfaction, and the potential for monetary compensation (table 2). The overall increase in the global burden of low back pain is almost entirely due to population increase and ageing in both high-income, low-income and middle-income countries, as opposed to increased prevalence. 1,68 Work disability Disability from low back pain is highest in working age groups worldwide (figure 3), 4,61 which is especially concerning in low-income and middle-income countries where informal employment is common and possibilities for job modification are almost completely absent. Further more, occupational musculoskeletal health policies, such as regulations for heavy physical work and lifting, are often absent or poorly monitored. 69 A survey of residents of an urban black community in Zimbabwe found that low back pain was among the top five reported primary health complaints, and reasons for activity limitation. 70 A survey among 500 farmers in rural Nigeria showed that more than half reduced their farming workload because of low back pain. 71 Thus, disability associated with low back pain might contribute to the cycle of poverty in poorer regions of the world. In high-income countries, differences in social compensation systems, not differences in occupational Published online March 21,

65 Series Symptom-related factors Previous episodes Back pain intensity Presence of leg pain Outcomes (predictor scale: association with low back pain disability) Chronic disabling pain* at 3 6 months; more vs less episodes: median LR 1 0 (range ); chronic disabling pain* at 12 months; more vs less episodes: median LR 1 1 (range ) Chronic disabling pain* at 3 6 months; high intensity pain vs non-high: median LR 1 7 (range ); chronic disabling pain* at 12 months; high intensity pain vs non-high: median LR 1 3 (range ) Chronic disabling pain* at 3 6 months; leg pain or radiculopathy vs no leg pain: median LR 1 4 (range ); chronic disabling pain* at 12 months; leg pain or radiculopathy vs no leg pain: median LR 1 4 (range ) Source of evidence Systematic review including nine longitudinal studies 63 Systematic review including eight longitudinal studies 63 Systematic review including ten longitudinal studies 63 Lifestyle factors Body mass Chronic disabling pain* at 3 6 months; BMI >25 or >27 vs lower BMI: median LR 0 91 (range ); chronic Systematic review including three longitudinal studies 63 disabling pain* at 12 months; BMI >25 or >27 vs lower BMI: median LR 0 84 (range ) Smoking Chronic disabling pain* at 3 6 months; current smoker vs not: median LR 1 2 (range ) Systematic review including three longitudinal studies 63 Physical activity Disability 1 5 years; significant association in one of five studies (no effect size reported) Systematic review including five longitudinal studies 64 Psychological factors Depression Mixed outcomes; significant associations with poor outcome in eight of 13 cohorts; OR (range) Systematic review including 13 longitudinal studies 65 Catastrophising Disability at 3 12 months; significant association in nine of 13 studies; high catastrophising: OR 1 56 (95% CI ); 0 6 scale: 7 63 ( ); 0 52 scale: 1 05 ( ); contribution to explained variance: 0 23% Systematic review including 13 longitudinal studies 66 Fear avoidance beliefs Social factors Physical work loads Education Compensation Work satisfaction Pain or activity limitation at 3 12 months; no pooled estimates; no systematic association between fear avoidance and outcome; poor work-related outcome at 3 12 months; elevated fear avoidance: OR (range) 1 05 (95% CI ) to 4 64 ( ; from four studies done by disability insurance companies); chronic disabling pain* at 3 6 months; high vs no fear avoidance: median LR 2 2 (range ); chronic disabling pain* at 12 months; median LR 2 5 (range ) Chronic disabling pain* at 3 6 months; higher vs lower physical work demands: median LR 1 2 (range ); chronic disabling pain* at 12 months; higher vs lower physical work demands: median LR 1 4 (range ) Chronic disabling pain* at 3 6 months; no college education or not college graduate vs more education: median LR 1 0 (range ); chronic disabling pain* at 12 months; no college education or not college graduate vs more education: median LR 1 1 (range ) Chronic disabling pain* at 3 6 months; compensated work injury or sick leave vs not compensated work injury or sick leave: median LR 1 3 (range ); chronic disabling pain* at 12 months; compensated work injury or sick leave vs not compensated work injury or sick leave: median LR 1 4 (range ) Chronic disabling pain* at 3 6 months; less vs more work satisfaction: median LR 1 1 (range ); chronic disabling pain* at 12 months; less vs more work satisfaction: median LR 1 5 (range ) Systematic review including 21 longitudinal studies 67 Systematic review including four longitudinal studies 63 Systematic review including four longitudinal studies 63 Systematic review including ten longitudinal studies 63 Systematic review including seven longitudinal studies 63 Systematic review including five longitudinal studies 63 The information provided in the table provides a broad overview and was not based on a systematic review of the literature. LR=positive likelihood ratio. BMI=body-mass index. OR=odds ratio. HR=hazard ratio. *Pain persistent beyond 3 months and at least moderately affecting ability to work or function. Table 2: Overview of selected predictors and their association with dichotomous outcomes of low back pain disability exposure or individual factors, are largely responsible for national differences in the rates and extent of work disability attributed to low back pain. 7 In Europe, low back pain is the most common cause of medically certified sick leave and early retirement. 72 However, work disability due to low back pain varies substantially among European countries. For example, in Norway and Sweden in 2000, short-term sickness absence rates in people with back pain were similar (5 1% and 6 4%, respectively), but the rate of longer-term medically certified sickness absence was very different (22% and 15%, respectively). 73 In the USA, low back pain accounts for more lost workdays than any other occupational musculoskeletal condition, 74 but although 58 of US workers filed a back-related claim in 1999, the comparable figure from Japan during the same year was only one of Social identity and inequality The effect of low back pain on social identity and inequality is substantial worldwide. Ethnographic interviews of villagers in Botswana found that low back pain and other musculoskeletal symptoms resulted in both economic and subsistence consequences as well as loss of independence and social identity because of inability to fulfil traditional and expected social roles in a society with harsh living conditions. 76 Froud and colleagues 77 reviewed 42 qualitative studies all from high-income countries, and found that many people living with low back pain struggled to meet their social expectations and obligations and that achieving them might then threaten the credibility of their suffering, with disability claims being endangered. Although those with back pain seek to achieve premorbid levels of health, many find with time that this aim is unrealistic and live with reduced expectations. 77 Likewise, MacNeela and colleagues 78 reviewed 38 separate qualitative studies, also from high-income countries, and found some common themes, including: worry and fear about the social consequences of chronic low back pain, hopelessness, family strain, social withdrawal, loss of job and lack of money, disappointment with healthcare encounters (in particular with general practitioners), 6 Published online March 21,

66 Series coming to terms with the pain, and learning selfmanagement strategies. Globally, low back pain contributes to inequality. In low-income and middle-income countries, poverty and inequality might increase as participation in work is affected. Furthermore, formal return-to-work systems are often not in place, and workers might be retrenched, placing more strain on family and community livelihoods. 69 In Australia, Schofield and colleagues 79 found that individuals who exit the workforce early as a result of their low back pain have substantially less wealth by age 65 years, even after adjustment for education. The median value of accumulated wealth for those who retire early because of low back pain is only AUS$5038 by the time they reach 65 years of age, compared with $ for those who remain in the workforce. 79 Cost of low back pain No relevant studies on costs associated with low back pain from low-income and middle-income countries were identified. Costs associated with low back pain are generally reported as direct medical (health-care) costs, and indirect (work absenteeism or productivity loss) costs. Only a few studies have reported other direct nonmedical costs, such as costs from transportation to appointments, visits to complementary and alternative practitioners, and informal help not captured by the health-care system, which means that most studies underestimate the total costs of low back pain (appendix). The economic impact related to low back pain is comparable to other prevalent, high-cost conditions, such as cardiovascular disease, cancer, mental health, and autoimmune diseases. 6 Replacement wages account for 80 90% of total costs, and consistently a small percentage of cases account for these. 80 Some of the observed variation in costs for low back pain over time might be explained by changes in disability legislation and healthcare practices. For example, in the Netherlands, costs associated with low back pain were substantially reduced between 1991 and 2007 after a change in legislation that reduced disability pensions and applied evidence-based criteria for medical practices. 7,81 Estimates of direct medical costs associated with low back pain are also all from high-income countries, with the USA having the highest costs, attributable to a more medically intensive approach and higher rates of surgery compared with other high-income countries (appendix). 8,82 In the UK in 2006, one in seven of all recorded consultations with general practitioners were for musculoskeletal problems with complaints of back pain being the most common (417 consultations per year for low back pain per registered persons), 83 and in South Africa, low back pain is the sixth most common complaint seen in primary health care. 84 In addition to conventional medicine, complementary and alternative medical approaches are popular with people who have low back pain. For example, DALYs Figure 3: Global burden of low back pain, in disability-adjusted life-years (DALYs), by age group, for 1990 and 2015 Data are from the Global Health Data Exchange. in the USA 44% of the population used at least one complementary or alternative health-care therapy in 1997; 85 and the most common reason was low back pain. 86 Natural history Low back pain is increasingly understood as a longlasting condition with a variable course rather than episodes of unrelated occurrences. 87 Around half the people seen with low back pain in primary care have a trajectory of continuing or fluctuating pain of low-tomoderate intensity, some recover, and some have persistent severe low back pain. 88 A systematic review 89 (33 cohorts; participants) provides strong evidence that most episodes of low back pain improve substantially within 6 weeks, and by 12 months average pain levels are low (6 points on a 100-point scale; 95% CI 3 10). However, two-thirds of patients still report some pain at 3 months; 67% (95% CI 50 83) and 12 months; 65% (54 75). 89,90 Recurrences of low back pain are common but a 2017 systematic review (seven studies; 1780 participants) found that research does not provide robust estimates of the risk of low back pain recurrence. The best evidence suggests around 33% of people will have a recurrence within 1 year of recovering from a previous episode. 91 Risk factors and triggers for episodes of low back pain Although the impact of low back pain in low-income and middle-income countries on systems and people differs from high-income countries, there seem to be fewer fundamental differences in the risk factors between regions. A systematic review 92 (eight cohorts; 5165 participants) found consistent evidence that people who have had previous episodes of low back pain are at increased risk of a new episode. Likewise, people with other chronic conditions, including asthma, headache, and diabetes, are more likely to report low back pain Age group (years) For the Global Health Data Exchange see healthdata.org/gbd-2016 See Online for appendix Published online March 21,

67 Series than people in good health (pooled ORs ). 93 People with poor mental health are also at increased risk. For example, a UK cohort study 94 (5781 participants) found psychological distress at age 23 years predicted incident low back pain 10 years later (OR 2 52, 95% CI ]. The Canadian National Population Health Survey 95 with 9909 participants found that pain-free individuals with depression were more likely to develop low back pain within 2 years than were people without depression (OR 2 9, 95% CI ). Mechanisms behind the coexistence of low back pain and other chronic diseases are not known, but systematic reviews of cohort studies indicate that lifestyle factors such as smoking, 96 obesity, 97,98 and low levels of physical activity 99 that relate to poorer general health are also associated with occurrence of low back pain episodes or development of persistent low back pain, although independent associations remain uncertain. A systematic review 93 (seven twin studies; participants) found the genetic influence on the liability to develop low back pain ranged from 21% to 67%, with the genetic component being higher for more chronic and disabling low back pain than for inconse quential low back pain. A comprehensive genetic epidemiological analysis of Danish twins (44% monozygotic and 56% dizygotic) found that heritability estimates for pain in different spinal regions were quite similar and there is a moderate to high genetic correlation between the phenotypes, which might indicate a common genetic basis for a high proportion of spinal pain. 100 An Australian case-crossover study (999 participants) showed that awkward postures (OR 8 0, 95% CI ), heavy manual tasks (5 0, ), feeling tired (3 7, ), or being distracted during an activity (25 0, ) were all associated with increased risk of a new episode of low back pain. 101 Similarly, work exposures of lifting, bending, awkward postures, and tasks considered physically demanding were also associated with an increased risk of developing low back pain in low-income and middle-income countries. 56,60 A systematic review (25 cohorts) showed that the effect of heavy workload on onset of low back pain ranged from OR 1 61 (95% CI ) to OR 4 1 ( ). 102 The existence of a causal pathway between these risk factors and low back pain, however, remains unclear. 103 Multifactoral contributors to persistent disabling low back pain In recent decades, the biopsychosocial model has been applied as a framework for understanding the complexity of low back pain disability in preference to a purely biomedical approach. Many factors including biophysical, psycho logical, social and genetic factors, and co morbidities (figure 1) can contribute to disabling low back pain (table 2). However, no firm boundaries exist among these factors and they all interact with each other. Thus, persistent disabling low back pain is not merely a result of nociceptive input. Although there are substantially fewer data from low-income and middleincome countries than from high-income countries, the available data suggest similar multifactorial contributors seem to be important in all countries. 104 Biophysical factors Although the role of biophysical impairments in the development of disabling low back pain is not fully understood, impairments are demonstrable in people with persistent low back pain. One example is that some people with persistent low back pain might have alterations in muscle size, 105 composition, 106 and coordination 107 that differ from those without pain. These changes could be more than merely a direct consequence of pain and are only partly affected by psychological factors. 108 Psychological factors Psychological factors are often investigated separately, but there is a substantial overlap of constructs such as depression, anxiety, catastrophising (ie, an irrational belief that something is far worse that it really is), and self-efficacy (ie, belief in one's ability to influence events affecting one's life). The presence of these factors in people who present with low back pain is associated with increased risk of developing disability even though the mechanisms are not fully understood (table 2). For example, in a UK cohort study of 531 participants, painrelated distress explained 15% and 28% of the variance in pain and disability, respectively. 109 The fear-avoidance model of chronic pain (including low back pain), which describes how fear of pain leads to the avoidance of activities and thus to disability, is well established. This model has more recently been expanded to capture the influence of maladaptive learning processes and disabling beliefs on pain perception and on behaviours, suggesting that pain cognitions have a central role in the development and maintenance of disability, and more so than the pain itself. 110 A systematic review, including 12 mediation studies, identified self-efficacy, psychological distress, and fear as intermediate factors explaining some of the pathway between having neck or back pain and developing disability. 111 The potential importance of self-efficacy is supported by a systematic review (83 studies; participants) of chronic pain conditions (23 low back pain studies) that found selfefficacy to be consistently associated with impairment and disability, affective distress, and pain severity. 112 Therefore, some chronic pain treatments have shifted away from aiming to directly alleviate pain to aiming to change beliefs and behaviours. 113 Social and societal factors Chronic disabling low back pain affects people with low income and short education disproportionally. In a UK study of 2533 people, life-time socioeconomic status 8 Published online March 21,

68 Series predicted disability due to any pain condition in older age (independent of comorbid conditions, psychological indicators and body-mass index (BMI); OR 2 04 (95% CI ). 114 Cross-sectional data from the USA (National Health Interview Survey , 5103 people) found that those with persistent low back pain were more likely to have had less than high-school education (2 27, ) and had an annual household income of less than US$ (2 29, ). 115 Suggested mechanisms for the effect of low education on back pain include environmental and lifestyle exposures in lower socioeconomic groups, lower health literacy, and health care not being available or adequately targeted to people with low education. 116 Also, being in routine and manual occupations and having increased physical workloads is associated with disabling low back pain (table 2). Central pain processing and modulation Nociceptive input is processed throughout the nervous system, including modulation within the spinal cord and supraspinal centres. In chronic pain, supraspinal centres can show varying levels of activation and can be recruited for activation (or not) in a dynamic fashion contingent on nociceptive drive, context, cognition, and emotion. If any of these factors change, the same nociceptive input can produce a different cerebral signature in the same patient. 117 A systematic review (27 studies; 1037 participants) identified moderate evidence that patients with chronic low back pain show structural brain differences in specific cortical and subcortical areas, and altered functional connectivity in pain-related areas following painful stimulation. 118 The clinical implication of these findings remains to be clarified. 117 Multivariable predictive models Pain intensity, psychological distress, and accompanying pain in the leg or at multiple body sites are identified as predictors across externally validated multivariable predictive models, which have been developed to identify people at particular risk of developing disabling low back pain (appendix). In a systematic review (50 studies; participants), the average amount of variance explained in seven development samples was 43%, indicating that most of the variation between individuals is due to unknown or unmeasured factors. 119 Limitations Despite advances in many aspects of understanding low back pain, including the burden, course, risk factors, and causes, some important limitations exist. Most evidence comes from high-income countries, and may or may not generalise to low-income and middleincome countries. Although many factors are associated with both the development of low back pain and the transition to persistent disabling pain, the underlying mechanisms, including the effect of co-occurring noncommunicable diseases, are poorly understood. Despite the burden of low back pain, research is often not a priority in low-income and-middle income countries, and thus the consequences of low back pain in these settings are largely unknown. The functional domains used in the GBD 2015 study do not take into account broader aspects of life, such as participation, wellbeing, social identity, carer burden, use of health-care resources, and work disability costs. In cost studies, a top-down approach is most often used and those might not capture all costs as seen from the individual point of view in specific contexts. Conclusion Low back pain is now the number one cause of disability globally. The burden from low back pain is increasing, particularly in low-income and middle-income countries, which is straining health-care and social systems that are already overburdened. Low back pain is most prevalent and burdensome in working populations, and in older people low back pain is associated with increased activity limitation. Most cases of low back pain are short-lasting and a specific nociceptive source cannot be identified. Recurrences are, however, common and a few people end up with persistent disabling pain affected by a range of biophysical, psychological, and social factors. Costs associated with health care and work disability attributed to low back pain are enormous but vary substantially between countries, and are related to social norms, health-care approaches, and legislation. Although there are several global initiatives to address the global burden of low back pain as a public health problem, there is a need to identify cost-effective and context-specific strategies for managing low back pain to mitigate the consequences of the current and projected future burden. Contributors JH and MU were part of the team that developed the original proposal for the series and coordinated production of papers. JH and MH led the drafting of this paper in collaboration with the other authors. AK, QL, and MU closely revised many sections. Thereafter all authors contributed to all sections of the paper and edited it for key intellectual content. JH, MJH, AK, JK, MLF, SG, RJS, QL, GP, and MU participated in the authors meeting, drafted different sections of the paper, and took part in discussions during the drafting process. All other authors have read and provided substantive intellectual comments to the draft and approved the final version of the paper. The Lancet Low Back Pain Series Working Group Steering Committee: Rachelle Buchbinder (Chair) Monash University, Melbourne, Australia; Jan Hartvigsen (Deputy Chair), University of Southern Denmark, Odense, Denmark; Dan Cherkin, Kaiser Permanente Washington Health Research Institute, Seattle, USA; Nadine E Foster, Keele University, Keele, UK; Chris G Maher, University of Sydney, Sydney, Australia; Martin Underwood, Warwick University, Coventry, UK; Maurits van Tulder, Vrije Universiteit, Amsterdam, Netherlands. Members: Johannes R Anema, VU University Medical Centre, Amsterdam, Netherlands; Roger Chou, Oregon Health and Science University, Portland, USA; Stephen P Cohen, Johns Hopkins School of Medicine, Baltimore, USA; Lucíola Menezes Costa, Universidade Cidade de Sao Paulo, Sao Paulo, Brazil; Peter Croft, Keele University, Keele, UK; Manuela Ferreira, Paulo H Ferreira, Damian Hoy, University of Sydney, Sydney, Australia; Julie M Fritz, University of Utah, Salt Lake City, USA; Stéphane Genevay, University Hospital of Geneva, Geneva, Switzerland; Published online March 21,

69 Series Douglas P Gross, University of Alberta, Edmonton, Canada; Mark Hancock, Macquarie University, Sydney, Australia; Jaro Karppinen, University of Oulu and Oulu University Hospital, Oulu, Finland; Bart W Koes, Erasmus MC, University Medical Center Rotterdam, Rotterdam, Netherlands; Alice Kongsted, University of Southern Denmark, Odense, Denmark; Quinette Louw, Stellenbosch University, Tygerberg, South Africa; Birgitta Öberg, Linkoping University, Linkoping, Sweden; Wilco Peul, Leiden University, Leiden, Netherlands; Glenn Pransky, University of Massachusetts Medical School, Worcester, USA; Mark Schoene, The Back Letter, Lippincott Williams & Wilkins, Newburyport, USA; Joachim Sieper, Charite, Berlin, Germany; Rob Smeets, Maastricht University, Maastricht, Netherlands; Judith A Turner, University of Washington School of Medicine, Seattle, USA; Anthony Woolf, Royal Cornwall Hospital and University of Exeter Medical School, Truro, UK. Declaration of interests See appendix for authors' declaration of interests. References 1 Hoy D, Bain C, Williams G, et al. A systematic review of the global prevalence of low back pain. Arthritis Rheum 2012; 64: Kamper SJ, Henschke N, Hestbaek L, Dunn KM, Williams CM. Musculoskeletal pain in children and adolescents. Braz J Phys Ther 2016; 16: Hartvigsen J, Christensen K, Frederiksen H. Back pain remains a common symptom in old age. a population-based study of 4486 Danish twins aged Eur Spine J 2003; 12: Global Burden of Disease, Injury Incidence, Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, : a systematic analysis for the Global Burden of Disease Study Lancet 2016; 388: Hoy DG, Smith E, Cross M, et al. Reflecting on the global burden of musculoskeletal conditions: lessons learnt from the global burden of disease 2010 study and the next steps forward. Ann Rheum Dis 2015; 74: Maniadakis N, Gray A. The economic burden of back pain in the UK. Pain 2000; 84: Anema JR, Schellart AJ, Cassidy JD, Loisel P, Veerman TJ, van der Beek AJ. Can cross country differences in return-to-work after chronic occupational back pain be explained? An exploratory analysis on disability policies in a six country cohort study. J Occup Rehabil 2009; 19: Deyo RA, Mirza SK, Turner JA, Martin BI. Overtreating chronic back pain: time to back off? J Am Board Fam Med 2009; 22: Foster NE, Anema JR, Cherkin D, et al. Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet 2018; published online March S (18) Buchbinder R, van Tulder M, Öberg B, et al. Low back pain: a call for action. Lancet 2018; published online March org/ /s (18) Dionne CE, Dunn KM, Croft PR, et al. A consensus approach toward the standardization of back pain definitions for use in prevalence studies. Spine 2008; 33: 95s Maher C, Underwood M, Buchbinder R. Non-specific low back pain. Lancet 2017; 389: Hartvigsen J, Natvig B, Ferreira M. Is it all about a pain in the back? Best Pract Res Clin Rheum 2013; 27: Hancock MJ, Maher CG, Latimer J, et al. Systematic review of tests to identify the disc, SIJ or facet joint as the source of low back pain. Eur Spine J 2007; 16: Maas ET, Ostelo RW, Niemisto L, et al. Radiofrequency denervation for chronic low back pain. Cochrane Database Syst Rev 2015: CD Kalichman L, Li L, Kim DH, et al. Facet joint osteoarthritis and low back pain in the community-based population. Spine 2008; 33: Maas ET, Juch JN, Ostelo RW, et al. Systematic review of patient history and physical examination to diagnose chronic low back pain originating from the facet joints. Eur J Pain 2017; 21: Dudli S, Fields AJ, Samartzis D, Karppinen J, Lotz JC. Pathobiology of Modic changes. Eur Spine J 2016; 25: Dudli S, Liebenberg E, Magnitsky S, Miller S, Demir-Deviren S, Lotz JC. Propionibacterium acnes infected intervertebral discs cause vertebral bone marrow lesions consistent with Modic changes. J Orthop Res 2016; 34: Brinjikji W, Diehn FE, Jarvik JG, et al. MRI Findings of disc degeneration are more prevalent in adults with low back pain than in asymptomatic controls: a systematic review and meta-analysis. Am J Neuroradio 2015; 36: Maatta JH, Karppinen J, Paananen M, et al. Refined phenotyping of Modic changes: imaging biomarkers of prolonged severe low back pain and disability. Medicine 2016; 95: e Brinjikji W, Luetmer PH, Comstock B, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. Am J Neuroradiol 2015; 36: Steffens D, Hancock MJ, Maher CG, Williams C, Jensen TS, Latimer J. Does magnetic resonance imaging predict future low back pain? A systematic review. Eur J Pain 2014; 18: Jarvik JG, Gold LS, Comstock BA, et al. Association of early imaging for back pain with clinical outcomes in older adults. JAMA 2015; 313: Wong JJ, Cote P, Sutton DA, et al. Clinical practice guidelines for the noninvasive management of low back pain: A systematic review by the Ontario Protocol for Traffic Injury Management (OPTIMa) Collaboration. Eur J Pain 2016; 21: Stochkendahl MJ, Kjaer P, Hartvigsen J, et al. National clinical guidelines for non-surgical treatment of patients with recent onset low back pain or lumbar radiculopathy. Eur Spine J 2018; 27: Bernstein IA, Malik Q, Carville S, Ward S. Low back pain and sciatica: summary of NICE guidance. BMJ 2017; 356: i Qaseem A, Wilt TJ, McLean RM, Forciea MA, for the Clinical Guidelines Committee of the American College of Physicians. Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2017; 166: Lin CW, Verwoerd AJ, Maher CG, et al. How is radiating leg pain defined in randomized controlled trials of conservative treatments in primary care? A systematic review. Eur J Pain 2014; 18: Verwoerd AJ, Mens J, El Barzouhi A, Peul WC, Koes BW, Verhagen AP. A diagnostic study in patients with sciatica establishing the importance of localization of worsening of pain during coughing, sneezing and straining to assess nerve root compression on MRI. Eur Spine J 2016; 25: Kongsted A, Kent P, Jensen TS, Albert H, Manniche C. Prognostic implications of the Quebec Task Force classification of back-related leg pain: an analysis of longitudinal routine clinical data. BMC Musculoskelet Dis 2013; 14: Chiu CC, Chuang TY, Chang KH, Wu CH, Lin PW, Hsu WY. The probability of spontaneous regression of lumbar herniated disc: a systematic review. Clin Rehabil 2015; 29: Chad DA. Lumbar spinal stenosis. Neurol Clin 2007; 25: Tomkins-Lane C, Melloh M, Lurie J, et al. Consensus on the clinical diagnosis of lumbar spinal stenosis: results of an international Delphi study. Spine 2016; 41: Schousboe JT. Epidemiology of vertebral fractures. J Clin Densitom 2016; 19: Downie A, Williams CM, Henschke N, et al. Red flags to screen for malignancy and fracture in patients with low back pain: systematic review. BMJ 2013; 347: f Henschke N, Maher CG, Refshauge KM, et al. Prevalence of and screening for serious spinal pathology in patients presenting to primary care settings with acute low back pain. Arthritis Rheum 2009; 60: Stolwijk C, van Onna M, Boonen A, van Tubergen A. The global prevalence of spondyloarthritis: A systematic review and meta-regression analysis. Arthritis Care Res 2015; 68: Rudwaleit M, van der Heijde D, Landewe R, et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009; 68: Sieper J, van der Heijde D. Review: Nonradiographic axial spondyloarthritis: new definition of an old disease? Arthritis Rheum 2013; 65: Published online March 21,

70 Series 41 Arnbak B, Hendricks O, Horslev-Petersen K, et al. The discriminative value of inflammatory back pain in patients with persistent low back pain. Scand J Rheumatol 2016; 45: Poddubnyy D, van Tubergen A, Landewe R, Sieper J, van der Heijde D. Defining an optimal referral strategy for patients with a suspicion of axial spondyloarthritis: what is really important? Ann Rheum Dis 2015; 74: e van Hoeven L, Luime J, Han H, Vergouwe Y, Weel A. Identifying axial spondyloarthritis in Dutch primary care patients, ages years, with chronic low back pain. Arthritis Care Res 2014; 66: Lewandrowski K, Anderson M, McLain R. Tumors of the Spine. In: Herkowitz H, Garfin S, Eismont F, Bell G, Balderston R, eds. Rothman-Simeone the spine. Philadelphia, PA: Elsevier Sounders; 2011: Schoenfeld AJ, Wahlquist TC. Mortality, complication risk, and total charges after the treatment of epidural abscess. Spine J 2015; 15: Fantoni M, Trecarichi EM, Rossi B, Mazzotta V, Di Giacomo G, Nasto LA, Di Meco E, Pola E. Epidemiological and clinical features of pyogenic spondylodiscitis. Eur Rev Med Pharmacol Sci 2012; 16: Akiyama T, Chikuda H, Yasunaga H, Horiguchi H, Fushimi K, Saita K. Incidence and risk factors for mortality of vertebral osteomyelitis: a retrospective analysis using the Japanese diagnosis procedure combination database. BMJ Open 2013; 3: e Trecarichi EM, Di Meco E, Mazzotta V, Fantoni M. Tuberculous spondylodiscitis: epidemiology, clinical features, treatment, and outcome. Eur Rev Med Pharmacol Sci 2012; 16: Kehrer M, Pedersen C, Jensen TG, Lassen AT. Increasing incidence of pyogenic spondylodiscitis: a 14-year population-based study. J Infect 2014; 68: Lavy C, James A, Wilson-MacDonald J, Fairbank J. Cauda equina syndrome. BMJ 2009; 338: b Abrahm JL. Assessment and treatment of patients with malignant spinal cord compression. J Support Oncol 2004; 2: Galukande M, Muwazi S, Mugisa DB. Aetiology of low back pain in Mulago Hospital, Uganda. Afr Health Sci 2005; 5: Underwood M, Buchbinder R. Red flags for back pain. BMJ 2013; 347: f American Academy of Family Physicians. Imaging for Low Back Pain (accessed Nov 1, 2017). 55 Calvo-Munoz I, Gomez-Conesa A, Sanchez-Meca J. Prevalence of low back pain in children and adolescents: a meta-analysis. BMC Pediatrics 2013; 13: Louw QA, Morris LD, Grimmer-Somers K. The prevalence of low back pain in Africa: a systematic review. BMC Musculoskelet Disord 2007; 8: Lemeunier N, Leboeuf-Yde C, Gagey O. The natural course of low back pain: a systematic critical literature review. Chiropract Man Ther 2012; 20: Hoy D, March L, Brooks P, et al. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis 2014; 73: Jackson T, Thomas S, Stabile V, Shotwell M, Han X, McQueen K. A Systematic review and meta-analysis of the global burden of chronic pain without clear etiology in low- and middle-income countries: trends in heterogeneous data and a proposal for new assessment methods. Anesth Analg 2016; 123: Garcia JB, Hernandez-Castro JJ, Nunez RG, et al. Prevalence of low back pain in Latin America: a systematic literature review. Pain Phys 2014; 17: Global Burden of Disease 2015 DALYs and HALE Collaborators. Global, regional, and national disability-adjusted life-years (DALYs) for 315 diseases and injuries and healthy life expectancy (HALE), : a systematic analysis for the Global Burden of Disease Study Lancet 2016; 388: Global Burden of Disease 2015 DALYs and HALE Collaborators. Global, regional, and national disability-adjusted life years (DALYs) for 306 diseases and injuries and healthy life expectancy (HALE) for 188 countries, : quantifying the epidemiological transition. Lancet 2015; 386: Chou R, Shekelle P. Will this patient develop persistent disabling low back pain? JAMA 2010; 303: Hendrick P, Milosavljevic S, Hale L, et al. The relationship between physical activity and low back pain outcomes: a systematic review of observational studies. Eur Spine J 2011; 20: Pinheiro MB, Ferreira ML, Refshauge K, et al. Symptoms of depression as a prognostic factor for low back pain: a systematic review. Spine J 2016; 16: Wertli MM, Eugster R, Held U, Steurer J, Kofmehl R, Weiser S. Catastrophizing-a prognostic factor for outcome in patients with low back pain: a systematic review. Spine J 2014; 14: Wertli MM, Rasmussen-Barr E, Weiser S, Bachmann LM, Brunner F. The role of fear avoidance beliefs as a prognostic factor for outcome in patients with nonspecific low back pain: a systematic review. Spine J 2014; 14: Hoy D, March L, Brooks P, Woolf A, Blyth F, Vos T, Buchbinder R. Measuring the global burden of low back pain. Best Pract Res Clin Rheum 2010; 24: Lucchini RG, London L. Global occupational health: current challenges and the need for urgent action. Ann Glob Health 2014; 80: Jelsma J, Mielke J, Powell G, De Weerdt W, De Cock P. Disability in an urban black community in Zimbabwe. Disabil Rehabil 2002; 24: Fabunmi AA, Aba SO, Odunaiya NA. Prevalence of low back pain among peasant farmers in a rural community in South West Nigeria. Afr J Med Med Sci 2005; 34: Bevan S, Quadrello T, McGee R, Mahdon M, Vavrovsky A, Barham L. Fit For work? Musculoskeletal disorders in the European workforce: fit For work Europe: The Work Foundation, Ihlebaek C, Hansson TH, Laerum E, et al. Prevalence of low back pain and sickness absence: a borderline study in Norway and Sweden. Scand J Public Health 2006; 34: US Bone & Joint Initiative. The Burden of Musculoskeletal Diseases in the United States, about/rights (accessed Nov, 2017). 75 Volinn E, Nishikitani M, Volinn W, Nakamura Y, Yano E. Back pain claim rates in Japan and the United States: framing the puzzle. Spine 2005; 30: Hondras M, Hartvigsen J, Myburgh C, Johannessen H. Everyday burden of musculoskeletal conditions among villagers in rural Botswana: a focused ethnography. J Rehabil Med 2016; 48: Froud R, Patterson S, Eldridge S, et al. A systematic review and meta-synthesis of the impact of low back pain on people s lives. BMC Musculoskelet Disord 2014; 15: MacNeela P, Doyle C, O Gorman D, Ruane N, McGuire BE. Experiences of chronic low back pain: a meta-ethnography of qualitative research. Health Psychol Rev 2015; 9: Schofield D, Kelly S, Shrestha R, Callander E, Passey M, Percival R. The impact of back problems on retirement wealth. Pain 2012; 153: Hashemi L, Webster BS, Clancy EA. Trends in disability duration and cost of workers compensation low back pain claims ( ). J Occup Environ Med 1998; 40: Lambeek LC, van Tulder MW, Swinkels IC, Koppes LL, Anema JR, van Mechelen W. The trend in total cost of back pain in The Netherlands in the period 2002 to Spine 2011; 36: Hansson TH, Hansson EK. The effects of common medical interventions on pain, back function, and work resumption in patients with chronic low back pain: a prospective 2-year cohort study in six countries. Spine 2000; 25: Jordan KP, Kadam UT, Hayward R, Porcheret M, Young C, Croft P. Annual consultation prevalence of regional musculoskeletal problems in primary care: an observational study. BMC Musculoskelet Disord 2010; 11: Mash B, Fairall L, Adejayan O, et al. A morbidity survey of South African primary care. PLoS One 2012; 7: e Wolsko P, Ware L, Kutner J, et al. Alternative/complementary medicine: wider usage than generally appreciated. J Altern Complement Med 2000; 6: Wolsko PM, Eisenberg DM, Davis RB, Kessler R, Phillips RS. Patterns and perceptions of care for treatment of back and neck pain: results of a national survey. Spine 2003; 28: Dunn KM, Hestbaek L, Cassidy JD. Low back pain across the life course. Best Pract Res Clin Rheum 2013; 27: Published online March 21,

71 Series 88 Kongsted A, Kent P, Axen I, Downie AS, Dunn KM. What have we learned from ten years of trajectory research in low back pain? BMC Musculoskelet Dis 2016; 17: da C Menezes Costa L, Maher CG, Hancock MJ, McAuley JH, Herbert RD, Costa LO. The prognosis of acute and persistent low-back pain: a meta-analysis. CMAJ 2012; 184: E Itz CJ, Geurts JW, van Kleef M, Nelemans P. Clinical course of non-specific low back pain: a systematic review of prospective cohort studies set in primary care. Eur J Pain 2013; 17: da Silva T, Mills K, Brown BT, Herbert RD, Maher CG, Hancock MJ. Risk of recurrence of low back pain: a systematic review. J Orthop Sports Phys Ther 2017; 47: Taylor JB, Goode AP, George SZ, Cook CE. Incidence and risk factors for first-time incident low back pain: a systematic review and meta-analysis. Spine J 2014; 14: Ferreira PH, Beckenkamp P, Maher CG, Hopper JL, Ferreira ML. Nature or nurture in low back pain? Results of a systematic review of studies based on twin samples. Eur J Pain 2013; 17: Power C, Frank J, Hertzman C, Schierhout G, Li L. Predictors of low back pain onset in a prospective British study. Am J Public Health 2001; 91: Currie SR, Wang J. More data on major depression as an antecedent risk factor for first onset of chronic back pain. Psychol Med 2005; 35: Shiri R, Karppinen J, Leino-Arjas P, Solovieva S, Viikari-Juntura E. The association between smoking and low back pain: a meta-analysis. Am J Med 2010; 123: Zhang TT, Liu Z, Liu YL, Zhao JJ, Liu DW, Tian QB. Obesity as a risk factor for low back pain: a meta-analysis. Clin Spine Surg 2016; 31: Shiri R, Karppinen J, Leino-Arjas P, Solovieva S, Viikari-Juntura E. The association between obesity and low back pain: a meta-analysis. Am J Epidemiol 2010; 171: Shiri R, Falah-Hassani K. Does leisure time physical activity protect against low back pain? Systematic review and meta-analysis of 36 prospective cohort studies. Br J Sports Med 2017; 51: Hartvigsen J, Nielsen J, Kyvik KO, et al. Heritability of spinal pain and consequences of spinal pain: a comprehensive genetic epidemiologic analysis using a population-based sample of 15,328 twins ages years. Arthritis Rheum 2009; 61: Steffens D, Ferreira ML, Latimer J, et al. What triggers an episode of acute low back pain? A case-crossover study. Arthritis Care Res 2015; 67: Heneweer H, Staes F, Aufdemkampe G, van Rijn M, Vanhees L. Physical activity and low back pain: a systematic review of recent literature. Eur Spine J 2011; 20: Kwon BK, Roffey DM, Bishop PB, Dagenais S, Wai EK. Systematic review: occupational physical activity and low back pain. Occup Med 2011; 61: Oncu J, Iliser R, Kuran B. Cross-cultural adaptation of the Orebro Musculoskeletal Pain Questionnaire among Turkish workers with low back pain. J Back Musculoskelet Rehabil 2016; 29: Goubert D, Oosterwijck JV, Meeus M, Danneels L. Structural changes of lumbar muscles in non-specific low back pain: a systematic review. Pain Phys 2016; 19: E985 E Sions JM, Elliott JM, Pohlig RT, Hicks GE. Trunk muscle characteristics of the multifidi, erector spinae, psoas, and quadratus lumborum in older adults with and without chronic low back pain. J Orthop Sports Phys Ther 2017; 47: Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 1996; 21: Dubois JD, Abboud J, St-Pierre C, Piche M, Descarreaux M. Neuromuscular adaptations predict functional disability independently of clinical pain and psychological factors in patients with chronic non-specific low back pain. J Electromyogra Kinesiol 2014; 24: Campbell P, Bishop A, Dunn KM, Main CJ, Thomas E, Foster NE. Conceptual overlap of psychological constructs in low back pain. Pain 2013; 154: Crombez G, Eccleston C, Van Damme S, Vlaeyen JW, Karoly P. Fear-avoidance model of chronic pain: the next generation. Clin J Pain 2012; 28: Lee H, Hubscher M, Moseley GL, et al. How does pain lead to disability? A systematic review and meta-analysis of mediation studies in people with back and neck pain. Pain 2015; 156: Jackson T, Wang Y, Wang Y, Fan H. Self-efficacy and chronic pain outcomes: a meta-analytic review. J Pain 2014; 15: Frost H, Klaber Moffett JA, Moser JS, Fairbank JC. Randomised controlled trial for evaluation of fitness programme for patients with chronic low back pain. BMJ 1995; 310: Lacey RJ, Belcher J, Croft PR. Does life course socio-economic position influence chronic disabling pain in older adults? A general population study. Eur J Public Health 2013; 23: Shmagel A, Foley R, Ibrahim H. Epidemiology of chronic low back pain in US adults: National Health and Nutrition Examination Survey Arthritis Care Res 2016; 68: Dionne CE, Von Korff M, Koepsell TD, Deyo RA, Barlow WE, Checkoway H. Formal education and back pain: a review. J Epidemiol Community Health 2001; 55: Roussel NA, Nijs J, Meeus M, Mylius V, Fayt C, Oostendorp R. Central sensitization and altered central pain processing in chronic low back pain: fact or myth? Clin J Pain 2013; 29: Kregel J, Meeus M, Malfliet A, et al. Structural and functional brain abnormalities in chronic low back pain: A systematic review. Semin Arthritis Rheum 2015; 45: Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific low back pain? A systematic review. Man Ther 2008; 13: Elsevier Ltd. All rights reserved Published online March 21,

72 Orthopedics & Biomechanics Thieme The Effect of Clinical Pilates on Functional Movement in Recreational Runners Authors Anna Laws, Sean Williams, Cassie Wilson Affiliation Department for Health, University of Bath, Bath, United Kingdom of Great Britain and Northern Ireland Key word pilates, functional, movement, injury, running accepted Bibliography DOI Published online: 2017 Int J Sports Med 2017; 38: Georg Thieme Verlag KG Stuttgart New York ISSN Correspondence Dr. Cassie Wilson University of Bath Department for Health Claverton Down BA2 7AY, Bath United Kingdom of Great Britain and Northern Ireland Tel.: + 44/122/ , Fax: + 44/122/ c.wilson3@bath.ac.uk AbstrACt Biomechanical imbalances and inefficient functional movements are considered contributing factors to running-related injuries. Clinical Pilates uses a series of exercises focused on retraining normal movement patterns. This study investigated whether a 6-week course of Clinical Pilates improves functional movement and thereby, potentially, reduces the risk of running-related injuries associated with movement dysfunction. A modified functional movement screen was used to analyze the functional movement ability of forty runners. Forty participants completed a 6-week course of Clinical Pilates delivered by a Clinical Pilates instructor. The movement screen was carried out 3 times for each runner: 6 weeks pre-intervention (baseline), within one week pre-intervention (pre) and within one week post-intervention (post). Repeated-measures analysis of variance and post-hoc tests found significant increases in scores between baseline and post (mean ± SD; 13.4 ± 2.4 vs ± 1.7, p < 0.01) and pre and post (mean ± SD; 13.5 ± 2.5 vs ± 1.7, p < 0.01), but no significant difference between baseline and pre (p = 0.3). A 6-week course of Clinical Pilates significantly improves functional movement in recreational runners, and this may lead to a reduction in the risk of running-related injuries. Introduction Running is considered a fundamental skill and a critical requirement for almost every sporting activity [31]. However, the complex and repetitive nature of running can make athletes prone to overuse injury [1, 37]. It is estimated that a staggering 80 % of runners may experience an overuse injury sometime during their running career [39]. Running-related injuries (RRI) are primarily lower limb injuries and are considered multifactorial in etiology [1, 19, 44]. Research has speculated that RRI may be due to runners primarily exercising in the sagittal plane of motion, suggesting that lack of movement in the frontal and transverse plane causes weakness of the hip abductors and external rotators, and thus makes runners more prone to muscle imbalance [16, 22, 24]. Lower limb muscle imbalance contributes to poor alignment and altered functional movement control, which potentially increases the risk of RRI [38, 40, 42]. Understanding the biomechanical function of the lower limb during running is important in identifying potential factors that relate to injury [31]. Physiotherapists are advised to take a preventative approach to RRI and consider modifiable risk factors for each individual runner [34]. Pearce [26] argues that athletes often succumb to injury as a result of correctable biomechanical imbalances. In agreement, Cook et al. [7] advise that a biomechanical analysis of fundamental movements should be incorporated into screening in order to target injury prevention. There is evidence in athletic populations that impaired functional movement contributes to injury risk [18], and functional movement screening (FMS) is a test proposed for determining whether an athlete has the essential movements needed to participate in sports activities with a decreased risk of injury [7]. It has been suggested that lower overall scores predict individuals who are at a greater risk of injury than those with higher scores and are utilizing compensatory movement patterns during their activities [7], which may lead to RRI. Screening therefore enables physiotherapists to address these imbalances to ensure that athletes are ready to participate safely in their sport [3]. Functional deficits observed during movement screening can be corrected by performing specific exercises [3]. Interventions for preventing or managing RRI have shifted away from isolated, local- 776 Laws A et al. The Effect of Clinical Int J Sports Med 2017; 38:

73 ized muscle strengthening exercises, and moved towards proximal strengthening and retraining functional movement control [2, 19, 29 27]. Clinical Pilates is a functional, dynamic intervention, and is essentially a mind-body centering technique that emphasizes the importance of movement control from a central stable core [20, 23]. The Pilates concept was initiated by Joseph Pilates who believed that muscle imbalance, poor habitual patterns of movement and compensatory movements were the main causes of injury and could be avoided through core strengthening [43]. Pilates concepts were used and exercises were modified leading to the development of the Australian Physiotherapy & Pilates Institute (APPI) Clinical Pilates method [APPI, The APPI Pilates method (2014). In Internet: (accessed 22 nd February 2016)]. The reported benefits of these proximal strengthening exercises include improvements in strength, flexibility, co-ordination, balance and proprioception [43]. Retraining the core should address functional deficits and promote functional capacity [17]. Although there is a growing body of evidence to support the use of proximal strengthening in rehabilitation, there is no evidence exclusively promoting Clinical Pilates as a proximal training program for improving functional movement control in an attempt to reduce the risk of injury. Therefore, the aim of this study was to investigate the effects of Clinical Pilates on functional movement in healthy recreational runners. It was hypothesized that a course of Clinical Pilates would significantly improve functional movement control in recreational runners and thereby, potentially, reduce the risk of RRI associated with movement dysfunction. Methods Participants 54 recreational runners from local running clubs participated in this investigation through volunteer sampling. Participants were only included if they met specific inclusion criteria to ensure that they were uninjured recreational runners (running 15 km or more per week), with no experience of Clinical Pilates and medically fit to take part in the exercise intervention. Ethical approval was granted from the University of Bath Research Ethics Approval Committee for Health, and the study met the ethical standards outlined by Harriss and Atkinson for the IJSM [13]. Protocol In order to investigate the effect of Clinical Pilates on functional movement, a modified version of the Functional Movement Screen (FMS) [7, 8] was used to assess the runners functional ability. The modified FMS (MFMS) was undertaken 3 times for each runner; 6 weeks before the Clinical Pilates intervention commenced (MFMSbaseline), within the week prior to the intervention commencing (MFMS pre ) and within one week following the completion of the intervention (MFMS post ) ( Fig. 1). The functional movement screen used in this study was a modified, sport-specific version based on 7 functional movements selected because they are associated with the physical and functional demands of running. Included in the MFMS were 5 exercises from MFMS baseline MFMS pre MFMS post 6 week Control Phase 6 week Clinical Pilates Intervention Fig. 1 Timeline of modified functional movement screening testing. the FMS [7, 8], which is considered a useful, generic, quantifiable method of analyzing basic movement abilities [7, 8, 30]. However, it has been argued that the FMS is not sport-specific and that the tests could be enhanced by working on variations of the skills to relate more closely to an athlete s particular sport [7, 8]. The deep squat, hurdle step and in-line lunge were selected as they were considered similar functional movements to a running stride [7, 21]. The trunk stability push-up and rotary stability tests were also included because global stability is arguably a necessary element of running [8, 21]. This study modified the FMS by excluding the straight leg raise (SLR) and shoulder mobility tests, as there is little evidence that shoulder mobility or SLR flexibility impacts on running biomechanics or injury risk [21]. In replacement, 2 tests were added to measure dynamic knee valgus angle. Several studies focusing on RRI identify increased dynamic knee valgus motion as a primary cause of RRI [5, 10, 16, 32, 38, 41]. The single leg squat (SLS) and drop jump test (DJT) are validated objective measures of dynamic knee valgus motion, and these were added to make the MFMS more specific to the biomechanical imbalances associated with RRI [15, 36, 38]. Table 1 shows the fundamental scoring criteria for the MFMS. The performance of each of the 7 functional movements was scored on a scale of 0 to 3, with 0 representing an inability to perform the movement due to pain and 3 representing a performance of the movement without any compensatory movements. Table 1 reports a simplified version of the comprehensive approach that was adopted by the MFMS scorers. This comprehensive protocol was designed by the researcher based on the FMS scoring protocol by Cook et al. [7, 8] to standardize scoring and improve reliability and repeatability of the study. Laws A et al. The Effect of Clinical Int J Sports Med 2017; 38:

74 Orthopedics & Biomechanics Thieme Reliability Beardsley & Contreras [3] praise FMS for having high levels of both inter-rater and intra-rater reliability. However, as FMS was modified for this study, a pilot study was conducted to test the inter-rater and intra-rater reliability of the MFMS measure.10 of the study participants were screened and scored using the MFMS. The researcher and research assistant independently screened each volunteer following the specific MFMS scoring protocol ( Table 1). The 10 volunteers were recorded during the original MFMS, and one week later the researcher reviewed each recording and re-scored each volunteer. Subsequently, ICC was used to assess reliability, where both inter-rater (0.98, 95 % CI: ) and intra-rater (0.99, 95 % CI: ) reliability were deemed to be excellent [28]. Clinical Pilates intervention The intervention phase of the study consisted of a 6-week course of Clinical Pilates. All participants attended a one-hour Clinical Pilates class once a week for 6 weeks. Classes were designed and delivered by a senior physiotherapist/appi Clinical Pilates instructor. The APPI Clinical Pilates method is established as a world leader in Pilates training for physiotherapists and is widely used in clinical practice in the UK. Clinical Pilates classes commonly run in blocks of 6 sessions. Therefore, for clinical relevance, a 6-week, progressive program was planned and delivered. The classes were progressive over the 6 weeks and focused on proximal strengthening and lumbar/pelvic stability with controlled movements selected from the repertoire of APPI exercises. Each class comprised a warm-up followed by core exercises and then a warm-down. At the beginning of each class verbal consent was given by all participants, and participants were asked if they had sustained any injuries or had any other reason not to take part and were excluded as appropriate. The instructor used visual, verbal and tactile cues in keeping with the APPI training method to promote concentration, breathing, centering, control, precision, flowing movement, integrated isolation and routine [APPI, The APPI Pilates method (2014). In Internet: appihealthgroup.com/uploads/files/1/resources/education-brochure.pdf; (accessed 22 nd February 2016)]. Statistical analysis Statistical analysis was performed using SPSS software statistical package (SPSS, version 22). Statistical significance was set at p < A one-way repeated-measures ANOVA compared the runners scores between MFMS baseline, MFMS pre and MFMS post. When a significant effect of time was found, Bonferonni post-hoc tests were used to determine the location of the variance. Results Of the 54 participants who took part in the study, 3 withdrew during the control phase (2 were injured and one left for unexplained reasons), and 11 withdrew during the intervention phase (4 injured, 4 unable to complete all Pilates sessions and 3 unexplained). The injured participants were withdrawn as the primary aim of this study was to assess the effect of Clinical Pilates intervention on healthy recreational runners. In addition, in some cases, it was considered unsafe for the participants to complete either the Clinical Table 1 Functional tests and scoring criteria for the modified functional movement screen. Functional Scoring Criteria (total/21) Movement 1. Deep squat 3 = Runner can perform the movement without any compensatory movements 2. Single leg squat 3. Hurdle step 2 = Runner can perform movement but utilizes poor mechanics and compensatory patterns to accomplish the movement 4. In-line lunge 5. Drop jump 1 = Runner cannot perform the movement even with compensatory movements 6. Trunk stability push-up 7. Rotary stability 0 = Runner experiences pain during the movement Pilates intervention or the functional movement screen. This resulted in a total of forty participants completing the protocol. Data presented are for these forty participants only. The results highlight a significant increase in scores post Clinical Pilates intervention between MFMS baseline and MFMS post (mean ± SD; 13.4 ± 2.4 vs ± 1.7, p < 0.01) and MFMS pre and MFMS post (mean ± SD; 13.5 ± 2.5 vs ± 1.7, p < 0.01) ( Fig. 2), with mean differences in scores of 3.60 (MFMS baseline vs. MFM- S post ; 95 % Confidence Intervals [CI] 4.03 to 3.17) and 3.48 (MFMS pre vs. MFMS post ; 95 % CI 3.92 to 3.03). During the control phase of the study there was no change in mean scores (p = 0.3; mean difference = 0.13; 95 % CI 0.27 to 0.02). Discussion The aim of the current study was to evaluate whether a 6-week course of Clinical Pilates affects functional movement control in recreational runners. The main findings of the study were that following a 6-week course of Clinical Pilates runners displayed a significant improvement in their functional movement ability, confirming the study s hypothesis. The clinical implications of this study are that Clinical Pilates has the potential to reduce the risk of RRI associated with movement dysfunction. Results from the study concur with other recent studies which have shown that exercises focused on proximal strengthening result in improved function [2, 9, 12, 22, 27]. The current study found that after a course of Clinical Pilates, a significant difference in MFMS scores was observed, demonstrating the training benefits of the Clinical Pilates intervention. In practice, researchers have generally identified 14 points as the ideal cut-off point for those at greater or less risk of injury [4, 11, 18, 25]. These studies have included military, team sports and athletic (running) populations. In this study, the group mean MFMS scores at baseline and pre-intervention were 13.4 and 13.5, respectively, while the post-intervention value increased to Out of the sample of 40 participants, 30 had individual scores of 14 at baseline and pre-intervention and following the intervention only 5 retained a score of 14 (all of whom scored 14), suggesting that the Pilates had the effect of moving the participants from a state where compensatory move- 778 Laws A et al. The Effect of Clinical Int J Sports Med 2017; 38:

75 Mean MFMS scores baseline pre post Fig. 2 Mean ( ± SD) modified functional movement screen scores across the 3 sessions. * denotes significantly different to baseline and pre intervention scores (P < 0.05). * ment patterns are likely being utilized, and the risk of injury is high, to a state of lowered risk [7]. Following the intervention, it was observed that the runners exhibited enhanced hip and knee control and improved lower limb alignment during the functional tests, resulting in reduced dynamic knee valgus and increased MFMS scores. Clinical Pilates focuses on retraining the core proximal stability muscles and emphasizes the importance of normal movement control [20, 23, 43]. This focused intervention appears to have addressed the functional deficits observed prior to the intervention and contributed to the biomechanical improvements observed post intervention. Several researchers hold the notion that lower limb alignment is an important component in running mechanics [9, 19, 27, 42]. Hip stability and dynamic postural stability are consistently reported as key contributing factors in maintaining lower limb alignment [5, 10, 27, 32, 33, 38, 41]. The hip abductors and external rotators eccentrically control hip abduction and internal rotation during the stance phase of running and consequently influence the dynamic Q angle of the lower limb [5, 10, 41]. This study has shown that Clinical Pilates improves dynamic hip and knee control, and may reduce frontal plane loading of the lower limb during running, thus improving functional movement control and potentially reducing the risk of RRI. This is consistent with research findings where poor lower limb biomechanics have been shown to lead to RRI [14, 21, 38]. It has been suggested that to successfully improve functional movement and running mechanics, it may be necessary to include neuromuscular training [38, 42]. This exercise concept is supported by Wallden [35] who favors facilitating change through active rehabilitation techniques. Stimulating the local core stability muscles (inner-unit) and then building upon the more phasic musculature (outer-unit) rebuilds the muscular system in the way that nature intended [35]. It could be argued that Clinical Pilates adheres to this model by activating the central core stability muscles: transverse abdomens, pelvic floor and multifidus and then building in controlled upper and lower limb movements [APPI, The APPI Pilates method (2014). In Internet: (accessed 22 nd February 2016)]. During Clinical Pilates classes, instructors give visual, verbal and tactile cues to feedback and reinforce neuromuscular movement control. The repetition of this sequence through a 6-week course of Clinical Pilates may have potential importance for the neural properties of the inner-unit musculature. Training runners to develop both proximal stability and distal mobility and teaching them to tolerate multi-directional movements provides them with a broad set of tools for minimizing the risk of injury [35]. It could be hypothesized that there is a strong element of neuromuscular training within Clinical Pilates, which may contribute to the significant improvement in functional movement ability displayed by the runners post intervention. There are, of course, methodological limitations to this study. It could be argued that a 0 3 scale is not a very sensitive scale for the MFMS and may fail to pick up subtle changes in functional movement control, potentially resulting in unreliable scores. However, this study is focused on gross functional movement control, and so the MFMS scale was deemed appropriate. Other assessment tools to analyze functional movement control through 3D motion capture have been used in previous studies and could possibly give a more objective, measurable value [9]. However, this approach has been criticized for limitations in precision and reliability [9]. Alternatively, the Performance Matrix Movement and Performance Screen by Cornerford & Mottram [6] may have greater specificity because it breaks down each movement and analyzes 3 dimensions of an uncontrolled movement: the site, the direction and the threshold at which the uncontrolled movement occurs. Although this may be more specific and sensitive, it was deemed much more time consuming and less clinically relevant given the focus on gross functional movement control in this study. The relatively small number of participants and the fact this was not a randomized control trial are also limitations or weaknesses of the study. To the authors knowledge, this study is the first to investigate the effect of Clinical Pilates on recreational runners and as such provides a good foundation for future studies into the use of Pilates for pre-habilitation or preventative exercise. Future studies should therefore look to investigate the effect of Pilates in reducing running-related injuries by adopting a longitudinal study design. The purpose of this study was to examine the effects of Clinical Pilates on functional movement in uninjured recreational runners. Results show that a 6-week course of Clinical Pilates significantly improves functional movement in recreational runners and potentially reduces the risk of RRI associated with movement dysfunction. Conflict of interest The author have no conflict of interest to declare. References [1] Agresta C, Slobodinsky M, Tucker C. Functional movement screen normative values for healthy distance runners. Int J Sports Med 2014; 35: [2] Baldon RDM, Serrrao FV, Silva R, Piva SR. Effects of functional stabilisation training on pain, function, and lower extremity biomechanics in women with patellofemoral pain: A randomized clinical trial. J Orthop Sports Phys Ther 2014; 44: Laws A et al. The Effect of Clinical Int J Sports Med 2017; 38:

76 Orthopedics & Biomechanics Thieme [3] Beardsley C, Contreras B. The functional movement screen: A review. Strength Cond J 2014; 36: [4] Chorba RS, Chorba DJ, Bouillion LE, Oevrmyer CA, Landis JA. Use of functional movement screen tool to determine injury risk in female collegiate athletes. N Am J Sports Phys Ther 2010; 5: [5] Clark NTM, Bourque RD, Schilling J. Treatment of patellofemoral pain syndrome in a track athlete. Int J Athl Ther Train 2014; 19: [6] Cornerford M, Mottram S. Kinetic control: the management of uncontrolled movement. Southampton: Elsevier; 2012 [7] Cook G, Burton L, Hoogenboom BJ, Voight M. Functional movement screening: The use of fundamental movements as an assessment of function, part 1. Int J Sports Phys Ther 2014; 9: [8] Cook G, Burton L, Hoogenboom BJ, Voight M. Functional movement screening: The use of fundamental movements as an assessment of function, part 2. Int J Sports Phys Ther 2014; 9: [9] Earl JE, Hoch AZ. A proximal strengthening program improves pain, function, and biomechanics in women with patellofemoral pain syndrome. Am J Sports Med 2011; 39: [10] Ferber R, Kendall KD, Farr L. Changes in knee biomechanics after a hip-abductor strengthening protocol for runners with patellofemoral pain syndrome. J Athl Train 2011; 46: [11] Garrison M, Westrick R, Johnson MR, Benenson J. Association between the functional movement screen and injury development in college athletes. Int J Sports Phys Ther 2015; 10: [12] Gildenhuys GM, Fourie M, Shaw BS, Toriola AL, Witthuhn J. Evaluation of pilates training on agility, functional mobility and cardiorespiratory fitness in elderly women. Afr J Phys Health Educ Recreat Dance 2013; 19: [13] Harriss DJ, Atkinson G. Ethical standards in sport and exercise science research: 2016 update. Int J Sports Med 2015; 36: [14] Herrington L. The effect of hip abductor muscle fatigue on frontal plane knee projection angle during step landing. Int J Athl Ther Train 2014; 19: [15] Hewett TE, Myer GD, Ford KR, Heidt RS, Colosimo AJ, Mclean SG, Van den Bogert AJ, Paterno MV, Succop P. Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes: A prospective study. Am J Sports Med 2005; 33: [16] Homan KJ, Norcross MF, Goerger BM, Prentice WE, Blackburn JT. The influence of hip strength on gluteal activity and lower extremity kinematics. J Electromyogr Kinesiol 2013; 23: [17] Key J. The core : understanding it, and retraining its dysfunction. J Bodyw Mov Ther 2012; 17: [18] Kiesel K, Plisky PJ, Voight ML. Can serious injury in professional footballers be predicted by a preseason functional movement screen? N Am J Sports Phys Ther 2007; 2: [19] Lack S. Are the biomechanics of your foot and hip affecting your knee joint? SportEX Medicine 2013; 55: [20] Lange C, Unnithan V, Larkam E, Latte PM. Maximizing the benefits of pilates-inspired exercise for learning functional motor skills. J Bodyw Mov Ther 2000; 4: [21] Louden JK, Parkerson-Mitchell AJ, Hildebrand LD, Teague C. Functional movement screen scores in a group of running athletes. J Strength Cond Res 2014; 28: [22] Lugo-Larcheveque N, Pescatello LS, Dugdale TW, Veltri DM, Roberts WO. Management of lower limb extremity malalignment during running with neuromuscular retraining of the proximal stabilisers. Curr Sports Med Rep 2006; 5: [23] McNeil W. Are movement screens relevant for pilates, circus or dance? J Bodyw Mov Ther 2014; 18: [24] Noehren B, Pohl MB, Sanchez Z, Cunningham T, Latterman C. Proximal and distal kinematics in female runners with patella femoral pain. Clin Biomech 2012; 27: [25] O Connor FG, Deuster PA, Davis J, Pappas CG, Knapik JJ. Functional movement screening: Predicting injuries in officer candidates. Med Sci Sports Exerc 2011; 43: [26] Pearce PZ. Prehabilitation: Preparing young athletes for sports. Curr Sports Med Rep 2006; 5: [27] Peters JSJ, Tyson NL. Proximal exercises are effective in training patellofemoral pain syndrome: A systematic review. Int J Sports Phys Ther 2013; 8: [28] Portney LG, Watkins MP. Foundations of clinical research: Applications to practice. 3rd ed. 2009Upper Saddle River, NJ: Prentice-Hall; [29] Sheerin KR, Hume PA, Whatman C. Effect of a lower limb functional lower limb exercise programme aimed at minimising knee valgus angle on running kinematics in young athletes. Phys Ther Sport 2012; 13: [30] Shultz R, Anderson SC, Matheson GO, Marcello B, Besier T. Test-retest and interrater reliability of the functional movement screen. J Athl Train 2013; 48: [31] Snache AG, Dorn TW, Williams GP, Brown NAT, Pandy MG. Lower-limb muscular strategies for increasing running speed. J Orthop Sports Phys Ther 2014; 44: [32] Taylor-Haas JA, Hugentobler JA, DiCesare CA, Lucas KCH, Bates NA, Myer GD, Ford KR. Reduced hip strength is associated with increased hip motion during running in young adult and adolescent male long-distance runners. Int J Sports Phys Ther 2014; 9: [33] Teng HL, Powers CM. Sagittal plane trunk posture influences patellofemoral joint stress during running. J Orthop Sports Phys Ther 2014; 44: [34] Van Gent R, Siem D, Van Middelkoop M, Van Os A, Bierma-Zeinstra M, Koes B. Incidence and determinants of lower extremity running injuries in long distance runners: a systematic review. Br J Sports Med 2007; 41: [35] Wallden M. Facilitating change through active rehabilitation techniques. J Bodyw Mov Ther 2013; 17: [36] Whatman C, Hume P, Hing W. Kinematics during lower extremity functional screening tests in young athletes Are they reliable and valid? Phys Ther Sport 2013; 14: [37] Wilk BR, Nau S, Valero B. Physical therapy management of running injuries using an evidence based functional approach. Am Med Athl Assoc J 2009; 22: 5 7 [38] Willy RW, Davis IS. The effect of a hip strengthening program on mechanics during running and during a single-leg squat. J Orthop Sports Phys Ther 2011; 41: [39] Wilson JD, Kernozek TW, Arndt RL, Reznichek DA, Straker JS. Gluteal muscle activation during running in females with and without patellofemoral pain syndrome. Clin Biomech 2011; 26: [40] Wilson JD, Petrowitz I, Butler RJ, Kernozek TW. Male and female gluteal muscle activity and lower extremity kinematics during running. Clin Biomech 2012; 27: [41] Woods JM, Panas N. Patellofemoral pain syndrome. American Fitness 2014; 32: [42] Wouters I, Almonroeder T, DeJarlais B, Lack A, Wilson JD, Kernozek TW. Effects of a movement training program on hip and knee joint frontal plane running mechanics. Int J Sports Phys Ther 2012; 7: [43] Yamato TP, Maher CG, Saragiotto BT, Hancock MJ, Ostelo RWJG, Cabral CMN, Menezes Costa LC, Costa LOP. Pilates for low back pain review. Cochrane Database Syst Rev 2015 Issue 7 [44] Yeung S, Yeung E, Gillespie L. Interventions for preventing lower limb soft tissue running injuries (review). Cochrane Database Syst Rev 2011 Issue Laws A et al. The Effect of Clinical Int J Sports Med 2017; 38:

77 Scand J Med Sci Sports 2014: : doi: /sms John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Muscle functional MRI analysis of trunk muscle recruitment during extension exercises in asymptomatic individuals E. M. D. De Ridder 1, J. O. Van Oosterwijck 1, A. Vleeming 2, G. G. Vanderstraeten 1, L. A. Danneels 1 1 Department of Rehabilitation Sciences and Physical Therapy, Faculty of Medicine and Health Sciences, Ghent University, Belgium, 2 Department of Anatomy, University of New England College of Osteopathic Medicine, Biddeford, Maine, USA Corresponding author: Lieven A. Danneels, Ghent University Hospital, Department of Rehabilitation Sciences and Physical Therapy, De Pintelaan 185, 3B3, B-9000 Ghent, Belgium. Tel: , Fax: , Lieven.Danneels@Ugent.be Accepted for publication 13 January 2014 The present study examined the activity levels of the thoracic and lumbar extensor muscles during different extension exercise modalities in healthy individuals. Therefore, 14 subjects performed four different types of extension exercises in prone position: dynamic trunk extension, dynamic static trunk extension, dynamic leg extension, and dynamic static leg extension. Pre- and post-exercise muscle functional magnetic resonance imaging scans from the latissimus dorsi, the thoracic and lumbar parts of the longissimus, iliocostalis, and multifidus were performed. Differences in water relaxation values (T2-relaxation) before and after exercise were calculated (T2-shift) as a measure of muscle activity and compared between extension modalities. Linear mixedmodel analysis revealed higher lumbar extensor activity during trunk extension compared with leg extension (T2- shift of 5.01 ms and 3.55 ms, respectively) and during the dynamic static exercise performance compared with the dynamic exercise performance (T2-shift of 4.77 ms and 3.55 ms, respectively). No significant differences in the thoracic extensor activity between the exercises could be demonstrated. During all extension exercises, the latissimus dorsi was the least activated compared with the paraspinal muscles. While all extension exercises are equivalent effective to train the thoracic muscles, trunk extension exercises performed in a dynamic static way are the most appropriate to enhance lumbar muscle strength. Up to 40% of the athlete population is affected by low back pain (LBP), which is the most common cause of lost playing time in professional sports (Bono, 2004). Sports-induced muscle imbalance within the trunk or hip muscles seems to be related to LBP (Renkawitz et al., 2006, 2008). Furthermore, there is considerable evidence suggesting that trunk muscle strength and endurance do not only play a key role in the prevention and treatment of LBP (Holmstrom et al., 1992; Luoto et al., 1995; Kuukkanen & Malkia, 1996; Moffroid, 1997; Holm & Dickinson, 2001; Mannion et al., 2001a, b; Kell & Asmundson, 2009), but are also related to sports performance (Smith et al., 2008; McGill, 2010). As decreased muscle performance implicates a lower fatigue threshold, the precision and control of movements is reduced, which results into a poorer sports performance (Durall et al., 2009). This implies that optimal functioning of the trunk extensors is beneficial for sports performance in athletes. Athletic training and LBP rehabilitation programs exist of different exercise regimes, which are performed to enhance trunk extensor muscle strength, endurance and spinal control, and will lead to decreased levels of pain and disability (Henchoz & So, 2008; Franca et al., 2010, 2012; Henchoz et al., 2010). While spinal control is optimized during stabilization and mobilization exercises, muscle strength and endurance are often enhanced by extension exercises of the trunk and/or the legs. These extension exercises are performed in a dynamic or dynamic static way, and specifically strengthen the thoracic and lumbar extensors. However, at present it is not clear to which extent the contraction modality (dynamic vs dynamic static and trunk extension vs leg extension) influences the activation of the thoracic and lumbar extensors. This is due to the scarce literature regarding this topic and the lacunas in existing studies. The few studies that exist have either compared muscle activation patterns between dynamic and dynamic static contractions or between trunk and bilateral leg extension exercises, and have reported conflicting results. Although these studies have used electromyography (EMG), more recently muscle functional magnetic resonance imaging (mfmri) has been used to determine the amount of muscle activity during exercise. Its main advantage compared with surface EMG (semg) is its superior spatial resolution, imaging deep and superficial muscles simultaneously at multiple levels and at both sides of the spine (Adams et al., 1992; Mayer et al., 2005; Cagnie et al., 1

78 De Ridder et al. 2009; Dickx et al., 2010a, b). Although fine-wire EMG can also be used to investigate the activity of deep muscles, it only provides an idea on electrical activity of a few motor units and is invasive in nature. Whereas EMG has been widely used to investigate thoracic and lumbar muscle activation, and lumbar muscle work has frequently been investigated during trunk extension, studies in which both thoracic and lumbar muscle activity during trunk and leg extension exercises is measured with mfmri are nonexistent. Therefore, this study was the first to evaluate simultaneously the amount of activity of the thoracic and lumbar muscles during standardized extension exercises with mfmri in healthy subjects. The present study examined: (a) the influence of different exercise modalities, i.e., trunk or leg extension, on the amount of the thoracic and lumbar extensor muscle activity by evaluating the T2-shift; and (b) whether the findings were influenced by the contraction modality of the exercise, i.e., dynamic or dynamic static contraction. It was hypothesized that the thoracic extensors, which are conjoining the thorax with the pelvic via long tendons, will be recruited more during trunk extension compared with leg extension, whereas leg extension may generate more activity of the lumbar extensors, which directly attach onto the lumbar vertebrae. Regarding the contraction modality, it was hypothesized that both the thoracic and the lumbar extensors will show bigger T2-shifts during the dynamic static exercise performance compared with the dynamic exercises. and two during the third session. The exercise sequence was randomized and determined by lottery. MRI scanning was performed before and immediately after performing each exercise modality. To prevent the potential influence of muscle fatigue, a rest period of 40 min was provided between the two extension exercise modalities. There were at least 7 days between the second and third session. The study protocol, information leaflet, and informed consent were approved by the local ethics committee. Extension exercises To perform the trunk extension exercises, subjects were installed in prone position on a variable angle chair, with their upper body in a 45 of trunk flexion. The superior border of the anterior iliac (SIAS) was positioned on the edge of the table and the ankles were strapped to the table. Hands were placed on the opposite shoulder (Fig. 1). The dynamic trunk extension implied that subjects raised their trunk to the horizontal in 2 s and returned to the start position in 2 s. During the dynamic static trunk extension, the trunk was raised to the horizontal in 2 s, the horizontal position was maintained for 5 s, after which the subject returned to 45 flexion in 2 s. To perform the leg extension exercises, subjects were installed in prone position with their lower body positioned at 45 flexion. The upper body was strapped at the level of the angulus inferior of the scapulae, hands were positioned under the forehead (Fig. 2). The dynamic leg extension consisted of a 2-s raise of the legs till the horizontal, followed by a period of 2 s to return to the start position. During the dynamic static leg extension, the legs were extended horizontally in 2 s, held in that position during 5 s, and lowered in 2 s. To reach the horizontal position, tactile feedback was given by a rope between the two vertical stands. A metronome (60 beats/ min) was used to ensure appropriate timing of the different movements. A set of 20 repetitions of each exercise modality was Materials and methods Subjects Fourteen subjects, eight men and six women, participated in this study. Subjects were characterized by a mean age of ± 3.19 years, mean height of ± 6.4 cm, mean weight of ± 12.5 kg, and body mass index (BMI) of ± 3.1 kg/ m 2, indicating normal weight. Study design An observational study to evaluate the recruitment of thoracic and lumbar extensor during different extension exercises was conducted on 14 healthy individuals. Subjects were recruited through adverts, which were spread among personal contacts of the researchers, the staff of Ghent University and Ghent University Hospital. If subjects experienced back pain recently, had a medical consultation concerning LBP in the past year, reported previous back surgery or spinal deformities, or when MRI was contradicted, they were not eligible for study participation. Each subject attended three sessions. The first session included a consultation of the information leaflet and signing of the informed consent, followed by anthropometric measurements and determination of the one repetition maximum (1-RM) for each extension exercise. Four different modalities of extension exercises (i.e., dynamic trunk extension, dynamic static trunk extension, dynamic leg extension, and dynamic static leg extension) were performed during the second and third sessions. Two exercise modalities were performed during the second session Fig. 1. Position trunk extension exercises. 2

79 Table 1. Individual load adjustments of the extension exercises at 60% 1-RM mfmri of recruitment of trunk muscles during extension Subjects Dynamic trunk extension Dynamic static trunk extension Dynamic leg extension Dynamic static leg extension Load Repetitions Load Repetitions Load Repetitions Load Repetitions Mean (kg) The resistant (positive) or assistant (negative) weight, relative to the weight of the trunk or legs and the maximum number of repetitions achieved, necessary to perform each extension modality at an exercise load of 60% of one repetition maximum (1-RM) are presented per subject. Weights are expressed in kg and are accurate up to 0.5 kg. Bold indicates mean adjusted weights (kg) to adjust the exercise load to 60% 1-RM. exercise weight. Each 1-RM test was executed in the same position, over the same range of motion, and with an identical timing as during the respective exercise modality. Afterwards, the exercise load (kg) corresponding to 60% of 1-RM, was estimated using the Holten diagram. This diagram describes the relation between the performed number of repetitions and the exercise intensity (Danneels et al., 2001b). Calculation of the individual exercise weight occurred identically as described by Dickx et al. (2010a), using the following formula: upper/lower body weight (kg) exercise load (60% 1-RM)/ exercise load determined on testing day (Holten diagram). The weight of the upper body is calculated as 70% of the total body weight and of the lower body as 30% of the total body weight. The total body weight was determined on a body scale during the first session. To adjust the exercise weight, the body was assisted via a load pulley system or extra weights were added (Table 1). Extra weight was added to the trunk by holding weight pockets against the chest by crossing their arms. In order to adjust the leg extension, exercise weight cuffs were tied around both thighs. Fig. 2. Position leg extension exercises. performed continuously at an exercise intensity of 60% of 1-RM. Thus to complete the dynamic static exercise, 180 s (20 repetitions 9 s) were needed, while the dynamic exercise condition was finalized in 80 s (20 repetitions 4 s). Determination of the exercise intensity (60% 1-RM) Minimum of 3 days before the second session took place, the individual 1-RM was indirectly determined by registering the maximum number repetitions participants could perform of each exercise modality with the weight of their upper/lower body as the Muscle functional MRI A 3-Tesla Trio Trim scanner (Siemens Erlangen, Germany) was used to assess changes in the relaxation time of muscle water (T2-relaxation time) as a result of muscle work during the extension exercises. The amount of muscle activity can be assessed by quantifying shifts in T2-relaxation times and is expressed as the T2-shift. (Meyer & Prior, 2000). To ensure a neutral spine position, subjects were placed symmetrically in supine position, with their head first. Two coils were used to partly cover the thoracic and lumbar spine: ventral, a flexible six-element body matrix coil centered on the belly button, and dorsal, a standard phased-array spine coil, was positioned. Two transversal slices corresponding the lower endplate of T12 and the lower endplate of L4 (Danneels et al., 2001a) were positioned as horizontal as possible on a sagittal view. A spin-echo multi-contrast sequence (Semc) was used for the acquisition of T2-weighted images. The following parameters were applied: repetition time 1000 ms, mm 2 matrix, 256 mm field of view (FOV), slice thickness 5 mm. Total scan time was 6 min 15 s. The images were obtained after a period of 15 min of prone lying (rest 3

80 De Ridder et al. Fig. 3. Image at lower endplate level T12. 1, latissimus dorsi; 2, longissimus thoracis pars thoracic; 3, iliocostalis lumborum pars thoracic. Statistical analysis SPSS 19.0 (IBM Corporation, Somers, New York, USA) was used to carry out statistical analyses. At first, baseline and post-exercise T2-values of the thoracic and lumbar muscles of the left and the right side were averaged, due to the symmetry of the exercises and the lack of significant side differences in T2-values (P < 0.05). In addition, the symmetry of the exercise was monitored by semg, the results which are published elsewhere confirmed the lack of significant side differences (De Ridder et al., 2013). Subsequently, descriptive statistics [means and standard deviation (SD)] were calculated for the anthropometric group characteristics and T2-values. To investigate the T2-shift values (i.e., the difference in T2-values between post-exercise and baseline) of the back muscles between different exercise modalities, a linear mixedmodel analysis was conducted. The following main factors were used: muscle (LD, IT, IL, LT, LL, MF), extension modality (trunk vs leg extension), and contraction type (dynamic vs dynamic static). In case one of the main factors was significant, separate mixed-models were conducted. Post-hoc comparisons were made when required and adjusted using a Bonferroni correction. Statistical significance for all tests was set at P 0.05 (CI 95%). Fig. 4. Image at lower endplate level L4. 1, multifidus; 2, longissimus thoracis pars lumborum; 3, iliocostalis lumborum pars lumborum. T2), and immediately following the exercises (exercise T2). The time span between the end of the exercise and the beginning of the scan ranged from 1 min 55 s to 2 min 30 s. Scanning was performed before and immediately after performance of every extension modality. Between two different exercise modalities, subjects rested in prone lying over a period of 40 min to ensure that the T2-values were able to decrease to baseline values. The MRI images were analyzed using Image J (Java-based version of the public domain NIH Image Software, Research Services Branch, National Institutes of Health, Washington, USA). Using the MRI analysis calculator plug-in a T2-value (in milliseconds per voxel) was calculated out of 16 echoes. Subsequently, the region of interest (ROI) was determined on all images by drawing the outlines of all muscles, avoiding visual fat, blood vessels, and connective tissue. This method has proven to be reliable in previous work (Danneels et al., 2000; Dickx et al., 2010b; D Hooge et al., 2013). At T12, the latissimus dorsi (LD) and the thoracic parts of the longissimus (LT) and iliocostalis (IT) were analyzed bilaterally (Fig. 3). At L4 the multifidus (MF) and the lumbar parts of the longissimus (LL) and iliocostalis (IL) were analyzed. Figure 4 presents the MF, LL, and IL on a mfmri image. Finally, the mean T2-value was derived for each ROI and used for further analysis. Results Recruitment of the trunk muscles The mixed-model analysis, which was used to examine the T2-shift in thoracic and lumbar muscles between and within the different extension exercise modalities, showed no significant interaction effects between the main factors, whereas the main factors muscle (P < 0.001) and extension modality (P = 0.045) had a significant effect on the T2-shift. No main effect for contraction type (P = 0.193) could be established. A post-hoc comparison between the different trunk extensors demonstrated that during all exercise modalities the LD was recruited significantly less compared with all other trunk extensors (Fig. 5). No differences between the other muscles could be established. Moreover, post-hoc data revealed that in general the mean T2-shift of all trunk muscles (mean of the sum of all T2-shift values) was significantly higher during trunk extension than during leg extension, regardless of the type of contraction (P = 0.045). Due to the significance of the factor muscle, reflecting possible anatomical and functional differences between the thoracic and lumbar muscles, two new mixed-models were conducted using the same factors, but one including the thoracic muscles (IT and LT) and one the lumbar muscles (IL, LL, and MF). Recruitment of the thoracic muscles An analysis of the thoracic muscles only was performed and showed no. three-way or two-way interaction effects between the main factors. Furthermore, the T2-shift of the IT was not significantly different to the T2-shift of the LT during the extension exercises (muscle P = 0.574). Moreover, neither the extension modality (trunk or leg extension), nor the contraction type 4

81 mfmri of recruitment of trunk muscles during extension Mean T2-shift Dynamic-static Dynamic Dynamic-static Dynamic leg trunk trunk leg Extension exercise modality (dynamic vs dynamic static) had a significant effect on the shift in T2-values of the thoracic muscles (P = and P = 0.591, respectively). The mean T2-values of the thoracic muscles during all exercise modalities are presented in Table 2. Recruitment of the lumbar muscles The analysis, which included solely the lumbar muscles, showed no three-way or two-way interaction effect between the main factors; however, a clear significant main effect of extension modality and contraction type was demonstrated. Lumbar muscles were recruited at a higher degree during the trunk extension exercises (T2-shift of 5.01 ms) compared with the leg extension exercises (P 0.001; T2-shift of 3.55 ms). Furthermore, the dynamic static extension exercises demanded more lumbar muscle work than the dynamic extension exercises (P 0.014; T2-shift 4.77 vs 3.78 ms). The mean T2-shift of the lumbar muscles during the different exercise modalities are displayed in Figure 5. Discussion Muscle LD LT IT LL IL MF Fig. 5. Mean T2-shift of the trunk muscles in response to each extension exercise modality. LD, latissimus dorsi; LT, longissimus thoracis pars thoracic; IT, iliocostalis lumborum pars thoracic; LL, longissimus thoracis pars lumborum; IL, iliocostalis lumborum pars lumborum; MF, multifidus; *, significant difference in mean T2-shift among the trunk muscles at P 0.05 level. (Bars: ± 1 standard deviation). The present study examined whether the amount of activity (estimated by the shift in T2-values) of the trunk extensor muscle extensors is influenced by different modalities of extension exercises, i.e., which Table 2. Shift of the T2 values of the trunk extensor muscles in response to each extension exercise modality Muscles Dynamic trunk extension Dynamic static trunk extension Dynamic leg extension Dynamic static leg extension T2 pre T2 post T2-shift T2 pre T2 post T2-shift T2 pre T2 post T2-shift T2 pre T2 post T2-shift LD ± ± ± ± ± ± ± ± ± ± ± ± 2.64 LT ± ± ± ± ± ± ± ± ± ± ± ± 5.89 IT ± ± ± ± ± ± ± ± ± ± ± ± 2.76 LL ± ± ± ± ± ± ± ± ± ± ± ± 1.34 IL ± ± ± ± ± ± ± ± ± ± ± ± 2.37 MF ± ± ± ± ± ± ± ± ± ± ± ± 0.96 Values in the table present the mean T2-value (in ms) and standard deviation (± SD) at rest (pre), after performing the exercise modality (post), and the difference between these two conditions (T2-shift) of the trunk muscles. IL, iliocostalis lumborum pars lumborum; IT, iliocostalis lumborum pars thoracis; LD, latissimus dorsi; LL, longissimus thoracis pars lumborum; LT, longissimus thoracis pars thoracic; MF, multifidus. 5

82 De Ridder et al. body part is extended (trunk or legs) and in which way the extension exercise is performed (dynamic or dynamic static). The difference in mean T2-shift of the thoracic and lumbar extensors, between the various exercise conditions, supports the hypothesis that the activity level of the back extensors is influenced by the manner in which an extension exercise is performed. These results implicate that although the exercise load of the different exercises was identical, the T2-shift of the lumbar muscles was higher during trunk extension exercises compared with leg extension exercises. The higher shift implies enhanced levels of (metabolic) activity within the lumbar muscles when performing a trunk extension compared with a leg extension exercise, which is the result of more activity of the lumbar muscles. Although previous studies already showed that the thoracic and lumbar extensors are activated during extension exercises (Mayer et al., 1999, 2002; Plamondon et al., 1999, 2002; Clark et al., 2002), to our knowledge this is the first study to compare differences in the activity level of thoracic and lumbar extensor muscles, and between different trunk and leg extension exercises. In addition, this is the first study on this topic using mfmri. The results confirm our previous findings, which were obtained using semg and using an identical exercise protocol on the same study population (De Ridder et al., 2013). The increased lumbar muscle activity during trunk extension in the present study, is inconsistent with the results of Plamondon et al. (2002). Those authors reported greater levels of lumbar extensor spinae muscles (LES) during performance of a dynamic prone-leg extension compared with a dynamic pronetrunk extension. However, we need to consider that Plamondon et al. (2002) used a different technique to evaluate lumbar muscle usage. While they used semg measures, which reflect the real-time neural muscle changes upon exercises, in the current study mfmri was used, which displays the acute activityinduced prolongation of T2-relaxation times of muscle water. Furthermore, inequalities in starting angle exist between the two studies, which could have a significant effect on the muscle force production, due to differences in muscle length (Mannion & Dolan, 1996). The present study was unable to demonstrate differences among the lumbar muscles, which suggest a similar action of these muscles during the extension exercises. This finding is in line with Danneels et al. (2002), who described that the MF and IL have a similar function during trunk movements. However, Ng et al. (1997, 1999) found that the MF was more activated than the IL or LL during isometric trunk and leg holding, respectively. The discrepancy between these results could be explained by dissimilarities in exercise modality. Ng et al. (1997, 1999) studied isometric exercises, while the current study examined dynamic and dynamic static contractions. As it has been demonstrated that the MF has a more stabilizing function compared with the IL and LL (Wilke et al., 1995), isometric contractions could be more sensitive to little variations in function. There are few other studies that used MRI to examine differences in lumbar muscle recruitment during trunk extension in a healthy population (Mayer et al., 2005; Dickx et al., 2010b). Mayer et al. (2005) demonstrated higher activity of the MF compared with the LES during dynamic trunk extension, and showed that there was a relationship between lumbar muscle contribution and exercise intensity. This relationship could explain previous described differences in study findings. More specifically, when a relative high-load exercise is performed, as was the case in the present study; the different muscles are recruited in a homogeneous way in order to obtain and maintain a relatively high force output. As other studies did not adjust or report the exact exercise intensity, we assume that these low-load exercises are more sensitive to objectify subtle differences in muscle activity. Further investigation in low-load conditions is recommended to verify this assumption. Besides the influence of the extended body part, we also showed clear differences in lumbar muscle usage between a dynamic and dynamic static performance of the extension exercises. Although the exercise intensity was identical in both cases, performing the extension exercises in a dynamic static way caused a higher T2-shift of the lumbar muscles compared with a dynamic contraction. The acute higher metabolic reactions in response to the dynamic static performance could support the findings of Danneels et al. (2001a), who studied the long-term effects of two training programs on the cross-sectional area (CSA) of the paravertebral muscles. They demonstrated that an increase in the CSA of the paravertebral muscles occurred after dynamic static muscle training, whereas dynamic training did not affect CSA. On the long term, the higher metabolic cost of dynamic static exercises may have triggered the volume-growing effect within the lumbar muscles, which resulted in a higher CSA. Because the contraction type (and the duration) of both modalities (dynamic vs dynamic static) differed, both conditions were tested separately in advance and the load (number of kilograms of resistance) was individually adapted to express 60% of 1-RM. Following this procedure the exercise load, which was applied in the dynamic static condition, was systematically lower than the load used in the dynamic version. This ensured that the exercise intensity of both exercises was identical. But although the perceived intensity was identical, it is apparent that the nature of the muscle 6

83 mfmri of recruitment of trunk muscles during extension contraction is different. We can assume that during a static muscle contraction, the arterial pressure is higher and the blood flow is lower compared with a dynamic muscle contraction (Masuda et al., 1999; Vanderthommen et al., 2003; Arimoto et al., 2005). In this way, the added static element during the dynamic static exercise condition could have resulted in an increased lactic acid accumulation compared with the dynamic condition. However, future exercise studies, in which mfmri is combined with lactate determination, are necessary to verify these assumptions. The higher lactate accumulation can support the higher T2-shift in this condition. Moreover, previous studies proved that changes in T2 depend on the duration of the exercise load (Jenner et al., 1994), resulting in a higher T2-value when the duration of the exercise is increased. We expected that the work of the thoracic muscles during trunk extension would have been higher compared with leg extension, and may be influenced by thecontraction type. The lack of difference in thoracicmuscle work between the exercise modalities in the present study did not support this hypothesis. The present study demonstrated that during all extension exercises the LD was less active compared with the other trunk muscles, which is in agreement with other studies and can be clarified by the main function of the LD, namely, arm movement. The lower activity levels of the LD are confirmed by our previous findings using semg in a similar protocol (De Ridder et al., 2013), and in line with the findings of Coorevits et al. (2008) who showed that the LD was the least fatigued muscle of the trunk and hip extensor muscles following isometric trunk extension. This study was limited to a small sample of healthy subjects. Although most MRI studies have a limited number of subjects, it would be useful to investigate a larger and more varied population. Furthermore, care should be taken with generalization of the recent study results. The current study did not investigate the trunk muscle activity patterns during a pure static exercise modality or exercises at lower intensities (i.e., stabilization exercises). It has been recommended that in the initial stage of spine-strengthening programs, subjects should be instructed to become aware of motor patterns and to recruit muscles in isolation at lower exercise intensities (Hibbs et al., 2008), but that an essential requirement is to progress to more functional and dynamic modalities (Akuthota & Nadler, 2004). Especially from an athletic perspective, extension exercises which comprise of a dynamic component and a high exercise load are usually preferred for strengthening the trunk muscles (Hibbs et al., 2008). Future studies may reveal which static exercise modality is the most appropriate to target the lumbar extensors during the early stages of LBP rehabilitation. Furthermore, to examine whether LBP or excessive sports participation will cause alterations in the recruitment of the trunk extensor muscles, it would be useful for future studies to examine these populations. If impaired recruitment patterns exist, it would be desirable to examine the efficacy of various extension exercise modalities in order to optimize the function of the trunk extensors. Our results demonstrated that during extension exercises the level of activity (shift in T2) of the lumbar extensors is influenced by the modality of the extension exercise, whereas the thoracic extensor activity is not. The highest activity of the lumbar muscles was found during the dynamic static trunk extension. Therefore, due to the need of a metabolic stimulus to enhance muscle strength, the dynamic static exercise performance and the trunk extension exercises physiologically seem the most appropriate to train the lumbar muscles in clinical practice, although this will need to be established by future studies. In order to strengthen the thoracic muscles, none of the exercise conditions seems to be superior. During all extension exercises, the LD was less active then the other trunk extensor muscles. Thus in case it is desirable to strengthen this muscle in addition to the trunk extensors, the previous described extensions exercises need to be complemented by more appropriate exercise which target this muscle. Perspective Optimal functioning of the trunk extensors is beneficial for sports performance in athletes and plays a key role in the prevention and treatment of LBP. In training programs, endurance and strength of the lumbar muscles are often enhanced using different extension exercise modalities. The present study examined how activity patterns of trunk muscles differ between several modalities, and which type of modality achieves the highest level of activation of the lumbar extensors. It was shown that the type of extension exercise did not influence the activity levels of the thoracic extensors, but does indeed determine the activity levels of the lumbar muscles. More specifically, it was shown that when trunk and leg extensions exercises comprise a dynamic and static component they will result in higher lumbar muscle activation than when performed pure dynamic. Furthermore, trunk extension exercises will result in higher levels of lumbar muscle activation compared with leg extension exercises. In conclusion, exercise programs which wish to efficiently optimize endurance and strength of the lumbar muscles should give preference to dynamic static trunk extension exercises. 7

84 De Ridder et al. Key words: Imaging, trunk extensor muscles, prone extension exercises, posterior muscle chain, spine, rehabilitation. Acknowledgements Jessica Van Oosterwijck is a postdoctoral research fellow funded by the Special Research Fund of Ghent University. References Adams GR, Duvoisin MR, Dudley GA. Magnetic-resonance-imaging and electromyography as indexes of muscle function. J Appl Physiol 1992: 73: Akuthota V, Nadler SF. Core strengthening. Arch Phys Med Rehabil 2004: 85: S86 S92. Arimoto M, Kijima A, Muramatsu S. Cardiorespiratory response to dynamic and static leg press exercise in humans. J Physiol Anthropol Appl Human Sci 2005: 24: Bono CM. Low-back pain in athletes. J Bone Joint Surg Am 2004: 86-A: Cagnie B, Barbe T, Vandemaele P, Achten E, Cambier D, Danneels L. MRI analysis of muscle/fat index of the superficial and deep neck muscles in an asymptomatic cohort. Eur Spine J 2009: 18: Clark BC, Manini TM, Mayer JM, Ploutz-Snyder LL, Graves JE. Electromyographic activity of the lumbar and hip extensors during dynamic trunk extension exercise. Arch Phys Med Rehabil 2002: 83: Coorevits P, Danneels L, Cambier D, Ramon H, Vanderstraeten G. Assessment of the validity of the Biering-Sorensen test for measuring back muscle fatigue based on EMG median frequency characteristics of back and hip muscles. J Electromyogr Kinesiol 2008: 18(6): Danneels LA, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE, Bourgois J, De Cuyper HJ. The effects of three different training modalities on the cross-sectional area of the paravertebral muscles. Scand J Med Sci Sports 2001a: 11: Danneels LA, Coorevits PL, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE, De CH. Differences in electromyographic activity in the multifidus muscle and the iliocostalis lumborum between healthy subjects and patients with sub-acute and chronic low back pain. Eur Spine J 2002: 11 (1): Danneels LA, Vanderstraeten GG, Cambier DC, Witrouw EE, De Cuyper HJ. CT imaging of trunk muscles in chronic low back pain 8 patients and healthy control subjects. Eur Spine J 2000: 9.4 (Aug): Danneels LA, Vanderstraeten GG, Cambier DC, Witvrouw EE, Bourgois J, Dankaerts W, De Cuyper HJ. Effects of three different training modalities on the cross sectional area of the lumbar multifidus muscle in patients with chronic low back pain. Br J Sports Med 2001b: 35: De Ridder EM, Van Oosterwijck JO, Vleeming A, Vanderstraeten GG, Danneels LA. Posterior muscle chain activity during various extension exercises: an observational study. BMC Musculoskelet Disord 2013: 14: 204. D Hooge R, Cagnie B, Crombez G, Vanderstraeten G, Achten E, Danneels L. Lumbar muscle dysfunction during remission of unilateral recurrent nonspecific low-back pain: evaluation with muscle functional MRI. Clin J Pain 2013: 29: Dickx N, Cagnie B, Achten E, Vandemaele P, Parlevliet T, Danneels L. Differentiation between deep and superficial fibers of the lumbar multifidus by magnetic resonance imaging. EurSpine J 2010a: 19: Dickx N, D hooge R, Cagnie B, Deschepper E, Verstraete K, Danneels L. Magnetic resonance imaging and electromyography to measure lumbar back muscle activity. Spine (Phila Pa 1976) 2010b: 35: E836 E842. Durall CJ, Udermann BE, Johansen DR, Gibson B, Reineke DM, Reuteman P. The effects of preseason trunk muscle training on low-back pain occurrence in women collegiate gymnasts. J Strength Cond Res 2009: 23: Franca FR, Burke TN, Caffaro RR, Ramos LA, Marques AP. Effects of muscular stretching and segmental stabilization on functional disability and pain in patients with chronic low back pain: a randomized, controlled trial. J Manipulative Physiol Ther 2012: 35: Franca FR, Burke TN, Hanada ES, Marques AP. Segmental stabilization and muscular strengthening in chronic low back pain: a comparative study. Clinics (Sao Paulo) 2010: 65: Henchoz Y, de Goumoëns P, Norberg M, Paillex R, So AK. Role of physical exercise in low back pain rehabilitation: a randomized controlled trial of a three-month exercise program in patients who have completed multidisciplinary rehabilitation. Spine (Phila Pa 1976) 2010: 35: Henchoz Y, So AKL. Exercise and nonspecific low back pain: a literature review. Joint Bone Spine 2008: 75: Hibbs AE, Thompson KG, French D, Wrigley A, Spears I. Optimizing performance by improving core stability and core strength. Sports Medicine 2008: 38: Holm SM, Dickinson AL. A comparison of two isometric back endurance tests and their predictability of first-time back pain: a pilot study. JNMS 2001: 9: Holmstrom E, Moritz U, Andersson M. Trunk muscle strength and back muscle endurance in construction workers with and without low-back disorders. Scand J Rehabil Med 1992: 24: Jenner G, Foley JM, Cooper TG, Potchen EJ, Meyer RA. Changes in magnetic resonance images of muscle depend on exercise intensity and duration, not work. J Appl Physiol (1985) 1994: 76: Kell RT, Asmundson GJ. A comparison of two forms of periodized exercise rehabilitation programs in the management of chronic nonspecific low-back pain. J Strength Cond Res 2009: 23: Kuukkanen T, Malkia E. Muscular performance after a 3 month progressive physical exercise program and 9 month follow-up in subjects with low back pain. A controlled study. Scand J Med Sci Sports 1996: 6: Luoto S, Heliovaara MH, Hurri H, Alaranta H. Static back endurance and the risk of low-back-pain. Clin Biomech 1995: 10: Mannion AF, Dolan P. The effects of muscle length and force output on the

85 mfmri of recruitment of trunk muscles during extension EMG power spectrum of the erector spinae. J Electromyogr Kinesiol 1996: 6(3): Mannion AF, Muntener M, Taimela S, Dvorak J. Comparison of three active therapies for chronic low back pain: results of a randomized clinical trial with one-year follow-up. Rheumatology 2001a: 40: Mannion AF, Taimela S, Muntener M, Dvorak J. Active therapy for chronic low back pain Part 1. Effects on back muscle activation, fatigability, and strength. Spine 2001b: 26: Masuda K, Masuda T, Sadoyama T, Inaki M, Katsuta S. Changes in surface EMG parameters during static and dynamic fatiguing contractions. J Electromyogr Kinesiol 1999: 9: Mayer JM, Graves JE, Clark BC, Formikell M, Ploutz-Snyder LL. The use of magnetic resonance imaging to evaluate lumbar muscle activity during trunk extension exercise at varying intensities. Spine 2005: 30: Mayer JM, Graves JE, Robertson VL, Pierra EA, Verna JL, Ploutz-Snyder LL. Electromyographic activity of the lumbar extensor muscles: effect of angle and hand position during Roman chair exercise. Arch Phys Med Rehabil 1999: 80: Mayer JM, Verna JL, Manini TM, Mooney V, Graves JE. Electromyographic activity of the trunk extensor muscles: effect of varying hip position and lumbar posture during Roman chair exercise. Arch Phys Med Rehabil 2002: 83: McGill S. Core training: evidence translating to better performance and injury prevention. Strength Cond J 2010: 32: Meyer RA, Prior BM. Functional magnetic resonance imaging of muscle. Exerc Sport Sci Rev 2000: 28 (2): Moffroid MT. Endurance of trunk muscles in persons with chronic low back pain: assessment, performance, training. J Rehabil Res Dev 1997: 34: Ng JKF, Chan CCH, Lui CYH, Sin AKY, Wong ISW, Yip JHH. Back and lower limb muscle activities in a prone leg holding exercise. J Back Musculoskelet 1999: 13: Ng JKF, Richardson CA, Jull GA. Electromyographic amplitude and frequency changes in the iliocostalis lumborum and multifidus muscles during a trunk holding test. Phys Ther 1997: 77: Plamondon A, Marceau C, Stainton S, Desjardins P. Toward a better prescription of the prone back extension exercise to strengthen the back muscles. Scand J Med Sci Sports 1999: 9: Plamondon A, Serresse O, Boyd K, Ladouceur D, Desjardins P. Estimated moments at L5/S1 level and muscular activation of back extensors for six prone back extension exercises in healthy individuals. Scand J Med Sci Sports 2002: 12: Renkawitz T, Boluki D, Grifka J. The association of low back pain, neuromuscular imbalance, and trunk extension strength in athletes. Spine J 2006: 6: Renkawitz T, Linhardt O, Grifka J. Electric efficiency of the erector spinae in high performance amateur tennis players. J Sports Med Phys Fitness 2008: 48: Smith CE, Nyland J, Caudill P, Brosky J, Caborn DNM. Dynamic trunk stabilization: a conceptual back injury prevention program for volleyball athletes. J Orthop Sports Phys Ther 2008: 38: Vanderthommen M, Duteil S, Wary C, Raynaud JS, Leroy-Willig A, Crielaard JM, Carlier PG. A comparison of voluntary and electrically induced contractions by interleaved 1H- and 31P-NMRS in humans. J Appl Physiol 2003: 94: Wilke HJ, Wolf S, Claes LE, Arand M, Wiesend A. Stability increase of the lumbar spine with different muscle groups. A biomechanical in vitro study. Spine (Phila Pa 1976) 1995: 20(2):

86 Manual Therapy (1998) 3(1), Harcourt Brace & Co. Ltd Review article Insufficient lumbopelvic stability: a clinical, anatomical and biomechanical approach to 'a-specific' low back pain A. L. Pool-Goudzwaard, A. Vleeming, R. Stoeckart, C. J. Snijders, J.M. A. Mens Research group 'musculoskeletal system': Department of Anatomy, Department of Biomedical Physics and Technology, Department of Rehabilitation Medicine, Erasmus University; Spine and Joint Centre, Rotterdam, The Netherlands SUMMARY. A clinical, anatomical and biomechanical model is introduced based on the concept that under postural load specific ligament and muscle forces are necessary to intrinsically stabilize the pelvis. Since load transfer from spine to pelvis passes through the sacroiliac (SI) joints, effective stabilization of these joints is essential. The stabilization of the SI joint can be increased in two ways. Firstly, by interlocking of the ridges and grooves on the joint surfaces (form closure); secondly, by compressive forces of structures like muscles, ligaments and fascia (force closure). Muscle weakness and insufficient tension of ligaments can lead to diminished compression, influencing load transfer negatively. Continuous strain of pelvic ligaments can be a consequence leading to pain. For treatment purposes stabilization techniques followed by specific muscle strengthening procedures are indicated. When there is a loss of force closure, for instance in peripartum pelvic instability, application of a pelvic belt can be advised. INTRODUCTION For modern society low back pain (LBP) is an expensive disease. The yearly prevalence varies from 15-20% in the USA to 25-40% in European countries, and the lifetime prevalence is as high as 60-90% (Van Tulder 1996). The minority of patients who recover within 3 months account for 75-90% of the total expense related to this health care problem, exceeding $60 billion per year in the USA (US Department of Health 1994). Notwithstanding the high prevalence, in 70-80% of cases, the cause of LBP is not clear. The authors wondered what the reason could be for such a deficiency in our understanding. The models used to understand and treat low back pain are generally based on descriptive anatomy. This branch of anatomy was developed to determine the A. L. Pool-Goudzwaard Pt, BSc, Department of Anatomy and Spine and Joint Centre, A. Vleeming PhD, Department of Anatomy and Spine and Joint Centre, R. Stoeckart PhD, Department of Anatomy, C. J. Snijders PhD, Department of Biomedical Physics and Technology, J.M. A. Mens MD, Department of Rehabilitation Medicine, Erasmus University Rotterdam, Faculty of Medicine and Allied Health Sciences, PO Box 1738, 3000 DR Rotterdam and Spine and Joint Centre, Westerlaan 10, 3016 CK, Rotterdam, The Netherlands. structures that comprise the body and to categorize them. Categories such as spine, pelvis and lower limbs are primarily based on bony anatomy. Functional anatomy of the locomotor system, which is strongly linked to biomechanics, attempts to explain how bones, ligaments and muscles operate as a system. Consequently the use of categories such as spine and pelvis can be misleading (Vleeming et al 1995a). For example 'spinal' muscles, like the multifidus muscle, are strongly connected to the pelvis and to the ligaments around the sacroiliac (SI) joints. The use of descriptive anatomy is not satisfactory in answering complex ques tions such as: what are the reasons why so many patients suffer from LBP? Or how do the spine, the pelvis and the lower limbs function as an integrated system? Using descriptive anatomic models, it is tempting to regard pain in the area of the SI joints as a separate syndrome and not as a low back or lumbar syndrome (Vleeming et al 1995). Could it be that better knowledge of the functional anatomy of the spine, the pelvis and lower limbs could lead to the understanding of the aetiology of so called 'aspecific' low back pain? To answer this question this research group has focused its investigations on the role of the SI joints. 12

87 Insufficient lumbopelvic stability I 3 WHY THE SACROILIAC JOINTS? Focusing on the functional anatomical relations between spine and pelvis we wondered why such flat joints typically occur at a site, where the transfer of large forces can be expected (Snijders et al 1993a,b). After all, joints with relatively flat surfaces are vulnerable to shear forces (Snijders et al 1993a). Would ankylosed SI joints not be more appropriate for force transfer? For many decades clinicians have been convinced the SI joints were not mobile, but this notion is not based on research findings. Especially in the last two decades research has proven otherwise; mobility in the SI joints is usual, even in old age (Egund et al 1978; Lavignolle et al 1983; Miller et al 1987; Sturesson et al 1989; Vleeming et al 1992). One advantage of mobile SI joints could be their ability to function as a shock absorber between the lower limbs and the spine (Snijders et al 1993a), while another advantage could be the afferent output of the joint capsule (proprioception), which is innervated by the dorsal rami of sacral nerves (Sl-S4) (Grob et al 1995). FORM CLOSURE Since the SI joints have to transfer large loads, it can be assumed that the shape of the joints is adapted to this task. The joint surfaces are relatively flat which is favourable for the transfer of compressive forces and bending moments (Snijders et al 1993a,b). However, as stated already, a relatively flat joint is vulnerable to shearforces. The SI joints are protected from these forces in three ways. Firstly, due to its wedge-shape the sacrum is stabilized by the innominates. Secondly, in contrast to normal synovial joints the articular cartilage is not smooth. Before birth the sacroiliac joint articular cartilage is already unusual in its irregularity (Sashin 1930; Bowen & Cassidy 1981). These modifications of the cartilage are more prominent in men than in women (Vleeming et al 1990). The gender difference may be related to childbearing and possibly to a different localization of the centre of gravity in relation to the SI joints. Thirdly, by studying frontal slides of intact joints of embalmed specimens we can show the presence of cartilage covered bone extensions protruding into the joint (Vleeming 1990; Dijkstra et al 1992), the so called ridges and grooves (Fig. 1). They seem irregular, but are in fact complementary, which serves a functional purpose. This stable situation with closely fitting joint surfaces, where no extra forces are needed to maintain the state of the system, given the actual load situation, we termed 'form closure' (Fig. 2a). (Vleeming 1990; Snijders et al 1993a,b). FORCE CLOSURE If the sacrum could fit in the pelvis with perfect form closure, mobility would be practically impossible. Fig. 1 Ridges and grooves in the sacroiliac joint that are covered with intact cartilage. Front Vleeming A. The Sacroiliac Joint. A clinical biomechanical and radiological study (dissertation) Rotterdam Erasmus University 1990 (reproduced with permission). However, during walking, mobility as well as stability in the pelvis must be optimal. Extra forces may be needed for equilibrium of the sacrum and the ilium during loading situations. How can this be reached? The principle of a Roman arch of stones resting on columns may be applicable to the force equilibrium of the SI joints (Snijders et al 1993a). Since the columns of a Roman arch cannot move apart, reaction forces in an almost longitudinal direction of the respective stones lead to compression and help to avoid shear. For the same reason, ligament and muscle-forces are needed to provide compression of the SI joint. Especially during unilateral loading of the legs this system has to become active. Due to compression of the SI joints, friction of the joint increases. This mechanism of compression of the SI joints due to extra forces, to keep an equilibrium, is called 'force closure' (Fig. 2b) (Vleeming 1990; Vleeming et al 1990a,b; Snijders et al l 993a,b). Force closure can be generated especially by structures with a fibre direction perpendicular to the SI joint and can be accommodated to the specific loading situation. Several 1998 Harcourt Brace & Co. Ltd Manual Therapy ( 1998) 3( 1 ), 12-20

88 14 ManualTherapy a C Fig. 2-Form closure (a) and force closure (b) lead to a selflocking mechanism (c). From: Snijders CJ et al. Transfer of Jumbosacral load to iliac bones and legs. Part 1. In: Proceedings of the First Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Reproduced by kind permission of the conference organisers, Rotterdam, The Netherlands. ligaments, muscles and even a fascia can contribute to force closure. The authors termed the shear prevention system, characterized by the combination of form and force closure, the 'selfbracing or selflocking mechanism' of the SI joint (Fig 2c) (Vleeming 1990; Vleeming et al 1990a,b; Snijders et al 1993a,b). The authors will now describe the contribution of ligaments, the thoracolumbar fascia and muscles to the selflocking mechanism. THE SELFLOCKING MECHANISM Ligaments Both interosseous and short dorsal sacroiliac ligaments originate from the sacrum and attach on to the ilium close to the joint surfaces. During nutation of the sacrum the tension in these ligaments increases, leading to more friction at the joint surfaces and hence more stability of the SI joints (Sturesson et al 1989). Nutation of the sacrum occurs during loading situations such as transferring from lying supine to sitting and standing (Egund et al 1978; Lavignolle 1983). The authors assume that tension of the sacrotuberous ligament (Fig. 3) can also stabilize the SI joints, since loading of the ligament in embalmed specimens showed a decrease of mobility in the SI joint (Vleeming et al l 989a,b; van Wingerden 1993). The sacrotuberous ligament originates from the dorsal side of the sacrum and attaches to the ischial tuberosity. Its tension can be influenced in four ways. Firstly, load tests on embalmed specimens showed increased tension in the sacrotuberous ligament by applying tension to the long head of the biceps femoris muscle (Vleeming et al 1989a,b; van Wingerden 1993). One explanation could be an anatomical feature, since in half of the specimens (n = 12) the fibres of the tendon of biceps femoris are continuous with the sacrotuberous ligament (Vleeming et al 1989a,b). Another explanation of this phenomenon is the tilting backwards of the ilium due to traction on the ischial tuberosity by increased tension in biceps femoris, increasing nutation in the SI joint leading to increased tension in the sacrotuberous ligament Manual Therapy (1998) 3( 1 ), (Vleeming et al 1989a,b; van Wingerden et al 1993). Secondly, the tension of the sacrotuberous ligament can be influenced by increased tension of the gluteus maximus and piriformis muscles because of the anatomical connection between the muscles and the ligament (Vleeming et al 1989a,b). Thirdly, the thoracolumbar fascia can increase the tension in the ligament due to the anatomical connections between the deep lamina of the superficial layer of the fascia and the sacrotuberous ligament (Vleeming et al 1989b). Finally, tension in the sacrotuberous ligament also increases during nutation (Vleeming et al 1989a). Apparently the sacrotuberous ligament can inhibit nutation. We wondered which ligament could inhibit contranutation. In our opinion, the long dorsal sacroiliac ligament (Fig. 4) can fulfill this task. It originates from the dorsal surface of the sacrum at the level of S2-S4 and attaches to the posterior superior iliac spine (PSIS) (Vleeming et al 1996). This ligament is easily palpable in the area directly caudal to the PSIS, although in an international survey among medical practitioners (Vleeming et al 1993; Vleeming et al 1996) less than 10% could identify the long dorsal sacroiliac ligament. The ligament is so Fig. 3-The Jong dorsal sacroiliac ligament. From: Vleeming A, Pool-Goudzwaard AL, Hammudoghlu D, Stoeckart R, Snijders CJ. The function of the long dorsal sacroiliac ligament. Spine 1996; 21(5): Reproduced by kind permission of the publishers Lippincott-Raven, Philadelphia Harcourt Brace & Co. Ltd

89 Insufficient lumbopelvic stability 15 Fig. 4---The sacrotuberous ligament. From: Vleeming A, Pool Goudzwaard AL, Hammudoghlu D, Stoeckart R, Snijders CJ. The function of the long dorsal sacroiliac ligament. Spine 1996; 21(5): Reproduced by kind permission of Lippincott-Raven, Philadelphia. solid and taut that it can easily be mistaken for a bony structure when palpated. The tension of the long dorsal sacroiliac ligament can be increased in two ways. Firstly, tension in this ligament strongly increases during incremental loading of the ipsilateral sacrotuberous ligament. The same is true, but to a lesser degree, for incremental loading of the ipsilateral part of the erector muscle. The tension of this ligament can be decreased by traction to the gluteus maximus muscle (Vleeming et al 1996). Apparently then, the sacrotuberous ligament and the long dorsal sacroiliac ligament have opposite functions. However, both ligaments also have a direct effect on each other. This is due to an anatomical connection between these ligaments caudal to the PSIS. Increased tension of the sacrotuberous ligament can directly increase tension of the long sacroiliac ligament, and vice versa (Vleeming et al 1996). In this way extensive slackening of both long dorsal sacroiliac ligament and sacrotuberous ligament is precluded. Such a mechanism could be essential for a flat joint that is susceptible to shear forces (Snijder et al 1993a,b). sacrotuberous ligament is connected to the deep layer (Vleeming et al 1995a). In transferring forces between spine, pelvis, and legs, the thoracolumbar fascia plays an important role, especially in rotation of the trunk and stabilization of the lower lumbar spine and SI joints (lumbopelvic stability). The authors assume that an increase in tension in the thoracolumbar fascia can lead to more compression on the SI joint, increasing force closure (Vleeming et al 1995a). The tension of the fascia can be increased in two ways, firstly, due to contraction of the muscles that are attached to the thoracolumbar fascia. Secondly, contraction of the erector spinae muscle and especially the multifidus muscle that can increase the tension by inflating it (the 'pump it up' phenomenon) (Bogduk & MacIntosh 1984; Vleeming et al 1995a). Due to the connections of the thoracolumbar fascia with upper and lower limb muscles, the fascia is capable of transmitting forces from the lower to the upper extremities and vice versa. Importantly the effect of contraction of the latissimus dorsi can be large because forces derived from its caudal part are fully transferred Thoracolumbar fascia The thoracolumbar fascia surrounds the dorsal muscles of the trunk. It proceeds from both sacrum and iliac bones and inserts on to the linea nuchae (Bogduk & MacIntosh 1984; Bogduk & Twomey 1987; Vleeming et al 1995a). The fascia forms an attachment for several upper limb and trunk muscles, e.g. the latissimus dorsi, gluteus maximus and the trapezius muscle are connected to the superficial fascial layer (Fig. 5). The transverse abdominal and internal oblique muscle are connected to the deep layer of the thoracolumbar fascia (Fig. 6) (Bogduk & MacIntosh 1984; Bogduk & Twomey 1987; Vleeming et al 1995a). In addition, the Fig. 5-The superficial layer of the thoracolumbar fascia and its attachments to: (A) the gluteus maximus; (B) the gluteus medius; (C) the external oblique; (D) the latissimus dorsi; (1) SIPS; (2) the sacrum. From: Vleeming A, Pool-Goudzwaard AL, Stoeckart R, Wingerden van JP, Snijders CJ. The posterior layer of the thoracolumbar fascia: its function in load transfer from spine to legs. Spine 1995; 20(1): Reproduced by kind permission of Lippincott-Raven, Philadelphia Harcourt Brace & Co. Ltd Manual Therapy ( 1998) 3(1 ), 12-20

90 16 ManualTherapy Fig The deep layer of the thoracolumbar fascia and its attachments to: (B) the gluteus medius; (E) attachments between the deep layer and the erector spinae muscle; (F) the internal oblique; (G) the serratus posterior inferior; (H) the sacrotuberous ligament; (I) SIPS; (2) sacrum. From: Vleerning A, Pool-Goudzwaard AL, Stoeckart R, Wingerden van JP, Snijders CJ. The posterior layer of the thoracolumbar fascia: its function in load transfer from spine to legs. Spine 1995a; 20(1): Reproduced by kind permission of Lippincott-Raven, Philadelphia. to the thoracolumbar fascia. The gluteus maximus and the latissimus dorsi merit special attention because they can conduct forces contralaterally since, at the levels of L4-S2, fibres of the superficial layer of the fascia cross the midline (Vleeming et al 1995a). Via the thoracolumbar fascia the gluteus and contralateral latissimus dorsi muscles are coupled. Recently it has been suggested that these muscles, normally categorized as 'hip' and 'arm' muscles could act as trunk rotators. EMG studies show that both muscles contract during rotation of the trunk (Basmaijan 1976; Mooney 1995). This implies that increase of force closure can be expected during rotation of the trunk (Mooney 1995). layer of the thoracolumbar fascia and the sacrotuberous ligament, which is connected to the long head of the biceps (Vleeming et al 1995b). Tension in this longitudinal sling will stabilize the SI joint in three ways. Due to contraction of the sacral part of the multifidus muscle the SI joint has a tendency to nutate (Snijders et al 1993b) increasing tension in the interosseous and short dorsal sacroiliac ligaments leading to more force closure of the SI joint (Vleeming 1990; Snijders et al 1993a). Secondly, these muscles can contribute to force closure by inflating the thoracolumbar fascia (Vleeming et al 1995a). Finally, contraction of the erector spinae muscle as well as the long head of the biceps femoris can help to increase force closure due to its anatomical connection with the sacrotuberous ligament. As described earlier, tension of the sacrotuberous ligament increases force closure. The posterior oblique sling (Fig. 7A) (Vleeming et al 1993; Vleeming 1995b) can be energized by the coupled function of the latissimus dorsi and the gluteus maximus muscle. These muscles function as synergists (Mooney 1995). Contraction can directly optimize stabilization of the SI joint. Force closure can also be increased indirectly, due to the anatomical connections of the gluteus maximus muscle and thoracolumbar fascia with the sacrotuberous ligament, as described previously (Vleeming et al 1989a,b, 1995a). The anterior oblique sling (Fig. 7B) (Snijders et al 1993b; Vleeming 1995b) can be energized by the external and internal abdominal muscles and also by the transversus abdominis muscle because of the connections between these muscles via the rectus sheath. These muscles can help to increase force closure (Snijders et al 1993b). As shown by an EMG study, a significant decrease of activity of the oblique abdominal muscles occurs as a result of leg crossing (Snijders et al 1995). A possible explanation of this phenomenon is an increase A B Musculature Several muscles are supposed to contribute to force closure of the SI joint. The authors have described three muscle slings - a longitudinal, a posteror oblique and an anterior oblique sling - that can be energized (Vleeming et al 1995b). The longitudinal sling consists of the combination of the multifidus muscle attaching to the sacrum, the deep Manual Therapy (1998) 3(1), Fig. 7A-The posterior oblique sling. (1) The latissimus dorsi; (2) the thoracolumbar fascia; (3) the gluteus maximus; (4) the iliotibial tract. Fig. 7B-The anterior oblique sling. (5) Linea alba; (6) external oblique muscle; (7) transverse abdominal muscle; (8) piriformis muscle; (9) rectus abdominis muscle; (10) internal oblique; (11) Jig. inguinale. From: Poster-presentation: 'Biomechanical model on the etiology of ' lower back pain'. Snijders CJ, Vleeming A, Stoeckart R, Starn A J. Figures on poster for Second World Congress on Biomechanics, Amsterdam Reproduced with permission Harcourt Brace & Co. Ltd

91 Insufficient lumbopelvic stability 17 of friction of the SI joint surfaces due to leg crossing. Less force closure by the oblique abdominals is needed, so crossing the legs seems to be a very functional habit (Snijders et al 1995). Electromyographic studies in healthy subjects by Hodges & Richardson (1996) show activity of the transversus abdominis muscle, a deep trunk muscle, occurring prior to any limb movement, this activity is different from other trunk muscles. Richardson & Jull (1995) concluded that this muscle has a primary responsibility for lumbar segmental stability. Another important component of the muscular lumbar segmental stability system is the multifidus muscle (Wilke et al 1995). It appears that a co-contraction motor programme of transversus and multifidus muscle is required for specific lumbar segmental stability (Richardson & Jull 1995). In the authors' opinion the multifidus and transversus abdominis muscles act not only as lumbar stabilizers but also as pelvic stabilizers. It might well be also that other muscles are involved, e.g. the pelvic floor and the respiratory diaphragm. INSUFFICIENT SELFLOCKING In general, the ligamentous structures surrounding the SI joints are assumed to be sufficient to stabilize the joints. The authors do not agree with this assumption since we have produced evidence to show that the ligaments alone are not capable of transferring lumbosacral load effectively from the spine to the iliac bones (Vleeming 1990; Vleeming et al 1990a,b). This is especially relevant for heavy loading situations and for conditions where sustained load resulting in creep occurs, such as in standing and sitting. According to the selflocking mechanism, resistance against shear forces is the result of the specific properties of joint articular surfaces and the compression of body weight (form closure), as well as muscle action and ligamentous force (force closure). This implies that several factors can lead to insufficient selflocking: force closure can decrease by changes in ligamentous tension, e.g. laxity of joint capsule and ligaments, reduced muscle strength or inadequate coordination between muscles. motor programme of the transversus abdominis muscle and multifidus muscle (Hides et al 1994; Hodges & Richardson 1996). Dysfunction of the transversus abdominis muscle has clearly been shown in back pain patients. The difference in timing of the transversus abdominis muscle activity after a few days pain is remarkable; the motor programme of the muscle changes (Hodges & Richardson 1996). Hides et al (1994) reported a unilateral wasting of the segmental cross sectional area of the multifidus muscle measured by ultra sound imaging technique in patients with acute unilateral low back pain. Inhibition due to perceived pain, via a long loop reflex which targeted the vertebral level of pathology to protect the damaged tissues, is the likely mechanism of wasting. Through this change of motor programming, coordination between the possible initiators of force closure, the transversus abdominis and the multifidus muscle, will be disturbed. In addition, laxity of the SI joint capsule can diminish optimal force closure. This is seen especially during pregnancy when hypermobility of the SI joint occurs, probably due to the presence of certain hormones leading to laxity of ligaments and joint capsule (Kristianson 1995). Due to hormonal laxity combined with muscular weakening and hence diminished compression selflocking can become insufficient, which can lead to peripartum pelvic pain (Snijders et al 1997; Snijders et al 1984; Mens et al 1995; Mens et al 1996). It has been assumed that during pregnancy women bend the upper body backwards to obtain equilibrium with the increasing weight in the abdomen (Snijders et al 1984). In fact the spine becomes straighter by a flattening of the lumbar curvature leading to a less nutated position of the SI joint (Egund et al 1978; Lavignolle et al 1983; Sturesson et al 1989). This relative countemutated position is especially seen late in pregnancy when women counterbalance the weight of the foetus (Snijders et al 1984). Diminished nutation can also be the effect of a pain withdrawal reaction, for example in women with a painful symphysis pubis following delivery (Mens et al 1996). In countemutation the SI joint is particularly less stable, which can enhance the instability problem. Some patients with peripartum pelvic pain show extreme pelvic hypermobility. As a result the stabilization of the SI joints becomes deficient. Decrease of force closure In addition to the diminished muscle power and/or unbalanced muscle function of several muscles (e.g. erector spinae, the gluteus maximus, the external and internal obliques and the latissimus dorsi muscle), insufficient ligamentous tension can lead to a decrease in force closure. Such an insufficiency of the ligaments can occur due to an unusual positioning of the SI joints. This can influence the afferent output of the joint capsule. There may be a casual relationship between different afferent output of the joint capsule and changes in the 1998 Harcourt Brace & Co. Ltd Consequences of insufficient selflocking An epidemiologic study showed the occurrence of long ligament pain syndrome (LLPS) in 21% of low back pain patients (Njoo 1996). The assessment of LLPS was related to pain within the boundaries of the long dorsal sacroiliac ligament directly palpable caudal to the PSIS. The interobserver agreement of establishing LLPS was found to be good. In a recent survey (n = 394) of peripartum pelvic pain patients, 42% indicated pain in the area caudal to the PSIS, within the boundaries of the long dorsal sacroiliac Manual Therapy (1998) 3( ]), 12-20

92 18 ManualTherapy ligament (Mens et al 1996). Increased tension of this ligament could clarify why pain is specifically felt directly caudal to the PSIS. Not only patients with peripartum pelvic instability but also patients with a-specific low back pain can show increased tension which can be the result of countemutation or diminished nutation. The authors assume that palpation of this ligament directly caudal to the PSIS is an important pain provocation site for assessment of ligamentous tension in patients with low back pain and lumbopelvic instability. According to the selflocking model, the SI joints become especially vulnerable to shear forces if loaded in countemutation. Countemutation, which is coupled to a supine position and flattening of the spine, can lead to abnormal loading, not only of the SI joints but also of the lumbar spine, the intervertebral joints and intervertebral discs. Based on the presented data low back disorders like sciatica, soft tissue lesions or herniation of discs, are not necessarily separate syndromes; they can be the results of insufficient stabilization of the pelvis and lower spine. Load transfer with an unstable SI joint can produce excessive loads on surrounding tissues and hence pain in local structures. Pain referral maps of the SI joint show pain occurrence in the lower back and the buttock sometimes radiating to one leg (Schwarzer et al 1995). CLINICAL ASSESSMENT AND THERAPY Clinial assessment When pelvic peripartum instability is involved, the pain can begin during pregnancy, often in the third month, or either directly or a few weeks after the delivery. For patients with low back pain and/or pain in the buttock sometimes radiating into one leg, it is important to know the history of pain, the exact site of pain and the spinal position that eases the pain. The information from this assessment can lead to more insight into how the SI joint and L5-S1 junction are involved and whether the long dorsal ligament is tensed. It is important to find answers to questions like: are the SI joints hypermobile? or is nutation of the SI joints diminished? However, in assessing the mobility of the SI joint caution is needed as mobility tests of the SI joint are not reliable; however, pain provocation tests seem to be more valid (Bernhard & Cassidy 1991; Laslett 1995; Schwarzer et al 1995; Njoo 1996). To assess stability of the SI joints clinicians can use the active straight leg raise test (ASLR) (Mens et al 1995). Lying supine, the patient is asked to lift one leg so that the heel lies 5 cm above the couch. The ASLR test is positive for an unstable pelvis if the patient is unable to lift the leg, or if the patient experiences diminished strength on one side. The test is repeated while the patient is wearing a pelvic belt, which has been shown Manual Therapy ( 1998) 3( I), to have a stabilizing effect on the pelvis (Mens et al 1996). In the case of impairment of the selflocking mechanism it will be easier to lift a leg wearing the belt. Another stabilizing effect can be reached by manual pressure on the lateral sides of the anterior superior iliac spines (ASIS) or contraction of the obliques, using the anterior oblique sling. The next stage in testing muscular lumbopelvic stability involves testing the strength of the transversus abdominis and multifidus muscles, the erector spinae, gluteus maximus, the oblique abdominal and the latissimus dorsi muscles, since these muscles can contribute to a proper force closure. In the case of peripartum pelvic pain patients, an additional assessment is possible using X-ray photography. After delivery the symphysis pubis is usually widened from the normal 4mm to 8mm (Anderson & Peterson 1944). A more specific relationship between pelvic instability and the X-ray photography is to look for a vertical shift of the pubic bone during standing on one leg: the Chamberlain technique. During weightbearing on one side, alternating left and right lower limbs, radiographs in the posteroanterior direction are made. Chamberlain suggested that the side of the high pubic bone was the abnormal side. However, this vertical shift is not, as Chamberlain describes, a shift of the pubic bone and ilium on the weight bearing side to a more cranial position, but a shift to a more caudal position on the non-weight bearing side (Mens et al 1996). A shift to a caudal position implies an unstable symphysis pubis and SI joint. Direct measurement of the mobility of the SI joints is still not possible; how ver, using colour doppler imaging images of the left and right SI joints can be compared (Buyruk et al 1995a,b). This non-invasive method seems to be reproducible. The outcome of this research shows that the most important criterium of hyper- and hypomobility of the SI joints is not mobility itself, but the difference in mobility between the left and right SI joints. However, it is still too early to draw clinical conclusions. Therapy By testing the muscular and articular lumbopelvic stability, on the basis of the ASLR, pain provocation tests and muscular strength and coordination it is possible to formulate a therapeutic plan. Firstly, a clinician should strive for optimal mobility of the SI joints. Diminished nutation or relative countemutation can increase instability and leads to heightened tension in the long sacroiliac ligaments. A simple technique, nutating the sacrum and tilting the ilium backwards, can give rise to an immediate relief of pain directly caudal of the PSIS, as Don Tigny (1992) has shown. Secondly, a patient can train muscular lumbopelvic stability by emphasizing a motor programme related specifically to the transversus abdominis and multifidus muscles. Through a stabilization programme, using an 1998 Harcourt Brace & Co. Ltd

93 Insufficient lumbopelvic stability 19 abdominal hollowing technique, both muscles become active, stabilizing the lumbopelvic region (Richardson & Jull 1995). This stabilization programme consists of three phases. During the first phase, a motor programme is recreated for both transversus abdominis and multifidus muscles. The aim is to contract these muscles without any movement of extremities and trunk. In the second phase both muscles have to keep the lumbopelvic region stabilized during movements of both extremities. During the third phase the transversus abdominis and multifidus muscles must be active during trunk movements. As latest research shows, this isolation, motor programming and stabilization of the lumbopelvic region can be trained effectively (Hides et al 1994; Richardson & Jull 1995). Finally, the strength and coordination of the erector spinae can be increased by extension of the trunk. The strength of the oblique abdominal muscles, the gluteus maximus and the latissimus dorsi can be increased during rotation of the trunk (Vleeming et al 1992). All these muscles can optimize force closure. However, it is very important that these muscles are not trained separately before stabilization of the lumbopelvic region is sufficient, by emphasizing the motor programme of the transversus abdominis and multifidus muscles. For instance, when the gluteus maximus muscle is trained, especially in case of hypermobile joints, asymmetrical movement of the pelvis will be the result which can provoke pain. During strengthening of the muscles the emphasis should be on training the three slings. For patients with peripartum pelvic instability due to hypermobility, therapy should include advice about activities of daily living. It is important for the patient to be advised on walking patterns, to climbing stairs, sitting, to lifting, etc., since strong contraction of the iliopsoas muscle should be avoided. Contraction of this muscle can lead to a large asymmetrical force placed on the SI joint, which can lead to more pain if the SI joints are insufficiently stabilized. As soon as stability increases by training stability, coordination and muscle strength, the patient will experience a less negative influence of the iliopsoas muscle. This stabilizing effect can also be reached by a pelvic belt (Mens et al 1996). The belt is probably too small to inhibit muscle activity but significantly increases force closure, as research on embalmed specimens has shown. Furthermore, it is necessary that a patient with peripartum pelvic instability is informed that too much rest can lead to loss of muscle strength, coordination, stability and of muscular and physical endurance. Too much rest is likely to increase the symptoms. Periods of short rest in the case of acute pain can be advised and activities that do not lead to an increase in pain can be continued. Acknowledgement With special thanks for financial support from the Dutch Association of Manual Therapists (NVMT). References Anderson R L, Peterson V L 1944 Clinical use of the Chamberlain technique in sacroiliac conditions. JAMA 124: Basmaijan JV 1976 Muscles Alive. Williams & Wilkins, Baltimore, Ch 13, Bernard TN, Cassidy JD 1991 The sacroiliac joint syndrome: pathophysiology, diagnosis and management. In: Frymoyer J W (ed). The Adult Spine: Principles and Practice. Raven Press, New York, Bogduk N, MacIntosh J E 1984 The applied anatomy of the thoracolumbar fascia. Spine 9: Bogduk N, Twomey LT 1987 Clinical Anatomy of the Lumbar Spine. Churchill Livingstone, Melbourne, Ch 8, Bowen V, Cassidy JD 1981 Macroscopic and microscopic anatomy of the sacro-iliac joints from embryonic life until the eighth decade. Spine 6: Buyruk HM, Stam HJ, Vleeming A, Lameris J S, Holland WP J, Snijders CJ 1995a Analysis of sacroiliac joint stiffness with colour doppler imaging on embalmed human pelvises. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Buyruk HM, Snijders CJ, Vleeming A, Lameris J S, Holland WP J, Stam HJ 1995b In vivo measurement of sacroiliac joint stiffness with colour doppler imaging. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Dijkstra PF, Vleeming A, Stoeckart R 1992 Complex motion tomography of the sacroiliac joint; an anatomical and roentgenological stdy. In: Vleeming A, Mooney V, Snijders CJ, Dorman T (eds). First Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Integrated function of the lumbar spine and sacroiliac joint. Rotterdam ECO, Don Tigny R L 1992 Sacroiliac dysfunction: recognition and treatment. In: Vleeming A, Mooney V, Snijders CJ, Dorman T (eds). First Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Integrated function of the lumbar spine and sacroiliac joint. Rotterdam ECO, Egund N, Ollson TH, Schmid H, Selvik G 1978 Movements in the sacroiliac joints demonstrated with roentgen stereofotogrammetry. Acta Radiologica 19: Grob KR, Neuhuber W L, Kissling RO 1995 Die innervation des Sacroiliacaal gelenkes beim menschen. Zeitschrift fur Rheumatologie 54: Hides J A, Stokes NJ, Saide M, Juli GA 1994 Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 19: Hodges PW, Richardson C A 1996 Inefficient muscular stabilization of the lumbar spine associated with low back pain: a motor control evaluation of transverse abdominis. Spine 21: Kristianson P 1995 S-Relaxin: a marker for back pain during pregnancy. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Laslett M 1995 The reliability of the selected pain tests for SI Joint pain pathology. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Lavignolle B, Vital J M, Senegas Jet al 1983 An approach to the functional anatomy of the sacroiliac joints in vivo. Anatomica Clinica 5: Mens J M A, Stam HJ, Vleeming A, Snijders C J 1995 Active straight leg raising. A clinical approach to the load transfer function of the pelvic girdle. In: Vleeming A, Mooney V, Snijders CJ, Dorman T (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Mens J MA, Vleeming A, Stoeckart R, Stam HJ, Snijders C J 1996 Understanding peripartum pelvic pain; implications of a patient survey. Spine 21: Miller J A, Schultz AB, Andersson GB 1987 Load displacement behaviour of sacro-iliac joint. Journal of Orthopaedic Research 5: Mooney V 1995 Evaluation and treatment of sacroiliac dysfunction Harcourt Brace & Co. Ltd Manual Therapy (1998) 3(1), 12-20

94 20 ManualTherapy 20 ManualTherapy In: Vleeming A, Mooney V, Donnan T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Njoo K H 1996 Nonspecific low back pain in general practice: a delicate point. Erasmus University Rotterdam dissertation Richardson CA, Jull GA 1995 Muscle control - pain control. What exercise would you prescribe? Manual Therapy I: 2-10 Schwarzer AC, April C N, Bogduk N 1995 The sacroiliac joint in chronic low back pain. Spine 20: Sashin D 1930 A critical analysis of the anatomy and the pathological changes of the sacroiliac joints. Journal of Bone and Joint Surgery 12: Snijders C J, Vleeming A, Stoeckart R 1993a Transfer of lumbosacral load to iliac bones and legs. Part I: Biomechanics of selfbracing of the sacroiliac joints and its significance for treatment and exercise. Journal of Clinical Biomechanics 8: Snijders CJ, Vleeming A, Stoeckart R 1993b Transfer of lumbosacral load to iliac bones and legs. Part 2: Loading of the sacroiliac joints when lifting in stooped posture. Journal of Clinical Biomechanics 8: Snijders CJ, Slagter AH E, Strik van R, Vleeming A, Stoeckart R, Stam HJ 1995 Why leg crossing? The influence of common postures on abdominal activity. Spine 20: Snijders CJ. Snijder JG N, Hoedt HT E 1984 Biomechanische modellen in het bestek van rugklachten tijdens de zwangerschap. Tijdschrift voor Sociale Gezondheidszorg 62: Snijders CJ, Seroo J M, Snijder JG M, Hoedt HT 1977 Change in fonn of the spine as a consequence of pregnancy. Digest 11th ICMBE, Ottawa: Sturesson B, Selvik G, Uden A 1989 Movements of the sacroiliac joints. A roentgen stereophotogrammetric analysis. Spine 14: US Department of Health and Hum;m Services. Agency for Health Care Policy and Research 1994 Acute Low Back Problems in Adults. Publication , Rockville, Maryland Van Tulder M 1996 Diagnostics and treatment of chronic low back pain in primary care. Free University, Amsterdam Vleeming A. Wingerden van JP, Dijkstra PF, Stoeckart R, Snijders SJ, Stijnen T 1992 Mobility of the Sijoints in old people: A kinematic and radiological study. Clinical Biomechanics 7: Vleeming A 1990 The Sacroiliac Joint. A clinical, biomechanical and radiological study (dissertation). Rotterdam Erasmus University Vleeming A, Stoeckart R, Volkers AC W, Snijders CJ 1990a Relation between form and function in the sacroiliac joint. Part 1: Clinical anatomical aspects. Spine 15(2): Vleeming A, Volkers AC W, Snijders CJ, Stoeckart R 1990b Relation between form and function in the sacroiliac joint. Part 2: Biomechanical aspects. Spine 15(2): Vleeming A, Wingerden van JP, Snijders CJ, Stoeckart R, Stijnen T 1989a Load application to the sacrotuberous ligament: influences on sacroiliac mechanics. Clinical Biomechanics 4: Vleeming A, Stoeckart R, Snijders CJ 1989b The sacrotuberous ligament: a conceptual approach to its dynamic role in stabilizing the sacroiliac joint. Clinical Biomechanics 4: Vleeming A, Pool-Goudzwaard A L, Stoeckart R, Wingerden van JP, Snijders CJ, Mens J MA 1995a The posterior layer of the thoracolumbar fascia: its function in load transfer from spine to legs. Spine 20(1): Vleeming A, Pool-Goudzwaard AL, Hammudoghlu D, Stoeckart R, Snijders C J, Mens J M A 1996 The function of the long dorsal ligament. Spine 21(5): Vleeming A, Buyrnk H M, Stoeckart R, Karamursel S, Snijders CJ 1992 Towards an integrated therapy for peripartum pelvic instability. American Journal of Obstetrics and Gynaecology 166(4): Vleeming A, Pool-Goudzwaard AL, Stoeckart R, Wingerden van JP, Snijders CJ 1993 Towards a better understanding of the etiology oflow back pain. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). First Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Vleeming A, Snijders CJ, Stoeckart R, Mens J MA 1995b A new light on low back pain. In: Vleeming A, Mooney V, Dorman T, Snijders CJ (eds). Second Interdisciplinary World Congress on Low Back Pain and its Relation to the SI Joint. Rotterdam ECO, Wilke HJ, Wolf S, Claes LE, Arand M, Wiesend A 1995 Stability increase of the lumbar spine with different muscle groups: a biomechanical in vitro study. Spine 20(2): van Wingerden JP, Vleeming A, Snijders CJ, Stoeckart R 1993 A functional-anatomical approach to the spine-pelvis mechanism: interaction between the biceps femoris muscle and the sacrotuberous ligament. European Spine Journal 2: Manual Therapy ( /998) 3( l ), Harcourt Brace & Co. Ltd

95 [ research report ] LISE R. STOLZE, DSc, MPT 1 STEPHEN C. ALLISON, PT, PhD 2 JOHN D. CHILDS, PT, PhD, MBA 3 Derivation of a Preliminary Clinical Prediction Rule for Identifying a Subgroup of Patients With Low Back Pain Likely to Benefit From Pilates-Based Exercise Low back pain (LBP) is a primary cause of disability in modern society, 65 yet the pathoanatomic cause of LBP cannot be identified in the majority of individuals. 32,61 Attempts to identify effective interventions for LBP have TTSTUDY DESIGN: Prospective cohort study. TTOBJECTIVE: To derive a preliminary clinical prediction rule for identifying a subgroup of patients with low back pain (LBP) likely to benefit from Pilates-based exercise. TTBACKGROUND: Pilates-based exercise has been shown to be effective for patients with LBP. However, no previous work has characterized patient attributes for those most likely to have a successful outcome from treatment. TTMETHODS: Ninety-six individuals with nonspecific LBP participated in the study. Treatment response was categorized based on changes in the Oswestry Disability Questionnaire scores after 8 weeks. An improvement of 50% or greater was categorized as achieving a successful outcome. Thirty-seven variables measured at baseline were analyzed with univariate and multivariate methods to derive a clinical prediction rule for successful outcome with Pilates exercise. Accuracy statistics, receiver-operator curves, and regression analyses were used to determine the association between standardized examination variables and treatment response status. TTRESULTS: Ninety-five of 96 participants completed the study, with 51 (53.7%) achieving a successful outcome. A preliminary clinical prediction rule with 5 variables was identified: total trunk flexion range of motion of 70 or less, duration of current symptoms of 6 months or less, no leg symptoms in the last week, body mass index of 25 kg/m 2 or greater, and left or right hip average rotation range of motion of 25 or greater. If 3 or more of the 5 attributes were present (positive likelihood ratio, 10.64), the probability of experiencing a successful outcome increased from 54% to 93%. TTCONCLUSION: These data provide preliminary evidence to suggest that the response to Pilates-based exercise in patients with LBP can be predicted from variables collected from the clinical examination. If subsequently validated in a randomized clinical trial, this prediction rule may be useful to improve clinical decision making in determining which patients are most likely to benefit from Pilates-based exercise. J Orthop Sports Phys Ther 2012;42(5): doi: / jospt TTKEY WORDS: classification, lumbar spine, Pilates-based exercise, stabilization been mitigated by the low methodological rigor of many of these studies. 45,75 Many researchers suggest that patients with LBP are not a homogeneous group and should be classified into subgroups of individuals who share similar clinical characteristics. 13,54 This type of classification system could guide diagnosis and treatment and improve overall decision making in the management of patients with LBP. 23 It may also improve research by denoting homogeneous subgroups for treatment outcomes studies. 12 Clinical prediction rules (CPRs) consist of combinations of variables obtained from self-report measures and historical and clinical examinations. One purpose is to assist with subgrouping patients into specific treatment-based classifications. Recently, CPRs have been shown to be useful in classifying patients with LBP who are likely to benefit from a particular treatment approach, such as spine manipulation and lumbar stabilization. 17,27,36 An advantage of CPRs is that they use the diagnostic properties of sensitivity, specificity, and likelihood ratios, so their interpretations can be readily applied to 1 Doctoral student (at time of study), Graduate Program in Orthopedics, Rocky Mountain University of Health Professions, Provo, UT; Physical Therapist, Steadman Hawkins Clinic Denver, Greenwood Village, CO; Affiliate Faculty Member, Regis University, Denver, CO. 2 Professor, Rocky Mountain University of Health Professions, Provo, UT; Associate Professor, Baylor University, Waco, TX. 3 Director of Musculoskeletal Research, Department of Physical Therapy (MSGS/SGCUY), 81st Medical Group, Keesler Air Force Base, Biloxi, MS. The protocol for this study was approved by the Institutional Review Board for the Rocky Mountain University of Health Professions. The primary author is an Educator for Polestar Pilates Education. The views expressed in this material are those of the authors and do not reflect the official policy or position of the US Government, the Department of Defense, or the Department of the Air Force. Address correspondence to Dr Lise Stolze, Steadman Hawkins Clinic Denver, 8200 East Belleview, Englewood, CO Lstolze@regis.edu journal of orthopaedic & sports physical therapy volume 42 number 5 may

96 [ research report ] individual patients. 53 Because CPRs are designed to improve decision making, it is important they be developed and validated according to rigorous methodological standards. McGinn et al 53 have suggested a 3-step process for developing and testing a CPR prior to widespread implementation of the rule in clinical practice. Once a CPR has been derived, validated, and shown to positively impact clinical behavior, it can be helpful in selecting the most effective treatment for an individual patient. Exercise therapy has been shown to be effective in decreasing pain and improving function in populations with chronic LBP. 34,75 Research has shown that specific exercise programs can be designed to be effective in certain subgroups of patients with LBP. 38,59 Evidence is emerging 14,47 in support of the belief popularized by McKenzie 54 that some patients with LBP respond to direction-specific exercise interventions that include end-range movements. Direction-specific exercises and stabilization exercises are 2 recognized treatment approaches for LBP outlined in the treatment-based classification system introduced by Delitto et al, 23 which aims to match treatments to specific patient categories. Some researchers support stabilization exercise regimens that improve strength of larger spinal muscles (erector spinae, oblique abdominals, and quadratus lumborum), 36,51,52 while others 38,44 have focused on the deep muscles of the spine (ie, multifidus, transversus abdominis), which have been shown to become neurologically inhibited with pain. 39,40 This has increased attention on stabilization programs that emphasize motor control and specific low-threshold training of these muscles. 38,59 Current studies, however, have questioned whether specific muscle retraining may be the most effective approach to stabilization. 15,43 The Pilates method of exercise is a unique mind-body exercise program developed by Joseph Pilates in the early 1900s. Pilates called his method Contrology, 64 and it became popular in the dance community and in dance medicine. The Pilates method incorporates movement principles that include both physical and cognitive elements: whole-body movement, attention to breath, balanced muscle development, concentration, control, centering, precision, and rhythm. 6 Clinical applications of the Pilates movement principles 3 (APPENDIX A) and exercises based on the Pilates method are being implemented by physical therapists as a therapeutic intervention. Pilates-based exercise uses movement enhancement techniques such as tactile and imagery cuing to reinforce the movement principles. The Pilates Reformer is an apparatus that provides an assisted environment through its pulley system and springs, grading movement from assistive to resistive and allowing nonpainful movement to begin early in the rehabilitation phase. Graded movement may be helpful in the treatment of fear-avoidance, 30,46 which can cause faulty motor patterns 7 and has been linked to chronic LBP. 21,63 Pilates-based exercises progress from basic gravity-eliminated movements to complex and functional movements requiring coordination and balance against gravity. It has been postulated that spine rehabilitation may need to focus more on coordination and less on actual strength or muscle torque, suggesting that isolated volitional strength of postural muscles is not as valuable as their coordinated integration. 66 Pilates-based exercise has characteristics of other exercise systems. It focuses on motor control of both global stabilizers and mobilizers as described by Comerford and Mottram 19 and could therefore be effective in the treatment of LBP for patients in both the stabilization and direction-specific categories outlined in the Treatment-Based Classification system. 23 It has been postulated that improved motor control that facilitates accurate anticipation of spinal loads may provide better protection to the intervertebral discs from the harmful effects of sudden loads. 59,60 Much like specific stabilization exercise therapy, 66 Pilates-based exercise emphasizes facilitation techniques, such as tactile cuing and imagery, to encourage skeletal alignment and breathing. Unlike specific stabilization exercise therapy, however, it does not attempt to facilitate conscious activation of any isolated muscle or muscle group. Some have incorporated conscious muscle activation techniques into a Pilates-based exercise program for a combined therapeutic intervention. 67 However, the automatic subconscious activation of low-threshold (local stabilizer 19 ) muscles during a Pilates-based exercise intervention alone, using movement enhancement techniques, warrants further investigation. Lim et al 45 concluded in a recently published systematic review that Pilatesbased exercise is superior to minimal intervention for pain, 67 but that current evidence does not establish superiority of Pilates to other forms of exercise for patients with LBP. Despite early evidence to support the effectiveness of Pilatesbased exercise, no studies to date have determined if there is a parsimonious set of physical, historical, and psychosocial characteristics that predict which patients may likely benefit from Pilates-based exercise. We hypothesized that a parsimonious set of factors would emerge from the clinical examination to identify patients with LBP who would be most likely to benefit from the Pilates-based exercise. METHODS Subjects This was a prospective cohort study in which 96 subjects with LBP were sequentially enrolled. Informed consent was obtained, and the rights of subjects were protected. The protocol for this study was approved by The Institutional Review Board of the Rocky Mountain University of Health Professions. Inclusion criteria included (1) current LBP with or without prior history of LBP, (2) modified Oswestry Disability Questionnaire (ODQ) score of 20% or greater, (3) age of 18 years or greater, and (4) referral to physical therapy by a physician or self-referred. Subjects were excluded if any of the following were present: (1) third trimester of preg- 426 may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

97 nancy; (2) 2 or more signs consistent with nerve root compression (positive straight leg raise test at an angle less than 45 or diminished lower extremity strength, sensory function, or deep tendon reflexes); (3) previous spinal fusion surgery; or (4) evidence of serious pathology (eg, acute spinal fracture, tumor, infection, etc). The desired minimum number of 95 participants for this study was determined using the rule of thumb approach for regression studies, 71 which specifies the need for 15 subjects per predictor variable in the final prediction model. Prior published studies suggest anticipating a 4- or 5-level CPR 27,36 and, therefore, an enrollment of 75 participants. Subsequently, a 15% dropout rate was estimated and added to the sample size, leading to the desired 95 participants. Study Setting The study was conducted primarily at the Steadman Hawkins Clinic in Denver, CO, where 92 subjects were recruited. Two subjects were recruited at Select Physical Therapy Center in Denver and 2 were recruited at Pinnacle Performance in Salt Lake City, UT. Physical Therapists Four licensed physical therapists participated in the examination and treatment of subjects in this study. All therapists had been trained in a comprehensive Pilates training program through Polestar Education ( All therapists received specific training and written instructions in the evaluation and Pilates intervention protocols. Each therapist had at least 2 years of individual experience in using Pilates-based exercise with patients. The primary researcher had 17 years of experience with this exercise approach. Examination Procedure Physical therapists administered a baseline standardized physical examination and collected data including demographic information. Pain at baseline was assessed using an 11-point (0-10) visual analog scale, with 0 representing no pain FIGURE 1. Supine hip and knee extension on the Pilates Reformer. FIGURE 2. Standing hip extension on the Pilates Reformer. and 10 representing emergency-room pain. The subjects also completed a pain diagram and the Fear-Avoidance Belief Questionnaire (FABQ). The FABQ has 2 subscales that measure fear-avoidance beliefs about work (7-item scale) and physical activity (4-item scale). The ODQ assesses disability related to LBP. The ODQ was administered at baseline and after 8 weeks of treatment, and served as the reference standard for determining the success of the treatment program. The health history intake included questions concerning mechanism of injury, nature of current symptoms, and prior episodes of LBP. Subjects were also asked about the distribution of symptoms for their current episode. The physical examination included measurement of lumbar spine range of motion (ROM) and total trunk flexion ROM using an inclinometer. 77 Aberrant motions during lumbar ROM were noted, including instability catch, 57 painful arc of motion, 22 Gowers' sign, or a reversal of lumbopelvic rhythm. 23 Supine straight leg raise and prone hip rotation ROM were measured using a single inclinometer. 77 Generalized ligamentous laxity was assessed on a 9-point scale described by Beighton and Horan. 8 Two special tests FIGURE 3. Prone spine extension on the Pilates Reformer. for lumbar spine instability were performed: the prone instability test 37 and the passive lumbar extension test. 1,42 A posterior/anterior lumbar spring test 48 was performed at each spinal level. Two strength tests, the active sit-up and active bilateral straight leg raise tests, were administered. 77 The extensor endurance test and the side support test 50 were performed to determine muscle endurance of the spinal extensors and lateral flexors. Operational definitions for components of the physical exam are provided in APPENDIX B. A total of 37 potential predictor variables were measured at the baseline assessment session. Treatment Treatment consisted of a standardized Pilates-based exercise program utilizing a Balanced Body Pilates Reformer (Balanced Body Inc, Sacramento, CA) (FIG- URES 1 through 3), with emphasis on tactile and imagery cuing. The Pilates-based exercises are listed in APPENDIX A and include modifications and progressions. One set of 8 to 10 repetitions per exercise was performed during each session. Therapists were instructed to use clinical reasoning skills to omit exercises entirely or to apply the appropriate regression or progression to an exercise. Criteria for exercise omission and modification included movement-direction preference 14,54 and subjective irritability levels. Modification or elimination of an exercise was applied if the subject was unable to perform the exercise in proper form or the subject s pain level increased by performing the exercise. The exercise was resumed at a later stage in the treatment if these difficulties were overcome. The journal of orthopaedic & sports physical therapy volume 42 number 5 may

98 [ research report ] Subjects with back pain screened for eligibility criteria, n = 119 Eligible, n = 96 Not eligible, n = 23 Agreed to participate, signed consent form, received and completed treatment, provided data at end point, n = 95 Dropped out first week due to increased neck pain, n = 1 Failed to score 10+ on ODQ, n = 17 Failed neurological screening, n = 2 Unable to comply with study protocol, n = 4 FIGURE 4. Flow diagram showing recruitment and retention. exercises consisted of a combination of spine stability and mobility movements. Subjects were seen twice per week for 8 weeks. Supplemental instructions were given to each subject, reinforcing the basic Pilates principles of breathing, skeletal alignment, and self-awareness in various relationships to gravity. Subjects were encouraged to practice finding neutral postures in supine, quadruped, sitting, and standing, especially as they encountered these postures in daily activities, and to pay attention to their breathing patterns. The supplemental instructions are described in APPENDIX C. After 8 weeks, a posttest ODQ was administered. Data Analysis Descriptive statistics were calculated to summarize the data. Individual variables from the self-report measures, history, and physical examination were tested for their univariate association with success using independent sample t tests for continuous variables and chi-square tests for categorical variables. Continuous predictor variables were dichotomized by establishing a cut score with receiver-operator curve analysis. This process computes sensitivity and specificity for multiple cut scores along the continuum of the scale, yielding coordinates for a plot so that the characteristics of the scale can be observed graphically. 33 Area under the receiver-operator curve was used as 1 measure of how well each continuous scale predictor performed in this respect. 24 Subjects were grouped according to success or nonsuccess with respect to treatment. Success with treatment was determined by percent change in disability scores on the ODQ after 8 weeks of Pilates-based exercise. Patients who experienced at least a 50% improvement were categorized as having a successful outcome. Those not achieving at least a 50% improvement were classified as having a nonsuccessful outcome. The minimum clinically important difference in ODQ score has been calculated as 5 to 6 points (10%-12% change). 10,23,29,56 Binary logistic regression analysis was used to filter the set of predictor variables further and to derive a multivariate model (CPR) that eliminated redundant or substantially correlated predictors. Potential predictors yielding P values less than or equal to.10 from the t tests and chi-square tests were entered into the logistic regression analysis using a forward stepwise procedure. The predictor variables were chosen for retention by the forward stepwise method if they had significant changes (P.05) in 2 log likelihood of the model, when added from the previous step of model development. For continuous scale variables and categorical scale variables with more than 2 levels, the raw (nondichotomized) variables were entered into the logistic regression analysis. Predictor variables that met the statistical criteria were ultimately accepted, based on their clinical plausibility. Sensitivity, specificity, and positive likelihood ratios (LRs) were calculated for all potential predictor variables. Once the number of predictor variables was determined, the CPR was developed by examining the accuracy statistics for various combinations of the retained variables. 68 Treatment success was defined as a 50% or greater reduction in ODQ score from baseline to completion of treatment. The goal for the final derivation of the CPR was to maximize the positive LR with a clinically sensible set of predictors at a level that yielded a viable proportion of subjects who were positive at that level. RESULTS Ninety-six subjects were enrolled in the study between February 2009 and September One subject dropped out after 1 week, due to increased neck symptoms, and these data 428 may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

99 TABLE 1 were excluded from the analysis. FIGURE 4 provides a flow diagram of subject recruitment and retention. Of the 95 subjects for which data were collected, 15 had prior physical therapy for their symptoms and 6 had previous Pilates exercise experience. Among the 95 subjects completing the study, 89 (93.7%) attended all 16 scheduled treatment sessions, while the History and Demographic Variables Assessed at Baseline* Demographic Variable All Subjects (n = 95) Success (n = 51) Nonsuccess (n = 44) P Value Age, y Body mass index, kg/m <.001 Sex (women), % Duration of symptoms, d 386 (12-16,842) 267 (12-16,842) 559 (38-15,167).003 Duration of symptoms, % mo >6 mo Number of prior episodes Distribution of symptoms Lumbar spine, % Buttock, % Thigh, % Lower leg/foot, % Prior history of LBP (yes), % No leg symptoms last week, % Abbreviation: LBP, low back pain. *Data are mean SD, except where specified for continuous variables, and percents for categorical variables. P values represent a 2 independent samples t test for continuous variables, where groups are defined as success and nonsuccess, and a chi-square test of association for categorical variables (success/nonsuccess is used as the reference criterion). Median and range. Missing data for this variable: total sample, n = 73; success, n = 39; nonsuccess, n = 34. TABLE 2 Self-Report Variables* Self-Report Variable All Subjects (n = 95) Success (n = 51) Nonsuccess (n = 44) P Value Pain rating FABQ work subscale (0-42) FABQ physical activity subscale (0-24) Baseline ODQ score (0-50) Change in ODQ score (0-50) <.001 Abbreviations: FABQ, Fear-Avoidance Belief Questionnaire (subscales where a higher score reflects a greater fear of movement and a greater avoidance of activities); ODQ, modified Oswestry Disability Questionnaire (where a higher score reflects greater disability). *Data are mean SD. P values represent 2 independent-samples t tests, where groups are defined as success and nonsuccess. Missing data for this variable: n = 84 for all subjects, n = 46 for success, n = 38 for nonsuccess. remaining 6 subjects attended at least 13 of the 16 sessions. Fifty-one (53.7%) subjects who completed the study experienced a 50% or greater improvement in their ODQ scores over 8 weeks. TABLE 1 provides subject descriptive information. TABLE 2 summarizes results for selfreport variables. Pain rating averaged a mean SD of on a 0-to-10- point scale, and the FABQ work and physical activity subscales averaged and , respectively. Eleven demographic, self-report, and physical exam variables (TABLE 3) were significantly associated with success and thus further analyzed as potential predictors (TABLE 4): FABQW less than or equal to 10, body mass index (BMI) of 25 kg/ m 2 or greater, total trunk flexion ROM of 70 or less, positive prone instability test, positive active sit-up test, duration of symptoms of 6 months or less, number of prior episodes of LBP less than or equal to 3, left- or right-side support of 30 seconds or longer, right or left hip average rotation of 25 or greater, and 2 variables from the self-report indicating that subjects currently experienced no lower extremity symptoms or that these symptoms were not currently bothersome. These variables were entered into the logistic regression modeling. Stepwise methods used in the modeling resulted in 5 predictors of success that were considered for the multivariate CPR (TABLE 5). The final model was statistically significant (P<.001), with reasonable goodness of fit (Hosmer-Lemeshow chi-square, 4.21; df = 8; P =.838). The best rule for predicting success (TABLE 6) was the presence of 3 or more of the 5 attributes (positive LR, 10.64; 95% confidence interval [CI]: 3.52, 32.14). Thirty-two of 35 subjects who were positive on 3 or more of the 5 criteria were in the successfuloutcome group (TABLE 5). Accuracy statistics were calculated for each threshold for number of attributes present (TABLE 6). A patient exhibiting 4 or more of the 5 attributes would have a 96% (95% CI: 61%, 100%) posttest probability of success and a positive LR of A patient exhibiting 3 or more of the 5 attributes would have a 93% (95% CI: 81%, 97%) posttest probability of success, and a positive LR of (95% CI: 3.52, 32.14). If a subject met fewer than 3 variables, the posttest probability of success was no greater than chance and did not inform the decision making about the use of the Pilates-based exercise. journal of orthopaedic & sports physical therapy volume 42 number 5 may

100 [ research report ] TABLE 3 DISCUSSION Physical Exam Variables Assessed at Baseline* Physical Exam Variable All Subjects (n = 95) Success (n = 51) Nonsuccess (n = 44) P Value Total trunk flexion ROM, deg Pelvic flexion ROM Lumbar flexion ROM Total trunk extension ROM, deg Left sidebending ROM, deg Right sidebending ROM, deg Average sidebending ROM left and right, deg SLR left lower extremity, deg SLR right lower extremity, deg Average hip rotation ROM right, deg Average hip rotation ROM left, deg Painful arc on return of trunk <.001 flexion, % Status change in symptoms with trunk movement (yes), % Beighton and Horan test (0-9) Prone instability test, %.07 No pain with PA pressure Pain with no relief (negative test) Pain with relief (positive test) Passive lumbar extension test (positive), % Lumbar segmental spring test (positive), % Active sit-up test (positive), % Active bilateral SLR (positive), % Extensor endurance test, s Side support test left, s Side support test right, s Abbreviations: PA, posterior to anterior; ROM, range of motion; SLR, straight leg raise. *Data are mean SD for continuous variables and percents for categorical variables. P values represent a 2 independent sample t test for continuous variables, where groups are defined as success and nonsuccess, and a chi-square test of association for categorical variables (success/nonsuccess was used as the reference criterion). Values are the average for hip internal and external rotation for the right hip and the left hip. Though Pilates-based exercise has gained in popularity as an option for the conservative management of LBP, evidence for its effectiveness is sparse and inconclusive. 45 Lim et al 45 concluded that the relatively low quality of existing studies and the heterogeneity of studies they reviewed suggest that results should be interpreted with caution. However, the few studies that have examined homogeneous subgroups of patients with specific exercise programs have been promising. 36,44,59 Hicks et al 36 established a preliminary CPR for success with stabilization exercises. The purpose of our study was to derive a preliminary CPR for identifying a subgroup of patients with LBP likely to benefit from the Pilatesbased exercise. In this study, the 54% pretest probability of success shifted to 96% with a positive LR of if a subject exhibited 4 or more of the 5 criteria in the preliminary prediction rule, and to 93% with a positive LR of if a subject exhibited 3 or more of the 5 criteria in the CPR. However, the 4+ level of the CPR was so specific that a relatively small percentage of subjects in the study (14%) met that criterion. Therefore, we selected the 3+ level as a clinically sensible threshold, because 42% of subjects presented with 3 or more positive tests. If this CPR can be validated with a randomized controlled trial, it is anticipated that recommending the 3+ level may help select appropriate treatment for about 40% of patients presenting for treatment who are similar to the subjects with LBP who participated in this study. Limitations One limitation to this study is that 81.1% of subjects were female. While this proportion accurately reflects the gender bias in the industry, 70 the consequence of this demographic may be that this CPR applies more to women than to men. There have been criticisms of CPR development 31 and suggestions for improving methodological quality. 9 This study incorporated research design elements intended to help ensure methodological quality for derivation of interventional CPRs. Our own assessment of the 18 quality criteria proposed by Beneciuk et al 9 resulted in a score of 67%, which was above the suggested threshold of 60% for a high-quality study. 9 Though our sample size was relatively small, we met the criterion of including at least 10 subjects with the outcome of interest for each predictor variable in the final model, a protection intended to avoid overfitting of multivariate models. 20 One specific criticism of previous CPR derivation studies is that cause-effect relationships cannot be inferred from single-arm trials. 31 However, we make no inferences regarding the efficacy or effectiveness of Pilates-based exercise for patients with LBP based on evidence 430 may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

101 from this current study, as results could potentially be attributable to natural history (ie, passage of time). Rather, this study follows the publication of multiple randomized controlled trials that have established Pilates-based exercise as a viable treatment option for patients with LBP. 4,45,67 In the validation process for this CPR, researchers should consider the broader set of predictor variables used to develop this CPR rather than limiting specific attention to only the final variables retained for the rule. Additionally, a validation study should include a longterm follow-up and comparison group to further investigate the predictive value of the variables in the preliminary CPR. If the rule is validated, an impact analysis of implementation of the rule on clinical practice patterns, outcomes, and cost of care should be investigated. Future research should also consider whether similar predictors emerge if the Pilatesbased exercise is delivered in a group setting versus an individual setting when considering cost of care. Predictor Variables This CPR includes 5 predictor variables that would require minimal time to assess as part of a comprehensive patient evaluation. BMI is a value calculated based on height and weight measurements and is significantly correlated with body fat content. 55 A BMI equal to or greater than 25 kg/m 2 is considered overweight and was a strong predictor of success (P<.001) for subjects in this study. Research has demonstrated that high BMI has a strong association with LBP, 35,69 and that being very overweight can change static and dynamic spine mechanics, including increased anterior pelvic tilt and limited thoracic flexion during forward-bend activities. 76 Both of these mechanical changes could adversely affect the stresses placed on the lumbar spine. Another predictor for success was total trunk flexion ROM of less than or equal to 70. Although the use of end-range motion measurements for outcome determinants in patients with LBP has been TABLE 4 Accuracy Statistics with 95% Confidence Intervals for Individual Predictor Variables Variable Sensitivity (95% CI) Specificity (95% CI) Positive Likelihood Ratio (95% CI) FABQW* (0.41, 0.68) 0.64 (0.49, 0.76) 1.51 (0.95, 2.40) Total trunk flexion ROM (0.17, 0.41) 0.95 (0.85, 0.99) 6.04 (1.45, 25.13) Active sit-up test positive 0.33 (0.22, 0.47) 0.86 (0.73, 0.94) 2.44 (1.06, 5.66) No leg symptoms in the last week 0.82 (0.70, 0.90) 0.57 (0.42, 0.70) 1.91 (1.33, 2.74) No distribution of symptoms in thigh/leg 0.55 (0.41, 0.68) 0.68 (0.53, 0.80) 1.73 (1.05, 2.84) BMI 25 kg/m (0.65, 0.88) 0.59 (0.44, 0.72) 1.92 (1.31, 2.81) Duration of symptoms 6 mo 0.47 (0.34, 0.60) 0.82 (0.68, 0.90) 2.59 (1.30, 5.17) Number of prior episodes (0.22, 0.47) 0.86 (0.73, 0.94) 2.44 (1.06, 5.66) Prone instability test positive 0.55 (0.41, 0.68) 0.57 (0.42, 0.70) 1.27 (0.84, 1.94) Side support test right or left 30 s 0.86 (0.60, 0.96) 0.60 (0.50, 0.70) 2.17 (1.54, 3.06) Right or left hip average rotation (0.45, 0.94) 0.57 (0.46, 0.87) 1.81 (1.18, 2.77) Abbreviations: BMI, body mass index; CI, confidence interval; FABQW, Fear-Avoidance Belief Questionnaire work subscale; ROM, range of motion. *Scores range from 0 to 42, with a higher score reflecting a greater fear of movement and a greater avoidance of activities with respect to work. Missing data (combined total cases present, n = 66). At least 1 side in the side support test is 30 s or longer. At least 1 hip with an average internal and external rotation of 25 or greater. TABLE 5 questioned, 73 such measurements are often used in physical therapy initial assessments to screen for other impairments. Limited motion may reflect pain-related fear, which often results in avoidance behavior that specifically limits or restricts motion of the lumbar spine, 74 even in the presence of lower FABQ scores. Evidence exists for a relationship between hip joint flexibility and LBP. 18,26 Hip joint ROM discrepancy was a variable in the preliminary CPR developed The Variables for the Clinical Prediction Rule* and the Number of Subjects in Each Group at Each Level Number of Predictor Variables Present Successful Outcome Group Nonsuccessful Outcome Group *No leg symptoms in the last week; body mass index 25 kg/m 2 ; total trunk flexion 70 ; left or right hip average rotation 25 (at least 1 hip with an average internal and external rotation of 25 or greater); duration of symptoms 6 months. Total cases present, n = 95 (1 subject had 0 criteria pres ent). Plus sign indicates or more. for spinal manipulation. 17,27 In our study, subjects who had a mean internal and external hip rotation ROM of less than or equal to 25 in either hip tended not to be successful with treatment. Restricted hip rotation ROM has been established as a clinical indicator of hip osteoarthritis 2,41 and may complicate the treatment of LBP in those with this comorbidity. We excluded subjects with signs of nerve root compression. However, subjects with distal symptoms in the absence journal of orthopaedic & sports physical therapy volume 42 number 5 may

102 [ research report ] TABLE 6 Accuracy Statistics With 95% Confidence Intervals for the 5 Levels of the Clinical Prediction Rule* Number of Predictor Variables Present Sensitivity Specificity Positive Likelihood Ratio Probability of Success (%) (0.01, 0.15) 0.99 (0.90, 1.00) 4.33 (0.21, 87.78) 84 (21-99) (0.15, 0.40) 0.99 (0.90, 1.00) (1.43, ) 96 (61-100) (0.58, 0.84) 0.93 (0.81, 0.99) (3.52, 32.14) 93 (81-97) (0.87, 1.00) 0.36 (0.22, 0.52) 1.51 (1.20, 1.90) 64 (58-69) (0.91, 1.00) 0.03 (0.00, 0.14) 1.02 (0.96, 1.09) 54 (53-56) Abbreviation: CI, confidence interval. *The probability of success is calculated using the positive likelihood ratios and assumes a pretest probability of 54%; total cases present, n = 95. Plus sign indicates or more. of positive neurologic signs were included in the study. Subjects who experienced leg symptoms within the week prior to enrollment tended not to succeed. This is consistent with studies demonstrating that up to 40% of patients with leg pain who are treated conservatively undergo delayed surgery. 58,62 Symptom duration of less than or equal to 6 months was a predictor of treatment success, even though 72% of all subjects in this study reported having symptoms for more than 6 months. Duration of symptoms for more than 12 weeks may be classified as chronic LBP 11 and is often associated with physical disabilities, psychological distress, depression, and inability to work. 72 Of those who remain disabled with back pain for more than 6 months, fewer than half return to work, 5 and they tend to have poor expectations for their back pain outcome. 28 The results in this study may be explained by the natural history of acute back pain, which is favorable without treatment, 25 even though its incidence of recurrence is about 40% in 6 months. 16 In addition, those who were not successful with treatment also experienced on average more than 3 (mean SD, ) prior episodes of LBP. It is likely that the group in this study of those who had persistent pain of more than 6 months in duration is a combination of subjects with continuous pain and those who had multiple recurrences. It is interesting to note that none of the variables testing trunk strength or stability were retained in the final model, given that they are theoretically key indications for a spine stabilization intervention. This suggests that the indication for the Pilates-based exercise may not be limited to those patients who have spine instability but may include a wider range of patients, such as those in whom reduced spine and extremity mobility is contributing to their LBP symptoms. CONCLUSION Five predictors collected from the clinical examination comprised a clinically sensible preliminary CPR to identify individuals with LBP who are likely to respond to treatment using Pilates-based exercise. These predictors were total trunk flexion ROM of 70 or less, duration of current symptoms of 6 months or less, no leg symptoms in the previous week, BMI of 25 kg/m 2 or greater, and left or right hip average rotation ROM of 25 or greater. If any 3 or more of the 5 attributes were present (which occurred in 42% of study subjects), the positive LR was 10.64, thus sufficient to yield a large shift from pretest to posttest probability for experiencing a successful treatment outcome. t KEY POINTS FINDINGS: A preliminary CPR with 5 variables was identified: total trunk flexion ROM less than or equal to 70, duration of current symptoms for 6 months or less, no leg symptoms in the previous week, BMI greater than or equal to 25 kg/m 2, and left or right hip average rotation of 25 or greater. If 3 or more attributes were present (positive LR, 10.64), the probability of experiencing a successful outcome increased from 54% to 93%. IMPLICATIONS: These data provide preliminary evidence to support the idea that the response to Pilates-based exercise in patients with LBP can be predicted from variables collected from the clinical examination. CAUTION: These results must be validated in a randomized controlled trial before clinicians can be confident that the CPR will be useful to improve clinical decision making in determining which patients are most likely to benefit from Pilates-based exercise. Insofar as 81.1% of subjects in this study were female, this CPR may apply more to women than to men. ACKNOWLEDGEMENTS: We would like to acknowledge Brent D. Anderson, PT, PhD, OCS, President, Polestar Physical Therapy and Polestar Pilates Education, Coral Gables, FL, for his contribution to the concept and design of the Pilates protocol in this study and his guidance as subject expert for the dissertation committee. REFERENCES 1. Alqarni AM, Schneiders AG, Hendrick PA. Clinical tests to diagnose lumbar segmental instability: a systematic review. J Orthop Sports Phys Ther. 2011;41: jospt Altman R, Alarcon G, Appelrouth D, et al. The 432 may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

103 American College of Rheumatology criteria for the classification REFERENCES and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991;34: Anderson B. Pilates rehabilitation. In: Davis CM, ed. Complementary Therapies in Rehabilitation: Evidence for Efficacy in Therapy, Prevention, and Wellness. 3rd ed. Thorofare, NJ: SLACK Incorporated; 2009: Anderson BA. Randomized clinical trial comparing active vs. passive approaches to the treatment of recurrent and chronic low back pain [thesis]. Miami, FL: University of Miami; Andersson GB. Epidemiological features of chronic low-back pain. Lancet. 1999;354: S (99) Appel C, Betz S, Bowen K, Ickes D-M, Anderson B. The PMA Pilates Certification Exam Study Guide. Miami, FL: Pilates Method Alliance; Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P. The influence of low back pain on muscle activity and coordination during gait: a clinical and experimental study. Pain. 1996;64: Beighton P, Horan F. Orthopaedic aspects of the Ehlers-Danlos syndrome. J Bone Joint Surg Br. 1969;51: Beneciuk JM, Bishop MD, George SZ. Clinical prediction rules for physical therapy interventions: a systematic review. Phys Ther. 2009;89: ptj Beurskens AJ, de Vet HC, Koke AJ. Responsiveness of functional status in low back pain: a comparison of different instruments. Pain. 1996;65: Bogduk N, McGuirk B. Medical Management of Acute and Chronic Low Back Pain: An Evidence- Based Approach. Amsterdam, The Netherlands: Elsevier; Bouter LM, van Tulder MW, Koes BW. Methodologic issues in low back pain research in primary care. Spine (Phila Pa 1976). 1998;23: Brennan GP, Fritz JM, Hunter SJ, Thackeray A, Delitto A, Erhard RE. Identifying subgroups of patients with acute/subacute nonspecific low back pain: results of a randomized clinical trial. Spine (Phila Pa 1976). 2006;31: dx.doi.org/ /01.brs a8 14. Browder DA, Childs JD, Cleland JA, Fritz JM. Effectiveness of an extension-oriented treatment approach in a subgroup of subjects with low back pain: a randomized clinical trial. Phys Ther. 2007;87: ptj Cairns MC, Foster NE, Wright C. Randomized controlled trial of specific spinal stabilization exercises and conventional physiotherapy for recurrent low back pain. Spine (Phila Pa 1976). 2006;31:E brs d 16. Carey TS, Garrett JM, Jackman A, Hadler N. Recurrence and care seeking after acute back pain: results of a long-term follow-up study. North Carolina Back Pain Project. Med Care. 1999;37: Childs JD, Fritz JM, Flynn TW, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study. Ann Intern Med. 2004;141: Cibulka MT, Sinacore DR, Cromer GS, Delitto A. Unilateral hip rotation range of motion asymmetry in patients with sacroiliac joint regional pain. Spine (Phila Pa 1976). 1998;23: Comerford MJ, Mottram SL. Functional stability re-training: principles and strategies for managing mechanical dysfunction. Man Ther. 2001;6: math Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med. 1993;118: Crombez G, Eccleston C, Vlaeyen JW, Vansteenwegen D, Lysens R, Eelen P. Exposure to physical movements in low back pain patients: restricted effects of generalization. Health Psychol. 2002;21: Cyriax J. Textbook of Orthopaedic Medicine. Volume 1: Diagnosis of Soft Tissue Lesions. Baltimore, MD: Williams & Wilkins; Delitto A, Erhard RE, Bowling RW. A treatmentbased classification approach to low back syndrome: identifying and staging patients for conservative treatment. Phys Ther. 1995;75: ; discussion Deyo RA, Centor RM. Assessing the responsiveness of functional scales to clinical change: an analogy to diagnostic test performance. J Chronic Dis. 1986;39: Deyo RA, Weinstein JN. Low back pain. N Engl J Med. 2001;344: org/ /nejm Ellison JB, Rose SJ, Sahrmann SA. Patterns of hip rotation range of motion: a comparison between healthy subjects and patients with low back pain. Phys Ther. 1990;70: Flynn T, Fritz J, Whitman J, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine (Phila Pa 1976). 2002;27: org/ /01.brs d 28. Foster NE, Bishop A, Thomas E, et al. Illness perceptions of low back pain patients in primary care: what are they, do they change and are they associated with outcome? Pain. 2008;136: Fritz JM, Irrgang JJ. A comparison of a modified Oswestry Low Back Pain Disability Questionnaire and the Quebec Back Pain Disability Scale. Phys Ther. 2001;81: George SZ, Fritz JM, Bialosky JE, Donald DA. The effect of a fear-avoidance-based physical therapy intervention for patients with acute low back pain: results of a randomized clinical trial. Spine (Phila Pa 1976). 2003;28: dx.doi.org/ /01.brs A2 31. Hancock M, Herbert RD, Maher CG. A guide to interpretation of studies investigating subgroups of responders to physical therapy interventions. Phys Ther. 2009;89: org/ /ptj Hancock MJ, Maher CG, Latimer J, et al. Systematic review of tests to identify the disc, SIJ or facet joint as the source of low back pain. Eur Spine J. 2007;16: org/ /s Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143: Hayden JA, van Tulder MW, Tomlinson G. Systematic review: strategies for using exercise therapy to improve outcomes in chronic low back pain. Ann Intern Med. 2005;142: Heuch I, Hagen K, Nygaard O, Zwart JA. The impact of body mass index on the prevalence of low back pain: the HUNT study. Spine (Phila Pa 1976). 2010;35: org/ /brs.0b013e3181ba Hicks GE, Fritz JM, Delitto A, McGill SM. Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program. Arch Phys Med Rehabil. 2005;86: Hicks GE, Fritz JM, Delitto A, Mishock J. Interrater reliability of clinical examination measures for identification of lumbar segmental instability. Arch Phys Med Rehabil. 2003;84: Hides JA, Jull GA, Richardson CA. Long-term effects of specific stabilizing exercises for firstepisode low back pain. Spine (Phila Pa 1976). 2001;26:E Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine (Phila Pa 1976). 1996;21: Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine (Phila Pa 1976). 1996;21: Hoppenfeld S, Hutton R. Physical Examination of the Spine and Extremities. New York, NY: Appleton-Century-Crofts; Kasai Y, Morishita K, Kawakita E, Kondo T, Uchida A. A new evaluation method for lumbar spinal instability: passive lumbar extension test. Phys Ther. 2006;86: org/ /ptj Koumantakis GA, Watson PJ, Oldham JA. Trunk muscle stabilization training plus general exercise versus general exercise only: randomized controlled trial of patients with recurrent low back pain. Phys Ther. 2005;85: Kumar SP. Efficacy of segmental stabilization exercise for lumbar segmental instability in patients with mechanical low back pain: a randomized placebo controlled crossover study. N Am J Med Sci. 2011;3: MORE INFORMATION Lim EC, Poh RL, Low AY, Wong WP. Effects of Pilates-based exercises on pain and disability in journal of orthopaedic & sports physical therapy volume 42 number 5 may

104 [ research report ] individuals with persistent nonspecific low back pain: a systematic review with meta-analysis. J Orthop Sports Phys Ther. 2011;41: dx.doi.org/ /jospt Lindstrom I, Ohlund C, Eek C, et al. The effect of graded activity on patients with subacute low back pain: a randomized prospective clinical study with an operant-conditioning behavioral approach. Phys Ther. 1992;72: ; discussion Machado LA, de Souza MS, Ferreira PH, Ferreira ML. The McKenzie method for low back pain: a systematic review of the literature with a meta-analysis approach. Spine (Phila Pa 1976). 2006;31:E brs Maher CG, Simmonds M, Adams R. Therapists conceptualization and characterization of the clinical concept of spinal stiffness. Phys Ther. 1998;78: McGill SM. Low back exercises: evidence for improving exercise regimens. Phys Ther. 1998;78: McGill SM, Childs A, Liebenson C. Endurance times for low back stabilization exercises: clinical targets for testing and training from a normal database. Arch Phys Med Rehabil. 1999;80: McGill SM, Cholewicki J. Biomechanical basis for stability: an explanation to enhance clinical utility. J Orthop Sports Phys Ther. 2001;31: McGill SM, Grenier S, Kavcic N, Cholewicki J. Coordination of muscle activity to assure stability of the lumbar spine. J Electromyogr Kinesiol. 2003;13: McGinn TG, Guyatt GH, Wyer PC, Naylor CD, Stiell IG, Richardson WS. Users guides to the medical literature: XXII: how to use articles about clinical decision rules. Evidence-Based Medicine Working Group. JAMA. 2000;284: McKenzie RA. The Lumbar Spine: Mechanical Diagnosis and Therapy. Waikanae, New Zealand: Spinal Publications; Mei Z, Grummer-Strawn LM, Pietrobelli A, Goulding A, Goran MI, Dietz WH. Validity of body mass index compared with other body-composition screening indexes for the assessment of body fatness in children and adolescents. Am J Clin Nutr. 2002;75: Nachemson AL. Advances in low-back pain. Clin Orthop Relat Res. 1985; Ogon M, Bender BR, Hooper DM, et al. A dynamic approach to spinal instability. Part I: sensitization of intersegmental motion profiles to motion direction and load condition by instability. Spine (Phila Pa 1976). 1997;22: Osterman H, Seitsalo S, Karppinen J, Malmivaara A. Effectiveness of microdiscectomy for lumbar disc herniation: a randomized controlled trial with 2 years of follow-up. Spine (Phila Pa 1976). 2006;31: org/ /01.brs O Sullivan PB, Phyty GD, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine (Phila Pa 1976). 1997;22: O Sullivan PB, Twomey L, Allison GT. Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. J Orthop Sports Phys Ther. 1998;27: Panjabi MM. Clinical spinal instability and low back pain. J Electromyogr Kinesiol. 2003;13: Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery versus prolonged conservative treatment for sciatica. N Engl J Med. 2007;356: NEJMoa Picavet HS, Vlaeyen JW, Schouten JS. Pain catastrophizing and kinesiophobia: predictors of chronic low back pain. Am J Epidemiol. 2002;156: Pilates JH, Miller WJ. Return to Life Through Contrology. Locust Valley, NY: J. J. Augustin; Prevalence and most common causes of disability among adults--united States, MMWR Morb Mortal Wkly Rep. 2009;58: Richardson C, Jull G, Hodges P, Hides J. Overview of the principles of clinical management of the deep muscle system for segmental stabilization. In: Richardson C, Jull G, Hodges P, Hides J, eds. Therapeutic Exercise for Spinal Segmental Stabilization in Low Back Pain: Scientific Basis and Clinical Approach. Edinburgh, Scotland: Churchill Livingstone; 1998: Rydeard R, Leger A, Smith D. Pilates-based therapeutic exercise: effect on subjects with nonspecific chronic low back pain and functional APPENDIX A disability: a randomized controlled trial. J Orthop Sports Phys Ther. 2006;36: dx.doi.org/ /jospt Sackett DL. The rational clinical examination. A primer on the precision and accuracy of the clinical examination. JAMA. 1992;267: Shiri R, Karppinen J, Leino-Arjas P, Solovieva S, Viikari-Juntura E. The association between obesity and low back pain: a meta-analysis. Am J Epidemiol. 2010;171: org/ /aje/kwp Sporting Goods Manufacturers Association. Pilates Training Participation Report Silver Spring, MD: Sporting Goods Manufacturers Association; Stevens J. Applied Multivariate Statistics for the Social Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates; Sullivan MJ, Reesor K, Mikail S, Fisher R. The treatment of depression in chronic low back pain: review and recommendations. Pain. 1992;50: Sullivan MS, Shoaf LD, Riddle DL. The relationship of lumbar flexion to disability in patients with low back pain. Phys Ther. 2000;80: Thomas JS, France CR. The relationship between pain-related fear and lumbar flexion during natural recovery from low back pain. Eur Spine J. 2008;17: s van Tulder M, Malmivaara A, Esmail R, Koes B. Exercise therapy for low back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine (Phila Pa 1976). 2000;25: Vismara L, Menegoni F, Zaina F, Galli M, Negrini S, Capodaglio P. Effect of obesity and low back pain on spinal mobility: a cross sectional study in women. J Neuroeng Rehabil. 2010;7:3. dx.doi.org/ / Waddell G, Somerville D, Henderson I, Newton M. Objective clinical evaluation of physical impairment in chronic low back pain. Spine (Phila Pa 1976). MORE INFORMATION PILATES REFORMER EXERCISES FOR LOW BACK PAIN Exercise, Pilates Nomenclature 6 Supine hip and knee extension (FIGURE 1), footwork Clinical Applications of the Pilates Spine Position Principles 3 Variations and Modifications Neutral Breathing, axial elongation, core control Vary hip, foot, and ankle MORE INFORMATION may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

105 APPENDIX A Exercise, Pilates Nomenclature 6 Spine Position Clinical Applications of the Pilates Principles 3 Variations and Modifications Supine shoulder extension, hundred (prep) Neutral, flexion Breathing, core control, shoulder girdle organization Increase (or decrease) knee and hip flexion, posterior pelvic tilt Supine long-leg femur arcs, feet in straps and long spine Neutral Breathing, core control Straps above knees, knees flexed, hip circumduction Supine bridging, pelvic lift Flexion Breathing, spine articulation Limit motion, neutral spine, single leg Quadruped hip extension, knee stretch Neutral Breathing, core control, shoulder girdle organization Tall kneeling shoulder extension, chest expansion Z-sit lateral flexion and rotation, mermaid Neutral Lateral flexion and rotation Breathing, core control, shoulder girdle organization Breathing, axial elongation, spine articulation, movement integration Prone spine extension (FIGURE 3) Extension Breathing, axial elongation, upper extremity alignment, spine articulation Standing hip extension (FIGURE 2) Neutral, extension Breathing, lower extremity alignment, movement integration Standing hip abduction, side splits Neutral Breathing, axial elongation, lower extremity alignment Modify spring tension, face head of Reformer to resist hip flexors Modify spring tension, seated on box Neutral hip position, limit range of motion Limit motion, adjust foot bar, adjust spring tension Reduce lower and upper extremity assistance Narrow stance, add support, flex hips and knees, adjust spring tension APPENDIX B Physical Measures Range of motion (ROM) Total trunk flexion ROM 77 OPERATIONAL DEFINITIONS FOR PHYSICAL MEASURES PROCEDURES Procedure The patient stands and an inclinometer is held at T12-L1. The patient is asked to reach down as far as possible toward the toes while keeping the knees straight. Pelvic flexion ROM 77 Same as above but inclinometer is placed at S2. Lumbar flexion ROM 77 Subtract pelvic flexion from total flexion. Total trunk extension ROM 77 The patient stands and an inclinometer is held at T12-L1. The patient is asked to arch backward as far as possible. Right and left sidebending ROM 77 The patient stands with an inclinometer aligned vertically in line with the spinous processes of T9 to T12. The patient is asked to lean over to 1 side as far as possible with the fingertips reaching down the side of the thigh. Right and left straight leg raise (SLR) ROM 77 Hip passive rotation ROM test 77 Muscle performance tests Side support test 49 The patient is supine. The inclinometer is positioned on the tibial crest just below the tibial tubercle. The lower extremity is raised passively by the examiner, whose other hand maintains the knee in extension. The lower extremity is raised slowly to the maximum tolerated SLR (not the onset of pain). The patient is lying prone. The knee is flexed to 90 and the lower leg is placed in vertical alignment. The inclinometer is placed on the distal aspect of the fibula and set at 0. Measurement of hip internal rotation and external rotation is recorded as the angle when the pelvis first begins to move. The patient is sidelying with lower extremities extended and the top foot in front of the lower foot. While resting on the elbow in contact with the table for support, the patient lifts the hips off the table, with only the elbow and feet remaining in contact with the table. The patient is instructed to hold this position as long as possible. The test is done for both sides, and the performance time is recorded in seconds. journal of orthopaedic & sports physical therapy volume 42 number 5 may

106 [ research report ] APPENDIX B Physical Measures Muscle performance tests (continued) Extensor endurance test 45 Active sit-up test 77 Active bilateral straight leg raise test 77 Special tests Prone instability test 50 Lumbar segmental spring testing for mobility 48 Passive lumbar extension test 38 Procedure The patient is asked to lie prone while holding the sternum off the floor for as long as possible. A small pillow is placed under the lower abdomen to decrease the lumbar lordosis. The patient also needs to maintain maximum flexion of the cervical spine and pelvic stabilization through gluteal contraction. The patient is asked to hold this position as long as possible, not to exceed 5 minutes. The performance time is recorded in seconds. The patient is supine and is asked to flex the knees to 90 and place the soles of the feet flat on the surface of the table. The examiner holds both feet down with 1 hand. The patient is instructed to reach up with the fingertips of both hands to touch (not hold) both knees and hold the position for 5 seconds. If the patient cannot maintain this position for 5 seconds, the test is positive. The patient is supine and is asked to lift both legs together 6 inches (15.24 cm) off the examining surface and hold that position for 5 seconds. Both heels and calves should be cleared from the examining surface. If the patient cannot maintain this position for 5 seconds, the test is positive. The patient lies prone with the body on the examining table and lower extremities over the edge and feet resting on the floor. While the patient rests in this position, the examiner applies posterior/anterior pressure to the lumbar spine. Any provocation of pain is reported. Then the patient lifts the lower extremities off the floor (the patient may hold table to maintain position) and a posterior/anterior compression is applied again to the lumbar spine. If pain is present in the resting position but subsides in the second position, the test is positive. The patient is prone. The L1 spinous process is contacted with the examiner s thenar eminence, and a posterior/anterior-directed force is applied. The procedure is repeated at each lumbar level. Mobility is judged as hypermobile or hypomobile. The subject is in the prone position; both lower extremities are elevated together to a height of about 30 cm from the bed while maintaining the knees extended and gently pulling the legs. The test was judged to be positive when, during elevation of both lower extremities, the subject complained of strong pain in the lumbar region, including low back pain, very heavy feeling on the low back, and feeling as if the low back was coming off, and such pain disappeared when the lower extremities returned to the initial position. APPENDIX C SUPPLEMENTAL INSTRUCTIONS Phase Instruction Clinical Application of the Pilates Principles 3 Phase 1: neutral spine, diaphragmatic breathing Phase 2: doorway stretch To be practiced in all functional positions, including: supine, hook-lying, prone, quadruped, sitting, and standing Stand in doorway, arms against doorframe, shoulders are abducted 90, elbows are flexed 90 Breathing, axial elongation, core control, shoulder girdle organization, lower and upper extremity alignment Breathing, axial elongation, core control, shoulder girdle organization, lower and upper extremity alignment 436 may 2012 volume 42 number 5 journal of orthopaedic & sports physical therapy

107 doi: /brain/awn154 Reorganization of the motor cortex is associated with postural control deficits in recurrent low back pain H. Tsao, 1 M. P. Galea 2 and P. W. Hodges 1 Brain (2008), 131, 2161^ NHMRC Centre of Clinical Research Excellence in Spinal Pain, Injury and Health, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane and 2 School of Physiotherapy, The University of Melbourne, Melbourne, Australia Correspondence to: Dr Paul Hodges, NHMRC Centre of Clinical Research Excellence in Spinal Pain, Injury and Health, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane, Qld 4072, Australia p.hodges@uq.edu.au Many people with recurrent low back pain (LBP) have deficits in postural control of the trunk muscles and this may contribute to the recurrence of pain episodes. However, the neural changes that underlie these motor deficits remain unclear. As the motor cortex contributes to control of postural adjustments, the current study investigated the excitability and organization of the motor cortical inputs to the trunk muscles in 11 individuals with and without recurrent LBP. EMG activity of the deep abdominal muscle, transversus abdominis (TrA), was recorded bilaterally using intramuscular fine-wire electrodes. Postural control was assessed as onset of TrA EMG during single rapid arm flexion and extension tasks. Motor thresholds (MTs) for transcranial magnetic stimulation (TMS) were determined for responses contralateral and ipsilateral to the stimulated cortex. In addition, responses of TrA totms over the contralateral cortex were mapped during voluntary contractions at 10% of maximum. MTs and map parameters [centre of gravity (CoG) and volume] were compared between healthy and LBP groups. The CoG of the motor cortical map of TrA in the healthy group was»2cm anterior and lateral to the vertex, but was more posterior and lateral in the LBP group.the location of the CoG and the map volume were correlated with onset of TrA EMG during rapid arm movements. Furthermore, the MT needed to evoke ipsilateral responses was lower in the LBP group, but only on the less excitable hemisphere.these findings provide preliminary evidence of reorganization of trunk muscle representation at the motor cortex in individuals with recurrent LBP, and suggest this reorganization is associated with deficits in postural control. Keywords: motor cortex; postural control; transcranial magnetic stimulation; abdominal muscles Abbreviations: LBP = low back pain; MVC = maximum voluntary contraction; MT = motor thresholds; TMS = transcranial magnetic stimulation; TrA = transversus abdominis Received April 23, Revised June 6, Accepted June 23, 2008 Introduction Described as a Western epidemic, low back pain (LBP) is the most common cause of work-related absence in western society (Blyth et al., 2001). Studies show 50 80% of adults in the general population will suffer this condition at some stage in their lives, and 15 30% will have LBP at any given time (Andersson, 1998). While many individuals will recover within 1 month (Pengel et al., 2003), most people will have recurrence of pain episodes within a 12-month period (Cassidy et al., 2005; Wasiak et al., 2006). A possible contributor to the persistence or recurrence of this condition is changes in postural control of the trunk muscles. Several studies have demonstrated delayed activation of the deep abdominal (Hodges and Richardson, 1996) and back muscles (Leinonen et al., 2001), and increased activity of superficial trunk muscles (Arendt-Nielsen et al., 1996; Radebold et al., 2001) in patients with recurrent LBP. Many of these changes persist after the resolution of symptoms (Hodges and Richardson, 1996) and have been argued to contribute to the recurrence of LBP episodes (Hodges and Moseley, 2003; Cholewicki et al., 2005). However, exactly how the organization of control of these responses in the motor system is changed with pain remains unclear. ß The Author (2008). Published by Oxford University Press on behalf ofthe Guarantors of Brain. Allrights reserved. For Permissions, please journals.permissions@oxfordjournals.org

108 2162 Brain (2008), 131, 2161^2171 H. Tsao et al. The motor cortex provides a critical contribution to postural control (see review Deliagina et al., 2008). For instance, stimulation of the motor cortex in standing cats induced both a flexion movement of the contralateral forelimb and an anticipatory postural change in the supporting forelimb (Gahéry and Nieoullon, 1978). In addition, data from human studies demonstrate that inhibition of the motor cortex can reduce postural activity of the trunk muscles associated with voluntary limb movements (Hodges et al., 2003). As cortical regions contribute to postural control, it could be speculated that deficits in postural activation, such as those observed in people with LBP, may be associated with changes in the excitability and organization of the motor cortex. There is a tremendous potential for areas of the brain, such as the motor and sensory cortices, to undergo an organizational change that was once thought only possible during early human development (Sanes and Donoghue, 2000). For instance, the motor cortex is extensively reorganized following stroke (Nudo and Milliken, 1996; Bütefisch, 2004). Furthermore, changes in motor cortex organization have been observed in conditions such as phantom limb pain (Flor et al., 1995; Karl et al., 2001) and complex regional pain syndrome (Krause et al., 2006; Maihöfner et al., 2007), where the central nervous system (CNS) remains largely intact. Few studies have examined the plasticity of the sensorimotor cortex in people with recurrent LBP. One study showed an expansion and shift in the representation of the lower back in the somatosensory cortex (Flor et al., 1997). Whether there are similar changes in the motor cortex of individuals with recurrent LBP remains unclear. The only available data suggest higher thresholds to evoke facilitation or inhibition of responses of the erector spinae muscles to transcranial magnetic stimulation (TMS) over the motor cortex compared to healthy individuals (Strutton et al., 2005). In that study, the change in threshold was related to the pain and functional disability experienced by LBP patients. However, it remains unclear whether changes in excitability are related to changes in organization at the motor cortex, or whether the cortical changes are associated with changes in postural control. This study investigated changes in postural activation of the deep abdominal muscle, transversus abdominis (TrA) in people with recurrent LBP. Feedforward postural activation of this muscle in association with arm movement is consistently delayed in these individuals compared to healthy controls (Hodges and Richardson, 1996; Hodges and Richardson, 1997). Although changes in trunk muscle activation are not restricted to the TrA, deficits in activation of this muscle provide an useful marker of motor control dysfunction as they are observed relatively consistently despite differences in LBP presentation. The study aimed to investigate the excitability and organization of cortical networks in the motor cortex that induce activation of TrA when excited by TMS in healthy individuals, and to compare these parameters to individuals with recurrent LBP. If changes in cortical parameters were observed, a further aim was to determine whether the extent of cortical reorganization was associated with changes in postural activation of the trunk muscles. Methods Participants Eleven right-handed individuals with recurrent non-specific LBP lasting longer than 3 months and 11 right-handed healthy individuals with no history of LBP were recruited (Table 1). Individuals were included in the LBP group if they experienced pain in the low back region with or without accompanying buttock pain and of sufficient intensity to have limited activities of daily living. This group was selected as previous studies have consistently shown delays in postural activation of TrA (Hodges and Richardson, 1996). Subjects in the LBP group had minimal pain at time of testing and symptoms were not aggravated by the experimental procedures, as variability in the level of pain during testing could increase the variability of the data. Subjects were excluded if they had any major circulatory, orthopaedic, neurological or respiratory conditions, a history or family history of epilepsy, recent or current pregnancies, previous surgery to the abdomen or back, or if they had undertaken any form of abdominal exercises in the preceding 12 months. The study conformed to the Declaration of Helsinki and was approved by the Institutional Medical Research Ethics Committee. EMG EMG activity of TrA was recorded bilaterally using intramuscular fine-wire electrodes (Teflon-coated stainless steel wire, 75 mm diameter with 1 mm of Teflon removed and tips bent back 1 and 2 mm to form hooks). Wires were threaded into a hypodermic needle and inserted with real-time ultrasound guidance (Hodges and Richardson, 1997). Pairs of surface electrodes (Ag/AgCl discs, Grass Telefactor, USA) were placed over the anterior and posterior deltoid muscles for assessment of an arm movement task (see below). The ground electrode was placed over the lower lateral rib cage. EMG data were pre-amplified 2000 times, band-pass filtered between 20 and 1000 Hz, and sampled at 2000 Hz using a Power1401 Data Acquisition System with Signal2 and Spike2 software (CED, UK). Table 1 Subject demographics (mean SD) Healthy (n =11) LBP (n =11) Age (years) Gender (M/F) 4/7 5/6 Height (cm) Weight (kg) Pain VAS (/10) ^ Pain duration (years) ^ Oswestry Disability Index (/100) ^ Pain visual analogue scale (VAS) represents the worst pain reported in the last month.

109 Motor cortex and postural control Brain (2008), 131, 2161^ TMS A single-pulse monophasic MagStim stimulator (Magstim Company, UK) was used to stimulate the motor cortex. A 7-cm figure-of-eight coil was placed with cross-over position over respective scalp sites and the coil handle oriented at 45 from the mid-sagittal plane to induce currents in an anterior-medial direction (Sakai et al., 1997). This coil orientation has been shown to evoke consistent contralateral responses from the abdominal muscles during submaximal voluntary contractions (Tunstill et al., 2001; Strutton et al., 2004; Tsao et al., 2008). The figure-of-eight coil provides better focality of stimulation compared to the standard circular coil and is more ideal for mapping of the motor cortex (Cohen et al., 1990; Brasil-Neto et al., 1992). However, our previous study showed that even at maximum stimulator output with the figure-of-eight coil, it was difficult to evoke ipsilateral responses in TrA during submaximal contractions, or evoke contralateral responses with the TrA muscle relaxed (Tsao et al., 2008). Thus, a 110 mm double-cone coil (Magstim Company, UK) was also used as this coil produces a stronger magnetic field and could induce consistent contralateral and ipsilateral responses of the trunk muscles (Davey et al., 2002; Strutton et al., 2005; Tsao et al., 2008). The double-cone coil was positioned perpendicular to the scalp site with induced current flowing in an anterior direction. EMG was recorded during maximum voluntary contraction (MVC) performed as a forced expiratory manoeuvre to determine targets for voluntary activation of TrA. Three repetitions, each lasting for at least 3 s, were completed with verbal encouragement. The contraction with the highest root-mean-square (RMS) EMG amplitude was identified and the peak RMS EMG over 1 s was recorded. The target for voluntary contraction of TrA was set at 10% MVC as this was a comfortable level for subjects to maintain and minimized the potential for fatigue. Feedback of real-time RMS EMG of TrA (averaged online over 200 ms windows on a duplicate channel) and the target level was displayed on a monitor. Excitability of the abdominal motoneurons is thought to be modulated throughout the respiratory cycle as a result of central respiratory drive potentials (Gill and Kuno, 1963; Sears, 1964), which would likely affect the amplitude of motor evoked potentials (MEPs; Lissens et al., 1995). To standardize motoneuron excitability, subjects ceased breathing at the end of normal expiration and maintained their glottis open prior to the TMS pulse. The phase of respiration was monitored online using a pressure cuff strapped to the chest to measure rib cage displacement (Hodges et al., 1997). Subjects sat comfortably in a reclined chair with arms wellsupported, hips flexed to 70, and knees flexed to 45. A tightfitting elastic cap was worn over the head and the location of the vertex was identified using the International 10/20 system (Jasper, 1958). The optimal location was identified (i.e. scalp site that induced the largest contralateral response in TrA) using the figureof-eight coil set at suprathreshold intensity (70 100% maximum stimulator output) during the activation of TrA to 10% MVC. Four motor thresholds (MT) for TrA were identified at the optimal location: (i) Active MT for contralateral responses using the figureof-eight coil. (ii) Active MT for contralateral responses using the double-cone coil. (iii) Active MT for ipsilateral responses using the double-cone coil. (iv) Resting MT for contralateral responses using the doublecone coil. MTs were defined as the minimum intensity to evoke five consecutive MEPs in TrA, and these were to be clearly discernible from background EMG activity (Mills and Nithi, 1997). Previous studies from our group have shown that TMS delivered over the optimal location (2 cm lateral to the midline in most subjects) minimizes concurrent stimulation of both motor cortices, and evokes contralateral responses that are faster than ipsilateral responses from the same hemisphere (Tsao et al., 2008). Topography of TrA responses to TMS over the contralateral motor cortex was examined using the figure-of-eight coil (Wassermann et al., 1992; Wilson et al., 1993). The coil intensity was set to 120% active MT, and stimulation was delivered over pre-marked scalp sites on a 5 5 cm grid oriented in a Cartesian system. Five stimuli were delivered at each scalp site during 10% MVC with an inter-stimulus interval of at least 5 s (Wilson et al., 1993). Early pilot trials in individuals with recurrent LBP showed that MEPs could also be induced when TMS was delivered at scalp sites posterior to the inter-aural line (i.e. the line that joins the left and right pre-auricular creases and pass through the vertex). Thus stimulation in the LBP group was extended 2 cm posterior to the vertex. This was not possible for the control group as data for that group were collected prior to the LBP group. This does not compromise the data as little or no activation was achieved by stimulation at or behind the inter-aural line in this group. Rapid arm movements Subjects performed single rapid arm movements during a choice reaction time task to induce a perturbation to the trunk for assessment of postural activation of TrA (Hodges and Richardson, 1997). Subjects stood comfortably and remained relaxed prior to movement of the left arm into flexion or extension to 45 as fast as possible in response to auditory tones triggered by the experimenter. Movement direction was indicated by distinct tones and emphasis was placed on the speed of arm movement rather than magnitude. Practice trials were included to familiarize subjects to the task. To ensure arm movements were performed in a similar manner between trials, the left arm was attached to a potentiometer at the wrist. This device restricted motion to the sagittal plane and measured angular displacement. Ten trials of arm flexion and extension were completed as this number of trials has been shown to yield sufficient repeatability of the data. Data analysis Data analysis was undertaken using MATLAB 7 (The Mathsworks, USA). For mapping, TrA activity from individual trials was fullwave rectified. Trials at each scalp site were averaged, and the onset and offset of the MEP (or onset of silent period) were visually identified from the averaged full-wave rectified traces. As recordings were made using intramuscular electrodes, peakto-peak amplitude is more variable due to recordings of a small population of motor units. Thus, the amplitude of TrA MEPs was measured as the RMS EMG amplitude between the onset and offset of the MEP, and background RMS EMG was removed (55 to 5 ms prior to stimulation). TrA EMG amplitudes were superimposed over respective scalp sites to produce a topographical map of responses of the muscle and normalized to the amplitude of the peak response. As current spread during

110 2164 Brain (2008), 131, 2161^2171 H. Tsao et al. magnetic stimulation enlarges the motor cortical map, it is difficult to define map boundaries due to small amplitude MEPs at the map edge (Uy et al., 2002). To minimize this problem, normalized values below 25% of the peak response were removed. Remaining responses were rescaled from 0 to 100%. Two parameters were calculated from the rescaled normalized maps. Map volume, which is a measure of the total excitability of cortical representation, was calculated as the sum of normalized MEP amplitudes recorded at all scalp sites where responses were evoked (Wassermann et al., 1992). The centre of gravity (CoG) was calculated using the formula: CoG ¼ P z i x i = P z i ; P z i y i = P z i ; where x i and y i are medial lateral and anterior posterior locations, and z i is amplitude (Wassermann et al., 1992). This measure gives an amplitudeweighted indication of map position. Although map volume can substantially change with changes in motor cortical excitability, studies have shown that the CoG is a valid and repeatable measure of motor cortical representation (Boroojerdi et al., 1999; Uy et al., 2002). Shifts in map position were also examined through calculation of the absolute distance between the location of the averaged CoG for the healthy group, and CoG for each individual in the healthy and LBP groups. For rapid arm movement tasks, the onsets of TrA and deltoid EMG were visually identified as the point at which the EMG increased above baseline levels. Visual identification of onset of EMG activation has been shown to be reliable and is preferred to computer-based methods as it is less affected by factors such as amplitude of background EMG or rate of increase in EMG activity (Hodges and Bui, 1996). Onsets of TrA EMG relative to that of the prime mover deltoid were calculated for the left and right TrA muscles. In addition, as onset of trunk muscle activity depends on the acceleration of the arm (Hodges, 2001); angular acceleration for each trial was calculated. Data of arm displacement at the shoulder were smoothed at 10 Hz and twice differentiated to yield angular acceleration. Peak acceleration was identified. Results MT Figure 1 shows group data for MTs during different stimulation conditions. For all subjects, MTs could be identified for all conditions over at least one motor cortex. However, contralateral responses using the figure-of-eight coil could not be evoked over one hemisphere in five subjects from the healthy group and four subjects from the LBP group, even at maximum stimulator output. Ipsilateral responses using the double-cone coil could not be evoked over one hemisphere in four healthy and one LBP individual. Responses evoked ipsilateral to the side of stimulation at the optimal location were on average 3 1 ms (mean SD) slower than MEPs evoked on the contralateral side. This confirms that ipsilateral responses were not due to concurrent excitation of the faster contralateral pathways from the opposite motor cortex, and originated from the stimulated motor cortex (Ziemann et al., 1999; Tsao et al., 2008). Using the double-cone coil, the lowest MT was found for responses contralateral to the stimulated cortex (main effect for condition P50.001; post hoc: P50.001; Fig. 1). No differences in MT were detected between the left and right TrA for any condition using the double-cone coil (main effect for muscle P = 0.84), or for MT using the figureof-eight coil (t = 0.35, P = 0.98). MT for ipsilateral responses in the LBP group was significantly lower compared to healthy controls (interaction between condition and group P = 0.009; post hoc: P = 0.04). Inspection of individual data showed a tendency for greater asymmetry of ipsilateral MTs between Statistical analysis Statistica 7 (Statsoft, USA) was used for statistical analysis. Map volume and CoG location (lateral and anterior to the vertex) were compared between groups (healthy versus LBP) and muscles (left versus right TrA) using a repeated-measures analysis of variance (ANOVA). Significant interactions were further analysed using post hoc Duncan s multiple range test. MTs identified using the double-cone coil were compared between groups, muscles and conditions (contralateral, ipsilateral and resting MT) with a repeated-measures ANOVA. MTs identified using the figure-of-eight coil were compared between groups and muscles. In addition, onset of TrA EMG relative to that of the prime mover deltoid was compared between individuals with and without recurrent LBP using repeated-measures ANOVA with two repeated measures [group and direction (flexion versus extension)] and one independent factor (muscle). To determine whether CoG, map volume and MT were associated with the relative onset of TrA EMG in the arm movement task, the relationship between MT and map parameters, and relative onset of TrA activation during arm movement tasks were examined using Pearson s correlation. Significance was set at P Fig. 1 Active and resting MTs for contralateral (contra) and ipsilateral (ipsi) responses in healthy and LBP group. Mean and 95% CI are illustrated. For double-cone coil, lower MT for ipsilateral responses in LBP group was observed compared to healthy controls ( P50.05). No differences were detected between groups for MT to elicit contralateral responses using the double-cone coil or figure-of-eight coils.

111 Motor cortex and postural control Brain (2008), 131, 2161^ hemispheres in most healthy individuals, but the lowest threshold was not always on the dominant or nondominant side. A lesser difference between hemispheres was observed for individual data for the LBP group. To quantify this observation, the absolute difference in MT between left and right TrA was calculated for each subject (Fig. 2A). There was a greater difference in MTs for ipsilateral responses compared to contralateral responses in the healthy group (P = 0.01), but not for the LBP group (P40.14). Furthermore, absolute difference in MTs for ipsilateral responses was greater for the healthy group compared to LBP participants (P = 0.043), i.e. less asymmetrical. To examine this further, ipsilateral MT for each individual were rearranged into either a lower or higher MT to evoke ipsilateral responses (Fig. 2B). When analysed in this way, the MT for the less excitable hemisphere was higher for the control subjects compared to those with LBP (P50.001), but there was no difference between groups in the MT for the more excitable hemisphere (P40.066). Together, these data suggest that the reduced MT of ipsilateral responses in the LBP group is likely due to reduced MT of a less excitable side. individuals (interaction for location and group P50.001; post hoc: P50.011, Fig. 4). No difference was observed between the left and right motor cortices (main effect for muscle P = 0.08). The TrA CoG in individuals with Motor cortical map Average maps of TrA responses to TMS for healthy and LBP groups are illustrated in Fig. 3. The time of onset and offset of contralateral MEPs were not different between healthy and LBP groups [main effect for group P = 0.63; mean SD onset 16 1 ms, mean SD offset 35 5ms (averaged across healthy and LBP groups)]. No differences in map volume were detected between the left and right hemispheres for either group (main effect for side P = 0.89). However, the volume of TMS maps for the LBP group was greater compared to healthy controls (healthy , LBP ; main effect for group P50.001). Furthermore, the location of TrA CoG for LBP was located posterior and lateral to the CoG location in healthy Fig. 3 Average normalized motor cortical maps for the healthy and LBP groups on the left and right hemisphere. Mean and SD of the CoG is displayed. The black cross represents the location of vertex, horizontal dotted line denotes the inter-aural line and vertical dotted line denotes the line that connects the nasion and inion (Calibration ^1cm). Fig. 2 (A) Absolute difference in MT between the left and right TrAs. Note absolute difference in MTwas greater for ipsilateral responses compared to that for contralateral responses in the healthy group. However, comparisons of the absolute difference between ipsilateral and contralateral MTs were not significant (NS) for the LBP group. (B) MT for ipsilateral responses in healthy and LBP group, re-arranged into the side with higher or lower MT. Mean and 95% CI are displayed. Note MT in LBP group on the less-excitable side is less than MT in the control group. ( P50.05).

112 2166 Brain (2008), 131, 2161^2171 H. Tsao et al. Fig. 4 Group (A) and individual data(b) of the CoG of TrA. Note that for group data, the location of TrA CoG in LBP group was located posterior and lateral to TrA CoG in healthy groups ( P50.05). This was observed in most subjects. recurrent LBP were located further away from the mean location of TrA CoG in healthy controls than the individual data for the control group (healthy: cm; LBP: cm; P50.001). Rapid arm movements Figure 5 illustrates the onset of TrA EMG relative to that of the deltoid during rapid arm flexion and extension tasks. When LBP individuals moved their arm rapidly into flexion or extension, activation of TrA EMG was significantly delayed compared to healthy individuals (interaction between group, muscle and direction: F = 7.8, P = 0.007; post hoc: P50.001). No differences were observed between onsets for the left and right TrA muscles (post hoc: P40.065). There was no difference in peak arm acceleration between healthy and LBP group for arm flexion (healthy: /s 2 ; LBP: /s 2 ; P = 0.32) or extension (healthy: /s 2 ; LBP: /s 2 ; P = 0. 41). This suggests the arm was moved in a similar manner by both groups. The location of TrA CoG was associated with onset of TrA EMG during rapid arm flexion and extension tasks (r40.37, P50.033; Fig. 6). That is, slower activation of TrA was likely when the CoG was located more posterior and lateral. Similar correlations were found for map volume; increased map area was associated with slower onset of TrA EMG (r40.57, P50.001; Fig. 7). There were no significant correlations between MTs and timing of TrA EMG onset (all: P40.17). Discussion This study demonstrates that LBP is associated with reorganization of the networks in the motor cortex associated with activation of the deep trunk muscle, TrA. Posterior and lateral shifts in the CoG and greater map volume were Fig. 5 Relative onset of TrA activation (mean and 95% CI) during rapid arm flexion and extension tasks for the healthy and LBP groups. Dotted line represents onset of prime mover activation. Data show slower activation of TrA in LBP group for both the left and right TrA muscles and during both arm flexion and extension tasks ( P50.05 between healthy and LBP group). observed in individuals with recurrent episodes of LBP compared to healthy individuals. A particularly novel finding was that the location of the CoG and volume of the TrA map at the motor cortex were related to the timing of onset of TrA EMG during a functional task. There were also changes in MT; compared to controls, threshold for ipsilateral responses on the less excitable side was reduced. As changes in control

113 Motor cortex and postural control Brain (2008), 131, 2161^ Fig. 6 Relationship between location of the CoG; (distance anterior and lateral from the vertex) and timing of TrA activation during arm flexion and extension. Linear regressions are shown with 95% CI. Vertical dotted line represents activation of prime mover deltoid. Circlesrepresentdataofindividualsfromthehealthygroup(white)andLBPgroup(black).DatashowedthatindividualswithslowerTrA activation (mostly individuals from LBP group) tended to havetra CoG located more posterior and lateral to the vertex. Fig. 7 Relationship between normalized map volume and timing of TrA activation during arm flexion and extension. Linear regressions are shown with 95% confidence interval. Vertical dotted line represents activation of prime mover deltoid. Circles represent data of individuals from the healthy group (white) and LBP group (black). There was a positive correlation between normalized map volume and relative onset of TrA activation. of TrA are consistently observed in individuals with recurrent LBP, these findings suggest that changes at the motor cortex may underlie or at least contribute to alterations in postural strategies. Reorganization of the motor cortex in recurrent LBP The current study confirms that topography of the representation of the abdominal muscles on the motor cortex can be evaluated using TMS. In healthy individuals, TrA representation was located at scalp sites 2 cm anterior and lateral to the vertex. This location is consistent with the optimal location used in previous studies to evoke responses in the more superficial abdominal muscles recorded with surface EMG electrodes (Fujiwara et al., 2001; Strutton et al., 2004). As expected, based on the distribution of body representation in the motor homunculus (Penfield and Boldrey, 1937; Woolsey et al., 1952), the representation of TrA was located medially compared to

114 2168 Brain (2008), 131, 2161^2171 H. Tsao et al. that of the upper limb (Wilson et al., 1993; Pascual-Leone et al., 1994), neck (Thompson et al., 1997) and facial muscles (McMillan et al., 1998). The CoG of TrA at the motor cortex in individuals with recurrent LBP was shifted posteriorly and laterally to that of healthy individuals. Shifts in the motor cortical representation of specific muscle(s) have been reported in people with other recurrent pain conditions. For instance, studies of patients with phantom limb pain following upper limb amputations demonstrated a shift in the optimal location to evoke responses in the facial muscles on the side of the amputation towards the representation of the missing hand (Karl et al., 2001). The CoG is argued to be a robust measure of motor cortical representation and corresponds closely to the area of high excitability of corticomotor neurons that project to the target muscle(s) (Wassermann et al., 1992). Furthermore, TMS CoG closely approximates the CoG identified from functional magnetic resonance imaging (Boroojerdi et al., 1999; Lotze et al., 2003). Thus, shifts in TrA CoG from TMS maps could imply changes in the structural or functional organization of cortical networks associated with activation of TrA at the motor cortex. As this shift was consistently observed in most individuals with recurrent episodes of LBP, we argue that these findings are unlikely to be related to cap displacement, variability in coil placement or orientation, or inaccurate identification of the vertex. Future studies that utilize brain MRI and navigated brain stimulation with TMS mapping are likely to reduce the variability of motor cortical maps and further validate the present findings. In addition, as evidence reveals reduced grey matter density of the prefrontal cortex in patients with chronic LBP (Apkarian et al., 2004), further studies are needed to examine whether these structural changes in the brain are associated with changes in functional organization of the motor cortex, or how this relates to changes in motor behaviour. Map volume was also increased in the LBP group compared to healthy individuals. This finding is similar to expansion in area of representation at the somatosensory cortex of patients with acute (Soros et al., 2001) and recurrent pain (Flor et al., 1997). However, increases in motor cortical map volume are not consistent with reduced map areas in patients with complex regional pain syndrome (Krause et al., 2006). This could be related to differences in the calculation of map area, as that study measured twodimensional spread of representation over scalp sites whereas we measured map volume. It is also possible that reduced map areas detected in patients with complex regional pain syndrome relate to the nature of injury sustained in this group of patients, all of which involved forearm fractures that were immobilized for a period of time following post-fracture. As reduced motor cortex representation has been demonstrated with immobilization and disuse (Liepert et al., 1995), reduced map area in that study could be associated with disuse rather than the presence of pain and injury. Increased map volume is difficult to interpret as stimulation of the cortical cells/interneurons with TMS is accompanied by current spread to produce TMS maps which are larger than the actual area of motor cortical cells that project to the target muscle (Roth et al., 1991; Mortifee et al., 1994; Thickbroom et al., 1998). It could be argued that increases in map volume in the LBP group could correspond to an increase in the total excitability of motor cortical cells and thus excitation by stimulation over a large area of the cortex (Wassermann et al., 1992). However, there were no differences in MT for contralateral MEPs at the optimal location between the LBP and healthy groups. Taken together, increased map volume without changes in MT at the optimal location suggest that alteration of motor cortical maps are not simply due to increased excitability of cortical cells, but involve more complex neural mechanisms that increase the area of the cortical networks involved in the activation of TrA. Representation at the motor cortex is associated with postural activation of the trunk muscles During rapid voluntary limb movements, the CNS initiates postural adjustments in advance of predictable perturbations to the body and these involve activation of trunk and limb muscles (Belen kii et al., 1967). As this activation is initiated before feedback is available, they must be preprogrammed by the nervous system (Bouisset and Zattara, 1981). In the current study, postural activation of TrA, in most trials, occurred either before or 550 ms after the onset of deltoid EMG. Taking into account electromechanical delay and the latency for nerve conduction, even the shortest latency response to feedback from limb movements cannot be initiated before 50 ms after the onset of deltoid EMG (Aruin and Latash, 1995). Thus, the current study confirms previous findings that feedforward activation of TrA is associated with voluntary limb movements in healthy individuals (Hodges and Richardson, 1997), and that activation of TrA is delayed in individuals with recurrent LBP (Hodges and Richardson, 1996). A novel finding was that the cortical reorganization of inputs to TrA was associated with onset of TrA EMG during rapid limb movements. This relationship is consistent with the observation that the motor cortex contributes, at least in part, to postural activation associated with limb movements (Gahéry and Nieoullon, 1978; Hodges et al., 2003). The present data provide evidence that changes in its organization at the motor cortex may contribute to deficits in feedforward postural control. As similar deficits in feedforward control have been demonstrated in other trunk muscles [for instance, the lumbar multifidus muscles (MacDonald et al., 2004)], the current findings suggest the potential for similar reorganization of

115 Motor cortex and postural control Brain (2008), 131, 2161^ neural networks at the motor cortex that contribute to the control of these muscles. Exactly how changes in motor cortical representation are associated with deficits in postural control remains speculative. One possibility is that reorganization in motor cortical map of TrA and potentially other muscles in patients with recurrent LBP could distort the coordination between muscles. Data from people with focal hand dystonia provide evidence that reduced ability to isolate finger movements is associated with reorganization (i.e. reduce differentiation of individual finger representations) of the sensorimotor cortex (Byl et al., 1996). Although TMS maps of other trunk or limb muscles were not evaluated in the current study, similar mechanisms may contribute to changes in postural control of the trunk muscles. Furthermore, several studies have reported atrophy of the paraspinal muscles in human and animal studies of pain and injury to the low back (Hides et al., 1994; Hodges et al., 2006). Thus, it is reasonable to speculate that these morphological changes may be associated with reorganization at the sensorimotor cortex. This warrants further investigation. In addition, the trunk muscles receive multiple projections from other supraspinal and spinal centres (see review Iscoe, 1998). For example, the reticulospinal and vestibulospinal neurons, which are intricately involved in postural control, have descending projections to the abdominal motoneurons (Miller et al., 1985; Miller et al., 1989). Changes in the excitability and/or organization of these regions of the CNS are also likely to contribute to changes in postural control of the trunk muscles in patients with recurrent LBP. Reduced MT for ipsilateral corticospinal projections to the abdominal muscles There was greater asymmetry in MT between the left and right hemispheres for ipsilateral responses compared to contralateral responses in control subjects. The lateralization of ipsilateral MT to one side is consistent with findings from other proximal (MacKinnon et al., 2004) and axial muscles (Strutton et al., 2004), and was unrelated to handedness. Interestingly, individuals with recurrent LBP did not demonstrate this asymmetry in ipsilateral MT. The MT to evoke responses in the left and right TrA in people with recurrent LBP was similar to the more excitable side in healthy individuals. One interpretation of this finding could be that coordination of the activation of the right and left abdominal muscles in healthy individuals is mediated by contralateral and ipsilateral projections from a single hemisphere, that is, the cortex with the lower threshold ipsilateral projections. In contrast, in people with recurrent pain the symmetrical excitability of ipsilateral projections suggests no preference for control by a single hemisphere. Although changes in MT for ipsilateral responses did not correlate with changes in timing of TrA activation, this does not exclude a contribution of ipsilateral projections to postural control of the trunk muscles. It has been argued that responses evoked ipsilateral to the stimulated cortex using TMS are likely to be mediated via slower conducting uncrossed polysynaptic corticospinal pathways that project via regions in the brain stem (Ziemann et al., 1999). These brainstem pathways are also intricately involved in the coordination of timing and magnitude of postural responses (Inglis and Macpherson, 1995; Prentice and Drew, 2001). Changes in excitability of these pathways may contribute to alterations in activation of the trunk muscles for postural control. However, whether the earliest component of postural responses is related to ipsilateral projections, or how changes in excitability are related to coordination of muscle activation remains unclear. Clinical implication The present findings suggest that deficits in postural control in a patient population are associated with reorganization of the motor cortex. In patients with recurrent LBP, these deficits in postural activity can be trained by skilled motor training that involves voluntary contractions of muscle(s) (Tsao and Hodges, 2007, 2008), and is associated with improvements in clinical outcomes (see review Ferreira et al., 2006). Skilled motor training induces greater plastic change at the motor cortex than strength training (see review Adkins et al., 2006). Thus it is reasonable to predict that reorganization of the motor cortex following skilled motor training may be associated with improved postural activity in patients with deficits in postural control. This requires further investigation. Acknowledgements Financial support was provided by the National Health and Medical Research Council of Australia (ID to HT and ID to PH) and the Physiotherapy Research Foundation (Grant No: 007/06). References Adkins DL, Boychuk J, Remple MS, Kleim JA. Motor training induces experience-specific patterns of plasticity across motor cortex and spinal cord. J Appl Physiol 2006; 101: Andersson GB. Epidemiology of low back pain. Acta Orthop Scand Suppl 1998; 281: Apkarian AV, Sosa Y, Sonty S, Levy RM, Harden RN, Parrish TB, et al. Chronic back pain is associated with decreased prefrontal and thalamic gray matter density. J Neurosci 2004; 24: Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P. The influence of low back pain on muscle activity and coordination during gait: a clinical and experimental study. Pain 1996; 64: Aruin AS, Latash ML. Direction specificity of postural muscles in feedforward postural reactions during fast voluntary arm movements. Exp Brain Res 1995; 103: Belen kii VY, Gurfinkel VS, Pal tsev YI. Control elements of voluntary movements. Biofizika 1967; 12: Blyth FM, March LM, Brnabic AJ, Jorm LR, Williamson M, Cousins MJ. Chronic pain in Australia: a prevalence study. Pain 2001; 89:

116 2170 Brain (2008), 131, 2161^2171 H. Tsao et al. Boroojerdi B, Foltys H, Krings T, Spetzger U, Thron A, Topper R. Localization of the motor hand area using transcranial magnetic stimulation and functional magnetic resonance imaging. Clin Neurophysiol 1999; 110: Bouisset S, Zattara M. A sequence of postural adjustments precedes voluntary movement. Neurosci Lett 1981; 22: Brasil-Neto JP, Cohen LG, Panizza M, Nilsson J, Roth BJ, Hallett M. Optimal focal transcranial magnetic activation of the human motor cortex: effects of coil orientation, shape of the induced current pulse and stimulus intensity. J Clin Neurophysiol 1992; 9: Bütefisch CM. Plasticity in the human cerebral cortex: lessons from the normal brain and from stroke. Neuroscientist 2004; 10: Byl NN, Merzenich MM, Jenkins WM. A primate genesis model of focal dystonia and repetitive stain injury: I. Learning-induced dedifferentiation of the representation of the hand in the primary somatosensory cortex in adult monkeys. Neurology 1996; 47: Cassidy JD, Cote P, Carroll LJ, Kristman V. Incidence and course of low back pain episodes in the general population. Spine 2005; 30: Cholewicki J, Silfies SP, Shah RA, Greene HS, Reeves NP, Alvi K, et al. Delayed trunk muscle reflex responses increase the risk of low back injuries. Spine 2005; 30: Cohen LG, Roth BJ, Nilsson J, Dang N, Panizza M, Bandinelli S, et al. Effects of coil design on delivery of focal magnetic stimulation. Technical considerations. Electroencephalogr Clin Neurophysiol 1990; 75: Davey NJ, Lisle RM, Loxton-Edwards B, Nowicky AV, McGregor AH. Activation of back muscles during voluntary abduction of the contralateral arm in humans. Spine 2002; 27: Deliagina TG, Beloozerova IN, Zelenin PV, Orlovsky GN. Spinal and supraspinal postural networks. Brain Res Rev 2008; 57: Ferreira PH, Ferreira ML, Maher CG, Herbert RD, Refshauge K. Specific stabilisation exercise for spinal and pelvic pain: A systematic review. Aust J Physiother 2006; 52: Flor H, Braun C, Elbert T, Birbaumer N. Extensive reorganization of primary somatosensory cortex in chronic back pain patients. Neurosci Lett 1997; 224: 5 8. Flor H, Elbert T, Knecht S, Wienbruch C, Pantev C, Larbig W. Phantom limb pain as a perceptual correlate of massive cortical reorganization in upper extremity amputees. Nature 1995; 375: Fujiwara T, Sonoda S, Okajima Y, Chino N. The relationships between trunk function and the findings of transcranial magnetic stimulation among patients with stroke. J Rehabil Med 2001; 33: Gahéry Y, Nieoullon A. Postural and kinetic co-ordination following cortical stimuli which induce flexion movements in the cat s limb. Brain Res 1978; 149: Gill PK, Kuno M. Excitatory and inhibitory actions on phrenic motoneurones. J Physiol 1963; 168: Hides JA, Stokes MJ, Saide M, Jull GA, Cooper DH. Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 1994; 19: Hodges PW. Changes in motor planning of feedforward postural responses of the trunk muscles in low back pain. Exp Brain Res 2001; 141: Hodges PW, Bui BH. A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography. Electroencephalogr Clin Neurophysiol 1996; 101: Hodges PW, Butler JE, Taylor JL, Gandevia SC. Motor cortex may be involved in feedforward postural responses of the deep trunk muscles. In: Lords, Menz HB, editors. Proceedings of International Society for Posture and Gait Research. Sydney: International Society for Posture and Gait Research; p Hodges PW, Gandevia SC, Richardson CA. Contraction of specific abdominal muscles in postural tasks are affected by respiratory maneuvers. J Appl Physiol 1997; 83: Hodges PW, Holm AK, Hansson T, Holm S. Rapid atrophy of the lumbar multifidus follows experimental disc or nerve root injury. Spine 2006; 31: Hodges PW, Moseley GL. Pain and motor control of the lumbopelvic region: effect and possible mechanisms. J Electromyogr Kinesiol 2003; 13: Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdominis. Spine 1996; 21: Hodges PW, Richardson CA. Feedforward contraction of transversus abdominis is not influenced by the direction of arm movement. Exp Brain Res 1997; 114: Inglis JT, Macpherson JM. Bilateral labyrinthectomy in the cat: effects on the postural response to translation. J Neurophysiol 1995; 73: Iscoe S. Control of abdominal muscles. Prog Neurobiol 1998; 56: Jasper HH. The ten-twenty electrode system of the International Federation. Electroencephalogr Clin Neurophysiol 1958; 10: Karl A, Birbaumer N, Lutzenberger W, Cohen LG, Flor H. Reorganization of motor and somatosensory cortex in upper extremity amputees with phantom limb pain. J Neurosci 2001; 21: Krause P, Forderreuther S, Straube A. TMS motor cortical brain mapping in patients with complex regional pain syndrome type I. Clin Neurophysiol 2006; 117: Leinonen V, Kankaanpaa M, Luukkonen M, Hanninen O, Airaksinen O, Taimela S. Disc herniation-related back pain impairs feed-forward control of paraspinal muscles. Spine 2001; 26: E Liepert J, Tegenthoff M, Malin J-P. Changes of cortical motor area size during immobilization. Electroencephalogr Clin Neurophysiol 1995; 97: Lissens MA, De Muynck MC, Decleir AM, Vanderstraeten GG. Motor evoked potentials of the abdominal muscles elicited through magnetic transcranial brain stimulation. Muscle and Nerve 1995; 18: Lotze M, Kaethner RJ, Erb M, Cohen LG, Grodd W, Topka H. Comparison of representational maps using fmri and TMS. Clin Neurophysiol 2003; 114: MacDonald DA, Moseley GL, Hodges PW. The function of the lumber multifidus in unilateral LBP. In: Vleeming A, Mooney V, Hodges P, Lee D, McGill S, Östgaard HC, et al., editors. Fifth Interdisciplinary World Congress on Low Back and Pelvic Pain. Melbourne: World Congress on Low Back and Pelvic Pain; p MacKinnon CD, Quartarone A, Rothwell JC. Inter-hemispheric asymmetry of ipsilateral corticofugal projections to proximal muscles in humans. Exp Brain Res 2004; 157: Maihöfner C, Baron R, DeCol R, Binder A, Birklein F, Deuschl G, et al. The motor system shows adaptive changes in complex regional pain syndrome. Brain 2007; 130: McMillan AS, Watson C, Walshaw D. TMS mapping of the cortical topography of the human masseter muscle. Arch Oral Biol 1998; 43: Miller AD, Ezure K, Suzuki I. Control of abdominal muscles by brain stem respiratory neurons in the cat. J Neurophysiol 1985; 54: Miller AD, Tan LK, Lakos SF. Brainstem projections to cats upper lumbar spinal cord: implications for abdominal muscle control. Brain Res 1989; 493: Mills KR, Nithi KA. Corticomotor threshold to magnetic stimulation: normal values and repeatability. Muscle and Nerve 1997; 20: Mortifee P, Stewart H, Schulzer M, Eisen M. Reliability of transcranial magnetic stimulation for mapping the human motor cortex. Electroencephalogr Clin Neurophysiol 1994; 93: Nudo RJ, Milliken GW. Reorganization of movement representations in primary motor cortex following focal ischemic infarcts in adult squirrel monkeys. J Neurophysiol 1996; 75: Pascual-Leone A, Cohen LG, Brasil-Neto JP, Hallett M. Non-invasive differentiation of motor cortical representation of hand muscles by mapping of optimal current directions. Electroencephalogr Clin Neurophysiol/Evoked Potentials Section 1994; 93: Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studies by electrical stimulation. Brain 1937; 60:

117 Motor cortex and postural control Brain (2008), 131, 2161^ Pengel LHM, Herbert RD, Maher CG, Refshauge KM. Acute low back pain: systematic review of its prognosis. Br Med J 2003; 327: Prentice SD, Drew T. Contributions of the reticulospinal system to the postural adjustments occurring during voluntary gait modifications. J Neurophysiol 2001; 85: Radebold A, Cholewicki J, Polzhofer GK, Greene HS. Impaired postural control of the lumbar spine is associated with delayed muscle response times in patients with chronic idiopathic low back pain. Spine 2001; 26: Roth BJ, Saypol JM, Hallett M, Cohen LG. A theoretical calculation of the electric field induced in the cortex during magnetic stimulation. Electroencephalogr Clin Neurophysiol 1991; 81: Sakai K, Ugawa Y, Terao Y, Hanajima R, Furubayashi T, Kanazawa I. Preferential activation of different I waves by transcranial magnetic stimulation with a figure-of-eight-shaped coil. Exp Brain Res 1997; 113: Sanes JN, Donoghue JP. Plasticity and primary motor cortex. Annu Rev Neurosci 2000; 23: Sears TA. The slow potentials of thoracic respiratory motorneurones and their relation to breathing. J Physiol 1964; 175: Soros P, Knecht S, Bantel C, Imai T, Wusten R, Pantev C, et al. Functional reorganization of the human primary somatosensory cortex after acute pain demonstrated by magnetoencephalography. Neurosci Lett 2001; 298: Strutton PH, Beith I, Theodorou S, Catley M, McGregor AH, Davey NJ. Corticospinal activation of internal oblique muscles has a strong ipsilateral component and can be lateralized in man. Exp Brain Res 2004; 158: Strutton PH, Theodorou S, Catley M, McGregor AH, Davey NJ. Corticospinal excitability in patients with chronic low back pain. J Spinal Disord Tech 2005; 18: Thickbroom GW, Sammut R, Mastaglia FL. Magnetic stimulation mapping of motor cortex: factors contributing to map area. Electroencephalogr Clin Neurophysiol 1998; 109: Thompson ML, Thickbroom GW, Mastaglia FL. Corticomotor representation of the SCM. Brain 1997; 120: Tsao H, Galea MP, Hodges PW. Concurrent excitation of the opposite motor cortex during transcranial magnetic stimulation to activate the abdominal muscles. J Neurosci Methods 2008; 171: Tsao H, Hodges PW. Immediate changes in feedforward postural adjustments following voluntary motor training. Exp Brain Res 2007; 181: Tsao H, Hodges PW. Persistence of improvements in postural strategies following motor control training in people with recurrent low back pain. J Electromyogr Kinesiol 2008; 18: Tunstill SA, Wynn-Davies AC, Nowicky AV, McGregor AH, Davey N. Corticospinal facilitation studied during voluntary contraction of human abdominal muscles. Exp Physiol 2001; 86: Uy J, Ridding MC, Miles TS. Stability of maps of human motor cortex made with transcranial magnetic stimulation. Brain Topogr 2002; 14: Wasiak R, Kim J, Pransky G. Work disability and costs caused by recurrence of low back pain: longer and more costly than in first episodes. Spine 2006; 31: Wassermann EM, McShane LM, Hallett M, Cohen LG. Noninvasive mapping of muscle representations in human motor cortex. Electroencephalogr Clin Neurophysiol 1992; 85: 1 8. Wilson SA, Thickbroom GW, Mastaglia FL. Transcranial magnetic stimulation mapping of the motor cortex in normal subjects. The representation of two intrinsic hand muscles. J Neurol Sci 1993; 118: Woolsey CN, Settlage PH, Meyer DR, Sencer W, Hamuy TP, Travis AM. Patterns of localization in the precentral and supplementary motor areas and their relation to the concept of a premotor area. Research Publications Association for Research in Nervous and Mental Disease 1952; 30: Ziemann U, Ishii K, Borgheresi A, Yaseen Z, Battaglia F, Hallett M, et al. Dissociation of the pathways mediating ipsilateral and contralateral motor-evoked potentials in human hand and arm muscles. J Physiol 1999; 518:

118 ARTICLE IN PRESS Manual Therapy 10 (2005) Original article Abdominal muscle recruitment during a range ofvoluntary exercises Donna M. Urquhart a,b,, Paul W. Hodges c,d, Trevor J. Allen b, Ian H. Story b a Department of Epidemiology and Preventive Medicine, Monash University, Central and Eastern Clinical School, Alfred Hospital, Commercial Rd, Melbourne, Victoria 3004, Australia b School of Physiotherapy, The University of Melbourne, Victoria, Australia c Prince of Wales Medical Research Institute, New South Wales, Australia d Department of Physiotherapy, The University of Queensland, Queensland, Australia Received 5 March 2003; received in revised form 11 August 2004; accepted 27 August Abstract Various exercises are used to retrain the abdominal muscles in the management oflow back pain and other musculoskeletal disorders. However, few studies have directly investigated the activity ofall the abdominal muscles or the recruitment ofregions of the abdominal muscles during these manoeuvres. This study examined the activity of different regions of transversus abdominis (TrA), obliquus internus (OI) and externus abdominis (OE), and rectus abdominis (RA), and movement ofthe lumbar spine, pelvis and abdomen during inward movement ofthe lower abdominal wall, abdominal bracing, pelvic tilting, and inward movement ofthe lower and upper abdominal wall. Inward movement ofthe lower abdominal wall in supine produced greater activity oftra compared to OI, OE and RA. During posterior pelvic tilting, middle OI was most active and with abdominal bracing, OE was predominately recruited. Regions of TrA were recruited differentially and an inverse relationship between lumbopelvic motion and TrA electromyography (EMG) was found. This study indicates that inward movement of the lower abdominal wall in supine produces the most independent activity oftra relative to the other abdominal muscles, recruitment varies between regions oftra, and observation ofabdominal and lumbopelvic motion may assist in evaluation ofexercise performance. r 2004 Elsevier Ltd. All rights reserved. Keywords: Exercises; Abdominal muscles; Transversus abdominis; Low back pain 1. Introduction A diverse range ofexercises is used clinically to retrain the trunk muscles. However, recruitment ofthe abdominal muscles during exercises that aim to restore motor control have not been clearly defined. Most studies have used surface electromyography (EMG) to investigate these techniques (Partridge and Walters, 1960; Kennedy, 1980; Richardson et al., 1990; Jull et al., 1995; Allison et al., 1998; O Sullivan et al., 1998; Vezina and Hubley- Kozey, 2000) and the results ofthe small number of intramuscular EMG studies are inconclusive (Carman Corresponding author. Tel.: ; fax: address: donna.urquhart@med.monash.edu.au (D.M. Urquhart). et al., 1972; Strohl et al., 1981; Goldman et al., 1987; De Troyer et al., 1990). For example, three different recruitment patterns were reported when six subjects were instructed to pull in their abdominal wall (De Troyer et al., 1990). A contemporary approach for low back pain (LBP) involves recruitment oftransversus abdominis (TrA) with minimal activity ofthe superficial abdominal muscles in the early stages ofrehabilitation. This approach is based on evidence that activity oftra contributes to spinal control (Cresswell et al., 1992; Hodges et al., 1999) and dysfunction of this muscle occurs in people with LBP (Hodges and Richardson, 1996b, 1998; Hodges, 2001). Although recruitment of TrA is emphasized initially, all ofthe trunk muscles are considered to be important for the restoration of normal function and progression involves strategies for X/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi: /j.math

119 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) re-education ofthe whole muscle system (Richardson et al., 1999). The efficacy of this method has been established in randomized control trials with acute and chronic LBP patients (Hides et al., 1996; O Sullivan et al., 1997b,c). The technique involves inward movement ofthe lower abdominal wall without movement of the spine or pelvis (Richardson et al., 1999). Surface EMG studies indicate that activity ofthe superficial abdominal muscles is minimal during this manoeuvre (Jull et al., 1995), and indirect measurements oftra activity with a pressure cuff under the abdomen to indicate movement ofthe abdominal wall, are related to direct EMG measures oftra motor control (Hodges et al., 1996a). However, no study has directly investigated TrA activity during this, or other exercise approaches. Other exercise strategies have also been argued to be beneficial in LBP management. Abdominal bracing (lateral flaring ofthe abdominal wall) (Kennedy, 1980) and posterior pelvic tilting have been proposed to improve lumbopelvic control by elevation ofintraabdominal pressure and by reduction ofthe lumbar lordosis, respectively (Kennedy, 1980; Vezina and Hubley-Kozey, 2000). However, there is controversy regarding the specific patterns ofabdominal muscle recruitment during these exercises. A recent review concluded that muscle activation patterns during pelvic tilting are not clearly defined in people with or without LBP (Vezina et al., 1998). An additional consideration is that there are differences in the morphology and recruitment ofregions of TrA and obliquus internus abdominis (OI) (Askar, 1977; Rizk, 1980; Hodges et al., 1999; Urquhart et al., 2001, 2004). Upper fascicles of TrA that attach to the rib cage are horizontal, and middle and lower fascicles that fuse with the thoracolumbar fascia and the iliac crest are inferomedial (Urquhart et al., 2001). Fibres ofupper TrA are also active with the opposite direction oftrunk rotation to lower and middle fibres (Urquhart et al., 2004), and activity oflower and upper fibres ofoi vary during posterior pelvic tilting (Carman et al., 1972). Although these reports suggest regional differences in activity ofthe abdominal muscles, their recruitment has not been comprehensively investigated during voluntary exercises. The aims ofthis study were to investigate recruitment ofregions ofthe abdominal muscles during exercises used in LBP management, and to determine if common clinical techniques, such as observation of abdominal, spinal and pelvic motion, assist differentiation ofpatterns ofabdominal muscle recruitment. 2. Methods 2.1. Subjects Seven subjects (4 male, 3 female), with a mean (SD) age, height, and weight of30(4) years, 174(9) cm, and 68(15) kg, participated in the study. Subjects were excluded ifthey had a history oflow back or leg pain that affected function in the preceding 2 years, or any abdominal, gastrointestinal, neurological or respiratory condition. All subjects had an average activity level, as determined by the habitual physical activity questionnaire (Baecke et al., 1982). Five subjects had performed the exercises previously and all subjects were involved in another study (Urquhart et al., 2004). All procedures were approved by the institutional research ethics committee and conducted in accordance with the declaration ofhelsinki Electromyography Recordings ofemg were made using bipolar finewire electrodes inserted into three regions ofthe abdominal wall under the guidance ofreal-time ultrasound imaging (5 MHz curved array transducer) (128XP/4, Acuson, Mountain View, CA). Electrodes were fabricated from two strands of Teflon-coated stainless steel wire (75 mm) (A-M Systems Inc., Everett, Washington, USA), with 1 mm ofteflon removed from the ends. The electrodes were threaded into a hypodermic needle ( mm) and the tips bent back 1 2 mm to form hooks. Electrodes were inserted into the upper region oftra (adjacent to the 8th rib), the middle region oftra, OI and obliquus externus abdominis (OE) (midway between the iliac crest and inferior border of the rib cage), and the lower region of TrA and OI (adjacent to the anterior superior iliac spine (ASIS)) (De Troyer et al., 1990; Cresswell et al., 1992; Hodges and Richardson, 1997; Urquhart et al., 2004). Pairs ofsurface EMG electrodes (Ag/AgCl discs, 1 cm diameter and 2 cm inter-electrode distance) were placed over rectus abdominis (RA), halfway between the umbilicus and the pubic symphysis. A ground electrode was placed on the iliac crest. EMG data were bandpass filtered between 50 Hz and 1 khz and sampled at 2 khz using a Power1401 data acquisition system and Spike2 software (Cambridge Electronic Design, Cambridge, UK). The data was exported and analysed using Matlab 6 (release 12; MathWorks, Natick, MA, USA) Video motion analysis A video motion analysis system was used to quantify displacement ofthe upper, middle and lower regions of the abdominal wall and movement ofthe lumbar spine and pelvis in prone. Data were captured with a digital video camera (Sony DCR TRV20, Tokyo, Japan), positioned 2 m away and perpendicular to the subject. A diffuse light source, placed under the subject s abdomen, and a black background were used to highlight the edge ofthe abdominal wall in the video image (Fig. 1). A marker was placed on the spinous

120 146 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) process ofthe L3 vertebrae and the left ASIS to allow measurement oflinear displacement ofthe lumbar spine and pelvis. The border ofthe upper and middle abdominal regions (lower border ofthe rib cage), and the middle and lower abdominal regions (upper border ofthe iliac crest) were also identified. Video data were transferred to computer and edited using imovie editing software (Apple Computer, Inc., Cupertino, CA). An edge detection program was written using Igor Pro (WaveMetrics Inc., Lake Oswego, USA) to measure displacement ofthe abdominal wall, and spine and pelvic motion was measured with NIH Image (National Institute ofhealth, Bethesda, MD, USA). Distances were calibrated to an object ofknown dimensions filmed in the same plane as the abdominal wall. Resolution was 0.5 mm. The motion parameters were found to be accurate and repeatable over a 24-h interval (ICC[2,1] =0.99) (Urquhart, 2002) Procedure Subjects were positioned in prone with raised supports placed underneath the xiphisternum and pubic symphysis (Fig. 1). This allowed the edge ofthe anterior abdominal wall to be visible. The spine was positioned in neutral and the hips were flexed to 451. In separate trials, subjects were positioned in supine with similar lumbar spine, hip and knee positions. Subjects were trained by physiotherapists, experienced in exercise prescription for the abdominal muscles, to perform four manoeuvres using standard instructions (Table 1); inward movement ofthe lower abdominal wall (Richardson et al., 1999), abdominal bracing (flaring ofthe lateral and anterior abdominal wall) (Kennedy, 1965, 1980), posterior pelvic tilting (posterior rotation ofthe pelvis), and combined inward movement ofthe lower and upper abdominal wall. Contemporary exercise interventions focus on low level contractions (Richardson et al., 1999), which is consistent with evidence that suggests low effort is sufficient to provide muscle stiffness required for joint control (Hoffer and Andreassen, 1981; Cholewicki and McGill, 1996). Thus, each task was performed with mild effort, which is equivalent to a rating of 2 on the Borg scale (Borg, 1982). Subjects were trained with instruction and verbal and tactile feedback until they were able to perform the manoeuvres correctly. Three repetitions were performed and the order of tasks was randomized. A trigger was activated by the subject to signal when they were relaxed (baseline) and had performed the task. Maximum voluntary isometric trunk flexion, ipsilateral and contralateral trunk rotation, and a maximal valsalva and forced expiratory manoeuvre were performed in supine for normalization of RA, OI, OE and TrA EMG, respectively (Hodges et al., 1999). The peak activity ofeach muscle across these tasks was selected for normalization. A submaximal isometric manoeuvre was performed as an alternative task for EMG normalization and involved elevation ofboth legs so that the heels were 5 cm from the supporting surface. trigger (A) (B) abdominal regional markers light source spine marker black background electrode insertion sites ASIS marker LED 2.5. Data processing The root mean square (RMS) EMG amplitude was calculated for 2 s at baseline and for 2 s during the manoeuvre (at the time indicated by the trigger). The mean displacement ofthe upper, middle and lower regions ofthe abdominal wall, and the motion ofthe spine and pelvis in the vertical and horizontal planes was also determined for these periods. EMG activity recorded during the maximal and submaximal tasks was used to normalize the RMS EMG amplitude. Although reduced variance has been reported with normalization ofsurface EMG recordings to a submaximal task (Allison et al., 1998), maximal efforts have been considered to provide more meaningful values for interpretation (Andersson et al., 1998; Burden and Bartlett, 1999) Statistical analysis Fig. 1. Experimental set-up. Subjects were positioned in prone with supports underneath the xiphisternum and pubic symphysis (A), and in supine with their hips flexed to 451 (B). A marker was placed on the spinous process ofthe L3 vertebrae and the left ASIS, and borders of the abdominal regions were marked. A black background was used and a light source was placed inferior to the abdominal wall. A two-way repeated-measures ANOVA was used to compare activity between exercise tasks and between muscles/regions. Duncan s multiple-range test was used for post-hoc analysis. To examine the association between EMG activity ofthe abdominal muscles and

121 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) Table 1 Standardized instructions used for the voluntary exercises Exercise Inward movement ofthe lower abdominal wall Inward movement ofthe lower and upper abdominal wall Abdominal bracing Posterior pelvic tilting Instructions a Breathe in and out. Gently and slowly draw in your lower abdomen below your navel without moving your upper stomach, back or pelvis. Breathe in and out. Gently and slowly draw in your lower and upper abdomen without moving your back or pelvis. Breathe in and out. Gently and slowly swell out your waist without drawing your abdomen inwards or moving your back or pelvis. Breathe in and out. Gently and slowly rock your pelvis backwards. a Subjects were also instructed to perform each exercise with mild effort (a rating of 2 on the Borg scale). Table 2 Standard deviation data for the RMS EMG amplitude of the abdominal muscles normalized to maximal (Mx) and submaximal (SMx) isometric voluntary contractions and results ofthe F max test (F) for comparison of the variance between these normalization techniques Muscle/region Abdominal exercise Lower (supine) Pelvic tilting Bracing Lower (prone) Lower/upper Mx SMx F Mx SMx F Mx SMx F Mx SMx F Mx SMx F LTrA S S S S S MTrA S S S S S UTrA S S S S S LOI S S S S S MOI NS S S S S OE S S NS S NS RA NS NS NS NS NS L lower; M middle; U upper; Lower (supine) inward movement ofthe lower abdominal wall in supine; pelvic tilting posterior tilting ofthe pelvis; bracing abdominal bracing; lower (prone) inward movement ofthe lower abdominal in prone; lower/upper inward movement ofthe lower and upper abdominal wall; NS non-significant; S significant (Po0.05). abdominal, spinal and pelvic motion, Pearson productmoment correlations were calculated. The F max statistic was used to investigate differences in variance between the mean RMS EMG for each muscle normalized to a maximal and submaximal task (Winer et al., 1991). Statistical significance was set at Results 3.1. EMG normalization Prior to analysis ofthe abdominal tasks, the maximal and submaximal EMG normalization methods were compared. There was greater variability in the mean RMS EMG amplitude with normalization to the submaximal procedure for all muscles except RA (Table 2). The standard deviations for the RMS EMG of lower and middle TrA were up to 180 times greater compared to the maximal normalization. Therefore, the intramuscular EMG data was normalized to EMG activity recorded during the maximal manoeuvre Comparison of abdominal muscle recruitment for each exercise There were differences in recruitment between the abdominal muscles during inward movement ofthe lower abdominal wall in supine, abdominal bracing and pelvic tilting (Po0.001) (Fig. 2A). In contrast, no difference between the abdominal muscles was observed with inward movement ofthe lower abdominal wall in prone (P40.05) and combined inward movement ofthe lower and upper abdominal wall (P40.009). During inward movement ofthe lower abdominal wall in supine, TrA EMG was 70%, 100% and 65% greater than that ofoi, OE and RA, respectively (Po0.01). Minimal activity ofoi, OE and RA (1.3%, 0.9%, 1.8%) was also observed for one subject. There were regional differences in TrA recruitment. Mean RMS EMG amplitude ofthe upper region was approximately halfthat ofthe lower and middle regions (Po0.001). In contrast, OI EMG was less than lower and middle TrA (Po0.02), but similar to RA and OE (P40.07). In addition, no difference in OI EMG was

122 148 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) Fig. 2. RMS EMG amplitude of abdominal muscles/regions during different exercise conditions (normalized to a maximal voluntary contraction). Mean (SD) RMS EMG oflower and middle TrA and OI, and upper TrA, OE and RA during inward movement ofthe lower abdominal wall (supine (lower supine) and prone (lower prone)), bracing, posterior tilting ofthe pelvis (pelvic tilting) and combined inward movement ofthe lower and upper abdominal wall (lower/upper). Note the greater and more independent activity oftra in supine compared to prone. Similarities in activation ofthe lower and middle regions of TrA, contrast with differences in activation of the upper region of the muscle. The standard deviations are large indicating variability in abdominal muscle recruitment between subjects. Po0.05. identified between regions ofthe muscle (P ¼ 0:3). Mean OE RMS EMG was negative, indicating reduction in activity from baseline. With abdominal bracing, OE EMG was greater than that ofupper TrA, lower OI, and RA (Po0.05). There was minimal activity ofupper TrA, and although there was a trend for differences in the EMG activity of regions oftra, this was not significant (lower TrA: P ¼ 0:07; middle TrA: P ¼ 0:051). There was also similar activity ofthe lower and middle OI during abdominal bracing (P ¼ 0:09). When subjects tilted their pelvis posteriorly, middle OI had greater activity compared to RA (P ¼ 0:03) and upper TrA (P ¼ 0:01). In contrast, there was no

123 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) difference between the abdominal muscles during inward movement ofthe lower abdominal wall (P40.05), and the lower and upper abdominal wall in prone (P40.09). However, there was a trend towards greater TrA activity compared to the other abdominal muscles Comparison of abdominal muscle recruitment between exercises Recruitment oflower and middle TrA, and OE differed between the exercise conditions (Po0.001) (Fig. 2B). Lower and middle TrA EMG was greater during inward movement ofthe lower abdominal wall in supine than other tasks (Po0.05). In contrast, OE EMG was greater during abdominal bracing than the other techniques (except pelvic tilting) (Po0.05). Activity of lower and middle OI, RA and upper TrA was similar between exercises and between the supine and prone positions Movement of the abdominal wall, spine and pelvis Mean abdominal displacement (mm) (A) Regional proportion of abdominal displacement (B) * * lower lower/upper bracing pelvic tilting upper middle lower lower lower/upper bracing pelvic tilting Fig. 3. Displacement ofregions ofthe abdominal wall. Abdominal displacement expressed as the mean (SD) ofabsolute movement (A) and the mean expressed as a proportion ofthe total abdominal movement (B) during inward movement ofthe lower abdominal wall in prone (lower), pelvic tilting, abdominal bracing, and combined inward movement ofthe lower and upper abdominal wall (lower/ upper). Note the differences in abdominal displacement between the exercise conditions. Po0.05. * Abdominal wall displacement differed between tasks (Po0.001), but not between the upper, middle and lower abdominal regions (P ¼ 0:1) (Fig. 3). Greater abdominal motion occurred during pelvic tilting compared to the other abdominal manoeuvres (Po0.001), and abdominal displacement with inward movement of the upper and lower abdominal wall was greater than abdominal bracing and inward movement ofthe lower abdominal wall in prone (Po0.002; Po0.001). The later two exercises did not differ in abdominal motion (P ¼ 0:5). Lumbar spine and pelvic motion was minimal and did not differ between tasks, with the exception of posterior pelvic tilting, in which greater spine and pelvic motion occurred (Po0.001) (Fig. 4). There was a high correlation between movement ofthe lumbar spine and pelvis (r ¼ 0:9), and a significant negative correlation between lumbopelvic motion and TrA EMG (as a proportion of total activity) was found (r ¼ 0:6) (Fig. 5). Although there was no significant correlation between displacement ofthe lumbopelvic region and OI and RA EMG, there was a positive correlation between OE EMG and lumbopelvic motion. In addition, there was a low to moderate correlation between movement ofthe abdominal wall and TrA EMG (r ¼ 0:4; Po0.05) (Fig. 5). Spinal displacement (mm) (A) Pelvic displacement (mm) (B) * lower lower/upper bracing pelvic tilting lower lower/upper bracing pelvic tilting Fig. 4. Mean (SD) displacement ofthe lumbar spine and pelvis. Movement ofthe lumbar spine (A) and pelvis (B) during inward movement ofthe lower abdominal wall in prone (lower), pelvic tilting, abdominal bracing, and combined inward movement ofthe lower and upper abdominal wall (lower/upper). Note greater movement ofthe pelvis and spine during posterior pelvic tilting. Po0.05. * * * * *

124 150 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) TrA EMG activity (increase) Abdominal displacement (increase) 4. Discussion This study presents several important findings. First, there were distinct patterns ofabdominal muscle recruitment between exercise tasks. Notably, the greatest and most independent activity oftra was recorded with inward movement ofthe lower abdominal wall in supine. Second, abdominal muscle activity was dependent on body position, with differential activity of TrA evident in supine, but not in prone. Third, there were regional differences in the recruitment of TrA, with greater activity ofthe lower and middle regions oftra compared to the upper region. Finally, activity oftra was greater relative to the other abdominal muscles when lumbopelvic motion was limited. These results have important implications for selection of exercise techniques, positions and strategies for assessment and retraining ofabdominal muscle function Methodological issues 0.1 Pelvic motion (decrease) Fig. 5. Association between EMG activity, abdominal displacement and lumbopelvic motion. A three-dimensional graph depicting the relationship between EMG activity ofall regions oftra (as a proportion ofthe total abdominal muscle activity) (y axis), maximal abdominal displacement (x axis), and pelvic motion (z axis). EMG activity oftra, relative to the other abdominal muscles, was greater when abdominal movement was performed without pelvic motion. Two methodological issues require consideration. Firstly, due to the invasive nature ofthe study only seven subjects were recruited. Although this number is relatively consistent with previous intramuscular EMG investigations, it is important to consider that this may limit the statistical power ofthe study. Second, data in this study were normalized to maximal voluntary contractions. Variability in the present study was less when data were normalized to maximal manoeuvres rather than submaximal tasks. Although this contrasts with a previous study (Allison et al., 1998), the differences may be explained by the use of surface EMG in that investigation Inward movement of the lower abdominal wall in supine The results suggest that recruitment oftra with minimal activity ofother abdominal muscles may be best achieved during inward movement ofthe lower abdominal wall. These findings agree with reports that TrA is most consistently active during a belly in manoeuvre (Strohl et al., 1981; Goldman et al., 1987; De Troyer et al., 1990), and that minimal superficial abdominal muscle activity occurs during this task (Jull et al., 1995; Richardson et al., 1995). The results are also consistent with an exercise approach for the management oflbp, which involves retraining the activity of TrA to be independent ofthe other abdominal muscles (Richardson et al., 1999). Three randomised control trials ofdi ferent subgroups have reported improvements in pain and function with exercise interventions that involve inward movement ofthe lower abdomen (Hides et al., 1996; O Sullivan et al., 1997b,c). These outcomes have been hypothesized to result from improved motor control of TrA (and multifidus). Each ofthese studies involved training in a variety ofpositions, including supine (O Sullivan et al., 1997b,c) and standing (Hides et al., 1996) in the early stages ofrehabilitation, and during functional activities as exercise retraining was progressed (Hides et al., 1996; O Sullivan et al., 1997b,c). Although it is unlikely that the improvements were solely due to changes in TrA function, this is the common feature of the interventions. As the results of the present study suggest that the ability to activate TrA may vary between positions and it cannot be confirmed that the same manoeuvre examined in the current study was implemented, further research is required to determine whether TrA activity can be changed with this intervention. Activation oftra with minimal superficial abdominal muscle activity has been argued to be an important feature ofinward movement ofthe lower abdominal wall. In this study mean EMG activity ofthese muscles was considerably less than that oftra. In addition, minimal activity ofoi, OE and RA in one subject suggests that it may be possible to activate TrA almost independently from the other abdominal muscles, at least with training during this task. There was no difference between OI, OE and RA during inward movement ofthe lower abdominal wall in supine. However, the slightly greater activity ofoi may have reached significance with a greater number of

125 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) subjects. Dowd (1992) reported similar findings using intramuscular EMG but did not record from TrA. In contrast, surface EMG studies have reported greater activity ofoi and/or OE relative to RA (O Sullivan et al., 1997a; Vezina and Hubley-Kozey, 2000). These differences may be explained by cross-talk from deeper and adjacent muscles, possibly resulting in overestimation ofthe superficial muscle activity. There were regional differences in TrA recruitment during inward movement ofthe lower abdominal wall in supine. This is a novel finding. Although activity ofupper and lower/middle TrA varies during trunk rotation (Urquhart et al., 2004) and repetitive limb movements (Hodges et al., 1999), no studies have identified regional differences during voluntary manoeuvres Inward movement of the lower abdominal wall in prone Unlike supine, there was no differentiation in abdominal muscle activity with inward movement of the lower abdominal wall in prone. This is consistent with previous studies that report differences in abdominal muscle activity between positions (Carman et al., 1972; Richardson et al., 1992). This may be due to the greater gravitational demand in prone, or reflexmediated activity ofthe superficial muscles in response to stretch. In addition, an individual s internal body representation has been shown to vary with the relative position ofbody segments, which may influence movement performance (Gurfinkel, 1994). The absence of differentiation of abdominal muscle activity in prone is not consistent with the use ofthis position for evaluation oftra activity in clinical practice (Richardson et al., 1999). Although this technique is widely referenced and the position used in this study differs in several characteristics to the clinical test (e.g. abdominal support), assessment in supine may be more optimal for future clinical and laboratory work Abdominal bracing Identification ofgreater OE activity than the other abdominal muscles with abdominal bracing differs from previous reports which indicate greater RA activity compared to the anterolateral abdominals (Richardson et al., 1995), and no difference between muscles (Allison et al., 1998). However, the results suggest that bracing would not be appropriate ifthe aim ofthe exercise is to preferentially activate TrA or OI Posterior pelvic tilt Similar to our data, Partridge and Walters (1960) reported greater activity ofoi than RA and OE with posterior pelvic tilt. However, other studies have found greater RA activity compared to the anterolateral abdominals (Richardson et al., 1995), and greater activity ofoe than RA (Vezina and Hubley-Kozey, 2000). In addition, similar activity ofoi and RA has been observed during this manoeuvre (Flint and Gudgell, 1965; Carman et al., 1972). Although these varying results may have been due to differences in the task, electrode placement or EMG normalization technique, they also provide evidence that body position may contribute to differences in abdominal muscle recruitment Comparison of abdominal muscle recruitment between exercises In contrast to OI and RA, activity oftra and OE differed between the tasks, with greater activity during inward movement ofthe abdominal wall and pelvic tilting, respectively. This is consistent with previous reports ofgreater OE EMG activity during posterior tilting ofthe pelvis compared to abdominal hollowing (drawing your navel up and in towards your spine) (Vezina and Hubley-Kozey, 2000). However, activity of RA (Vezina and Hubley-Kozey, 2000) and the oblique abdominals (Richardson et al., 1992) has also been reported to vary between these manoeuvres. Differences between studies may be due to variation in the level of effort. It is important to note that activity of lower OI followed a similar pattern to that of TrA. Although there was no difference in OI activity between the exercises, this may have been due to insufficient statistical power that resulted from the small number ofsubjects used in this invasive study Abdominal, lumbar spine and pelvic movement Although abdominal wall movement differed between the tasks, there was no variation in the displacement between regions ofthe abdominal wall. This may be due to the small size ofthe displacement. However, there was trend towards greater movement ofthe lower region during inward movement ofthe lower abdominal wall. This finding is consistent with clinical observations (Richardson et al., 1999). Recruitment oftra and the combined activity ofoi, OE and RA (as a proportion oftotal abdominal muscle activity) was found to vary linearly with the amplitude oflumbar spine and pelvic displacement. This is consistent with clinical hypotheses and indicates that activation oftra is more independent ifthere is no pelvis or spinal motion (Richardson et al., 1999). There was also a trend for TrA EMG to be related to abdominal wall movement. This agrees with previous reports ofa relationship between pressure change (as measured with an air-filled cuff) associated with inward

126 152 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) displacement ofthe abdominal wall, and function of TrA, recorded as EMG onsets associated with arm movement (Hodges et al., 1996a). Thus, TrA is more likely to represent a greater proportion oftotal abdominal activity when abdominal movement occurs with limited lumbopelvic motion Clinical implications This study has implications for abdominal muscle retraining in clinical practice. The results provide further evidence to validate inward movement ofthe lower abdominal wall in the rehabilitation oftra in LBP patients. The findings may also assist in selection of exercises for assessment and retraining of the other abdominal muscles. For instance, pelvic tilting is likely to produce greater activity ofmiddle OI relative to upper TrA and RA, and abdominal bracing recruits OE with less activity ofupper TrA, lower OI and RA. In addition, incorrect strategies used to mimic the required task may also be identified. To activate TrA independently from the other abdominal muscles, it would be important to discourage movement ofthe upper abdomen, bracing ofthe abdominal wall, or posterior tilting ofthe pelvis. These results also emphasize the importance of observation for assessment of muscle function. For instance, motion ofthe abdominal wall and lumbopelvic region may assist in the determination ofthe muscle recruitment strategy. Furthermore, these results indicate that abdominal muscle recruitment may be influenced by patient positioning. Differential recruitment of TrA may be improved in supine compared to prone, indicating that assessment and re-education ofabdominal muscle function in a range of positions should be considered. However, further research is required to determine whether similar strategies are used by people with LBP and to develop improved strategies for restoration ofmotor control. Acknowledgments This work was supported by the Australian Physiotherapy Association (Victorian Branch) and the National Health and Medical Research Council. The authors wish to thank Lorimer Moseley for his assistance with data collection and Beryl Kennedy for discussions on methodology. References Allison GT, Godfrey P, Robinson G. EMG signal amplitude assessment during abdominal bracing and hollowing. Journal of Electromyography and Kinesiology 1998;8(1):51 7. Andersson EA, Ma Z, Thorstensson A. Relative EMG levels in training exercises for abdominal and hip flexor muscles. Scandinavian Journal ofrehabilitation Medicine 1998;30(3): Askar OM. Surgical anatomy ofthe aponeurotic expansions ofthe anterior abdominal wall. Annals ofthe Royal College ofsurgeons ofengland 1977;59(4): Baecke JAH, Burema J, Frijters JER. A short questionnaire for the measurement ofhabitual physical activity in epidemiological studies. The American Journal ofclinical Nutrition 1982;36(5): Borg GA. Psychophysical bases ofperceived exertion. Medicine and Science in Sports and Exercise 1982;14(5): Burden A, Bartlett R. Normalisation ofemg amplitude: an evaluation and comparison ofold and new methods. Medical Engineering and Physics 1999;21(4): Carman DJ, Blanton PL, Biggs NL. Electromyographic study ofthe anterolateral abdominal musculature utilising indwelling electrodes. American Journal ofphysical Medicine 1972;51(3): Cholewicki J, McGill SM. Mechanical stability ofthe in vivo lumbar spine: implications for injury and chronic low back pain. Clinical Biomechanics 1996;11(1):1 15. Cresswell AG, Grundstrom H, Thorstensson A. Observations on intraabdominal pressure and patterns ofabdominal intra-muscular activity in man. Acta Physiologica Scandinavica 1992;144(4): De Troyer A, Estenne M, Ninane V, Van Gansbeke D, Gorini M. Transversus abdominis muscle function in humans. Journal of Applied Physiology 1990;68(3): Dowd C. An electromyographic study ofthe abdominal muscles in various exercise positions. Masters Thesis, University ofsouth Australia, Adelaide, Flint MM, Gudgell J. Electromyographic study ofabdominal muscular activity during exercise. The Research Quarterly 1965;36(1): Goldman JM, Lehr RP, Millar AB, Silver JR. An electromyographic study ofthe abdominal muscles during postural and respiratory manoeuvres. Journal ofneurology, Neurosurgery and Psychiatry 1987;50(7): Gurfinkel VS. The mechanisms ofpostural regulation in man. Soviet Scientific Reviews. Section F. Physiology and General Biology 1994;7: Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after resolution of acute, first-episode low back pain. Spine 1996;21(23): Hodges PW, Richardson CA, Jull GA. Evaluation ofthe relationship between laboratory and clinical tests oftransversus abdominis function. Physiotherapy Research International 1996a;1(1): Hodges PW, Richardson CA. Inefficient muscular stabilization ofthe lumbar spine associated with low back pain. A motor control evaluation oftransversus abdominis. Spine 1996b;21(22): Hodges PW, Richardson CA. Feedforward contraction of transversus abdominis is not influenced by the direction ofarm movement. Experimental Brain Research 1997;114(2): Hodges PW, Richardson CA. Delayed postural contraction of transversus abdominis in low back pain associated with movement ofthe lower limb. Journal ofspinal Disorders 1998;11(1): Hodges PW, Cresswell A, Thorstensson A. Preparatory trunk motion accompanies rapid upper limb movement. Experimental Brain Research 1999;124(1): Hodges PW. Changes in motor planning of feedforward postural responses ofthe trunk muscles in low back pain. Experimental Brain Research 2001;141(2): Hoffer J, Andreassen S. Regulation of soleus muscle stiffness in premammillary cats: intrinsic and reflex components. Journal of Neurophysiology 1981;45(2):

127 ARTICLE IN PRESS D.M. Urquhart et al. / Manual Therapy 10 (2005) Jull GA, Richardson CA, Hamilton CA, Hodges PW, Ng JK-F. Towards the validation ofa clinical test for the deep abdominal muscles in back pain patients. In: Proceedings ofninth Biennial Conference of the Manipulative Physiotherapists Association of Australia, Gold Coast, Australia, p Kennedy B. Muscle-bracing technique for utilising intra-abdominal pressure to stabilise the lumbar spine. Australian Journal of Physiotherapy 1965;11(1): Kennedy B. An Australian programme for management of back problems. Physiotherapy 1980;66(4): O Sullivan PB, Twomey L, Allison G, Sinclair J, Miller K, Knox J. Altered patterns ofabdominal muscle activation in patients with chronic low back pain. Australian Journal ofphysiotherapy 1997a;43(2):91 8. O Sullivan PB, Twomey L, Allison G, Taylor JR. Specific stabilising exercise in the treatment ofchronic low back pain with a clinical and radiological diagnosis oflumbar segmental instability. In: Proceedings ofmanipulative Physiotherapists Association of Australia Tenth Biennial Conference, Melbourne, Australia, 1997b. p O Sullivan PB, Twomey LT, Allison GT. Evaluation ofspecific stabilizing exercise in the treatment ofchronic low back pain with radiologic diagnosis ofspondylolysis or spondylolisthesis. Spine 1997c;22(24): O Sullivan PB, Twomey L, Allison GT. Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. Journal oforthopaedic and Sports Physical Therapy 1998;27(2): Partridge MJ, Walters CE. Participation ofthe abdominal muscles in various movements ofthe trunk in man. An electromyographic study. The Physical Therapy Review 1960;39(12): Richardson CA, Toppenberg R, Jull G. An initial evaluation of eight abdominal exercises for their ability to provide stabilisation for the lumbar spine. Australian Journal of Physiotherapy 1990; 36:6 11. Richardson CA, Jull G, Toppenberg R, Comerford M. Techniques for active lumbar stabilisation for spinal protection: A pilot study. Australian Journal ofphysiotherapy 1992;38(2): Richardson CA, Jull GA, Richardson BA. A dysfunction of the deep abdominal muscles exists in low back pain patients. In: Proceedings ofworld Confederation ofphysical Therapists, Washington DC, p Richardson CA, Jull GA, Hodges PW, Hides JA. Therapeutic exercise for spinal segmental stabilization in low back pain: Scientific basis and clinical approach. London: Churchill Livingstone; Rizk NN. A new description ofthe anterior abdominal wall in man and mammals. Journal ofanatomy 1980;131(3): Strohl KP, Mead J, Banzett RB, Loring SH, Kosch PC. Regional differences in abdominal muscle activity during various maneuvers in humans. Journal ofapplied Physiology 1981;51(6): Urquhart DM, Barker PJ, Hodges PW, Story IH, Briggs CA. Regional morphology oftransversus abdominis and internal oblique. In: Proceedings ofthe Musculoskeletal Physiotherapy Australia Twelfth Biennial Conference, Adelaide, Australia, Urquhart DM, Hodges PW, Story IH. Regional recruitment of transversus abdominis in trunk rotation. European Spine Journal 2004; p. 35. Urquhart, DM. Regional variation in morphology and recruitment of the abdominal muscles: Implications for control and movement of the lumbar spine and pelvis. PhD Thesis, School ofphysiotherapy, The University ofmelbourne, Vezina MJ, Hubley-Kozey CL, Egan DA. A review ofthe muscle activation patterns associated with the pelvic tilt exercise used in the treatment oflow back pain. The Journal ofmanual and Manipulative Therapy 1998;6(4): Vezina MJ, Hubley-Kozey CL. Muscle activation in therapeutic exercises to improve trunk stability. Archives ofphysical Medicine and Rehabilitation 2000;81(10): Winer BJ, Brown DR, Michels KM. Statistical principles in experimental design, 3rd ed. New York: McGraw-Hill; 1991.

128 Journal of Anatomy J. Anat. (2014) 225, pp doi: /joa The functional coupling of the deep abdominal and paraspinal muscles: the effects of simulated paraspinal muscle contraction on force transfer to the middle and posterior layer of the thoracolumbar fascia A. Vleeming, 1,2 M. D. Schuenke, 1 L. Danneels 2 and F. H. Willard 1 1 Department of Anatomy, University of New England College of Osteopathic Medicine, Biddeford, ME, USA 2 Department of Rehabilitation Sciences and Physiotherapy, University of Ghent, Ghent, Belgium Abstract The thoracolumbar fascia (TLF) consists of aponeurotic and fascial layers that interweave the paraspinal and abdominal muscles into a complex matrix stabilizing the lumbosacral spine. To better understand low back pain, it is essential to appreciate how these muscles cooperate to influence lumbopelvic stability. This study tested the following hypotheses: (i) pressure within the TLF s paraspinal muscular compartment (PMC) alters load transfer between the TLF s posterior and middle layers (PLF and MLF); and (ii) with increased tension of the common tendon of the transversus abdominis (CTrA) and internal oblique muscles and incremental PMC pressure, fascial tension is primarily transferred to the PLF. In cadaveric axial sections, paraspinal muscles were replaced with inflatable tubes to simulate paraspinal muscle contraction. At each inflation increment, tension was created in the CTrA to simulate contraction of the deep abdominal muscles. Fluoroscopic images and load cells captured changes in the size, shape and tension of the PMC due to inflation, with and without tension to the CTrA. In the absence of PMC pressure, increasing tension on the CTrA resulted in anterior and lateral movement of the PMC. PMC inflation in the absence of tension to the CTrA resulted in a small increase in the PMC perimeter and a larger posterior displacement. Combining PMC inflation and tension to the CTrA resulted in an incremental increase in PLF tension without significantly altering tension in the MLF. Paraspinal muscle contraction leads to posterior displacement of the PLF. When expansion is combined with abdominal muscle contraction, the CTrA and internal oblique transfers tension almost exclusively to the PLF, thereby girdling the paraspinal muscles. The lateral border of the PMC is restrained from displacement to maintain integrity. Posterior movement of the PMC represents an increase of the PLF extension moment arm. Dysfunctional paraspinal muscles would reduce the posterior displacement of the PLF and increase the compliance of the lateral border. The resulting change in PMC geometry could diminish any effects of increased tension of the CTrA. This study reveals a co-dependent mechanism involving balanced tension between deep abdominal and lumbar spinal muscles, which are linked through the aponeurotic components of the TLF. This implies the existence of a point of equal tension between the paraspinal muscles and the transversus abdominis and internal oblique muscles, acting through the CTrA. Key words: abdominal muscles; erector spinae; middle lamina; multifidus; posterior lamina; spine; thoracolumbar fascia; transversus abdominis. Introduction Correspondence Mark D. Schuenke, Department of Anatomy, University of New England College of Osteopathic Medicine, 11 Hills Beach Rd, Biddeford, ME 04005, USA. T: ; E: mschuenke@une.edu Accepted for publication 11 July 2014 Article published online 20 August 2014 Stabilization and movement of the lumbosacral spine is contingent on the complex interaction between muscles, ligaments and fascia surrounding the torso. The thoracolumbar fascia (TLF) represents a girdling structure consisting of several aponeurotic and fascial layers that separates the paraspinal muscles from the muscles of the posterior 2014 Anatomical Society

129 448 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. abdominal wall. Understanding the complex function of the TLF and its associated fascial compartments is critical to anatomical and biomechanical analysis, and implementation of effective treatment in patients with lumbopelvic pain. The TLF envelops the back muscles from the sacral region, through the thoracic region, and plays an important role in posture, load transfer and respiration (Gracovetsky et al. 1981, 1985; Bogduk & MacIntosh, 1984; Carr et al. 1985; Tesh et al. 1987; Vleeming et al. 1995; Barker & Briggs, 1999, 2007; Hodges et al. 2003; Barker et al. 2004, 2006; Schuenke et al. 2012; Willard et al. 2012). The TLF is comprised of three layers of which both the fibrous posterior layer (PLF) and the middle layer (MLF) have a significant biomechanical function (Gracovetsky et al. 1981, 1985; Bogduk & MacIntosh, 1984; Carr et al. 1985; Tesh et al. 1987; Hukins et al. 1990; Vleeming et al. 1995; Barker & Briggs, 1999; Barker et al. 2004, 2006; Urquhart & Hodges, 2007; Gatton et al. 2010; Schuenke et al. 2012; Willard et al. 2012). The delicate anterior layer merely represents the thin transversalis fascia lining the deep surface of transversus abdominus and the quadratus lumborum muscles (Willard et al. 2012). Superficial lamina of the PLF The PLF is a composite of superficial and deep laminae of connective tissue. The superficial lamina derives from the aponeurosis of the latissimus dorsi (LD; Bogduk & MacIntosh, 1984; Tesh et al. 1987; Vleeming et al. 1995; Barker & Briggs, 1999; Gatton et al. 2010; Willard et al. 2012) and is part of a collective sheath of fascia, bridging from the first rib down to the xiphoid process anteriorly, and from the cranial base to the sacrum posteriorly (Stecco et al. 2009; Willard et al. 2012). This fascial sheath contains muscles such as the pectoralis major and minor, rhomboid major and minor, trapezius, serratus anterior (Barker et al. 2004; Willard et al. 2012) and the expansive LD, reaching far caudally and forming the superficial lamina of the PLF. In addition, the aponeurosis of this muscle partially crosses the midline to connect to the fascia of the contralateral gluteus maximus muscle (GM; Bogduk & MacIntosh, 1984; Vleeming et al. 1995; Barker et al. 2004). Deep lamina of the PLF The deep lamina of the PLF extends from the spinous processes to the transverse processes, and is distinct from both the superficial lamina of the PLF and the middle layer of the MLF. Cranially, the deep lamina most likely begins on the occipital bone and extends caudally to its fusion with the superficial lamina over the sacrum. The lateral margin of the deep lamina of the TLF is located at the common intersection of the hypaxial (e.g. ventral trunk muscles) and epaxial (paraspinal) muscles (Willard et al. 2012). Several authors have studied the deep lamina of the PLF (Bogduk & MacIntosh, 1984; Tesh et al. 1987; Vleeming et al. 1995; Barker & Briggs, 1999). Bogduk & MacIntosh (1984) described the deep lamina as having alternating bands of dense fibers, which they termed accessory ligaments and proposed that the deep lamina stems most likely from the crossed fibers of the aponeurosis tendon of the LD. Vleeming et al. (1995) and Barker & Briggs (1999) describe the same fascial bands; however, typically characterizing the deep lamina of the PLF as being mainly formed by the aponeurosis tendon of the serratus posterior inferior muscle. The arrangement of a fascial compartment in the lumbar spine, created by a fascial sheath encapsulating the paraspinal muscles, has been noted or illustrated by numerous authors (Spalteholz, 1923; Schaeffer, 1953; Hollinshead, 1969; Grant, 1972; Farfan, 1973; Gracovetsky et al. 1977; Bogduk & MacIntosh, 1984; Clemente, 1985; Tesh et al. 1987; Vleeming et al. 1995; Barker et al. 2004; Gatton et al. 2010). Standring (2008) described a designated osteofascial compartment for the paraspinal muscles. Many authors cited above utilize the deep lamina of the PLF to describe the inner posterior wall of this encapsulating sheath and the MLF to describe the anterior wall. However, most of these descriptions are based on the assumption that the deep lamina of the PLF is a longitudinally oriented, flat fascial sheath (Willard et al. 2012). The lateral border of the deep lamina contributes to the lateral raphe (Schaeffer, 1953; Bogduk & MacIntosh, 1984; Tesh et al. 1987; Vleeming et al. 1995; Barker et al. 2004). Spalteholz (1923) describes the lateral border as curving around the paraspinal muscles to join anteriorly to the MLF. Tesh et al. (1987) describe the deep lamina as encircling the paraspinal muscles. Likewise, Carr et al. (1985) measured intra-compartmental TLF pressure and concluded that sustained pressure within the TLF is only possible if the paraspinal muscles are enclosed by a continuous fascial sheath. A recent study has examined the extent of the deep lamina and confirmed that it forms a sheath surrounding the paraspinal muscles, which has been termed the paraspinal retinacular sheath (PRS). This sheath represents the innermost part of the deep lamina of the PLF (Schuenke et al. 2012), and is attached to the spinous process posteriorly and the transverse process anteriorly. Laterally, the PRS forms a junction with the common transversus tendon (CTrA), deriving from the inner oblique muscle below the transverse process of L3 (Bogduk & MacIntosh, 1984; Barker et al. 2004; Urquhart & Hodges, 2005) but mainly from the transversus abdominus. The CTrA forms a strong anchor or seam for the transmission of force between the abdominal muscles anteriorly and the paraspinal muscles posteriorly (Fig. 1). The CTrA and the PRS enclose a triangular-shaped, fatfilled space, the lateral interfascial triangle (LIFT), which derives from the bifurcation of the anterior and posterior lamina of the CTrA, and the portion of PRS that spans 2014 Anatomical Society

130 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 449 Fig. 1 Modified with permission from fig. 4, Schuenke et al. (2012). A schematic and simplified view of the bifurcation of the common transversus tendon (CTrA) and the paraspinal retinacular sheath (PRS), creating the lumbar interfascial triangle (LIFT). The CTrA bifurcates into anterior and posterior laminae. The anterior lamina contributes to the middle layer of the thoracolumbar fascia (MLF). The posterior lamina contributes to the deep lamina of the posterior layer of the thoracolumbar fascia (PLF). The junction of the CTrA with the PRS creates the LIFT. dplf, deep lamina of PLF; splf, superficial lamina of PLF. between them. This LIFT may act as a fulcrum distributing laterally mediated tension, to balance different viscoelastic moduli, along either the MLF or PLF. The presence of the LIFT explains why an externally ridged-union of dense connective tissue is formed, called the lateral raphe (Bogduk & MacIntosh, 1984; Schuenke et al. 2012). Three studies specifically analyzed lateral force transfer through the CTrA to the TLF. Tesh et al. (1987) were the first to simulate intra-compartmental pressure (ICP) within the paraspinal compartment. They replaced the paraspinal muscles with foam dowels to create a static tension within the TLF. Their study focused primarily on tension in the PLF. Barker et al. (2004) studied lumbar neutral zone movement, and applied both static and cyclical loading to lumbar segments. They analyzed the effect of CTrA pull to the TLF, to examine the significance of the transverse abdominus and internal oblique muscles to segmental stability of the spine. The authors simulated tension through the CTrA, and found tension increasing primarily through the MLF as the mechanical pull to the CTrA was directed transversely (parallel to the MLF), without simulation of paraspinal contraction of the paraspinal muscle compartment (PMC). Hodges et al. (2003), in a porcine study, analyzed pull through the transverse abdominus muscle, exclusively focusing on the force transfer through the MLF as the PLF was cut. The study concluded that tensing the fascia (MLF) produces an extension moment. To better understand low back and pelvic girdle pain, it is essential to develop a detailed understanding of how abdominal and spinal muscles cooperate to influence lumbopelvic motion and postural stability. Specifically, how activation of the middle parts of the transverse abdominus and internal oblique muscles influence force transfer to the PLF and MLF. The aim of the present study is to analyze the effect of incrementally raising inflation within the PMC (simulating paraspinal contraction), without and combined with simultaneous CTrA tension (simulating transverse abdominus/internal oblique contraction), on force transfer through the PLF and MLF. To the authors knowledge, this is the first study in which incremental ICP of the PMC is examined. The results could provide a better understanding of the relationship between dysfunction of the paraspinal muscles in patients with low back pain (LBP), in combination with force transfer from the deep abdominal muscles via the CTrA to the outer perimeter of the PMC, influencing lumbar stability. Materials and methods Specimen characteristics and preparation Seven embalmed (70% isopropyl alcohol, 2% phenol, 1% formaldehyde) human specimens (three male, four female; years) were studied. On one specimen, the skin and superficial fascia had been dissected before axial sectioning. In comparison to the other six specimens, there were no statistically significant differences, and the data of seven specimens were pooled. A total of 14 axial slabs were sectioned (approximately 2 cm thickness) using an industrial band saw (Hobart 5801; Troy, OH, USA). Prior to sectioning, the lumbar region of each cadaver was assessed with a C-arm fluoroscope (Exposcope 7000; Ziehm Imaging, Orlando, FL, USA) to identify planes that contain transverse processes bilaterally. The transverse processes were marked by the transverse placement of a needle, and axial sections made between the needle positions. Left and right PMCs of all seven specimens were analyzed individually, resulting in N = 14. None of the samples revealed evidence of lumbosacral pathology or surgical procedures in the lumbar region. Conducting the measurements at the level of the transverse processes is essential, because the MLF loses its insertion at inter-transverse levels in order to create a passageway for the dorsal neurovasculature. Only axial sections through levels L2 and L3 were used in this study, because 2014 Anatomical Society

131 450 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. sections including L1 contained rib fragments. Similarly, sections through the L4 level were not included, because they contained portions of the iliac crest. Objectives A To test the hypothesis that changes of ICP within the PMC (mimicking incremental contraction of paraspinal muscles) alters the load transfer between the PLF and MLF. In order to test this, the following took place. a The perimeter of the left and right PMC (from transverse process to spinous process) was measured at three stages of ICP without tension to the CTrA. b Using the same pressure increments (as in 1A), the perpendicular straight-line distance without CTrA tension was measured from the lateral tip of the transverse process to the posterior border of the PLF, to analyze posterior displacement of the PLF (Fig. 2). B To test the hypothesis that with tension of the CTrA and incremental PMC pressure, the fascial tension is primarily transferred to the PLF, rather than the MLF. In order to test this, measurements similar to those described in 1A and 1B were repeated with 8.5 N tension being exerted bilaterally through the CTrA. a b Load cells were used to measure unidirectional tension along anterior and posterior CTrA laminae with CTrA tension (Fig. 3a). This was repeated under three incremental stages of TLF compartmental pressure. Only one load cell was used in experiment 2A, to measure unidirectional tension for each anterior lamina and posterior lamina of the CTrA, analyzing load transfer, respectively, to the MLF and PLF, during CTrA tension while incrementally inflating the PMCs (Fig. 3b). After finalizing the experiments (2A), it became obvious that the PLF became significantly distended. It is reasonable to expect that during each incremental inflation, the rubber tubes (simulating paraspinal contraction) impose a posteriorly-directed force on the PLF. However, for most experiments, the load cells were oriented to quantify force in the anterolateral direction during CTrA tension, to differentiate force transfer between the MLF and PLF. Subsequently, the load cell would not directly measure inflation-induced tension in the posterior direction because inflation has the biggest effect on the PLF, partially minimizing the force in the unidirectional load cell with each incremental inflation. Therefore, to quantify this inflation-induced, posteriorly-oriented force, an additional set of experiments (bidirectional tension measurements along the posterior CTrA lamina with CTrA tension) was conducted bilaterally on two axial slabs (measuring in total four compartments per inflation condition). One load cell was positioned as before along the PLF in the posteromedial direction, additionally a second load cell was positioned along the PLF in an anterolateral direction (Fig. 3c). The CTrA was tensed with the same load of 8.5 N along the CTrA and the same incremental inflation stages as in experiment 2A. Testing sequence The following methods were used (details outlined below). a Simulate tensioning of CTrA with paraspinal muscles intact followed by analysis of perimeter changes of the PMC. b Remove paraspinal muscles and insert a custom-made butyl inflation device with valves into the right and left PMCs (see Fig. 3a). c Inflate incrementally the PMC to simulate contraction of paraspinal muscles and concurrently simulate contraction of transverse abdominus/internal oblique by tensing the CTrA. d Analyze both the perimeter changes of the PMC, and measure the force differential between MLF and PLF with load cells. Testing methods Preventing vertebral displacement Performed in all experiments The vertebral body of the cadaveric slab was clamped to a customized baseboard, to prevent movement of the vertebra, but not to impede any soft tissue movement. Also, the board was designed in Fig. 2 Analyzing posterior and lateral displacement of the borders of the TLF compartment with incremental inflation. Beads (black circles) were affixed to the PMC in order to track movement of individual points. Posterior displacement of the posterior border was measured on a perpendicular straight line from the lateral-most point of the transverse process to the posterior border of the PLF (Method 4; indicated by black crosses). This line was then used as a reference line for measuring medial-to-lateral displacement of the PMC (SLD lat ). This was measured from the perpendicular straight line to the lateral-most point of the PMC (indicated by white crosses). These measurements were done with (Method 1a) and without (as shown) CTrA tension Anatomical Society

132 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 451 a such a way to permit insertion of inflatable tubes through the PMC (Fig. 3a). A customized compression bar was tightened in place over the vertebral body thereby eliminating unwanted motion of the bony structures (Fig. 3a). b Marking the perimeter of the PMC In order to track perimeter changes of the fascial compartments, copper beads (2 8 mm; Sigma-Aldrich, USA) were affixed (methyl 2- cyanoacrylate Loctite; Henkel, USA) at approximately 1.5-cm intervals along the anterior and posterior lamina of the CTrA and the PRS (Fig. 2). Care was taken to ensure that beads adhered only to fascia and not to adjacent muscles. Method 1. Loading the CTrA Method 1a The left and right CTrA were pulled anterolaterally (mimicking the curved shape and direction of the transverse abdominus and internal oblique muscles of each individual specimen) to generate bilateral forces of 8.5 N. Two load cells (LCMFD-10N; Omega Engineering, Stamford, CT, USA) measured the generated tension. A tension of 8.5 N was selected, based on the work of Barker et al. (2004) who demonstrated that the CTrA in cadavers could strain at 10 N. c Fig. 3 (a) Experimental apparatus design. Cadaveric slab (A) is placed on a wooden platform with holes to accommodate inflatable tubes (C) attached to a positive displacement pump (note: jagged cut-out is to demonstrate spatial context). To prevent vertebral rotation, a crossbar (B) is placed across the vertebral body and clamped down using wingnuts on threaded bolts. A hemostatic clamp (D) attaches the common aponeurosis of the transversus abdominis and internal oblique muscles the CTrA to a constant load. Alligator clips attach the anterior and posterior laminae of the common aponeurosis to load cells (E) that connect to load cell meters with digital display (F). (b) Experimental apparatus design. Magnified view of alligator clip placement for differentiating MLF/PLF force transfer (Method 5). Alligator clips are attaching the anterior and posterior laminae of the common aponeurosis to load cells (not shown). Hemostatic clamp is attaching the common aponeurosis to a known load (not shown). For simplicity, the inflatable tubes are not shown in this image. (c) Experimental apparatus design. Magnified view of alligator clip placement for analyzing the effect of inflation on PLF force transfer (Method 6). Alligator clips are attached along the posterior laminae of the common aponeurosis in opposing directions. Hemostatic clamp is attaching the common aponeurosis to a known load (not shown). For simplicity, the inflatable tubes are not shown in this image. Method 1b The CTrA was loaded with 8.5 N of tension anterolaterally, via selflocking hemostats, simulating the normal function of the CTrA. Tension load cells were either attached, respectively, to the anterior and posterior lamina of the CTrA (Fig. 3b), or both attached to the posterior lamina in opposite directions (described in objective 2B; Fig. 3c). Method 2. Simulation of paraspinal muscle contraction Paraspinal muscles were carefully removed so as not to damage the surrounding fascia. Custom-made inflatable butyl rubber tubes (uninflated diameter 2.54 cm, length 10 cm) were placed inside the right and left PMC. To simulate contraction of the paraspinal muscles, the tubes were inflated. Due to inter- and intra-specimen variation of paraspinal muscle size, a standard initial pressure could not be used. Instead, the first inflation (Inf1) was the minimum pressure required to hold the tube in the compartment. Subsequent inflations were set at 1.5-cm increments above the tube circumference of the initial inflation. All circumferential measurements were recorded using a soft, tailor s tape measure immediately above the superior surface of the axial section. The average intra-tube circumferences for Inf1, Inf2 and Inf3 were cm, cm, cm, respectively. Intra-tube pressure was also measured using Vernier LabPro (Vernier Software and Technology, Beaverton, OR, USA). The average intra-tube pressures for Inf1, Inf2 and Inf3 were 79.1 mm Hg, 99.8 mm Hg and mm Hg, respectively. A study of healthy young individuals showed that the mean submaximal muscle contraction pressure was 175 mm Hg, during isometric and concentric extension exercises (Styf, 1987). Compared with the present study, the average inflation pressure is 50% less. Method 3. Measuring the perimeter of the TLF The pre-tensed (neutral) position of the copper beads was imaged using a C-arm fluoroscope. During tensioning the CTrA with 8.5 N, a second image was captured. In order to compare the positioning 2014 Anatomical Society

133 452 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. of the beads between the neutral and tensed CTrA, the opacity of the tensed image was reduced to 40%. The tensed image was superimposed to the neutral image using Adobe Illustrator (Adobe Systems, San Jose, CA, USA). For every incremental inflation, new pre-tensed and tensed images were captured and compared. During superimposition, vertebral processes from each image were aligned. Subsequently, a curved line was drawn, connecting all of the beads for a given inflation condition to measure perimeter changes (NIH ImageJ software) around the PMC. There was no statistical difference between the right and left MLF, nor for the right and left PLF, for a given inflation condition. Therefore, lengths of the left and right MLF and PLF were pooled for each inflation condition. Method 4. Analyzing posterior and lateral displacement of the PMC The straight-line perpendicular distance from the lateral-most tip of the transverse process to the posterior part of the PLF in the x-plane (Fig. 2) was measured using ImageJ software, with MTrackJ plug-in (Meijering et al. 2012). The distance from the aforementioned straight line to the lateral-most point of the border of the PMC was also measured. These measurements were taken under the same inflation increments as described in Method 3, and performed both without CTrA tension and with 8.5 N CTrA tension. Method 5. Differentiating MLF/PLF force transfer resulting from CTrA tension Tension load cells were attached to the anterior and posterior laminae of the CTrA, using alligator clips (Fig. 3b). The anterior lamina is continuous with the MLF, and the posterior lamina is continuous with the PLF. A reading of each load cell was recorded without CTrA tension (neutral). A load (8.5 N, after accounting for frictional and drag components) was then applied anterolaterally on to the CTrA, using self-locking hemostats, and was suspended on a pulley. Each step of incrementally pressurizing the PMC, with the inflatable tubes, was recorded with load cells, analyzing the relative force transfer between MLF and PLF. The differential in load transfer through the MLF and PLF of the CTrA is calculated by: (Method 2). There were no statistical differences in perimeter length between the right and left MLF, or the right and left PLF for any given inflation condition. Consequently, measurements of the right and left MLF were pooled and likewise for the PLF (Fig. 4). The perimeter length of the MLF did not change significantly with inflation (P = 0.78). In contrast, the length of PLF increased significantly with inflation (P = 0.046). Post hoc analyses revealed that the length of the PLF increases significantly between the muscles intact condition (before inflation) and Inf3 (P = 0.012). There was a trend towards significance between muscles intact and Inf2 (P = 0.051), and between Inf1 and Inf3 (P = 0.092). In the muscles intact condition and all three inflation increments, the perimeter of the PLF is significantly longer than the perimeter of the MLF (P = 0.001). Analyzing posterior and lateral displacement of the PLF Without CTrA tension The perpendicular straight-line distance in the posterior direction (SLD post ) was measured from the lateral tip of the transverse process to the PLF (Method 4; see Fig. 2). The SLD post increases with inflation (P = ). Post hoc analyses indicate that the SLD post in the muscles intact condition is significantly shorter than the SLD post in Inf1 (P = ), Inf2 (P = ) and Inf3 (P = ). The SLD post does not significantly differ between the three inflation conditions (Inf1 vs. Inf2, P = 0.52; Inf1 vs. Inf3, P = 0.18; Inf2 vs. Inf3, P = 0.49). The aforementioned line drawn perpendicular to the lateral tip of the transverse process (described above; see Fig. 2) was then used as a reference line from which to measure the straight-line lateral displacement (SLD lat ) to the lateral-most point of the PMC. The SLD lat did not MLF tension ¼ MLF tensed MLF Neutral PLF tension ¼ PLF tensed PLF Neutral Method 6. Analyzing the effect of inflation on PLF force transfer To quantify the subtractive effect of incremental inflation on specifically PLF tension, two axial slabs (measuring four compartments per inflation condition) were tested. One alligator clip connected to a load cell was positioned as before along the PLF in the posteromedial direction. Another alligator clip connected to a second load cell was positioned along the PLF in an anterolateral direction. The same CTrA load (8.5 N) and incremental inflations (as described in Method 2) were used (see Fig. 3c). Results Fascial perimeter without CtrA tension The perimeter of each MLF and PLF was measured (Method 3) with muscles intact and at each inflation increment Fig. 4 The length of fibrous middle layer (MLF) and fibrous posterior layer (PLF) at each increment of ICP without CTrA tension. There is statistical significance (*) between the muscles intact and Inf3 conditions, and ( ) between MLF and PLF for a given ICP Anatomical Society

134 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 453 significantly differ between the muscle intact condition and any of the inflation increments (P = 0.68). With CTrA tension To determine the net effect of CTrA tension (Method 1a) and inflation (Method 2) of the PMC (i.e. co-contraction of transverse abdominus/internal oblique and paraspinal muscles), the perpendicular SLD post of the baseline condition (Method 4; no CTrA tension, no PMC pressure) was compared with 8.5 N CTrA tension with incremental inflation. For example: SLD diff ¼ SLD 0N; mm intact SLD 8:5N; Inf1 Predictably, when the CTrA is tensed with no pressure in the PMC, the PLF moves anteriorly (Fig. 5). However, with each incremental inflation the amount of anterior movement due to CTrA tension is significantly reduced (P = ). In fact, the anterior movement due to CTrA tension is negated by the inflation-induced posterior movement such that the net displacement of the PLF is in the posterior direction. The amount of CTrA-induced anterior movement of the PLF is also significantly reduced in inflations Inf2 (P = 0.024) and Inf3 (P = ), relative to Inf1. There was no significant difference between Inf2 and Inf3 (P = 0.32). Similarly, to determine the net effect of CTrA tension and inflation of the PMC (i.e. co-contraction of transverse abdominus/internal oblique and paraspinal muscles), the SLD lat of the baseline condition (no CTrA tension, no PMC pressure) was compared with 8.5 N CTrA tension at each inflation increment (Fig. 6). With muscles intact (no PMC pressure), 8.5 N CTrA tension resulted in lateral movement of the SLD lat relative to Fig. 5 Straight-line distance from the tip of the transverse process to the PLF under different inflation increments (see also Fig. 2): comparison of baseline condition (mm intact, no CTrA tension) with 8.5 N CTrA tension and incremental inflation. A positive value indicates anterior displacement. A negative value indicates posterior displacement. *Significant difference from mm intact. Significant difference from inf1. TLF, thoracolumbar fascia. Fig. 6 Perpendicular straight-line distance from the most lateral point of the PMC, projected to the line running from the tip of the transverse process to the PLF under different inflation increments (see also Fig. 2): comparison of baseline condition (mm intact, no CTrA tension) with 8.5 N CTrA tension and incremental inflation. A positive value indicates lateral displacement. A negative value indicates medial displacement. *Significantly different from mm intact. TLF, thoracolumbar fascia. muscles intact without CTrA tension. With inflation (pressure in PMC), the amount of lateral movement due to CTrA tension is significantly reduced (P = 0.047). In fact, the lateral movement due to CTrA tension is negated by the inflation-induced medial movement such that the net displacement of the TLF is in the medial and posterior directions. This supports the concept that co-contraction of the transverse abdominus/internal oblique and the paraspinal muscles transfer forces mainly through the PLF, hence posteriorly girdling the lumbar spine. Post hoc analyses indicated that the amount of CTrA-induced medial movement of the PLF is significantly reduced during inflations Inf2 (P = 0.03) and Inf3 (P = ), relative to mm intact. There was no significant difference between Inf1 and mm intact (P = 0.14). There were no significant differences in SLD lat between the three inflation increments. To determine whether a relationship exists between the inflation-induced posterior displacement vs. medial displacement of the walls of the PMC, a medial displacementto-posterior displacement ratio was calculated for each inflation increment and CTrA tension. The ratios were 0.18, and for Inf1, Inf2 and Inf3, respectively. This indicates a consistent medial displacement of the lateral wall of the PMC of approximately 0.18 cm for each 1 cm of posterior wall displacement. Common transversus tendon tension in the absence of a pressurized PMC (mm intact) results in an anterolateral displacement of the PMC (black bars in Figs 5 and 6; 2014 Anatomical Society

135 454 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. gray line in Fig. 7b). However, this CTrA tension-dependent anterolateral displacement is counteracted by incremental pressure of the PMC (e.g. simulating paraspinal muscle contraction), resulting in an overall posteromedial movement of the PMC (gray bars in Figs 5 and 6; dashed line in Fig. 7b). In this case, a theoretical point of equal tension between the paraspinal muscles and transverse abdominus and internal obliques pull through the CTrA is attained. Differentiating MLF/PLF force transfer: unidirectional tension measurement along anterior and posterior CTrA laminae in the presence of tension of the CTrA and inflation of the PMC The amount of force transmitted through MLF and PLF (Method 5) was measured while applying 8.5 N force through the CTrA (Method 1b) at each inflation increment (Method 2). At each level of inflation, a significantly greater a b Fig. 7 (a) Effect of inflation and tension on the PMC. Sample-specific example with comparisons of: (i) muscles intact (solid line) compared with inflation 1 (large dashed line), inflation 2 (small dashed line) and inflation 3 (dotted line) without tension in the common tendon of the transversus abdominis and internal oblique muscles (CTrA); (ii) muscles intact (solid line) compared with inflation 1 (large dashed line), inflation 2 (small dashed line) and inflation 3 (dotted line) with CTrA tension; (iii) muscles intact (solid line) to inflation 1 (large dashed line) without CTrA tension, arrows indicate the direction of movement that results from inflation; (iv) inflation 1 without and with CTrA tension, arrows indicate the direction of movement that results from CTrA tension; (v) inflation 2 without and with CTrA tension, arrows indicate the direction of movement that results from CTrA tension; (vi) inflation 3 without and with CTrA tension, arrows indicate the direction of movement that results from CTrA tension. (b) Effect of inflation and tension on the PMC. Composite drawing of the generalized effects of inflation on the PMC based on the mean (across all samples) displacement of beads from the muscles intact (gray line) to inflation 3 (dashed line). The PMC is pushed posteriorly and medially, as indicated by the arrows Anatomical Society

136 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 455 proportion of the force was transmitted through the PLF, relative to the MLF (P = ; Fig. 8). The force transmission through the MLF did not differ significantly between each increment of inflation (P = 0.74). Conversely, force transmission through the PLF is elevated at Inf1 and then gradually declines with further incremental inflation (P = 0.013). Post hoc analyses indicate that force transmission through the PLF in Inf3 is significantly lower than Inf1 (P = ), which will be further elucidated in Fig. 9. There was also a trend toward significance between Inf1 and Inf2 (P = 0.087). There was no significance between Inf2 and Inf3 (P = 0.2). It should be noted that it was not possible to obtain this measurement in the muscles intact condition, because there was no space to place the alligator clips. The data presented in Fig. 8 (Method 5) correspond with the data in Fig. 5 (Method 4). Because the load cells are Fig. 8 Unidirectional tension measurements along anterior and posterior CTrA laminae with CTrA tension and inflation. The amount of CTrA-mediated force directed through the fibrous posterior layer (PLF) appears to decrease with incremental PMC pressure, yet the amount of CTrA-mediated force directed through the fibrous middle layer (MLF) appears to be minimal and basically unaffected by incremental PMC pressure. *Indicated statistical significance between Inf1 and Inf3. Fig. 9 Effects of inflation and tension in the common tendon of the transversus abdominis and internal oblique (CTrA) on the PLF. With incremental inflation, posteriorly directed forces (black bars) decline and anteriorly directed forces (gray bars) increase. The sum of dual load cells measurements of anteriorly and posteriorly directed forces through the PLF does not differ across the three inflation increments. unidirectional, they only measured tension in the direction of the CTrA. Inflation moved the posterior part of the PLF in the opposite direction from the load cell measuring CTrA tension (Fig. 5), thereby having a subtractive effect on force measured through the PLF (Fig. 8). Bi-directional tension measurements analyzing the effect of inflation on PLF force transfer with CTrA tension To better understand the subtractive effect that PMC inflation had on the unidirectional tension measured in the CTrA (described above, observed in Fig. 8), two unidirectional load cells were placed on the PLF: (i) one measured tension in the anterolateral direction (i.e. in the direction of pull for the transverse abdominus and internal obliques through the CTrA); (ii) the other measured tension in the posteromedial direction (i.e. along the PLF toward the spinous process; Method 6). An 8.5 N load was then suspended from the CTrA, as described previously (Method 1b). As inflation increased, the amount of measured posteromedially oriented tension in the PLF decreased(black bars, Fig. 9). Conversely, the measured amount of anterolaterally oriented tension in the PLF increased as a result of the tension of the CTrA, but only with increased inflation (gray bars, Fig. 9). The sum of anterolateral and posteromedial forces did not differ between levels of inflation. These data confirm that with increased inflation of the PMC, the tension progressively increases on the CTrA. Summary of results Incremental inflation of the PMC, in the absence of tension on the CTrA, produced a minimal but significant increase in length of the PLF accompanied by its posterior displacement (Fig. 4). However, incremental inflation within the PMC does not alter MLF length (Fig. 4) or generate any displacementofthelateralwallofthepmc. In the absence of inflation within the PMC (mm intact), increasing CTrA tension results in anterior and lateral movement of the borders of the compartment (black bars in Figs 5 and 6; gray line in Fig. 7b). However, when CTrA tension is coupled with inflation, the net displacement of the borders of the compartment is posterior and slightly medial (gray bars in Figs 5 and 6; dashed line in Fig. 7b). Similarly, as pressure within the PMC increases, tension through the CTrA is significantly counteracted (Figs 8 and 9). Tension of the CTrA is predominately passed through the PLF, with very little impact on the MLF, regardless of the level of pressure in the PMC (Fig. 8). Collectively, this implies that an adequate paraspinal muscle contraction can counter the tension created by transverse abdominus and internal oblique contraction and vice versa. Further, it implies the existence of a point of equal tension between the paraspinal muscles and the transverse abdominus and internal 2014 Anatomical Society

137 456 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. oblique muscles acting through the CTrA producing increased force closure and self-bracing of the spine (Vleeming et al. 1990a,b). Discussion To the best of the authors knowledge, this is the first study in which simulated incremental contraction of the paraspinal muscles within the TLF compartment is combined with applied tension to the CTrA of the transverse abdominus and internal oblique muscles. This study was undertaken to examine the relative force transfer through the MLF and PLF of the PMC. The analysis of perimeter changes in the PMC in the setting of increasing compartment pressures, combined with measurements of load transfer from the CTrA through the PLF and MLF lamina, produced the following results: (i) in the absence of PMC inflation (mimicking the lack of muscle contraction), CTrA tension results in anterior and lateral movement of the PLF (black bar in Figs 5 and 6); (ii) the combination of PMC inflation (mimicking muscle contraction via increased pressure) with increased CTrA tension substantially displaces the PLF in a posterior and slightly medial direction. With each incremental inflation, the displacement triggered by force transfer from the CTrA to the PLF leveled off, indicating that a point of equal tension between paraspinal muscles and CTrA is reached. At this point of equilibrium, pressure within the PMC counters the pull created by the CTrA and further displacement ceases. This indicates that with increasing PMC inflation, the angle of load transfer between the anterolateral pull of CTrA to the PLF is further optimized, creating an increasingly linear pull to the PLF. This is in contrast to the MLF, where the angle of pull from the CTrA to the MLF becomes less optimal with increased PMC inflation. In this study, incrementally increased pressure in the PMC was used to mimic the volume growing effect of muscle contraction. The pressure-generated displacement of the PLF was coupled on the transverse and sagittal planes. For each 1 cm of expansion in the posterior direction (sagittal axis) there was a 0.18 cm movement inward (medial direction) on the transverse axis. This indicates that the lateral border of the TLF is not expanding and is even slightly displaced medially, compared with a much larger posterior displacement of the PLF. This tension effect could be explained by the fact that submaximal inflation (Inf3) of the TLF compartment is counteracting the strain of pulling the CTrA. In this case, a theoretical point of equalized tension between paraspinal muscles and the transverse abdominus and internal oblique muscles has been attained. These results show that increasing tension on the CTrA (mimicking deep abdominal muscle contraction), combined with PMC inflation, transfers significantly more load to the PLF in comparison to the MLF thereby girdling the spine posteriorly. On average, the maximal posterior expansion of the PLF obtained when comparing a neutral position (muscles intact in the PMC and no tension on the CTrA) with that of maximal stress on the PLF (level 3 inflation of the PMC with 8.5 N tension on the CTrA) was an increase of 1.56 cm. From a biomechanical prospective, the shift of the load to the PLF places the strain on the thick lumbar spinous processes and supraspinous ligaments rather than passing this stress through the MLF to the much thinner lumbar transverse process. In addition, this strain is partially transmitted to the contralateral site of the PLF (Vleeming et al. 1995; Barker & Briggs, 1999). The results of this study could improve our understanding of LBP patients, showing that in the setting of paraspinal and deep abdominal muscle weakness or dysfunction, tension to the PLF potentially can unbalance the girdling relationship between the PLF and the abdominal muscles. Comparisons to previous studies Tesh et al. (1987) created an innovative in vitro study to analyze whether an extension moment could be generated in the lumbar spine by simulating lateral pull to the PLF and indirectly to the supraspinous ligaments. To simulate compression within the PMC, two foam dowels were inserted bilaterally into the emptied PMC of two sequential lumbar segments. An incremental loading sequence was applied to the CTrA to create tension up to a maximum of 98 N. The results of the Tesh et al. (1987) study confirmed that by applying tension to the PLF, the lumbar spine is stabilized; however, to a lesser degree than the higher values reported in a related study by Gracovetsky and co-workers (Gracovetsky et al. 1981; Gracovetsky, 1985). The insertion of foam dowels, as applied in Tesh s study, created an unknown static tension, keeping the dowels in place but did not mimic various levels of paraspinal contractions. In the present study, custom-made butyl inflatable tubes were inserted, in order to simulate three stages of contraction/inflation (inf), with and without CTrA tension. Barker et al. (2006) investigated the effects of placing tension on the CTrA PLF/MLF couple on segmental lumbar stiffness during flexion and extension movements. In their study,thetensionplacedonthe CTrA at vertebral level L3 was transmitted almost twice as effectively to the MLF as compared with the PLF. Barker et al. (2006) cautiously comments while segmental studies indicate that the PLF resists flexion and it has a greater moment arm than the MLF, its midline attachments are relatively unthickened and mobile, so its efficiency in influencing segmental motion may be dependent on the activation of the paraspinal muscles. The present study confirms that by mimicking activation of the paraspinal muscles, combined with CTrA pull, the PLF becomes preferentially tensed while the MLF is barely influenced (Fig. 8) Anatomical Society

138 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 457 To effectively analyze the transfer of load between MLF and PLF subsequent to tension placedonthectra,itiscritical to replicate the normal compartmental pressure in the PMC as closely as possible, thus some form of compartmental inflation has to be employed. In addition, it is necessary to deliver the load from the CTrA in a manner that best mimics its in vivo angle. For this reason, the current study used the angle of the CTrA normally found in the individual specimen under study; this angle created by the anterolateral trajectory of the CTrA resembled a hoop surrounding the abdomen. Tesh et al. (1987) and Barker et al. (2006) used a synthetic angle of 90 from the midline. A transverse or horizontal pull to the CTrA, as used in these two studies, creates a non-physiological linear strain, predominantly in line with the fiber direction of the MLF. Barker et al. (2006) applied a 20 N strain to the CTrA tendon to simulate moderate transverse abdominus tension. The present study applies on average 8.5 N load to the CTrA. Although this is a lower force than used in the study by Barker et al. (2006), this force was chosen to match the fact that the current study used only one vertebral body slab, whereas Barker et al. (2006) used a thicker slab containing two vertebral bodies. More than 8.5 N force could rupture the relatively small CTrA associated with a singular vertebral level. Analyzing the extension moment of the PLF It has been postulated by Gracovetsky and co-workers (Gracovetsky et al. 1977, 1981; Gracovetsky, 1985) that, if lateral force through the CTrA is applied to the TLF when its lateral border is restrained, such as by inflation of the PMC, a longitudinal tension will develop in the PLF. Assuming that the PLF behaves like an ideal net structure, the potential of the PLF to transform horizontal CTrA tension into a longitudinal tension at the level of the spinous process will enable an extension moment in the lumbar spine (Gracovetsky et al. 1977, 1981; Gracovetsky, 1985; Macintosh & Bogduk, 1987; Tesh et al. 1987; Hukins et al. 1990; Dolan et al. 1994; Adams & Dolan, 2007). In this manner, a mechanical conversion (gain) takes place between forces acting on the PLF. This was coined the hydraulic amplifier mechanism (Gracovetsky et al. 1977). This gain is the ratio between longitudinal tension of the PLF and horizontal pull through the CTrA, and was found to be dependent on the inclination angle of the collagen fibers of the PLF between its lateral border and the midline (Tesh et al. 1987). Although several values have been calculated for the optimal angle (Bogduk & MacIntosh, 1984; Vleeming et al. 1995; Barker & Briggs, 1999), a comparison between various anatomical studies of the superficial lamina of the PLF (Willard et al. 2012) confirmed an average fiber direction of up to 40 from horizontal and a cross-hatched fiber appearance below T12. In this sense, the PLF resembles the annulus fibrosis in which tension in the crossed fibers acts to contain the pressure of the intervertebral disc (Macintosh & Bogduk, 1987; Tesh et al. 1987). Restraining lateral expansion of the TLF is both an effect of the cross-hatched orientation of the collagen fibers in the PLF, alternating between the superficial and deep lamina (Bogduk & MacIntosh, 1984; Tesh et al. 1987; Hukins et al. 1990; Vleeming et al. 1995; Barker & Briggs, 1999), and the lateral tension from the CTrA (Tesh et al. 1987; Hodges et al. 2003; Barker et al. 2004, 2006). In the present study, CTrA tension combined with PMC inflation restrains the lateral border and even creates a small medial displacement. This restraint of the lateral border matches well with a reported four times stronger lateral strength of the PLF, compared with longitudinal strength (Tesh et al. 1987). In vivo, a circumferential hoop tension occurs during lifting, by raising intra-abdominal pressure (IAP) and/or contracting the transverse abdominus and internal oblique muscles (Hodges & Richardson, 1996, 1999; Urquhart & Hodges, 2005), and thereby creating a horizontal force to the PLF. Using a biomechanical model, Hukins et al. (1990) showed that limiting radial expansion of the PMC due to contraction of the deep abdominal muscles could increase the tension within the TLF compartment by 30%, leading to a proportional increase in the extensor moment exerted by paraspinal muscles, as suggested by Gracovetsky and co-workers (Gracovetsky et al. 1977; Gracovetsky, 1985). The extension moment of the PLF depends on the distance to the instantaneous center of rotation of the vertebral body, which is calculated on average as 6 8 cm, possessing an excellent moment arm to resist flexion (Gracovetsky et al. 1981, 1985; McGill & Norman, 1985; Tesh et al. 1987; Hukins et al. 1990; Dolan et al. 1994; Barker et al. 2006; Adams & Dolan, 2007; Gatton et al. 2010). The present study shows an average posterior displacement of the PLF of 1.56 cm, when the PMC is submaximally inflated (Inf3); this represents a substantial increase of the moment arm of the PLF. In addition, this extensor moment becomes significantly larger in flexed postures (Gracovetsky et al. 1981; Tesh et al. 1987; Dolan et al. 1994; Adams & Dolan, 2007) and will also increase tension over the lumbosacral area, affecting stability of the pelvis (Vleeming et al. 2012). It has been calculated that the PLF could resist longitudinal forces exceeding 1 kn (Adams & Dolan, 2007). However, it is reasonable to expect expansion of the PMC to be substantially less in healthy trained individuals, with stronger paraspinal muscles, and hence stronger aponeuroses of the paraspinal muscles and PLF, resulting in an earlier and larger increase of pressure within the PMC. Dolan & Adams (1993) reported that in lifting weights the total extensor moment is unrelated to the electromyogram (EMG) activity of the paraspinal muscles. Less than 25% of this passive extensor moment comes from intervertebral discs and ligaments. The rest of this passive force appears to arise from non-contractile tissues, such as the PLF, the supraspinous ligaments and the surrounding fascial 2014 Anatomical Society

139 458 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. tissues within the erectors as well as raised IAP. Bogduk & MacIntosh (1984) report a significant flexion resistant role of the PLF; nonetheless, they argue that the hydraulic amplifier mechanism does not occur, because of the thinness and inconsistency of the deep lamina of the PLF. However, a recent study revealed that the paraspinal muscles are enclosed by an intact fascial sheath, the PRS (Schuenke et al. 2012). The PRS is a circular extension of the deep layer of the PLF, extending between the spinous process and the transverse process. The PRS contributes to the LIFT, formed by the anterior and posterior laminae of the CtrA (Fig. 1; Schuenke et al. 2012). This triangle becomes the intermediary point of force transfer, between deep abdominal muscles and the paraspinal muscles within the TLF. Hence, the LIFT equalizes tension between adjacent pressure vessels like the hypaxial abdominal and epaxial TLF fascial compartments. Therefore, encapsulation of the paraspinal muscles by this retinaculum makes it possible to create pressure within the PMC. The results of the present study suggest that a mechanical conversion (gain) is achieved in which horizontal tension is transformed into longitudinal tension, as proposed by Gracovetsky (1985) and Hukins et al. (1990). Impairment of muscles and fascia contributing to the TLF Various studies of abdominal and lumbar paraspinal muscles in LBP patients have reported significant pathological modifications in structure and function (Parkkola et al. 1993; Hides et al. 1994, 1995; Mooney et al. 1997; Danneels et al. 2000; Kader et al. 2000; Mengiardi et al. 2006; Dickx et al. 2008, 2010; Chan et al. 2012). Deconditioning, fatty involution, size and shape changes and alterations in motor control for the abdominal and paraspinal muscles, particularly the deep lumbar multifidus, have been defined in patients with LBP. All of these conditions have the potential to alter the effective load transfer characteristics of the TLF. Even in well-conditioned athletes such as ballet dancers, examination of those with low back and hip pain revealed alterations in cross-sectional area (CSA) of the multifidus muscle correlating with the side of the pain (Gildea et al. 2013). Besides changes in size of the paraspinal muscles, the stiffness of these muscles is altered in patients with LBP. Ultrasound elastography studies confirmed that the stiffness of the multifidus muscle in certain postures increases in male LBP patients, compared with asymptomatic male controls, indicating adaptation of muscle contractile function (Chan et al. 2012). Another study, using the same methodology, quantified shear plane motion within the deep and superficial lamina of the PLF, demonstrating that the shear strain between these layers was on average reduced by 20% in female and male patients with LBP. A sexual dimorphism was also found in this study, indicating that males had significantly lower shear strain values compared with females (Langevin et al. 2011). The authors suggest that this reduction of strain could be attributed to intrinsic connective tissue pathology and chronic inflammation of the PLF (Langevin et al. 2011). Moreover, mechanical testing of isolated PLF samples of humans and animals shows an increase in stiffness with deformation, when stretched for long periods (Tesh et al. 1987; Yahia et al. 1992; Vleeming et al. 1995; Schleip et al. 2012). In addition, patients with LBP have demonstrable changes in the histological characteristics of the PLF (Bednar et al. 1995). The present study demonstrates that robust paraspinal muscle contractions are required within the PMC to enable the pressure increase necessary to alter the geometric shape of this fascial structure. Compartment inflation leads especially to a posterior displacement of the erector spinae aponeurosis and PLF, thus enhancing the moment arm for spinal extension. However, when this expansion is combined with deep abdominal muscle contraction, the CTrA will transfer tension to the PLF and vice versa, thereby bracing and girdling the paraspinal muscles. Simultaneously, the lateral border of the PMC is restrained from lateral displacement and, in fact, is displaced slightly medially serving to maintain compartmental integrity around the paraspinal muscles. It could be assumed that changes in paraspinal muscle structural (Danneels et al. 2000) and functional (Dickx et al. 2010) properties in patients with chronic LBP could reduce the posterior displacement of the PLF and increase the compliance of the lateral border, thereby diminishing an effect of increased abdominal muscle tension. Maintaining ICP is critical to normal function of the muscles (Kjellmer, 1964; Garfin et al. 1981; Carr et al. 1985; Styf, 1987; Styf & Lysell, 1987). There appears to be an optimal pressure either above or below which there is loss of efficacy in function. Compartmental pressures vary depending on activity, increasing significantly with exercise (Kjellmer, 1964). In addition, compartmental pressure is not constant throughout the full range of motion, but may vary considerably with specific postures and activities (Carr et al. 1985). Specifically, from a neutral pressure on vertical, relaxed standing, PMC pressure in normal volunteers raises slightly on extension, drops slightly on flexion but returns to normalonfullflexion(90 ) of the spine (Carr et al. 1985). Opening a fascial compartment, thereby reducing its ICP, decreases muscle contraction force (Garfin et al. 1981). It has also been reported that increased pressure can be recorded in lumbar fascial compartments of a cohort of patients with LBP (Styf & Lysell, 1987). The present study shows that depending on the level of activation of the paraspinal and deep abdominal muscles, intra-abdominal and intra-compartmental PMC pressures affect each other through the tightening of the CTrA. This phenomenon was also indirectly studied by Carr et al. (1985), showing 2014 Anatomical Society

140 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 459 a strong increase of ICP within the PMC when the IAP was increased by applying the Valsalva maneuver. The studies of Carr et al. (1985) and Schuenke et al. (2012) have demonstrated that lumbar paraspinal muscles are contained within the TLF compartment that will support increased pressure. It is noteworthy that patients with LBP consistently show higher ICP values within this compartment, associated with increased flexion and loading as compared with healthy controls (Konno et al. 1994). These findings in patients could be an indication that a diminished volume filling function of the multifidus requires more work from its residual fibers and from associated muscles such as the longissimus and the iliocostalis. Less inflation of the PMC reduces posterior PLF displacement and alters the tension transfer between the PLF and CTrA. Conversely, continuous activation of paraspinal muscles in patients with LBP increases pressure within the PMC. It could be hypothesized that either diminished or too much tension of the paraspinal muscles, converted to the CTrA, could lead to an inhibitory response of the deep abdominal muscles, resulting in dysfunctional deep abdominal muscles as frequently reported in patients with chronic LBP (Hodges, 2008). The flexion relaxation response (FRR) During increased spinal flexion, diminution of EMG activity of the paraspinal muscles frequently occurs in healthy persons (Fick, 1911; Floyd & Silver, 1951). This phenomenon is termed the FRR (Andersson et al. 1977). The ability to diminish lumbar spinal muscle activity during increased flexion could be a consequence of tone in healthy robust spinal muscles sufficiently inflating the TLF compartment. Kaigle et al. (1998) studied patients with chronic LBP and healthy volunteers during dynamic spinal flexion and extension exercises. The results showed that both intervertebral motion and trunk mobility were significantly restricted in patients compared with healthy individuals. The FRR in full flexion of the controls showed a 78% decrease in myoelectric activity, compared with patients with LBP showing an average reduction of 13% and most patients showing no reduction. The authors conclude that the FRR occurs exclusively in subjects reaching a complete segmental intervertebral flexion, considerable before full trunk flexion was achieved. The authors comment that ongoing spinal muscle activation in the patient group during flexion precludes full intervertebral motion, tensing the ligaments, thereby preventing the FRR (Kaigle et al. 1998). The present study shows that in the absence of sufficient PMC inflation, CTrA tension results in anterior and lateral movement of especially the PLF. This suggests that decreased CSA of the paraspinal muscles, with less volume load to fill the PMC, combined with or without CTrA tension, will displace the TLF anteriorly and laterally, thereby reducing the extension moment of the PLF. The diminished efficacy of the extensor moment could be another reason why persistent muscle activation, compensating for less tension within the TLF compartment, precludes full intervertebral motion during deep flexion. Trunk and extremity muscles contributing to PLF tension In the present study, the biomechanical properties of the PMC were studied exclusively in the axial plane. However, the superficial and deep lamina of the PLF acts as an intermediary in transferring loads in three directions: between the upper and lower limbs; between left and right sides of the body; and between the abdominal wall and the lumbopelvic spine (Vleeming et al. 1995, 2012; Barker & Briggs, 1999). The superficial lamina of the PLF is formed mainly by the aponeurosis of the LD, additional connections to the deeper lamina of the PLF arise from the serratus posterior inferior muscles (Bogduk & MacIntosh, 1984; Vleeming et al. 1995; Barker & Briggs, 1999; Barker et al. 2004). This merging of abdominal and arm/trunk muscles into both laminae of the PLF creates a combined axial and frontal plane myofascial sling (Willard et al. 2012). Combinations of different trunk movements, like lumbar flexion/extension, lateral flexion and rotation, will generate specific directional tension to the PMC. For example, lumbopelvic flexion passively increases tension of the PLF (Dolan et al. 1994; Barker et al. 2004), which can be enhanced by combinations of trunk rotation and lateral flexion. Transversus tendon muscle action is uniquely associated with increased postural demand, and contributes to general spine stabilization when the trunk is exposed to moderate flexion and extension moments (Crommert et al. 2011). Uni- or bilateral contractions of the deep abdominal muscles, with or without activation of the paraspinal muscles, cause different tensioning patterns to the PMC. As an example, during abdominal crunch exercises, the paraspinal and deep abdominal muscles will act antagonistically: flexing the pelvis and spine in sitting or standing, together with posterior trunk displacement, requires an eccentric contraction of the abdominal muscles combined with a relaxation of the paraspinal muscles. Progressively flexing the spine and pelvis with robust contraction of deep abdominal muscles will generate increased tension transfer from the CTrA to the PLF, girdling the spine in flexion. Flexing the spine is a typical example of opening up the kinematic chain and therefore diminishing form closure of the spine, counteracted by increased force closure (Vleeming et al. 1990a,b) by tensing the CTrA and PLF. In this scenario, deep abdominal muscles activity generates tension to the most superficial, posterior part of the PMC compartment. Besides trunk muscles, several studies have clearly demonstrated that activity of the extremity muscles, like the LD, tense the PLF, both ipsi- and contralaterally, and reach as 2014 Anatomical Society

141 460 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. far as the contralateral GM muscle, even affecting resting tone of the contralateral hip (Bogduk & MacIntosh, 1984; Tesh et al. 1987; Vleeming et al. 1995; Mooney et al. 1997; Barker & Briggs, 1999; Barker et al. 2004; Willard et al. 2012; Carvalhais et al. 2013). It is obvious that a multi-vector analysis will be needed to appreciate how paraspinal muscles activity in conjunction with deep abdominal muscles, and extremity muscles such as the LD and GM, effect the tension of the PLF and hence PMC pressure. A flexible myofascial axial ring between the thorax and pelvis, replicating ribs Vertebral stability in the thorax is greatly increased by interconnecting ribs to the spine and sternum; likewise in the pelvis, in which laterally flaring innominates could be regarded as fused ribs stemming from the sacrum and running anteriorly to the symphysis pubis (Tidwell & Carpenter, 2005; Gracovetsky, 2007). In humans, between the lower thorax and upper pelvis a greater space is present compared with the great apes. The latter have a restricted flexible zone between the rib cage and pelvis, because of their cranially extending pelvis and far caudally reaching lower ribs, restricting lumbar motion (Lovejoy, 2007). Humans, especially women, have a reduced height of the pelvis (Vleeming et al. 2012), creating more space between the thorax and pelvis, and this feature permits increased spinal mobility like flexion/extension and lateral flexion/rotation in the spine (Lovejoy, 2007). Stability and mobility are opposing states in joints. To stabilize the lumbar spine, especially between L1 and L4 levels without ribs, the transverse abdominus and internal oblique muscles and their fascial aponeurosis resemble a flexible myofascial ring between the thorax and pelvis. This ring runs posteriorly from the transverse abdominus and internal oblique muscles, to the CTrA and via the MLF connecting to the transverse process, mimicking a dorsal rib construction. Anteriorly, the abdominal muscles fuse with the rectus abdominus fascia. The rectus muscle itself represents a contractible version of the inert bony sternum of the thorax or symphysis of the pelvis. Posteriorly, the PLF girdles the lumbar erector spinae and multifidus muscles. This axial abdominal lumbar myofascial ring, between the thorax and pelvis, generates muscles contraction to compensate for the lack of rib stability. In essence, tension in this myofascial can be adjusted by altering PMC pressure. The present study has demonstrated that co-contraction of the paraspinal muscles and the deep abdominal muscles is capable of creating this girdling effect. Methodological considerations Sectioning the axial slabs results in diminished longitudinal tension, inherent in an intact TLF. This applies particularly to the PLF and the underlying common aponeurosis of the paraspinal muscles. This could be considered a limitation of the study. However, the present study is designed to inflate and pressurize the PMC and simultaneously tension the CTrA, which will increase both longitudinal and lateral tension. Another consideration is that studying elderly specimens generally is accompanied by a thinner fascia and aponeurosis. Therefore, expansion of the PMC as measured in the present study could be substantially larger in comparison to healthy individuals with robust paraspinal and deep abdominal muscles and a stronger PMC. The muscular strength difference in healthy persons could lead to an earlier PMC pressure, with less posterior displacement of the PLF compared with the specimen. Concluding remarks The present study demonstrates that robust paraspinal muscle contractions are required within the PMC to enable pressure increases sufficient to alter the geometric shape of this fascial structure. PMC inflation (mimicking paraspinal muscle contraction and generating pressure increase) results in a posterior displacement of the erector spinae aponeurosis and the closely associated PLF. However, when this expansion is combined with simulated deep abdominal muscle contraction, the CTrA will transfer tension almost exclusively to the PLF, thereby bracing and girdling the paraspinal muscles. Simultaneously, the lateral border of the PMC is moving slightly medially serving to maintain the PMC integrity. The present study shows an average posterior displacement of the PLF of 1.56 cm, when the PMC is submaximally inflated. This represents a substantial increase of the extensor moment arm of the PLF. Conversely, dysfunctional paraspinal muscles will reduce the pressure surge and hence the posterior displacement of the PLF. As a consequence, the compliance of the lateral border is increased, diminishing the effects of CTrA tension. When both the paraspinal and deep abdominal muscles are weak or dysfunctional, pressure increase and expansion of the PMC will be minimal. This study shows a critical co-dependent mechanism between deep abdominal and lumbar spinal muscles linked to each other, especially through the PLF. Acknowledgements The authors would like to express appreciation to Mr Cole Southworth, Ms Dawn Whalen and Mr Innocent Ndzana for assisting in specimen preparation and data collection. The authors would also like to express appreciation to Mr Oran Suta for his artistic expertise and assistance on the figures. Lastly, the authors wish to express appreciation to Mr Hank Wheat for his knowledge and assistance in the design and construction of the research apparatuses Anatomical Society

142 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. 461 References Adams MA, Dolan P (2007)How to use the spine, pelvis, and legs effectively in lifting. In: Movement, Stability & Lumbopelvic Pain: Integration of Research and Therapy. (eds Vleeming A, Mooney V, Stoeckart R), pp Edinburgh: Churchill Livingstone. Andersson GB, Ortengren R, Herberts P (1977) Quantitative electromyographic studies of back muscle activity relatated to posture and loading. Orthop Clin North Am 8, Barker PJ, Briggs CA (1999) Attachments of the posterior layer of lumbar fascia. Spine 24, Barker PJ, Briggs CA (2007) Anatomy and biomechanics of the lumbar fasciae: implications for lumbopelvic control and clinical practice. In: Movement, Stability & Lumbopelvic Pain: Integration of Research and Therapy. (eds Vleeming A, Mooney V, Stoeckart R), pp Edinburgh: Churchill Livingstone. Barker PJ, Briggs CA, Bogeski G (2004) Tensile transmission across the lumbar fasciae in unembalmed cadavers: effects of tension to various muscular attachments. Spine 29, Barker PJ, Guggenheimer KT, Grkovic I, et al. (2006) Effects of tensioning the lumbar fasciae on segmental stiffness during flexion and extension: Young Investigator Award winner. Spine 31, Bednar DA, Orr FW, Simon GT (1995) Observations on the pathomorphology of the thoracolumbar fascia in chronic mechanical back pain. A microscopic study. Spine 20, Bogduk N, MacIntosh JE (1984) The applied anatomy of the thoracolumbar fascia. Spine 9, Carr D, Gilbertson L, Frymoyer J, et al. (1985) Lumbar paraspinal compartment syndrome. A case report with physiologic and anatomic studies. Spine 10, Carvalhais VO, Ocarino Jde M, Araujo VL, et al. (2013) Myofascial force transmission between the latissimus dorsi and gluteus maximus muscles: an in vivo experiment. J Biomech 46, Chan ST, Fung PK, Ng NY, et al. (2012) Dynamic changes of elasticity, cross-sectional area, and fat infiltration of multifidus at different postures in men with chronic low back pain. Spine J 12, Clemente CD (1985) Gray s Anatomy of the Human Body. Philadelphia: Lea & Febiger. Crommert ME, Ekblom MM, Thorstensson A (2011) Activation of transversus abdominis varies with postural demand in standing. Gait Posture 33, Danneels LA, Vanderstraeten GG, Cambier DC, et al. (2000) CT imaging of trunk muscles in chronic low back pain patients and healthy control subjects. Eur Spine J 9, Dickx N, Cagnie B, Achten E, et al. (2008) Changes in lumbar muscle activity because of induced muscle pain evaluated by muscle functional magnetic resonance imaging. Spine 33, E983 E989. Dickx N, Cagnie B, Parlevliet T, et al. (2010) The effect of unilateral muscle pain on recruitment of the lumbar multifidus during automatic contraction. An experimental pain study. Man Ther 15, Dolan P, Adams MA (1993) The relationship between EMG activity and extensor moment generation in the erector spinae muscles during bending and lifting activities. J Biomech 26, Dolan P, Mannion AF, Adams MA (1994) Passive tissues help the back muscles to generate extensor movements during lifting. J Biomech 27, Farfan HF (1973) Mechanical Disorders of the Low Back. Philadelphia: Lea and Febiger. Fick R (1911) Handbuch der Anatomie und Mechanik der Gelenke. Jena: Gustav Fischer. Floyd WF, Silver PH (1951) Function of erectores spinae in flexion of the trunk. Lancet 1, Garfin SR, Tipton CM, Mubarak SJ, et al. (1981) Role of fascia in maintenance of muscle tension and pressure. J Appl Physiol Respir Environ Exerc Physiol 51, Gatton ML, Pearcy MJ, Pettet GJ, et al. (2010) A threedimensional mathematical model of the thoracolumbar fascia and an estimate of its biomechanical effect. J Biomech 43, Gildea JE, Hides JA, Hodges PW (2013) Size and symmetry of trunk muscles in ballet dancers with and without low back pain. J Orthop Sports Phys Ther 43, Gracovetsky S (1985) An hypothesis for the role of the spine in human locomotion: a challenge to current thinking. J Biomed Eng 7, Gracovetsky S (2007) Is the sacroiliac joint an evolved costovertebral joint? In: Movement, Stability and Lumbopelvic Pain: Integration and Research. (eds Vleeming A, Mooney V, Stoeckart R), pp Edinburgh: Churchill Livingstone. Gracovetsky S, Farfan HF, Lamy C (1977) A mathematical model of the lumbar spine using an optimized system to control muscles and ligaments. Orthop Clin North Am 8, Gracovetsky S, Farfan HF, Lamy C (1981) The mechanism of the lumbar spine. Spine 6, Gracovetsky S, Farfan H, Helleur C (1985) The abdominal mechanism. Spine 10, Grant JCB (1972) An Atlas of Anatomy. Baltimore: Williams & Wilkins. Hides JA, Stokes MJ, Saide M, et al. (1994) Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 19, Hides JA, Richardson CA, Jull GA (1995) Magnetic resonance imaging and ultrasonography of the lumbar multifidus muscle. Comparison of two different modalities. Spine 20, Hodges PW (2008) Transversus abdominis: a different view of the elephant. Br J Sports Med 42, Hodges PW, Richardson CA (1996) Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 21, Hodges PW, Richardson CA (1999) Transversus abdominis and the superficial abdominal muscles are controlled independently in a postural task. Neurosci Lett 265, Hodges P, Kaigle Holm A, Holm S, et al. (2003) Intervertebral stiffness of the spine is increased by evoked contraction of transversus abdominis and the diaphragm: in vivo porcine studies. Spine 28, Hollinshead WH (1969) Anatomy for Surgeons: The Back and Limbs. New York: Hoeber-Harper. Hukins DW, Aspden RM, Hickey DS (1990) Thoracolumbar fascia can increase the efficiency of the erector spinae muscles. Clin Biomech (Bristol, Avon) 5, Kader DF, Wardlaw D, Smith FW (2000) Correlation between the MRI changes in the lumbar multifidus muscles and leg pain. Clin Radiol 55, Anatomical Society

143 462 Paraspinal force transfer to thoracolumbar fascia, A. Vleeming et al. Kaigle AM, Wessberg P, Hansson TH (1998) Muscular and kinematic behavior of the lumbar spine during flexion-extension. J Spinal Disord 11, Kjellmer I (1964) An indirect method for estimating tissue pressure with special reference to tissue pressure in muscle during exercise. Acta Physiol Scand 62, Konno S, Kikuchi S, Nagaosa Y (1994) The relationship between intramuscular pressure of the paraspinal muscles and low back pain. Spine 19, Langevin HM, Fox JR, Koptiuch C, et al. (2011) Reduced thoracolumbar fascia shear strain in human chronic low back pain. BMC Musculoskelet Disord 12, Lovejoy CO (2007) Evolution of the human lumbopelvic region and its relationship to some clinical deficits of the spine and pelvis. In: Movement, Stability and Lumbopelvic Pain: Integration and Research. (eds Vleeming A, Mooney V, Stoeckart R), pp Edinburgh: Churchill Livingstone. Macintosh JE, Bogduk N (1987) 1987 Volvo award in basic science. The morphology of the lumbar erector spinae. Spine 12, McGill SM, Norman RW (1985) Dynamically and statically determined low back moments during lifting. J Biomech 18, Meijering E, Dzyubachyk O, Smal I (2012) Methods for cell and particle tracking. Methods Enzymol 504, Mengiardi B, Schmid MR, Boos N, et al. (2006) Fat content of lumbar paraspinal muscles in patients with chronic low back pain and in asymptomatic volunteers: quantification with MR spectroscopy. Radiology 240, Mooney V, Gulick J, Perlman M, et al. (1997) Relationships between myoelectric activity, strength, and MRI of lumbar extensor muscles in back pain patients and normal subjects. J Spinal Disord 10, Parkkola R, Ryt okoski U, Kormano M (1993) Magnetic resonance imaging of the discs and trunk muscles in patients with chronic low back pain and healthy control subjects. Spine 18, Schaeffer JP (1953) Morris s Human Anatomy. New York: McGraw-Hill. Schleip R, Duerselen L, Vleeming A, et al. (2012) Strain hardening of fascia: static stretching of dense fibrous connective tissues can induce a temporary stiffness increase accompanied by enhanced matrix hydration. J Bodyw Mov Ther 16, Schuenke MD, Vleeming A, Von Hoof T, et al. (2012) Anatomical constituents of lumbar myofascial load transfer: an outline of the lateral raphe and the lumbar interlaminar triangle. J Anat 221, Spalteholz W (1923) Hand Atlas of Human Anatomy. Philadelphia: J.B. Lippincott. Standring S, ed. (2008) Gray s Anatomy, 14th edn, 1551 pp. Philadelphia: Churchill Livingstone. Stecco A, Masiero S, Macchi V, et al. (2009) The pectoral fascia: anatomical and histological study. J Bodyw Mov Ther 13, Styf J (1987) Pressure in the erector spinae muscle during exercise. Spine 12, Styf J, Lysell E (1987) Chronic compartment syndrome in the erector spinae muscle. Spine 12, Tesh KM, Dunn JS, Evans JH (1987) The abdominal muscles and vertebral stability. Spine 12, Tidwell V, Carpenter K eds) ((2005) Thunder-Lizards: The Sauropodomorph Dinosaurs (Life of the Past). Bloomington, IN: Indiana University Press. Urquhart DM, Hodges PW (2005) Differential activity of regions of transversus abdominis during trunk rotation. Eur Spine J 14, Urquhart DM, Hodges PW (2007) Clinical anatomy of the anterolateral abdominal muscles. In: Movement, Stability & Lumbopelvic Pain: Integration of Research and Therapy. (eds Vleeming A, Mooney V, Stoeckart R), pp Edinburgh: Churchill Livingstone. Vleeming A, Stoeckart R, Volkers AC, et al. (1990a) Relation between form and function in the sacroiliac joint. Part I: clinical anatomical aspects. Spine 15, Vleeming A, Volkers AC, Snijders CJ, et al. (1990b) Relation between form and function in the sacroiliac joint. Part II: biomechanical aspects. Spine 15, Vleeming A, Pool-Goudzwaard AL, Stoeckart R, et al. (1995) The posterior layer of the thoracolumbar fascia. Its function in load transfer from spine to legs. Spine 20, Vleeming A, Willard FH, Schuenke MD, et al. (2012) The sacroiliac joint: anatomy, function, and clinical implications. J Anat 221, Willard FH, Vleeming A, Schuenke MD, et al. (2012) The thoracolumbar fascia: anatomy, function and clinical considerations. J Anat 221, Yahia L, Rhalmi S, Newman N, et al. (1992) Sensory innervation of human thoracolumbar fascia. An immunohistochemical study. Acta Orthop Scand 63, Anatomical Society

144 Research Report The Definition and Application of Pilates Exercise to Treat People With Chronic Low Back Pain: A Delphi Survey of Australian Physical Therapists Cherie Wells, Gregory S. Kolt, Paul Marshall, Andrea Bialocerkowski C. Wells, BAppSci(Physio), MManip Ther, Discipline of Physiotherapy, Faculty of Health, University of Canberra, University Drive, Bruce, Australian Capital Territory, Australia At the time of the study, Ms Wells was a PhD student at School of Science and Health, University of Western Sydney, Penrith, New South Wales, Australia. Address all correspondence to Ms Wells at: cherie.wells@canberra.edu.au. G.S. Kolt, PhD, School of Science and Health, University of Western Sydney. P. Marshall, PhD, School of Science and Health, University of Western Sydney. A. Bialocerkowski, PhD, Griffith Health Institute, Griffith University, Gold Coast, Queensland, Australia. [Wells C, Kolt GS, Marshall P, Bialocerkowski A. The definition and application of Pilates exercise to treat people with chronic low back pain: a Delphi survey of Australian physical therapists. Phys Ther. 2014;94: ] 2014 American Physical Therapy Association Published Ahead of Print: October 31, 2013 Accepted: October 28, 2013 Submitted: January 31, 2013 Background. Pilates exercise is recommended for people with chronic low back pain (CLBP). In the literature, however, Pilates exercise is described and applied differently to treat people with CLBP. These differences in the definition and application of Pilates exercise make it difficult to evaluate its effectiveness. Objective. The aim of this study was to establish consensus regarding the definition and application of Pilates exercise to treat people with CLBP. Methods. A panel of Australian physical therapists who are experienced in treating people with CLBP using Pilates exercise were surveyed using the Delphi technique. Three electronic questionnaires were used to collect the respondents opinions. Answers to open-ended questions were analyzed thematically, combined with systematic literature review findings, and translated into statements about Pilates exercise for people with CLBP. Participants then rated their level of agreement with these statements using a 6-point Likert scale. Consensus was achieved when 70% of the panel members strongly agreed, agreed, or somewhat agreed (or strongly disagreed, disagreed, or somewhat disagreed) with an item. Results. Thirty physical therapists completed all 3 questionnaires and reached consensus on the majority of items. Participants agreed that Pilates exercise requires body awareness, breathing, movement control, posture, and education. It was recommended that people with CLBP should undertake supervised sessions for 30 to 60 minutes, twice per week, for 3 to 6 months. Participants also suggested that people with CLBP would benefit from individualized assessment and exercise prescription, supervision and functional integration of exercises, and use of specialized equipment. Limitations. Item consensus does not guarantee the accuracy of findings. This survey reflects the opinion of only 30 physical therapists and requires validation in future trials. Conclusion. These findings contribute to a better understanding of Pilates exercise and how it is utilized by physical therapists to treat people with CLBP. This information provides direction for future research into Pilates exercise, but findings need to be interpreted within the context of study limitations. Post a Rapid Response to this article at: ptjournal.apta.org 792 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

145 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain Chronic low back pain (CLBP) is defined as back pain of more than 12 weeks duration between the lower ribs and above the gluteal folds, with or without leg pain. 1 Chronic low back pain is a highly prevalent and disabling condition 2 4 that places a significant economic burden on society due to costs associated with treatment and the difficulty people have in returning to work. 3 5 Exercise has been shown to reduce pain and disability in people with CLBP. 6 8 According to current evidence, improvements are similar regardless of the type of exercise It is recommended, however, to consider the rationale underlying exercise approaches when prescribing an exercise program for people with CLBP. 12 This approach will assist in individually tailoring exercise programs for maximal effectiveness. 13,14 Pilates is a form of exercise that may be appropriate for people with CLBP Our recent systematic review of peer-reviewed literature described Pilates as a mind-body exercise that focuses on strength, core stability, flexibility, muscle control, posture, and breathing. 20 Several of these features of Pilates have been reported as effective in exercise programs for people with CLBP, such as mind-body therapies that encourage mental re-focusing and breathing and exercises that work on flexibility, strength, and motor control ,21 Furthermore, people with CLBP may benefit from exercises that address the control of posture and stabilizing muscles of the trunk. 12,22 25 Our review of systematic reviews, however, suggested that evidence of the effectiveness of Pilates exercise in people with CLBP is inconclusive. 26 This finding was due to the limited number, variable methodological quality, and small sample sizes of primary studies. 26 The heterogeneity of primary studies in terms of the population, intervention, comparison, and outcome measures also limits the strength of research findings, as pooling results of these studies in a meta-analysis is inappropriate. 26,27 The validity of research findings trials in relation to Pilates exercise also requires examination. 20,26 Clinical trials differ in their description of Pilates exercise and technique and in suggested program parameters, equipment, and levels of supervision for people with CLBP ,26 These variations of the definition and application of Pilates exercise create confusion around essential identifying features of Pilates exercise and make it difficult to apply findings in clinical practice. 26 A Delphi survey of a panel of Australian physical therapists was consequently undertaken to establish consensus regarding the definition and application of Pilates exercise to treat people with CLBP. The findings of this study will assist in the design of future Pilates exercise trials and in interpretation of existing findings. 20,26 The research questions of this Delphi survey were: 1. How is Pilates exercise defined in relation to people with CLBP? 2. What is the ideal Pilates exercise design, in terms of parameters, level of supervision, and equipment, for people with CLBP? 3. What principles are used to guide safe prescription and progression of Pilates exercise in people with CLBP? Method Design Overview A Delphi survey is a technique used to obtain group consensus from a panel of experts. 28,29 It involves a series of questionnaires, where panel members rank the relative importance or relevance of features under study. With each questionnaire round, panel members are provided with de-identified group feedback and a statistical summary of group findings. If desired, panel members can change their responses in subsequent rounds. Agreement among panel members is determined by consensus, voting, or through averaging of results. 28,30,31 Delphi surveys are frequently used in medical, health, and nursing research to explore topics with limited or conflicting research evidence. 28,30 A Delphi survey minimizes group bias by providing relative anonymity, where only the researchers are aware of the source of panel member comments and direct interaction among panel members does not occur. 28,29 National sampling also is convenient and costeffective, as panel members can be surveyed in different locations at different times. 29,30 Recruitment Participants were recruited via purposive sampling, where a panel of experts was selected based on their knowledge of and experience with the subject, their availability, and their interest and communication skills. 30 This method of recruit- Available With This Article at ptjournal.apta.org efigure: Overview of Delphi Survey Process eappendix 1: Delphi Survey: Questionnaire 1 eappendix 2: Delphi Survey: Questionnaire 2 eappendix 3: Delphi Survey: Questionnaire 3 June 2014 Volume 94 Number 6 Physical Therapy f 793 Downloaded from by guest on 02 August 2018

146 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain ment ensured that Delphi survey findings were based on informed opinions and that maximal participation rates were achieved. 30 Snowballing techniques also were used to identify potential panel members. Snowballing consists of participants nominating or recommending others to be involved in the study based on knowledge of the inclusion criteria of the study. 29 Using snowballing techniques of recruitment can increase both the size and diversity of the population sample. The recruitment process began with the primary researcher (C.W.) ing an invitation to participate to physical therapists who were likely to meet the selection criteria. This included research project information and informed consent and screening forms. Participants were invited to contact the primary researcher by or phone to discuss the project. Participants also were encouraged to forward the project information to interested physical therapists who they thought would meet the selection criteria. Interested participants then faxed or ed their completed screening and consent forms to the principal researcher. Once screening and consent forms were received and checked, participants were formally recruited into the study. Data collection commenced once a minimum of 30 participants were recruited. Selection Criteria To be included in the expert panel, participants needed to: 1. Be registered to currently practice as a physical therapist without restrictions in Australia with the Physiotherapy Board of Australia. The decision to include only registered physical therapists in Australia was to guarantee similar standards of practice of participants, as training and standards vary internationally Treat people with CLBP with Pilates exercise at least weekly or have published research on Pilates exercise and CLBP in a peer-reviewed journal. The decision to include people who were knowledgeable and experienced in using Pilates exercise to treat people with CLBP was to increase the usefulness of responses. 31,33 3. Be able to commit to completing at least 3 rounds of the Delphi survey, which may span 4 months. To do this, participants needed to be proficient in use of the English language, be computer literate, have access to and the Internet, and be able to commit time to complete the questionnaires. Survey Process The Delphi survey involved 3 electronic questionnaires provided over 4 months (March July 2012) (efigure, available at ptjournal.apta.org). An electronic survey was chosen over a paper-and-pen questionnaire due to the increased likelihood of greater participation rates and fewer missing data. 34 Conducting an electronic survey also provided an economical and efficient means of collecting data from a geographically dispersed sample. 29,30,34 Question- Pro software (QuestionPro Inc, Seattle, Washington) was selected as the electronic survey tool. 35,36 Participants were ed electronic links to each questionnaire and given individual login details to complete responses. Individual login details ensured security of information and prevented duplication of responses. Participants were requested to complete each questionnaire within 2 weeks. reminders were sent to participants who had not responded at 1 week and the day before the due date. If participants were not able to complete the questionnaires within the 2 weeks, they were provided with additional reminders and extra time to respond. Once at least 30 responses to a questionnaire had been received, participants who had not provided answers were not given the opportunity to answer any subsequent questionnaires. Questionnaire Development Questionnaire 1. The first questionnaire consisted primarily of open-ended questions to allow participants to express opinions without the provision of leading information (eappendix 1, available at ptjournal.apta.org). This method reduces response bias. 37 Multiplechoice questions (MCQs) were used to efficiently collect demographic information regarding the expert panel. 38 Responses to open-ended questions in the first questionnaire were summarized qualitatively using thematic analysis, a method for identifying, analysing and reporting patterns (themes) within data. 39 Two researchers (C.W., A.B.) were involved in this process to ensure validity and consistency of the approach. Themes identified from participant responses then were translated in statements about Pilates exercise and people with CLBP. These statements were utilized in the development of the second questionnaire. Questionnaire 2. The second questionnaire was developed from consideration of themes identified within responses to the first questionnaire and findings from systematic reviews on the definition and effectiveness of Pilates exercise in people with CLBP (eappendix 2, available at ptjournal.apta.org). 29,40 Participants were requested to rank their level of agreement with a num- 794 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

147 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain ber of statements regarding Pilates exercise in people with CLBP using a 6-point Likert response scale ( strongly agree, agree, somewhat agree, somewhat disagree, disagree, and strongly disagree ). A 6-point Likert scale was selected because it has been shown to be valid, reliable, and suitable for use with educated individuals. 41,42 The neutral category of the Likert scale was intentionally omitted from the scale to discourage ambivalence in responses. 42,43 The Likert scale of responses was used to identify areas of consensus or nonconsensus among the expert panel members. Prior to the commencement of this study, consensus was defined as when 70% to 100% of the panel members strongly agreed, agreed, or somewhat agreed (or strongly disagreed, disagreed, or somewhat disagreed) with an item. This definition of consensus was based on previously reported designs If the percentage of agreement or disagreement was 60% to 69%, the panel was considered to be approaching consensus for that question, as 60% agreement is considered by some authors to be appropriate for consensus. 44 If the percentage of agreement or disagreement was less than 60%, however, it was concluded that consensus had not been reached. Open-ended questions also were provided to ensure participants were able to express any further thoughts or opinions. Themes identified in these responses then were translated into questions for the third questionnaire. In addition, MCQs were used to collect information on exercise parameters and level of supervision in a time-efficient manner. 38 Questionnaire 3. The final questionnaire consisted only of questions requiring responses with the Likert response scale or MCQs (eappendix 3, available at ptjournal.apta.org). These responses determined the final level of consensus regarding several items. 28,30 Any questions that did not reach consensus during the second questionnaire were repeated in the final questionnaire. 28,30 Those items that gained consensus, however, were removed. Additional themes identified in open-ended questions in the second questionnaire were included to ensure thorough exploration of participant opinions. Participants also received a summary of de-identified responses from the second questionnaire. This summary was used to stimulate personal reflection on responses. 30 The summary of items with and without consensus was accompanied by percentages of agreement and disagreement. Data Analyses Participant information. The number of participant responses for each questionnaire was summated and monitored for dropouts. Participation rates, the time delay in returning questionnaires, and the number of reminders needed to maintain at least 30 responses in each round were monitored because they may indicate participant fatigue. 30 Demographic data regarding participants were summarized using descriptive statistics. Open-ended questions. Responses to open-ended questions in the first 2 questionnaires were summarized qualitatively using thematic analysis. 39 The number of identified themes was noted, and these themes were used to generate questions for subsequent questionnaires. Likert response scale questions. For questions with a Likert response scale, the number of responses of strongly agree, agree, or somewhat agree were summated and expressed as a percentage of agreement. Similarly, the number of responses of strongly disagree, disagree, or somewhat disagree were summated and expressed as a percentage of disagreement. MCQs. For MCQs related to exercise parameters and the level of supervision, the percentage of participants who selected each answer was interpreted as the percentage of agreement. Items with and without consensus. Items with and without consensus were identified in the final 2 questionnaires, where consensus was defined as when the percentage of agreement or disagreement for questions was 70% or greater. 28,30 Monitoring of any change in consensus for repeat questions in the 2 questionnaires was undertaken to observe any variation in the panel s views over time. 45 Strength of agreement or disagreement. Responses on the 6-point Likert scale were translated into numerical scores to understand the strength of agreement of participants regarding different questions. 44 A score of 1 represented strongly agree, a score of 2 represented agree, and so on, until a score of 6 represented strongly disagree. The median score and interquartile range of responses for these questions were then calculated. The median score was chosen over the mean due to the tendency of responses to converge with a Delphi survey. 40 Items where the median score indicated that participants strongly agreed were considered to be particularly important. Results Participant Recruitment Survey participants were recruited over February and March 2012 using purposive and snowballing sampling June 2014 Volume 94 Number 6 Physical Therapy f 795 Downloaded from by guest on 02 August 2018

148 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain techniques. 46,47 One hundred fiftythree invitations to participate were ed to potential participants by the primary researcher. Nine potential participants ed the researcher to decline to participate in the study, as they did not meet the selection criteria, and another invitation was returned due to an incorrect address. Thirty-seven physical therapists who met the selection criteria provided informed consent to participate. Of the 37 participants who received the first questionnaire, 33 (89.1%) responded. Of the 33 participants who received the second questionnaire, 31 (93.9%) responded. Of the 31 participants who received the third questionnaire, 30 (96.7%) responded. A high participation rate, therefore, was achieved, where 30 (81.1%) out of 37 participants completed the 3 questionnaires. 46 The use of snowballing in recruitment, however, restricted our ability to calculate an initial response rate, as the number of invitations sent to potential participants was unknown. 46,47 By the final questionnaire, 5 reminders and a 6-week period were required to ensure 30 participant responses were received. One of the reasons for this finding might have been participant fatigue with the Delphi survey process. 30 The 7 participants who did not return questionnaires did not provide reasons for dropping out of the study. Participant Demographics Of the 37 individuals who consented to participate in the study, 33 returned the first questionnaire. As such, detailed demographics apart from the selection criteria information were not collected from the 4 participants who did not respond to this questionnaire. When examining the demographics, the majority of participants were female physical therapists (91.0%) working in private practice (100%). The highest qualification held by the majority of participants was a bachelor s degree in physical therapy (54.5%) (a standard professional [entry-level] qualification required for registration as a physical therapist in Australia), and other participants had completed postgraduate qualifications in physical therapy (45.5%). Thirty-one participants (94.0%) had undertaken formal Pilates training, including courses run by Dance Medicine Australia (52.0%) and Polestar Pilates (30.0%). The 2 participants (6.0%) who did not undertake specific Pilates training reported having significant physical therapy work experience (18 years) and a physical therapy postgraduate degree. In terms of Pilates training, it is possible that these 2 participants had learned principles of Pilates exercise informally in the workplace or during their general university training. The mean (SD) age of the 33 participants was 33.8 (8.1) years while their mean number of years of physical therapy postgraduate experience was 10.9 (7.7) years. Approximately 80% of the participants reported that 20% or more of their clients experienced CLBP. Moreover, 67% of participants reported use of Pilates exercise to treat people with CLBP greater than 50% of the time. Participants were drawn from those who practiced physical therapy in the 6 states of Australia: New South Wales (36.4%), Western Australia (27.3%), Queensland (12.1%), South Australia (12.1%), Victoria (9.1%), and Tasmania (3.0%). Participants who completed all 3 questionnaires (n 30) and those who did not (n 3) had similar demographics, especially in relation to sex, workplace setting, qualifications, and usual clinical practice. Some differences were noted, however, in relation to the participants age, years of physical therapy experience, location of practice, and Pilates training. For example, the mean (SD) age of participants who did not complete all 3 questionnaires was 27.0 (3.6) years compared with 34.4 (8.1) years for participants who completed all 3 questionnaires, and their mean (SD) years of physical therapy experience was 4.0 (2.0) compared with 11.6 (7.8). The majority of participants who did not complete all 3 questionnaires practiced in Western Australia (66.7%) and trained with Polestar Pilates (66.7%), whereas a greater percentage of participants who completed all questionnaires practiced in states other than Western Australia (96.7%) and trained with Dance Medicine Australia Pilates (53.3%) rather than Polestar Pilates (26.7%). Thematic Analysis of Questionnaires From 18 different open-ended questions, a total of 192 themes were identified. These themes were used to generate questions regarding Pilates exercise by people with CLBP for subsequent questionnaires. Items of Consensus and Nonconsensus After 3 questionnaires, consensus levels of agreement were reached in regard to 91.7% (176/192) of the questions. Consensus was not obtained, however, in regard to 8.3% (16/192) of the questions. A summary of items of consensus and nonconsensus is provided below relative to research questions of this study. How is Pilates exercise defined in relation to people with CLBP? Consensus was reached on 97.1% (33/34) of questions related to identifying features of Pilates (Tab. 1). Identifying features of Pilates exercise that were particularly important included body awareness, breathing, control, education, individualized 796 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

149 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain exercises, movement control, and posture. Participants approached consensus regarding the question of fatiguing being part of Pilates exercise. Consensus was reached on 78.9% (15/19) of essential components of Pilates exercise programs for people with CLBP (Tab. 2). Essential components of particular importance included the use of therapist encouragement and feedback, functional integration of Pilates principles, incorporation of home exercises, client self-correction, and therapist reassessment. Consensus was not reached in regard to the prescription of a set number of exercises and incorporation of rest and cool-down exercise. What is the ideal Pilates exercise design, in terms of parameters, level of supervision, and equipment, for people with CLBP? Consensus was reached within a range of values on 100% of questions regarding ideal Pilates exercise parameters and supervision for people with CLBP. Participants agreed that supervised exercise sessions for people with CLBP should last between 30 and 60 minutes (100% agreement), should be undertaken at a frequency of 2 sessions per week (73.3% agreement), and should be conducted for a period of 3 to 6 months (83.4% agreement). The rationale reported by participants that underlie these parameters was to ensure clients remember their exercises, use the correct technique, successfully correct motor patterns, strengthen weak muscles, and achieve functional goals. These parameters also were thought to enable the reduction, prevention, and self-management of symptoms and fear-avoidance behavior and to maximize client enjoyment, motivation, and adherence within the con- Table 1. Identifying Features of Pilates Exercises in Relation to People With Chronic Low Back Pain Item With Consensus c Percentage of Agreement Median (Q1, Q3) a Strength of Agreement b 1. Body awareness (1.0, 1.0) Strongly agree 2. Breathing (1.0, 2.0) Strongly agree 3. Control (1.0, 2.0) Strongly agree 4. Education (1.0, 2.0) Strongly agree 5. Individualized (1.0, 1.0) Strongly agree 6. Movement control (1.0, 2.0) Strongly agree 7. Posture (1.0, 1.0) Strongly agree 8. Measured (1.0, 2.8) Strongly agree/agree 9. Mindfulness (1.0, 2.8) Strongly agree/agree 10. Concentration (2.0, 3.0) Agree 11. Coordination (1.0, 2.5) Agree 12. Core stability (1.0, 3.0) Agree 13. Direction preference (2.0, 3.0) Agree 14. Endurance (2.0, 2.5) Agree 15. Flexibility (2.0, 3.0) Agree 16. Goal orientated (1.0, 2.0) Agree 17. Graded (1.0, 2.0) Agree 18. Low impact (2.0, 3.0) Agree 19. Mind-body connection (1.0, 3.0) Agree 20. Muscle balance (1.0, 2.0) Agree 21. Precision (2.0, 3.0) Agree 22. Proprioception (1.0, 2.0) Agree 23. Relaxation (2.0, 3.0) Agree 24. Self-paced (2.0, 3.0) Agree 25. Supervised (1.0, 2.5) Agree 26. Structured (2.0, 3.0) Agree 27. Cognitive-behavioral therapy (2.0, 3.0) Somewhat agree 28. Flow (2.0, 3.0) Somewhat agree 29. Functional (2.3, 4.0) Somewhat agree 30. Holistic (2.0, 3.0) Somewhat agree 31. Pain-free (2.0, 3.8) Somewhat agree 32. Specific exercise (2.0, 3.8) Somewhat agree 33. Strength (2.0, 3.0) Somewhat agree Approaching Consensus d 34. Fatiguing (2.3, 4.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. d 60% 69% of participants agreed. June 2014 Volume 94 Number 6 Physical Therapy f 797 Downloaded from by guest on 02 August 2018

150 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain Table 2. Essential Components of Pilates Exercises for People With Chronic Low Back Pain With Consensus c Item fines of availability and budget (100% agreement). The level of supervision recommended by participants was 1 client per therapist at the start of the program (80.0% agreement) and 2 to 4 clients per therapist after 2 weeks (100% agreement). Participants agreed that these supervision levels allowed individual exercise prescription, progression, and monitoring of technique and ensured prevention of pain and injury, client selfmanagement, and a reduction in Percentage of Agreement Median (Q1, Q3) a Strength of Agreement b 1. Encouragement (1.0, 2.0) Strongly agree 2. Feedback on technique d (1.0, 1.0) Strongly agree 3. Functional integration (1.0, 1.0) Strongly agree 4. Home exercises (1.0, 2.0) Strongly agree 5. Reassessment (1.0, 1.0) Strongly agree 6. Client self-correction (1.0, 2.0) Strongly agree/agree 7. Balance exercises (1.0, 2.0) Agree 8. Contraction of stabilizing muscles of the lower back (2.0, 3.0) Agree 9. Education (1.0, 2.0) Agree 10. Equipment use (2.0, 3.0) Agree 11. Low load, high repetitions (2.0, 2.0) Agree 12. Pelvic-floor screening (1.0, 3.0) Agree 13. Strengthening exercises (2.0, 3.0) Agree 14. Stretching exercises d (2.0, 3.0) Agree 15. Warm-up exercises d (2.0, 3.8) Somewhat agree Approaching Consensus e 16. Minimum of 5 different exercises d (2.0, 4.0) Agree/somewhat agree 17. Rest between exercises d (3.0, 4.0) Somewhat agree Without Consensus f 18. Cool-down exercise d (3.0, 4.0) Somewhat agree 19. Maximum of 10 different exercises d (2.3, 4.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. d Asked in second and third questionnaires. e 60% 69% of participants agreed. f 0% 59% of participants agreed. dependence on the therapist over time (100% agreement). Consensus was reached on 67.9% (19/28) of questions related to essential Pilates equipment (Balanced Body, Sacramento, California) for people with CLBP (Tab. 3). The Reformer and mirror were considered to be especially important. Consensus was not reached in regard to use of Chi, Franklin, massage, and prop balls, Ladder Barrel, Magic Circle, suspension trainer, vibration machine, and video analysis. There was 90.9% agreement (10/11) regarding the rationale underlying the use of Pilates equipment for people with CLBP. Participants agreed that Pilates equipment provides proprioceptive and visual feedback, assists in the maintenance of spinal posture, and increases the functional relevance and variation of exercises. Pilates equipment also can provide adjustable resistance, opportunities for progression, and complement home exercises. Participants did not reach consensus regarding the cost of equipment influencing use. What principles are used to guide safe prescription and progression of Pilates exercise in people with CLBP? Participants reached consensus on 100% of questions related to individualization of programs for people with CLBP (Tab. 4). Factors that are particularly important to consider included client goals, functional requirements, irritability, specific movement or activity fears, and body awareness. Participants also reached consensus on 100% of questions related to exercise progression for people with CLBP (Tab. 5). Participants agreed that progression of exercises should primarily involve an increase in exercise complexity, replication of a relevant sport or functional activity, and functional integration of exercise principles. Participants reached consensus on 94.7% (18/19) of questions regarding the principles of Pilates exercise prescription for people with CLBP (Tab. 6). Principles of particular importance included conducting an initial assessment; educating clients regarding the purpose of Pilates exercise and chronic pain mechanisms; prescribing functionally relevant exercises according to client needs, ability, irritability, and pathology; supervising sessions, monitoring quality of technique, and encour- 798 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

151 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain aging breathing with movement; challenging fear-avoidance belief systems; and regularly reassessing symptoms and functional outcomes. Consensus was not reached in regard to teaching traditional Pilates principles. There was 100% consensus by participants that following prescription principles will ensure treatment outcomes (eg, improved posture, movement control, and function; decreased fear of movement; correction of maladaptive movement patterns; increased activation of appropriate muscles) are reached. Repeated Questions Items where consensus was not obtained in the second questionnaire were repeated in the third questionnaire. A total of 15 items that related to the essential components of Pilates exercise (n 7), exercise parameters (n 3), Pilates equipment (n 4), and prescription principles (n 1) were repeated (Tabs. 1, 2, 3, and 6). With repeat questioning, consensus of these items was obtained for 40.0%, including items relating to essential components (n 3) and ideal parameters of Pilates exercise for people with CLBP (n 3). Concluding the Delphi Survey The decision was made to finish the Delphi survey after 3 questionnaires. This decision was based on the analysis of the number and importance of items without consensus (16/192) and potential participant fatigue in responding to multiple questionnaires. 30 Discussion Findings In this Delphi survey, 30 physical therapists reached consensus on the majority of items relating to the definition and application of Pilates exercise in people with CLBP (Tabs. 1, 2, 3, 4, 5, and 6). After 3 rounds of Table 3. Ideal Pilates Exercise Equipment for People With Chronic Low Back Pain Item With Consensus c Percentage of Agreement questionnaires, consensus levels of agreement were reached for 91.7% (176/192) of questions. Items that did not reach consensus related to identifying features of Pilates (1/34), essential components of Pilates Median (Q1, Q3) a Strength of Agreement b 1. Mirror (1.0, 1.0) Strongly agree 2. Reformer (1.0, 2.0) Strongly agree 3. Exercise sheet (2.0, 3.0) Agree 4. Fitball (2.0, 3.0) Agree 5. Foam rollers (2.0, 3.0) Agree 6. Mat (1.0, 3.0) Agree 7. Pillows (2.0, 3.0) Agree 8. Raised bench/step (2.0, 3.0) Agree 9. Real-time ultrasound (2.0, 3.5) Agree 10. Resistance bands (1.0, 2.0) Agree 11. Towels (1.3, 3.0) Agree 12. Trapeze table (1.0, 2.0) Agree 13. Wunda chair (2.0, 3.5) Agree 14. Balance disk (2.0, 3.0) Somewhat agree 15. Educational books (2.0, 4.0) Somewhat agree 16. Hand weights (2.0, 3.0) Somewhat agree 17. Pressure biofeedback pillow (2.0, 3.5) Somewhat agree 18. Step barrel/spine corrector (2.0, 4.0) Somewhat agree 19. Balance board (2.0, 3.0) Somewhat agree Approaching Consensus d 20. Ladder Barrel e (2.0, 4.0) Somewhat agree 21. Magic Circle e (3.0, 4.0) Somewhat agree 22. Massage ball (2.0, 4.0) Somewhat agree 23. Prop ball e (2.0, 4.0) Somewhat agree Without Consensus f 24. Chi ball (3.0, 4.0) Somewhat agree 25. Franklin ball (2.0, 4.0) Somewhat agree 26. Suspension trainer (3.0, 4.0) Somewhat disagree 27. Vibration machine e (3.0, 4.0) Somewhat agree 28. Video analysis (3.0, 4.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. d 60% 69% of participants agreed. e Asked in the second and third questionnaires. f 0% 59% of participants agreed. (4/19), essential forms of equipment (9/28) and rationale for use (1/11), and exercise prescription principles (1/19) (Tabs. 1, 2, 3, and 6). June 2014 Volume 94 Number 6 Physical Therapy f 799 Downloaded from by guest on 02 August 2018

152 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain Table 4. Individualization of Pilates Exercise Programs for People With Chronic Low Back Pain With Consensus c Item Definition of Pilates exercise. Participants agreed that the 7 components of Pilates exercise identified in a recent systematic review of the literature (ie, breathing, posture, flexibility, movement control, strength, core stability, and a mindbody connection) were relevant to people with CLBP. 20 Breathing, movement control, and posture were considered to be particularly Percentage of Agreement Median (Q1, Q3) a Strength of Agreement b 1. Body awareness (1.0, 2.0) Strongly agree 2. Client goals (1.0, 1.0) Strongly agree 3. Functional requirements (1.0, 1.0) Strongly agree 4. Irritability (1.0, 2.0) Strongly agree 5. Specific movement or activity fears (1.0, 2.0) Strongly agree 6. Chronicity of symptoms (2.0, 3.0) Agree 7. Client availability (1.0, 2.0) Agree 8. Client commitment (2.0, 3.0) Agree 9. Client financial capacity (1.0, 2.0) Agree 10. Client motivation (1.5, 2.5) Agree 11. Flexibility (2.0, 3.0) Agree 12. Functional limitations (1.0, 2.0) Agree 13. Intensity of pain (1.5, 2.5) Agree 14. Movement control (1.0, 2.0) Agree 15. Muscle strength (2.0, 3.0) Agree 16. Pain management (2.0, 2.8) Agree 17. Pain-relieving exercise (1.0, 2.0) Agree 18. Pathology (1.0, 2.0) Agree 19. Pelvic-floor muscle dysfunction (1.0, 2.0) Agree 20. Posture (1.0, 2.0) Agree 21. Previous Pilates experience (2.0, 3.0) Agree 22. Previous treatment and effect (1.3, 2.0) Agree 23. Psychosocial factors (1.0, 2.0) Agree 24. Cardiovascular fitness (2.0, 3.0) Somewhat agree 25. Medications (2.0, 3.0) Somewhat agree 26. Previous exercise or sports experience (2.0, 3.0) Somewhat agree 27. Time of day (2.0, 3.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. important, as indicated by the high median score of agreement. The relative importance of other identifying features and essential components, however, warrants further examination (Tabs. 1 and 2). Exercise parameters, levels of supervision, and equipment. Consensus findings provide specific guidelines for using Pilates to treat people with CLBP. When comparing these parameters with those used in research trials, the length and frequency of Pilates exercise sessions have often been appropriate; however, the duration of exercise programs (ie, 6 8 weeks) has been too short ,26 Given that the total number of sessions and exercise hours may be associated with effect sizes in exercise trials for people with CLBP, it may be important that future trials maximize outcomes by ensuring Pilates interventions are 3 to 6 months in duration. 48 Consensus findings also provide direction regarding the essential equipment and levels of supervision for using Pilates to treat people with CLBP. The majority of Pilates exercise trials have not utilized equipment in their programs for people with CLBP ,26 Given survey findings, however, future research should investigate the benefits of programs with and without use of equipment (Tab. 3). Similarly, supervision levels need to be carefully considered in future trials, as they may influence exercise effectiveness in people with CLBP. 14,49 Prescription principles. Participants agreed on several principles for prescription of Pilates exercise that are similar to principles of other exercise approaches that are effective in people with CLBP. For example, participants agreed that exercises should be individually tailored and supervised and include stretching and strengthening. 13,14 Pilates exercises also should focus on trunk muscle strength, endurance, and coordination; respect the directional preferences of clients; and include cognitive-behavioral therapy, education, and feedback. 12,13,50,51 The importance of other items of consensus relating to the individualization, prescription, and progression of exercises needs to be verified by sub- 800 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

153 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain sequent clinical research (Tabs. 4, 5, and 6). Participants did not reach consensus on the importance of teaching people with CLBP traditional Pilates principles. Our systematic review of the literature showed that traditional Pilates principles, such as centering, concentration, and precision, were not mentioned in published studies of CLBP participants, suggesting they may not be important. 20 Nevertheless, when examining consensus findings regarding identifying features of Pilates exercise, traditional principles of concentration, precision, flow, control, and breathing were included. 52,53 Although the traditional principle of centering was not specifically mentioned by participants, it could be that it is incorporated in the idea of core stability. 20,53 Future research should clarify the importance of traditional principles for people with CLBP, given this conflicting finding. Strengths This is the first Delphi survey, to our knowledge, that has developed consensus on the definition and application of Pilates exercise in people with CLBP according to 30 Australian physical therapists. Although there are no universal guidelines in the literature regarding appropriate sample sizes for Delphi surveys, a sample size of 30 participants can be argued as adequate, given the participants were homogenous. 30,54 It also has been reported that having more than 30 participants may not increase the quality of results but instead may increase management or attrition problems. 55 In this study, findings were minimally affected by attrition, as 4 of the 7 participants who dropped out did not return any questionnaires. 45 In Delphi surveys, the representativeness of samples is indicated by the qualities of the expert panel Table 5. Methods of Progression of Pilates Exercises for People With Chronic Low Back Pain With Consensus c Item rather than its numbers. 31 This finding is because nonprobability sampling, such as purposive and snowballing techniques, is used to recruit participants who can provide wellconsidered responses based on specialized knowledge and experience. 56 The credibility of findings in this survey, therefore, is enhanced by participants training, education, and experience in prescribing Pilates exercise to treat people with CLBP. 28,30,31 All participants were registered to practice physical therapy in Australia, which ensured similar baseline university education and competency in treating people with CLBP. 32 Several physical therapists (45.5%) also had undertaken further Percentage of Agreement Median (Q1, Q3) a Strength of Agreement b 1. Increase in exercise complexity (1.0, 2.0) Strongly agree 2. Functional integration of exercise (1.0, 1.5) Strongly agree principles 3. Replicate functional tasks or sport (1.0, 2.0) Strongly agree 4. Activation of stabilizing lower back muscles combined with limb movement (1.0, 2.0) Agree 5. Activation of stabilizing lower back muscles combined with breathing (1.0, 3.0) Agree 6. Decrease base of support (2.0, 3.0) Agree 7. Include movements outside of movement direction preference of client (1.0, 3.0) Agree 8. Incorporate segmental spinal movement (1.0, 2.5) Agree 9. Increase in exercise duration (1.0, 2.5) Agree 10. Increase in exercise load or resistance (2.0, 3.0) Agree 11. Increase in exercise repetitions (1.5, 2.0) Agree 12. Progress toward feared movements (1.0, 2.0) Agree 13. Reduce supervision and feedback (1.0, 2.0) Agree 14. Increase speed of exercise (2.0, 4.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. postgraduate physical therapy study, which may indicate advanced physical therapy knowledge and skills. 57,58 The average length of physical therapy work experience was greater than 10 years, which may indicate expert physical therapy status. 59,60 To be involved in this study, participants needed to use Pilates exercise at least weekly to treat people with CLBP. The majority of participants reported that at least 20% of their clients per week presented with CLBP and that they used Pilates to treat these clients more than 50% of the time. Although there was likely to be variation among participants in terms of formalized Pilates training and experience, 94.0% of the participants had undertaken some form of June 2014 Volume 94 Number 6 Physical Therapy f 801 Downloaded from by guest on 02 August 2018

154 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain Table 6. Principles of Pilates Exercise Prescription for People With Chronic Low Back Pain With Consensus c Item Pilates training outside of their entrylevel university physical therapy study. The Delphi survey design also enhanced the quality and integrity of participant responses. Participants had several opportunities to express and qualify their opinions with multiple rounds of questionnaires, repetition of questions without consensus, and use of open-ended, multiple-choice, and Likert response Percentage of Agreement Median (Q1, Q3) a Strength of Agreement b 1. Conduct an initial assessment (1.0, 1.0) Strongly agree 2. Consider client irritability (1.0, 1.5) Strongly agree 3. Consider client pathology (1.0, 2.0) Strongly agree 4. Educate regarding purpose of Pilates exercises (1.0, 2.0) Strongly agree 5. Educate regarding chronic pain mechanisms (1.0, 2.0) Strongly agree 6. Encourage breathing with movement (1.0, 2.0) Strongly agree 7. Monitor quality of technique (1.0, 2.0) Strongly agree 8. Prescribe exercises according to client needs and ability (1.0, 1.0) Strongly agree 9. Prescribe functionally relevant exercise (1.0, 2.0) Strongly agree 10. Regularly reassess symptoms and (1.0, 2.0) Strongly agree functional outcomes 11. Supervise exercise sessions (1.0, 2.0) Strongly agree 12. Challenge fear-avoidance belief systems (1.0, 2.0) Strongly agree 13. Consider movement direction preference of client (1.0, 3.0) Agree 14. Encourage muscle balance (1.0, 2.0) Agree 15. Ensure exercises do not increase (1.0, 2.0) Agree or cause pain 16. Gradually increase difficulty of exercises (1.0, 2.0) Agree 17. Start exercises in neutral spine position (2.0, 3.5) Agree 18. Ensure exercise variation (2.0, 4.0) Somewhat agree Approaching Consensus d 19. Teach traditional Pilates exercise principles e (3.0, 4.0) Somewhat agree a Q1 25th percentile, Q3 75th percentile. Scores are on a scale from 1 to 6, where 1 strongly agree, 2 agree, 3 somewhat agree, 4 somewhat disagree, 5 disagree, and 6 strongly disagree. b Qualitative descriptor of median score. c 70% 100% of participants agreed. d 60% 69% of participants agreed. e Asked in second and third questionnaires. scale questions. 30,31 In addition, the provision of de-identified group summary responses and the relative anonymity of participant responses encouraged participants to reflect on their answers and respond honestly without pressure from other group members. 28,30 The validity of findings also was enhanced by the clear, methodical, and consistent manner by which participant responses were summarized, analyzed, and interpreted. The accuracy of thematic analysis of open-ended questions was improved by more than one researcher being involved. 30 Consensus was clearly defined a priori as 70% participant agreement or disagreement, which is similar to other levels of consensus in the literature. 28,30 A comparison of median scores for questions with a Likert response scale assisted in organizing items of consensus in order of importance. 44,61 Finishing the survey after 3 rounds was supported by the relatively small number and importance of items without consensus (16/192). For example, obtaining consensus about all 28 potential forms of equipment is unlikely to be helpful, particularly when 19 pieces of equipment have had already been confirmed as ideal by the participants. Limitations These Delphi survey results reflect the perspectives of 30 physical therapists registered to practice in Australia who use Pilates exercise at least weekly to treat people with CLBP. The external validity of findings, therefore, is limited, as physical therapists from other countries and non physical therapist Pilates practitioners may have different but equally important views that have not been incorporated. 29 Moreover, inclusion of a more heterogeneous sample of experts, including alternative medicine Pilates practitioners, may have ensured that a greater spectrum of opinions were considered. 30 Only 30 physical therapists participated in this Delphi survey, which means that findings may be skewed, as only a proportion of Australian physical therapists experienced in the use of Pilates exercise in people with CLBP gave their opinion. 30 Selection and response bias are likely to be present where physical thera- 802 f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

155 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain pists who met the selection criteria were not invited to participate, did not agree to participate, or did not follow through in completing questionnaires. 29,30 It also should be noted that 2 of the participants had not undertaken any formalized Pilates training, which may limit the validity of their responses. Their extensive physical therapy experience and postgraduate physical therapy training, however, suggest expert status in treating people with common musculoskeletal conditions such as CLBP. 2 5 The findings of this study also could be compromised due to different definitions of CLBP and aspects of Pilates exercise being used by participants. For example, CLBP is usually described as pain in the lumbar region lasting more than 12 weeks; however, at times subacute and recurrent LBP have been classified together with CLBP. 62,63 Similarly, the mind-body feature of Pilates exercise could refer to the psychological impact of physical exercise, or a combination of behavioral, psychological, social, and spiritual approaches to treatment. 64,65 Future research, therefore, should provide definitions of terms to be used by participants in a Delphi survey. The Delphi technique itself has inherent limitations. The iterative and de-identified group feedback process has the potential of encouraging participants to agree, even though participants do not directly interact with each other. 30,31 This process can lead to researcher and participant bias. Delphi survey findings can only be considered as expert opinion and are not considered high in the hierarchy of evidence compared with primary studies. 66 Finally, a consensus of findings does not mean the group conclusion is correct. 30 These findings, therefore, need to be validated and tested in subsequent clinical research. Implications This Delphi survey provides potentially valuable information for interpreting the results of clinical trials that investigate the effectiveness of Pilates exercise in people with CLBP. For example, the validity of definitions of Pilates exercise and the optimization of exercise design and prescription can be evaluated through comparison with consensus items. This comparison may provide an indication of the ecological validity of evidence available, from the perspective of 30 Australian physical therapists who regularly use Pilates exercise to treat people with CLBP. Items of consensus relating to the definition and application of Pilates exercise could be used to direct future research and clinical practice. The efficacy of Pilates exercise in people with CLBP then could be evaluated in a consistent manner according to the perspectives of Australian physical therapists expressed in this survey. Future research also should examine items without consensus, such as the use of different types of equipment, and those that are conflicting, such as ensuring Pilates exercises are pain-free and challenge fear-avoidance behavior. It must be remembered, however, that findings of this Delphi survey represent the opinions of 30 Australian physical therapists who are experienced in the use of Pilates exercise to treat people with CLBP. Exploration into how physical therapists define and use Pilates exercise to treat people with CLBP differently across the globe may provide interesting insights, as would investigation into how non physical therapist Pilates practitioners use Pilates exercise to treat people with CLBP. Ms Wells, Dr Kolt, and Dr Bialocerkowski provided concept/idea/research report. All authors provided writing. Ms Wells provided data collection. Ms Wells and Dr Bialocerkowski provided data analysis. Ms Wells and Dr Marshall provided project management. Dr Kolt, Dr Marshall, and Dr Bialocerkowski provided consultation (including review of manuscript before submission). Ethical approval to conduct the Delphi survey was provided by the Human Research Ethics Committee of the University of Western Sydney. DOI: /ptj References 1 Charlton JE. Core Curriculum for Professional Education in Pain. 3rd ed. Seattle, WA: International Association of the Study of Pain (IASP) Press; Freburger JK, Holmes GM, Agans RP, et al. The rising prevalence of chronic low back pain. Arch Intern Med. 2009;169: Gore M, Tai KS, Sadosky A, et al. Use and costs of prescription medications and alternative treatments for patients with osteoarthritis and chronic low back pain in community settings. Pain Pract. 2012; 12: Hoy D, Brooks P, Blyth F, Buchbinder R. The epidemiology of low back pain. Best Pract Res Clin Rheumatol. 2010;24: Dageniais S, Caro J, Haldeman S. A systematic review of low back pain cost of illness studies in the United States and internationally. Spine J. 2008;8: Koes BW, van Tulder M, Lin CW, et al. An updated overview of clinical guidelines for the management of non-specific low back pain in primary care. Eur Spine J. 2010; 19: Pillastrini P, Gardenghi I, Bonetti F, et al. An updated overview of clinical guidelines for chronic low back pain management in primary care. Joint Bone Spine. 2012;79: Van Middelkoop M, Rubinstein SM, Kuijpers T, et al. A systematic review on the effectiveness of physical and rehabilitation interventions for chronic non-specific low back pain. Eur Spine J. 2011;20: Koumantakis G, Watson P, Oldham J. Trunk muscle stabilization training plus general exercise versus general exercise only: randomised controlled trial of patients with recurrent low back pain. Phys Ther. 2005;85: May S, Johnson R. Stabilisation exercises for low back pain: a systematic review. Physiotherapy. 2008;94: Van Middelkoop M, Rubinstein SM, Verhagen AP, et al. Exercise therapy for chronic nonspecific low-back pain. Best Pract Res Clin Rheumatol. 2010;24: Macedo LG, Latimer J, Maher CG, et al. Effect of motor control exercises versus graded activity in patients with chronic nonspecific low back pain: a randomized control trial. Phys Ther. 2012;92: June 2014 Volume 94 Number 6 Physical Therapy f 803 Downloaded from by guest on 02 August 2018

156 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain 13 Liddle SD, Baxter GD, Gracey JH. Exercise and chronic low back pain: What works? Pain. 2004;107: Hayden JA, van Tulder MW, Tomlinson G. Systematic review: strategies for using exercise therapy to improve outcomes in chronic low back pain. Ann Intern Med. 2005;142: La Touche R, Escalante K, Linares MT. Treating non-specific chronic low back pain through the Pilates method. J Bodyw Mov Ther. 2008;12: Aladro-Gonzalvo AR, Araya-Vargas GA, Machado-Diaz M, Salazar-Rojas W. Pilatesbased exercise for persistent, nonspecific low back pain and associated functional disability: a meta-analysis with meta-regression. J Bodyw Mov Ther. 2012;17: Lim ECW, Poh RLC, Low AY, Wong WP. Effects of Pilates-based exercises on pain and disability in individuals with persistent nonspecific low back pain: a systematic review with meta-analysis. J Orthop Sports Phys Ther. 2011;41: Pereira LM, Obara K, Dias JM, et al. Comparing the Pilates method with no exercise or lumbar stabilisation for pain and functionality in patients with chronic low back pain: systematic review and metaanalysis. Clin Rehabil. 2012;2: Posadzki P, Lizis P, Hagner-Derengowska M. Pilates for low back pain: a systematic review. Complement Ther Clin Pract. 2011;17: Wells C, Kolt GS, Bialocerkowski A. Definition of Pilates: a systematic review. Complement Ther Med. 2012;20: Sherman KJ, Cherkin DC, Erro J, et al. Comparing yoga, exercise and a self-care book for chronic low back pain. Ann Int Med. 2005;143: Dankaerts W, O Sullivan P, Burnett A, Straker L. Differences in sitting postures are associated with nonspecific chronic low back pain disorders when patients are subclassified. Spine. 2006;31: Ferreira PH, Ferreira ML, Maher CG, et al. Changes in recruitment of transversus abdominis correlate with disability in people with chronic low back pain. Br J Sports Med. 2010;44: Macedo LG, Maher CG, Latimer J, McAuley JH. Motor control exercise for persistent, nonspecific low back pain: a systematic review. Phys Ther. 2009;89: Wallwork T, Stanton W, Freke M, Hides J. The effect of chronic low back pain on size and contraction of the lumbar multifidus muscle. Man Ther. 2009;14: Wells C, Kolt GS, Marshall P, et al. Effectiveness of Pilates exercise in treating people with chronic low back pain: a systematic review of systematic reviews. BMC Med Res Methodol. 2013;13:7. 27 Slavin RE. Best evidence synthesis: an intelligent alternative to meta-analysis. Clin Epidemiol. 1995;48: Keeney S, Hasson F, McKenna H. Consulting the oracle: ten lessons from using the Delphi survey technique in nursing research. J Adv Nurs. 2006;53: Portney L, Watkins MP. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River, NJ: Pearson Education; Keeney S, Hasson F, McKenna H. The Delphi Technique in Nursing and Health Research. Chichester, West Sussex, United Kingdom: Wiley-Blackwell; Powell C. The Delphi technique: myths and realities. J Adv Nurs. 2003;41: Higgs J, Smith M, Webb G, et al. Contexts of Physiotherapy Practice. Sydney, New South Wales, Australia: Elsevier-Churchill Livingstone; Novakowski N, Wellar B. Using the Delphi technique in normative planning research: methodological design considerations. Environ Plan A. 2008;40: Smyth JD, Dillman D, Christian LM, Mcbride M. Open-ended questions in web surveys: can increasing the size of answer boxes and providing extra verbal instructions improve response quality? Public Opin Q. 2009;73: Carter-Pokras O, McClellan L, Zambrana RE. Surveying free and low-cost survey software. J Natl Med Assoc. 2006;98: QuestionPro: online research made easy. Available at: com. Accessed January 13, Choi BCK, Pak AWP. A catalog of biases in questionnaires. Prev Chronic Dis. 2005;2: A13. Available at: issues/2005/jan/04_0050.htm. Accessed January 13, Collins J. Education techniques for lifelong learning: writing multiple-choice questions for continuing medical education activities and self-assessment modules. Radiographics. 2006;26: Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3: Hsu C, Sandford BA. The Delphi technique: making sense of consensus. Practical Assessment, Research and Evaluation. 2007;12:1 8. Available at: pareonline.net/pdf/v12n10.pdf. Accessed January 13, Preston CC, Colman AM. Optimal number of response categories in rating scales: reliability, validity, discriminating power, and respondent preferences. Acta Psychologica. 2000;104: Weijters B, Cabooter E, Schillewaert N. The effect of rating scale format on response styles: the number of response categories and response category labels. International Journal of Research in Marketing. 2010;27: Nowlis SM, Kahn BE, Dhar R. Coping with ambivalence: the effect of removing a neutral option on consumer attitude and preference judgements. J Consum Res. 2002; 29: Ferguson FC, Brownlee M, Webster V. A Delphi study investigating consensus among expert physiotherapists in relation to the management of low back pain. Musculoskeletal Care. 2008;6: Hsu C, Sandford BA. Minimizing nonresponse in the Delphi process: how to respond to non-response. Practical Assessment, Research and Evaluation. 2007;12: 1 8. Available at: pdf/v12n17.pdf. Accessed January 13, American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 7th ed Available at: cfm?section Standard_Definitions2& Template /CM/ContentDisplay.cfm &ContentID Accessed July 29, Callegaro M, DiSogra C. Computing response metrics for online panels. Public Opin Q. 2008;72: Ferreira ML, Smeets RJEM, Kamper SJ, et al. Can we explain heterogeneity among randomized clinical trials of exercise for chronic back pain? A metaregression analysis of randomized controlled trials. Phys Ther. 2010;90: Marshall P, Murphy B. Self-report measures best explain changes in disability compared with physical measures after exercise rehabilitation for chronic low back pain. Spine. 2008;33: Delitto A, George SZ, van Dillen LR, et al. Low back pain. J Orthop Sports Phys Ther. 2012;42:A1 A Sveinsdottir V, Eriksen HR, Reme SE. Assessing the role of cognitive behavioural therapy in the management of chronic nonspecific back pain. J Pain Res. 2012; 5: Latey P. Updating the principles of the Pilates method: part 2. J Bodyw Mov Ther. 2002;6: Friedman P, Eisen G. The Pilates Method of Physical and Mental Conditioning. 10th ed. London, United Kingdom: Penguin Books; Skulmoski G, Hartman FT, Krahn J. The Delphi method for graduate research. Journal of Information Technology Education. 2007;6:1 21. Available at: Skulmoski212.pdf. Accessed January 13, de Villiers MR, de Villiers PJ, Kent AP. The Delphi technique in health sciences education research. Med Teach. 2005;27: Glässel A, Kirchberger I, Kollerits B, et al. Content validity of the extended ICF Core Set for stroke: an international Delphi survey of physical therapists. Phys Ther. 2011;91: Donohoe HM, Needham RD. Moving best practice forward: Delphi characteristics, advantages, potential problems, and solutions. International Journal of Tourism Research. 2009;11: Australian Qualifications Framework Council. Australian Qualifications Framework. 2nd ed. Available at: aqf.edu.au/wp-content/uploads/2013/05/ AQF-2nd-Edition-January-2013.pdf Accessed January 13, f Physical Therapy Volume 94 Number 6 June 2014 Downloaded from by guest on 02 August 2018

157 Definition and Application of Pilates Exercise to Treat Chronic Low Back Pain 59 Stathopoulos I, Harrison K. Study at master s level by practicing physiotherapists. J Physiother. 2003;89: Doody C, McAteere M. Clinical reasoning of expert and novice physiotherapists in an outpatient orthopaedic setting. J Physiother. 2002;88: Wilde VE, Ford JJ, McMeeken JM. Indicators of lumbar zygapophyseal joint pain: survey of an expert panel with the Delphi technique. Phys Ther. 2007;87: Wasiak R, Young AE, Dunn KM, et al. Back pain recurrence: an evaluation of existing indicators and direction for future research. Spine. 2009;34: Pengel HM, Maher CG, Refshauge KM. Systematic review of conservative interventions for subacute low back pain. Clin Rehabil. 2002;16: Sherman R, Hickner J. Academic physicians use placebos in clinical practice and believe in the mind-body connection. J Gen Intern Med. 2008;23: Atin JA. Mind-body therapies for the management of pain. Clin J Pain. 2004;20: Howick J. Oxford Centre for Evidence- Based Medicine: Levels of Evidence. Available at: o Accessed January 13, June 2014 Volume 94 Number 6 Physical Therapy f 805 Downloaded from by guest on 02 August 2018

158

159

160

161

162

163 HEALTH AUGUST SEPTEMBER 2010 THE ROLE OF PILATES IN LUMBAR SPINE INSTABILITY TRAINING Glenn Withers, founder of the Australian Physiotherapy and Pilates Institute (APPI), explains how his methods can help populations with lower back pain Over 65% of the clients that present to my practices are reporting low back pain (LBP) and one of the reasons Pilates has become so popular is its perceived ability to help. The recurrence of LBP is most widely attributed to the nature of instability, which is often debated as a cause due to its challenges in diagnosis. In the instance of clinical instability, the treatment of choice is specific rehabilitation of the segmental spinal muscles that attach directly to the spinal segment/s that are moving too much. Hides 1 showed that by activating the local multifidus muscle of the lumbar spine, in co-activation with the transverses abdominus muscle, the recurrence rate of acute back pain 12-month post specific rehabilitation was reduced by as much as 50%. THE APPI METHOD However, more recent studies have suggested that this static stabilisation approach is not as effective as once thought. 2 Indeed, these recent reviews are advocating a more well-rounded rehabilitation programme that takes into account the many aspects of LBP that affect individuals. It is now widely accepted that rehabilitation programmes for LBP should also encompass the emotional and functional effects of pain in addition to the mechanical pain. 3 The conclusions of studies into LBP rehabilitation conducted over the past ten years are advocating a rehabilitation approach that restores normal movement as well as emotional and functional well being. This approach leads us back to the popularity of the Pilates method. Joseph Pilates developed his own set of exercises designed to improve strength and posture. His exercises focused on the development of a strong central core of abdominal muscles to enable more efficient movement. 4 Pilates believed that injuries were caused by imbalances in the body and habitual patterns of movement. He observed that when a person had a weak or malaligned area, they overcompensated, or overdeveloped another area to achieve a certain functional movement. He developed three primary objectives for his exercises: 1. The correction of malaligned movement 2. Movement re-education 3. The prevention of injury recurrence 5 However, despite common objectives, there are limitations to the application of traditional Pilates in the management of LBP. For example, several traditional Pilates exercises involve long lever movements with the spine in high degrees of lumbar and cervical flexion. At a scientific level, increased lever length is known to place increased pressure on the spine. 6 Therefore, the founders of the APPI Method have deemed these selective traditional movements too stressful on the spine when being used in the clinical management of LBP. The above example highlights the limitations of applying traditional Pilates to the LBP population. However, there are many Pilates objectives and movements that are ideal for managing LBP once modified according to known scientific and research principles. A widely-used modified Pilates approach comes from the Australian Physiotherapy and Pilates Institute (APPI). The APPI Pilates Method was developed by analysing each of the 34 traditional movements and breaking each one down into four or six levels suitable for the LBP clinical population. Analysis was based on spinal stabilisation, biomechanical and anatomical research. Through this analysis, the APPI also felt that several of the traditional movements were inappropriate for the LBP population (see right). In these pictures you can see that the spine is placed in an extreme degree of flexion, with full bodyweight and gravity being loaded through the cervical spine with the lumbar spine in full loaded flexion. These end-range positions and extreme loads to the spinal discs are some of the most common precursors to a spinal disc injury. The APPI method to LBP rehabilitation consists of a five stage programme based on pain, pathology and function. Importantly, the end goal is functional restoration. Along each of the five stages, a client is given an exercise that replicates their functional goal. This exercise is initially performed in non-weight bearing positions and gradually progressed to weight bearing and more functional positions based on the client's ability to maintain their neutral spinal position. In addition, the core of the APPI method is a focus on clinical reasoning of each and every exercise. To illustrate the role of APPI Pilates in rehabilitation amongst the LBP population, consider the common 20 FITPRO NETWORK AUGUST SEPTEMBER _21 Pilates_alt.indd 20 8/7/10 12:15:57

164 HEALTH presentation of a middle- aged client reporting low grade backache that is acerbated by running on a treadmill. On examination, this presentation will lack the required hip extension needed to push off effectively and then compensate by hyperextending or hinging in the lumbar spine. This hinging movement is seen as a form of instability whereby the motion of this hinge or shear can gradually wear away at the discs in the lower back. If the client can learn to stabilise the lumbar spine by correctly activating the muscles that directly attach to the motion segment in question, then they will be more likely to regain the independent hip extension needed to run pain free. In applying this common complaint to the APPI fivestage programme (below) we can start to ensure a full functional recovery. Step 1: Activation of the centre: Here the client is unloaded so gravity is not affecting the load on the spine. The client is taught to activate the lower abdominals, in association with the multifidus, pelvic floor and diaphragm 7. It is important to note here that in the APPI Method the client is asked to think of activating only one of these muscles, not to activate one and then the other as is common in Pilates teaching. This is based on the findings of Sapsford et al who showed that activation of the pubococcygeus and TrA occurs in co-activation at the required sub-maximal level for tonic retraining of the postural fibres required for stability. ONE-LEG STRETCH, LEVEL ONE Step 2: Closed-chain progression: The client is challenged by controlling the centre as a closed-chain lever is added. The exercise one leg stretch is chosen here as it specifically replicates the muscle memory required for the walking and running motion. SCISSORS, LEVEL ONE Step 3: Open chain progression: The client is now progressed to an open chain movement, whereby the weight of the leg is added to the challenge of stabilising the lumbar spine. Here the scissors level one movement is chosen as it replicates the challenges of the walking or running motion. SHOULDER BRIDGE, LEVEL ONE Step 4: Spinal Articulation: In this stage the spine is moved to ensure that synergy between the local muscles (those that attach to and stabilise the spine) and the global muscles (those that move the spine and extremities) is retrained. This synergy is vital as this is how the body functions normally in everyday activities. Shoulder bridge level one is selected as this movement requires the gluteal muscle activation to generate a relative hip extension moment (therefore, helping to retrain the primary problem that was initially at fault) as well as control segmental spinal movement. It is important to note that the APPI adaptation of the shoulder bridge movement is different from traditional Pilates. The APPI adaption of this exercise focuses on the initial gluteal activation to commence the hip extension movement rather than a pure focus on the hamstring muscles. This is to avoid the client incorrectly either over-recruiting the hamstring muscles or reinforcing poor timing of activation between the gluteals and hamstring muscle groups, which can cause a major imbalance that is commonly seen in LBP. Step 5: Function: In this stage, the principles of the previous four stages are applied to any functional movement that the therapist sees fit. This final stage is not Pilates-specific, but sport, or function-specific. In the example of running, a lunge squat may be used to teach the lower limb biomechanics of loading from one limb to the other, while encouraging the hip extension required to run pain free. Importantly, in this stage the client is weight bearing, as this is required for the functional task. Pilates enjoys worldwide popularity and is an achievable exercise method for a variety of populations. In using Pilates in the treatment of lumbar instability, it is important to ensure that the relevant research is applied to the movements and the effects of pain acknowledged in the rehabilitation phase. The application of Pilates in the low back pain population requires ongoing assessment, monitoring and adaption of the movements and their responses to ensure that the best outcomes are achieved. fn GLENN WITHERS Glenn Withers is the founder of the Australian Physiotherapy & Pilates Institute, director of Pilates Art Physiotherapy and London Sports Medicine Centres AUGUST SEPTEMBER 2010 FITPRO NETWORK 20_21 Pilates_alt.indd 21 8/7/10 12:16:00