Cervical Muscle Area Measurements in Whiplash Patients: Acute, 3, and 6 Months of Follow-up

Transcription

1 CME JOURNAL OF MAGNETIC RESONANCE IMAGING 36: (2012) Original Research Cervical Muscle Area Measurements in Whiplash Patients: Acute, 3, and 6 Months of Follow-up Erika J. Ulbrich, MD, 1,2 * Ramon Aeberhard, MD, 2 Sylvia Wetli, MD, 2 Andre Busato, PhD, 3 Chris Boesch, MD, PhD, 4 Heinz Zimmermann, MD, 5 Juerg Hodler, MD, PhD, 2 Suzanne E. Anderson, MD, PhD, 2,6 and Matthias Sturzenegger, MD 7 Purpose: To investigate the role of the cervical spine muscles in whiplash injury. We hypothesized that (i) cervical muscle hypotrophy would be evident after a 6-month follow-up and, (ii) cervical muscle hypotrophy would correlate with symptom persistence probably related to pain or inactivity. Materials and Methods: Ninety symptomatic patients (48 females) were recruited from our emergency department and examined within 48 h, and at 3, and 6 months after a motor vehicle accident. MRI cross-sectional muscle area (CSA) measurements were performed bilaterally of the cervical extensor and sternocleidomastoid muscles using transverse STIR (Short Tau inversion Recovery) sequences at the C2 (deep and total dorsal cervical extensor muscles), C4 (sternocleidomastoid muscles) and C5 (deep and total dorsal cervical extensor muscles) levels. Two blinded raters independently performed the measurements at each time point. First, CSA changes over time were analyzed and, second, CSAs were correlated with clinical outcomes (EuroQuol, Whiplash Disability Score, neck pain intensity [VAS], cervical spine mobility). Results: There was a high agreement of CSA measurements between the two raters. Women consistently had smaller CSAs than men. There were no significant 1 Department of Radiology, University Hospital, Zurich, Switzerland. 2 Department of Diagnostic, Interventional and Pediatric Radiology, Inselspital, University Hospital, and University of Berne, Switzerland. 3 Institute of Social and Preventive Medicine, Health Services Research, Berne, Switzerland. 4 Department of Clinical Research and Radiology, University of Bern, Berne, Switzerland. 5 Department of Emergency, Inselspital, Bern University Hospital, and University of Bern, Switzerland. 6 School of Medicine, Medical Imaging, The University of Notre Dame Sydney, Darlinghurst Campus, New South Wales, Australia. 7 Department of Neurology, Inselspital, University Hospital, and University of Berne, Switzerland. Contract grant sponsor: National Research Programme; Contract grant number: NRP 53, Musculoskeletal Health-Chronic Pain ; Contract grant sponsor: The Swiss National Science Foundation; Contract grant number: ; Contract grant sponsor: Von Hevesy Foundation; Contract grant sponsor: Inselspital Research Foundation. *Address reprint requests to: E.J.U., Department of Radiology, University Hospital Zurich, R amistr. 100, 8091 Zurich, Switzerland. erikajulbrich@googl .com Received November 14, 2011; Accepted June 28, DOI /jmri View this article online at wileyonlinelibrary.com. changes of CSAs over time at any of the three levels. There were no consistent significant correlations of CSA values with the clinical scores at all time points except with the body mass index. Conclusion: Our results do not support a major role of cervical muscle volume in the genesis of symptoms after whiplash injury. Key Words: whiplash injury; cervical muscles; muscle cross-sectional area; MRI; clinical scores J. Magn. Reson. Imaging 2012;36: VC 2012 Wiley Periodicals, Inc. SEVERAL ARGUMENTS SUPPORT a protective role of cervical muscles for symptomatic whiplash injury, such as, e.g., women with smaller muscles suffering more frequently from symptoms or symptoms being rarely observed in sport activities with trained athletes like rugby. In a previous study comparing 38 consecutive whiplash patients in the early post-acute setting and 38 age- and sex-matched controls, we could not find significant differences of cross-sectional areas (CSAs) in dorsal cervical extensor and sternocleidomastoid muscles using MRI (1), findings that did not support a protective role. These findings support the validity of postacute measurements (within 48 h) for appropriate baseline to compare with subsequent longterm changes. The role of cervical muscles in the longterm symptom evolution so far has not been evaluated. Therefore, in the present study, we first followed the same cervical muscle CSAs 3 and 6 months after the injury and then related the CSA acute and follow-up measurements to clinical outcome measures. We attempted to evaluate two hypotheses: (i) that cervical muscle hypotrophy would be evident after a 6- month follow-up and, (ii) that cervical muscle hypotrophy would be most significant in patients with persistent symptoms, probably related to pain or inactivity. MATERIALS AND METHODS This prospective study was approved by the local Ethical Committee. Written informed consent was obtained from all patients before study inclusion. VC 2012 Wiley Periodicals, Inc. 1413

2 1414 Ulbrich et al. Figure 1. Planning of the sagittal and transverse images of the cervical spine MRI. a: Planning of sagittal images on a coronal localizer image. b: Planning of two slabs of transverse images. The 20 transverse images per slab were planned on the sagittal sequence and were performed with two overlapping slabs which were perpendicular to the posterior surface of the vertebral body in the middle of the respective slab. Subjects Ninety acute symptomatic whiplash patients (48 females, 42 males) were consecutively recruited within a 3-year period from the emergency department and underwent MR examinations of the cervical spine within 48 h and 3 and 6 months after injury. Study eligibility was ascertained with standardized questionnaires. Whiplash injury was defined as an injury mechanism with acceleration/deceleration of the head relative to the trunk caused by a rear-end collision in a motor vehicle accident (MVA). An injury mechanism according to the above definition included a detailed history confirmed by police protocol. Additional inclusion criteria included age older than 18 years and injury-related symptoms consisting of at least neck pain (whiplash injury Grade I or II (Quebec Task force classification) (2). Patients were excluded if they had a fracture of the spine, direct head injury, previous surgery or neurosurgery to the head or spine, previous history of whiplash injury with persisting pain, musculoskeletal inflammatory disorders or other severe illness (with continuous pain or reduction of working ability), pre-existing head and neck pain, psychiatric disorders, drug or alcohol abuse, primary tumors and metastases of the head and neck, claustrophobia, presence of a cardiac pacemaker (and other implants preventing MR imaging), and pregnancy. During the study period, patients were treated by their family physicians, all received analgesics and physical treatments according to their symptoms. study protocol were used for this sub-study to measure muscle cross-sectional areas. Parameters for the sagittal TIRM sequence consisted of repetition time (TR) 4860 ms, echo time (TE) 28 ms, inversion time 150 ms, flip angle 180, turbo factor 7, total acquisition time 5:47. Two transversal TIRM sequences with TR 5110 ms, TE 28 ms, inversion time 150 ms, flip angle 180, turbo factor 7, and total acquisition time 4:48 each were combined to cover the region of interest. All images were acquired on a pixel matrix with a 210 mm 158 mm field of view (FOV) and a slice thickness of 3 mm. A dedicated neck coil and a spine array coil were used with the patient in the supine position. Saturation pulses placed over the upper airway region were used for reducing breathing and vessel-related artifacts. Sagittal and transverse slices of the cervical spine were obtained from the midpoint of the cerebellum through to the second thoracic vertebral level to include the entire cervical extensor musculature. The 19 sagittal images were centered on the spinal cord at the C4 level and planned as shown in Fig. 1a. The 20 transverse images per slab were planned on the sagittal sequence and were performed with two overlapping slabs which were perpendicular to the posterior surface of the vertebral body in the middle of the respective slab (Fig. 1b). Three slabs were used if required to cover the entire region. The tilt of the transverse images was thus adapted to the vertebral body orientation. Each transverse slab included 20 images. MRI Acquisition and Protocol MR images were obtained using a SONATA 1.5 Tesla (T) magnet (Siemens Medical Solutions, Erlangen, Germany). Sagittal and transverse STIR/TIRM (Short Tau Inversion Recovery/Turbo Inversion Recovery Magnitude) sequences obtained as part of the larger Imaging Analysis Muscle CSAs were taken as surrogate markers for muscle volumes (3,4). Two blinded specially trained raters (R.A., S.W.) independently performed the measurements on a Picture Archiving and Communication System (PACS, Philips Easy Vision PACS Viewing and

3 Cervical Muscle in Whiplash Patients 1415 Figure 2. Female whiplash patient, imaged in the acute setting, with the cross sectional area (CSA) measurements indicated for the different muscle groups at different levels. a: Transverse STIR/TIRM at C2 level: deep cervical extensor muscles on both sides (left) and total dorsal cervical extensor muscles on both sides (right). b: Transverse STIR/TIRM at C4 level: sternocleidomastoid muscles on both sides. c: Transverse STIR/TIRM at C5 level: deep cervical extensor muscles on both sides (left) and total dorsal cervical extensor muscles on both sides (right). d: Sagittal STIR/TIRM with line markers identifying the 3 levels (C2, C4, C5), where the measurements were performed. Reporting Workstation). The sternocleidomastoid and deep as well as total cervical extensor muscles were manually traced using irregular regions of interest (ROIs). Each side of the neck was evaluated separately. Measurements were obtained at the following levels: (i) C2 level: inferior oblique and rectus capitis posterior major muscles (C2deep_sum) as well as total dorsal cervical extensor muscles (C2tot_sum) (see Fig. 2a); (ii) C4 level: sternocleidomastoid muscles (C4sterno_sum) (see Fig. 2b); and (iii) C5 level: semispinalis cervicis plus multifidus plus interspinal and spinal muscles (C5deep_sum), as well as total dorsal cervical extensor muscles (C5tot_sum) (see Fig. 2c). The middle portion of each vertebral body (C2, C4, C5) localized in the sagittal sequence (see Fig. 2d) was the landmark used for measurement of the CSAs. The anatomical drawings illustrate the different muscles of the neck on C2 level (Fig. 3a) and C5 level (Fig. 3b). Clinical Characteristics All patients had a clinical examination and completed the following questionnaires in the acute setting, and at 3- and 6-month follow-up: EuroQuol 5D (5), specific Whiplash Disability Questionnaire (WDQ) (6), pain rating with a Visual Analog Scale (VAS). The following items were taken for follow-up analysis and correlation with CSAs: assessment of subjective health status (indicated by patients on a scale from 0 to 100, with 100 meaning best health state and 0 worst); Whiplash Disability Score (a sum score calculated on the basis of a standardized questionnaire [WDQ] with 13 questions each graded 0 to 10, with 0 meaning no symptoms and 130 maximum symptoms); neck pain intensity assessed with a Visual Analog Scale (VAS, 0 no pain, 100 maximum pain); range of motion of cervical spine (active head rotation, to the left and to the right added, in degrees). Chronic patients were defined as patients with symptoms (at

4 1416 Ulbrich et al. Figure 3. Anatomic axial drawing at C2 and C5 level illustrating the dorsal cervical muscles. a: C2 level: (a) inferior oblique, (b) rectus capitis posterior major, (c) semispinalis capitis, (d) longissimus capitis, (e) splenius, (f) trapezius, (g) sternocleidomastoid, (h) digastricus, (i) scalenus medius, (k) scalenus posterior, (l) longus colli, (m) longus capitis, (n) scalenus anterior. b: C5 level: (a) multifidus, (b) interspinalis, (c) spinalis, (d) semispinalis cervicis, (e) semispinalis capitis, (f) longissimus capitis, (g) longissimus cervicis, (h) iliocostalis cervicis, (i) splenius, (k) trapezius, (l) levator scapulae, (m) scalenus posterior, (n) scalenus medius, (o) scalenus anterior, (p) longus capitis, (q) longus colli, (r) sternocleidomastoideus. least comprising neck pain) persisting for 6 months or more (7). Statistical Analysis Intraclass correlation of CSAs between raters were calculated for all muscles and at all times of measurements (7). Data of both raters were pooled for subsequent analysis based on these results. Differences between base-line data and follow-up data at 3 and 6 months were calculated for CSAs, clinical outcomes and body mass index (BMI) and paired t-tests were used to test the hypotheses of zero differences. Differences between gender groups and chronic and cured patient groups were assessed using Wilcoxon rank tests. Linear associations between variables were analyzed with Pearson correlation coefficients. The level of significance was set at P < 0.05 throughout the study, and SAS 9.2 (SAS Institute Inc., Cary, NC) was used for all calculations. RESULTS A total of 90 whiplash patients with an average age of 37.1 years (mean age, 37.1; median, 36.3; SD, 14.8) were recruited (48 women, mean age: 36.5, median: 28.6, SD: 17.1; 42 men, mean age: 37.8, median: 38.3, SD: 11.6). There was a significant difference between women and men for age (P ¼ 0.045) and BMI (P < 0.001) without any change during the follow-up. CSA Measurements There was a large variability of measures typical for muscle CSA. However, an analysis for repeatability of individual muscle measurements showed high intraclass-correlation values (ICC between and 0.977). Pooled CSA values for left and right side of both raters for the measured muscles (C2deep_sum, C2tot_sum, C4sterno_sum, C5deep_sum, C5tot_sum) over time (0, 3, 6 months) are summarized in Table 1. There was no significant side difference of all measured CSA values at any time point. There were no significant differences in CSA values of any muscle at any level between baseline and 3 months of follow-up. With the exception of the sternocleidomastoid, there were no significant differences of CSA values between baseline and 6 months of follow-up. The CSA of the sternocleidomastoid was 2% larger at 6-month followup than at baseline. Taken together, we could not demonstrate any significant changes of CSAs over time which is demonstrated in Figure 4 illustrating the CSA differences for the different spine levels and patient groups (Fig. 4, Table 1). Women had significantly smaller average CSA values than men for all muscle areas at all time points (baseline, 3, 6 months). However, there was no gender influence on CSA evolution over time. Women (23.4, SD 4.33) also demonstrated a significantly lower mean value of BMI than men (25.2, SD 3.40) (P < ). Irrespective of gender the BMI remained almost constant over the 6 months (Table 2). There were significant correlations between BMI and CSA values of all measured muscle areas at any level and any time point. There were also highly significant correlations between the different CSA measurements (Pearson correlation coefficient between 0.62 and 0.90, P < ) with highly constant values over time supporting validity of measurements.

5 Table 1 Mean CSA Values (6SD) C2deep_sum a C2tot_sum a C4sterno_sum a C5deep_sum a C5tot_sum a 0 Month Men (154.96) (486.59) (222.27) (105.90) (732.52) Women (135.75) (355.22) (142.88) (88.37) (499.02) All (221.30) (777.54) (255.40) (169.47) ( ) 3 Months Men (171.55) (542.60) (210.28) (122.81) (622.62) Women (140.76) (449.99) (129.50) (96.88) (495.09) All (227.65) (810.98) (265.47) (178.10) (999.56) 6 Months Men (173.54) (563.83) (217.96) (117.79) (682.65) Women (123.92) (370.42) (144.82) (79.67) (424.02) All (219.67) (796.08) (269.40) (176.36) ( ) a sum means CSA of left- and right-sided muscles added (values in mm 2 ). Figure 4. Testing the difference of CSA values [in mm 2 ] between baseline and 3 months (diff_3_0) and baseline and 6 months (diff_6_0). Values are upper and lower confidence limits. Numbers of patient groups are given in the first graph. Definition of muscle areas: a: C2deep_sum inferior oblique muscle plus rectus capitis posterior major muscle at C2 level. b: C2tot_sum total dorsal cervical muscles at C2 level. c: C4sterno_sum sternocleidomastoid muscle at C4 level. d: C5deep_ sum semispinalis cervicis muscle plus multifidus muscle plus interspinal muscle and plus spinal muscle at C5 level. e: C5tot_sum total dorsal cervical muscles at C5 level

6 1418 Ulbrich et al. Table 2 Characteristics and Clinical Scores (6SD) BMI Health state a Whiplash score b Neck pain c Motion range d 0 Month Men (3.40) (13.93) (29.64) (22.01) (50.92) Women (4.33) (21.31) (30.66) (24.80) (43.85) All (4.01) (19.19) (30.48) (24.14) (47.07) 3 Months Men (3.30) (20.75) (27.46) (21.32) (30.38) Women (4.35) (19.37) (25.18) (25.79) (25.05) All (3.98) (19.93) (26.10) (24.33) (27.55) 6 Months Men (3.56) (17.77) (20.63) (20.30) (32.65) Women (4.20) (16.36) (23.44) (26.04) (34.95) All (4.03) (16.99) (22.11) (23.69) (33.79) a EuroQuol (100 is best). b Specific Whiplash Disability Questionnaire (0 is best, 130 is worst). c Visual Analog Scale, 100 is maximum pain intensity. d Neck rotation (sum of rotation to the left and to the right) (values in degrees). Clinical Outcomes: Health State (EuroQuol), Whiplash Disability Score (Whiplash Disability Questionnaire), Neck Pain Intensity, and Range of Motion Mean values of the four clinical scores are given in Table 2. There was a significant improvement of subjective health state (0.0014/<0.0001), whiplash score (<0.0001/<0.0001), neck pain intensity (<0.0001/ <0.0001), and cervical spine range of motion (<0.0001/<0.0001) in the time course (P values for comparison of 3- and 6-month findings with baseline each) without any gender differences. Correlations Between Clinical Outcomes and CSA Values Health state showed no correlation with any CSAs at any time point except for C2tot_sum at baseline. There were no correlations of Whiplash Score with any CSA values at any time point. Neck pain intensity correlated only with C2deep_ sum at baseline. The cervical range of motion again showed no significant correlation with any CSA value at any time point. Role of Chronic Whiplash-Associated Symptoms on CSA There were six patients with persistent symptoms (neck pain as a minimum) at 6 months of follow-up. Not unexpectedly, chronic patients were characterized by significantly poorer health status and higher whiplash disability scores at 6 months compared with the cured. As expected, significant differences were found also for neck pain (by definition, since pain defined the chronic state) at 3 months and at 6 months, but not for range of motion. Comparing the 6 symptomatic with the 84 asymptomatic patients, there were no significant differences of CSA values (Fig. 4). The variability of CSA measures within the chronic group was large due to the small sample (n ¼ 6). DISCUSSION Studies evaluating the role of cervical muscles regarding symptom evolution or developing chronic symptoms (whiplash-associated disorder, WAD) after whiplash injuries are rare (8). To our knowledge, the course of cervical muscle volumes and their correlation with symptoms in patients suffering from whiplash injuries during a 6-month follow-up has not yet been evaluated. We hypothesized (i) that cervical muscle hypotrophy if any would be evident after a 6-month follow-up, and (ii) that cervical muscle hypotrophy would be most significant in patients with persistent symptoms probably related to pain or inactivity. However, our results did not confirm these two hypotheses: (i) there were no changes of cervical muscle CSA during the first 6 months in patients with acute symptomatic whiplash injury, and (ii) there was no consistent correlation of muscle CSA with any of the clinical outcome variables. Several mechanisms may cause different changes of muscle volumes after whiplash injury: (i) Following a MVA traumatic muscle edema may cause swelling which is expected to decrease during follow-up. Our results, however, do not reflect such a mechanism unless traumatic edema occurred only after the first MR examination (after 48 hours). (ii) Inactivity due to persistent pain in chronic whiplash patients may lead to structural changes of cervical spine muscles. Elliot et al (9) demonstrated significantly larger amounts of fatty infiltration in all extensor muscles of chronic whiplash patients compared with healthy controls. In this study, only female patients were included and examined at a single (chronic) time point. Again, we could not confirm a decrease of muscle volumes over time as there usually is associated fatty infiltration. Because we did not measure fat content in our study,

7 Cervical Muscle in Whiplash Patients 1419 we cannot comment on changes of muscular tissue constituents over time (10). A significant correlation between CSA values and BMI was found, which might be explained by higher fat content. Elliot et al (9), however, did not report any influences between BMI and the fat to muscle ratio as measured on T1- weighted spin-echo images. We have previously found an increase of CSAs with older age. One possible explanation for this is increasing fat content of muscles in elderly persons. However, Elliot et al (10) did not find any correlation between fatty infiltration in the upper cervical extensor muscles and age or inactivity in healthy controls. Possibly, differences in study populations explain this. In another study (4), Elliot demonstrated that cervical extensor muscle volumes in healthy controls did not correlate with age except for the rectus capitis posterior and splenius capitis muscles. Therefore, in whiplash patients, the various changes (muscle edema, fatty infiltration, and muscle hypotrophy) probably affect not all muscle groups the same way and at different time points. Furthermore, they may cancel out each other depending on the method of muscle investigation. The large variability of measures known for muscle CSA and also present in neck muscles, certainly hampers the detection of small changes. In the present study, women had significant smaller muscle volumes than men for all measured muscles at all levels analogous to the findings of Matsumoto et al (11) and Okada et al (12). The fact that women suffer more often from whiplash injury (13) is well known. The smaller muscle volumes may therefore be a predisposing factor for women to suffer from symptomatic whiplash injuries after a MVA. However, there was no difference in the course over the 6 months between men and women. All clinical scores (subjective health state, whiplash disability score, neck pain intensity, range of motion of cervical spine) significantly improved over time in women as well as in men. We could not find any consistent correlation of CSA values with the clinical scores. There was also no significant difference of the CSA values between the 6 patients with persistent symptoms and those 84 who recovered. These last two findings are analogous to those of Matsumoto et al in a small patient group with a 10-year follow-up (14). However, the chronic whiplash patient group was too small for robust statistical comparisons, and the large variability of CSA measures in this group (Fig. 3) is explained by the small sample size (n ¼ 6). With a prospective strict study design, it is difficult to get larger patient samples for a long-term follow-up. Our 7% of patients with persistent symptoms is in the lower range of the wide range reported in the literature (15,16). Furthermore, there is no uniform definition of the so-called chronic or late whiplash syndrome. The large CSA variability found in the chronic group most probably is due to the small subject number, but might also indicate a more complex pathology. Kristjansson (17) found a significant reduction in CSA of the cervical multifidus at the C4 level in (chronic) whiplash patients compared with healthy controls. To the contrary, Elliot et al (3,4) reported significant larger CSA of the cervical multifidus muscles at all spinal levels (C3 C7) in chronic whiplash patients when compared with healthy controls. However, the cervical semispinal muscles were smaller in chronic whiplash patients than in controls. Elliot et al (3) showed only weak relationships between age, BMI, duration of symptoms, NDI (neck disability index) scores, and volumes of some muscles. Thus, the different muscles or muscle groups may have variable influences on symptoms. There is no consensus regarding muscle volumes in chronic whiplash patients over time, because results presented by different authors vary considerably. Furthermore, the optimal MR imaging protocol for muscle evaluation has still to be defined. In our study, the highly significant correlations between the different CSA measurements with highly constant values over time support validity of measurements performed. This study has limitations, mainly concerning methodology: (i) Especially at the C2 level, it is not possible to image certain muscles in a true cross-sectional way due to their orientation. Consequently, theoretically a small change in head position might introduce a change in muscle area during follow-up measurements. (ii) Some patients were not positioned with a head rotation of precisely 0, which would have been difficult especially in the acute phase. This may have caused slight left-to-right asymmetries. (iii) The small number of patients with long-term persistent symptoms does not allow definite conclusions with respect to late whiplash syndrome. (iv) We cannot exclude that the interval from injury to evaluation (6 months) is too short. (v) Our imaging protocol did not include fat measurements. In conclusion, our results to date do not support a major role of cervical muscle volumes in the genesis of symptoms after whiplash injury. ACKNOWLEDGMENTS Many thanks to our study coordinators (Nathalie Sollberger, Sandra Stiegler); to our MRI technicians (Jeanette Pfeiffer, Karolina Dobrowolska, Verena Beutler, and Karin Zwygart); the computer technicians of the Department of Radiology, University Hospital, Inselspital, Bern; and Professors Vincent Im Hof, Peter Vock, and Juergen Triller. Origin of work: Department of Diagnostic, Interventional and Pediatric Radiology, Inselspital, University Hospital, and University of Berne, Freiburgstrasse, CH-3010 Bern, Switzerland. This project was supported by the National Research Programme of the Swiss National Science Foundation. REFERENCES 1. Ulbrich EJ, Anderson SE, Busato A, et al. Cervical muscle area measurements in acute whiplash patients and controls. J Magn Reson Imaging 2011;33: Spitzer WO, Skovron ML, Salmi LR, et al. Scientific monograph of the Quebec Task Force on Whiplash-Associated Disorders: redefining whiplash and its management. Spine (Phila Pa 1976) 1995;20(Suppl):1S 73S.

8 1420 Ulbrich et al. 3. Elliott J, Jull G, Noteboom JT, Galloway G. MRI study of the cross-sectional area for the cervical extensor musculature in patients with persistent whiplash associated disorders (WAD). Man Ther 2008;13: Elliott JM, Jull GA, Noteboom JT, Durbridge GL, Gibbon WW. Magnetic resonance imaging study of cross-sectional area of the cervical extensor musculature in an asymptomatic cohort. Clin Anat 2007;20: Deyo RA, Battie M, Beurskens AJ, et al. Outcome measures for low back pain research. A proposal for standardized use. Spine (Phila Pa 1976) 1998;23: Pinfold M, Niere KR, O Leary EF, Hoving JL, Green S, Buchbinder R. Validity and internal consistency of a whiplash-specific disability measure. Spine (Phila Pa 1976) 2004;29: Fleiss J. The design and analysis of clinical experiments. New York: John Wiley and Sons; Carroll LJ, Hogg-Johnson S, van der Velde G, et al. Course and prognostic factors for neck pain in the general population: results of the Bone and Joint Decade Task Force on Neck Pain and Its Associated Disorders. Spine (Phila Pa 1976) 2008; 33(Suppl):S75 S Elliott J, Jull G, Noteboom JT, Darnell R, Galloway G, Gibbon WW. Fatty infiltration in the cervical extensor muscles in persistent whiplash-associated disorders: a magnetic resonance imaging analysis. Spine (Phila Pa 1976) 2006;31:E847 E Elliott JM, Galloway GJ, Jull GA, Noteboom JT, Centeno CJ, Gibbon WW. Magnetic resonance imaging analysis of the upper cervical spine extensor musculature in an asymptomatic cohort: an index of fat within muscle. Clin Radiol 2005;60: Matsumoto M, Okada E, Ichihara D, et al. Changes in the crosssectional area of deep posterior extensor muscles of the cervical spine after anterior decompression and fusion: 10-year follow-up study using MRI. Eur Spine J 2012;21: Okada E, Matsumoto M, Ichihara D, et al. Cross-sectional area of posterior extensor muscles of the cervical spine in asymptomatic subjects: a 10-year longitudinal magnetic resonance imaging study. Eur Spine J 2011;20: Dolinis J. Risk factors for whiplash in drivers: a cohort study of rear end traffic crashes. Injury 1997;28: Matsumoto M, Ichihara D, Okada E, et al. Cross-sectional area of the posterior extensor muscles of the cervical spine in whiplash injury patients versus healthy volunteers - 10 year follow-up MR study. Injury 2012;43: Poorbaugh K, Brismee JM, Phelps V, Sizer PS Jr. Late whiplash syndrome: a clinical science approach to evidencebased diagnosis and management. Pain Pract 2008;8:65 87; quiz Scholten-Peeters GG, Verhagen AP, Bekkering GE, et al. Prognostic factors of whiplash-associated disorders: a systematic review of prospective cohort studies. Pain 2003;104: Kristjansson E. Reliability of ultrasonography for the cervical multifidus muscle in asymptomatic and symptomatic subjects. Man Ther 2004;9:83 88.