PATELLOFEMORAL PAIN syndrome (PFPS) describes

183 Delayed Onset of Electromyographic Activity of Vastus Medialis Obliquus Relative to Vastus Lateralis in Subjects With Patellofemoral Pain Syndrome Sallie M. Cowan, GradDip, Kim L. Bennell, PhD, Paul W. Hodges, PhD, Kay M. Crossley, GradDip, Jenny McConnell, GradDip, MBiomedEng ABSTRACT. Cowan SM, Bennell KL, Hodges PW, Crossley KM, McConnell J. Delayed onset of electromyographic activity of vastus medialis obliquus relative to vastus lateralis in subjects with patellofemoral pain syndrome. Arch Phys Med Rehabil 2001;82:183-9. Objective: To determine whether electromyographic (EMG) onsets of vastus medialis obliquus (VMO) and vastus lateralis (VL) are altered in the presence of patellofemoral pain syndrome (PFPS) during the functional task of stair stepping. Design: Cross-sectional. Setting: University laboratory. Patients: Thirty-three subjects with PFPS and 33 asymptomatic controls. Interventions: Subjects ascended and descended a set of stairs 2 steps, each 20-cm high at usual stair-stepping pace. EMG readings of VMO and VL taken on middle stair during step up (concentric contraction) and step down (eccentric contraction). Main Outcome Measures: Relative difference in onset of surface EMG activity of VMO compared with VL during a stair-stepping task. EMG onsets were determined by using a computer algorithm and were verified visually. Results: In the PFPS population, the EMG onset of VL occurred before that of VMO in both the step up and step down phases of the stair-stepping task (p.05). In contrast, no such differences occurred in the onsets of EMG activity of VMO and VL in either phase of the task for the control subjects. Conclusion: This finding supports the hypothesized relationship between changes in the timing of activity of the vastimuscles and PFPS. This finding provides theoretical rationale to support physiotherapy treatment commonly used in the management of PFPS. Key Words: Patellofemoral pain syndrome; Knee; Muscles; Electromyography; Rehabilitation. 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation From the School of Physiotherapy, University of Melbourne, Melbourne (Cowan, Bennell, Crossley); Prince of Wales Medical Research Institute, Sydney (Hodges); and McConnell and Clements Physiotherapy, Sydney (McConnell), Australia. Accepted in revised form May 9, 2000. Supported by grants from the Physiotherapy Research Foundation and the Victorian Branch of the Australian Physiotherapy Association. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the author(s) is/are associated. Reprints requests to Sallie M. Cowan, School of Physiotherapy, University of Melbourne, 200 Berkeley St, Carlton, Vic, 3010, Australia, e-mail: s.cowan@ pgrad.unimelb.edu.au. 0003-9993/01/8202-6046$35.00/0 doi:10.1053/apmr.2001.19022 PATELLOFEMORAL PAIN syndrome (PFPS) describes anterior or retropatellar knee pain in the absence of other pathology. Clinically, the condition presents as diffuse anterior or retropatellar knee pain exacerbated by activities such as stair climbing, prolonged sitting, squatting, and kneeling. 1 PFPS is a common complaint in the sporting and general populations especially in which repetitive lower limb loading is involved. 2,3 Although the development of PFPS is multifactorial, abnormal lateral tracking of the patella has been proposed as a contributing factor 4 ; this may increase patellofemoral contact pressure and precipitate pathology in the patellofemoral articular cartilage. 5,6 One proposed mechanism for abnormal patellar tracking is an imbalance in the activity of the vastus medialis obliquus (VMO) relative to the vastus lateralis (VL). 5,7 This could be caused either by a reduction in the force-producing capabilities of the VMO 8 or altered temporal control of VMO and VL activity in PFPS sufferers. 7 Altered onset of the VMO may be of particular importance because in the asymptomatic population it has been hypothesized that VMO must be activated earlier than VL to optimally track the patella due to VMO s smaller cross-sectional area and VL s predominantly laterally directed force. 9 Controversy exists in the literature as to the normal relationship between the timing of electromyographic (EMG) activity of the VMO and VL, and whether this differs in the population with PFPS. 7,9-15 This controversy may be caused by a number of factors. First, studies have used different methods to determine EMG onset 7,9-15 and have often not expressed such onset of the vastimuscles directly, instead representing it as a percentage of the gait cycle 15 or in relation to force changes. 14 Second, a number of studies have investigated reflex tasks that have questionable clinical significance. 7,10,11 Last, studies have often had small sample sizes that have limited the power of their results. 11,13 Thus, this cross-sectional study sought to compare the EMG onset of the VMO and VL directly during the functional task of stair stepping in a group of subjects with PFPS and an asymptomatic population. METHODS Subjects Thirty-three subjects (11 men, 22 women) diagnosed with PFPS on the basis of clinical examination by an experienced musculoskeletal physiotherapist, and 33 asymptomatic controls (13 men, 20 women) were recruited for the study. The inclusion and exclusion criteria were based on those used in other PFPS studies. 16-19 Subjects in the PFPS group were included if they had anterior or retropatellar knee pain reported on at least 2 of the following activities: prolonged sitting, ascending or descending stairs, squatting, running, kneeling, and hopping/ jumping. In addition they were included if they had pain on patella palpation, symptoms for at least 1 month, an average pain level of 3cm on a 10-cm visual analog scale (VAS), and an insidious onset of symptoms unrelated to a traumatic inci-

184 ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan dent. All subjects were aged 40 years or younger, to reduce the likelihood of osteoarthritic changes in the patellofemoral joint. Subjects were excluded if they had signs or symptoms of other, even coexisting, pathologies. The exclusion criteria were a recent history (within 3mo) of knee surgery; a history of patellar dislocation or subluxation; or clinical evidence of meniscal lesion, ligamentous instability, traction apophysitis around the patellofemoral complex, patellar tendon pathology, chondral damage, osteoarthritis, or spinal referred pain. The pain characteristics of the PFPS subjects are presented in table 1. The asymptomatic control subjects were recruited from the University of Melbourne School of Physiotherapy. Subjects were examined by an experienced musculoskeletal physiotherapist and were excluded if they had any history of lower limb pathology or other disorder that might interfere with the kinetics or kinematics of knee motion. The mean standard deviation age, height, weight, and body mass index (BMI) of the PFPS subjects were 27.0 8.1yr, 171.1 9.3cm, 69.1 15.9kg, and.23.04kg/m 2, respectively. The mean age, height, weight, and BMI of the control subjects were 23.6 4.9yr, 169.8 11.9cm, 64.6 10.9kg, and.22.02kg/m 2, respectively. There were no significant differences between groups for these variables assessed by using an independent t test. The study was approved by the University of Melbourne Human Research Ethics Committee. All subjects provided written informed consent. EMG Recordings EMG activity of the VMO and VL was recorded by using surface electrodes. Silver/silver chloride electrodes a were placed over the muscle bellies of the VMO and VL with an interelectrode distance of 22mm (fig 1). The electrode for the VMO was placed over the muscle belly approximately 4cm superior to and 3cm medial to the superomedial patella border, and orientated 55 to the vertical. The electrode for the VL was placed 10cm superior and 6 to 8cm lateral to the superior border of the patella, and orientated 15 to the vertical. 20,21 The ground electrode was placed over the tibial tubercle. Before electrode placement, the skin was shaved, swabbed with alcohol, and gently abraded with sandpaper to reduce the electrical impedance to less than 5K. Movement Analysis Movement of the lower limb in the sagittal plane was measured by using a PEAK movement analysis system b to identify the concentric and eccentric phases of the stair-stepping task and the time taken to complete the phase. Reflective skin markers b were placed over bony landmarks as described by Tully and Stillman. 22 The landmarks were the lateral malleolus, the anterior surface of the head of the fibula, the iliotibial band at the level of the superior border of the patella, and the lateral Table 1: PFPS Group Pain and Disability Characteristics Characteristic Mean SD Range Time since onset of symptoms (mo) 42.2 49.9 1 144 Worst pain in past weekon descending 5.4 2.4 1 10 stairs (VAS) (cm) Worst pain in past weekon ascending 5.4 2.5 0 10 stairs (VAS) (cm) Worst pain in past week(vas) (cm) 7.1 1.6 4 10 Average pain in past week(vas) (cm) 4.3 1.2 2 7 Abbreviations: SD, standard deviation. Fig 1. Surface EMG electrode placement for the VMO and VL. thigh at the junction of the proximal 1 3 and distal 2 3 of a line joining the tip of the greater trochanter to the midpoint of the lateral knee joint line. Movement data were sampled at 50Hz by using a single camera c placed perpendicular to the center of the lateral side of the stair apparatus. Stair Apparatus The dimensions of the stairs were based on a previous study 21 and consisted of a 60-cm platform with 2 steps 20-cm high on both sides. The stairs were placed in the center of a 5-meter walkway. Procedure Subjects stood 1.8 meters from the lower step apparatus and, when instructed, ascended and descended the stairs at a rate of 96 steps per minute as paced by an external metronome. d This rate approximates usual stair-stepping pace, was identical to that used previously, 21 and was aimed at increasing the repeatability of the stair-stepping task because it is not known if differing walking speeds affect muscle onset times. Subjects completed at least 5 practice trials (each trial involved a single ascent and descent of the stair apparatus) to ensure that they were able to step in time with the metronome and were able to contact the middle step with the leg to be tested. Recordings of EMG e activity of the VMO and VL were made during the stance phase on the middle stair during ascent (concentric contraction) and descent (eccentric contraction), for 5 consecutive trials. During the stair-stepping task, the PFPS subjects had an average pain measured on a 10-cm VAS of 2.6 2cm (range, 0 6cm), whereas the control subjects had no pain. EMG data were preamplified (10 times) distal to the surface electrodes, band-pass filtered between 20 to 500Hz, sampled at 1000Hz and 12-bit A-D converted. e Data Analysis The EMG data were full-wave rectified and low-pass filtered at 50Hz. A computer algorithm f was used to identify the onset of EMG activity of each of the muscles. The algorithm identified the point at which the EMG signal deviated by more than 3 standard deviations (SDs) for a minimum of 25ms, above the

ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan 185 baseline level (averaged over 200ms before the commencement of the trial). The rectified unfiltered EMG data were visually checked to verify the onsets identified by the computer. The sampling rate of the EMG allowed a resolution of 1ms. Before data analysis, 50 traces were randomly selected in which the onset was not obscured by movement artefact or noise, and a number of different algorithms were compared with visually identified EMG onsets. The 25ms/3SD-combination was found to deviate least from the visual onset identified and, thus, to be the most accurate for the data analyzed (r.99, y-intercept.80ms, p.001). EMG onsets were identified from individual trials and averaged over the 5 repetitions. The relative difference in the time of onset of EMG activity of the VMO and VL was quantified during concentric and eccentric contraction by subtracting the EMG onset of the VMO from that of the VL. Reliability The reliability of the determination of EMG onset timing difference of the VMO and VL in the stair-stepping task was examined in 10 healthy subjects tested on 2 occasions 1 week apart. 23 The reliability was found to be excellent in both the concentric (intraclass correlation coefficient [ICC] model 3 for test-retest and averaged over 5 trials (ICC 3,5.91) and eccentric (ICC 3,5.96) phases of stair stepping. The standard error of measurement was found to be low at 6.22ms (95% confidence interval [CI] 12.20ms) for the concentric and 5.90ms (95% CI 11.50ms) for the eccentric phases of stair stepping. Statistical Analysis All data were analyzed by using StatView SE Graphics Software. h The data were assessed for normality by calculating values for kurtosis and skewness. A 2-way analysis of variance was used to determine whether there was a difference in the EMG onset timing difference value between the PFPS and control groups and between contraction types. An independent 1-group t test was used to determine if the concentric and eccentric onset timing difference differed significantly from 0 in the PFPS group and the control group (ie, whether the VMO and VL onset occur simultaneously). The PFPS and control groups were then divided into 3 subgroups in which the EMG onset of (1) VMO preceded VL by more than 10ms, (2) VMO followed up VL by more than 10ms, and (3) those in which the difference in EMG onset of VMO compared with VL was less than 10ms. The chi-square statistic was used to compare these 3 groups in the PFPS and control subjects. RESULTS When subjects with no history of PFPS either stepped up onto the step (concentric task) or down from the step (eccentric task) the EMG onsets of the VMO and VL were almost synchronous (figs 2A, 3A). In contrast, when the identical tasks were performed by the subjects with PFPS, the onset of VL preceded that of VMO (figs 2B, 3B). The means and SDs of the concentric and eccentric EMG onset timing differences in the PFPS and control subjects are Fig 2. EMG data of a representative subject from the (A) control and (B) PFPS group in the concentric phase of the stair-stepping task. Note in the control subject the onset of the VMO and VL occur at approximately the same time, whereas in the subject with PFPS the onset of the VL occurs before VMO.

186 ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan Fig 3. EMG data of a representative subject from the (A) control and (B) PFPS group during the eccentric phase of the stair-stepping task. Note in the control subject the onset of the VMO and VL occur almost synchronously, whereas in the subject with PFPS the onset of the VL occurs before VMO. presented in figure 4. Although there was large variation in individual onset times, there was a difference in the onset of VMO relative to VL between the PFPS and control groups during both the eccentric and concentric tasks (F 1,64 11.64, p.001). There was no difference between the eccentric and concentric tasks for either group (F 1,64 0.00, p.5) and no interaction effect between the tasks or groups (F 1,64 1.11, p.5). There was no difference in the time taken to complete the concentric or eccentric phases of the stair-stepping task (t 64 1.61, p.05; t 64 0.75, p.05, respectively). When the PFPS subjects stepped up onto or down from the step there was a timing difference between the onsets of the VMO and VL (fig 4). One group independent t tests showed that in the PFPS subjects the EMG onset for the VMO occurred after that of VL in both the concentric and eccentric phases of stair stepping (t 32 3.13, p.005; t 32 4.58, p.005). In contrast, in the control subjects there were no differences in the EMG onsets for the VMO and VL in either phase of the stair-stepping task (t 32 0.16, p.05, t 32.61, p.05, respectively). Figure 5 shows the percentage of subjects in each group in which the onset of the VL preceded, followed, or occurred at the same time as VMO. In the PFPS group, the onset of VL preceded that of VMO by more than 10ms in most subjects. However, in a proportion of subjects, the onset of the VMO preceded that of VL by more than 10ms (12% in both the concentric and eccentric phases). In the control subjects, the order of EMG onset of VMO and VL varied in relatively even proportions. The chi-square statistic showed that in both phases of stair stepping the onset of VL was more likely to precede VMO in the PFPS subjects than in the control subjects ( 2 2 10.36, p.05; 2 2 19.65, p.05). DISCUSSION There are 2 main findings from our study. First, the results show a difference in the EMG onset of the VMO relative to that of the VL in subjects with PFPS compared with asymptomatic controls. Second, there was large individual subject variation for onsets within each group. These findings have important clinical implications for the possible role of timing differences between the VMO and VL in the cause of PFPS. The EMG Onset of VL Occurred Before VMO in Patients With PFPS The mean difference in EMG onset of the VMO compared with the VL in the PFPS subjects was 15.80ms in the concentric and 19.39ms in the eccentric phases of stair stepping. These differences are several orders of magnitude larger than previously shown differences of between 0.5ms and 5.6ms in other tasks 7,9,10 and may be explained by differences in task characteristics. We used a functional task, whereas previous studies

ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan 187 Fig 4. Mean difference in EMG onsets of the VMO and VL for the PFPS and control groups in the concentric and eccentric phases of the stair-stepping task. The midpoint of each line indicates the mean EMG onset timing difference. The SDof the mean is also indicated. Note in the control subjects the average EMG onset of the VMO and VL occurs almost at the same time in both phases of stair stepping. In contrast, the EMG onset of the VMO occurs after that of the VL in both the concentric and eccentric phases of the stair-stepping task in the subjects with PFPS. that have reported significant differences investigated reflex activity or knee extension. Functionally it is not known whether a difference of 15 to 20ms has a significant biomechanic effect on the functioning of the patellofemoral joint. Further research is required to evaluate the specific clinical effects of a deficit of this magnitude and to determine whether changes of this magnitude would be clinically identifiable. The finding that the EMG onset of the VMO occurred after the VL in subjects with PFPS contrasts with previous results that found no difference between these onsets. 11,13-15 The failure to show such a difference might have been due to a number of factors, including insufficient subject numbers, the tasks tested, and the methods used to determine and represent EMG onset of the vastimuscles. The determination of EMG onset may be of particular importance. The use of a computer algorithm to determine muscle onset has been reported to improve the reliability of muscle onset determination and to reduce the need for experimenter experience. 24,25 However, the time identified by a computer algorithm as the onset of EMG activity is influenced by several factors, including the amount of EMG background activity, the rate of EMG increase, and the presence of artefacts. 25 Thus, to ensure the validity of computerderived EMG onsets, it is recommended that each trace be inspected visually to ensure movement artefact or other interference is not incorrectly identified as a muscle onset. 25 In previous studies, only one used a combination of a computer algorithm and visual analysis to identify EMG onset. 11 Others failed to define the technique used 14 or used only a computer algorithm. 15 The study by Powers et al 15 defined muscle onset as the time at which the amplitude exceeded 5% of a maximal voluntary quadriceps contraction. The results were not checked visually and, thus, artefacts may have affected the computergenerated muscle onsets. Additionally the use of a maximal isometric contraction is problematic in PFPS subjects owing to the possible presence of pain that may affect the patients ability to perform a maximal contraction. The finding that on average the onset of EMG activity of the VL occurred before that of VMO in subjects with PFPS has important clinical implications. First, the results provide evidence for 1 proposed factor in the cause of PFPS. For many years it has been argued that abnormal patella tracking might be a cause of PFPS. 4 Our findings provide evidence of an imbalance in the timing of activation of the VMO and VL that has been argued to be 1 factor that may lead to a change in the tracking of the patella. 5,7 Although further validation would be required to confirm this link between timing and changes to patella mechanics, the results are consistent with the presence of a change in motor control rather than a strength change. Deficits in the timing of muscle activity have been identified in other musculoskeletal conditions such as lower back pain in which the EMG activity of transversus abdominis has been shown to be delayed compared with a control group. 26 Second, the results provide a theoretical rationale to support physiotherapy treatment commonly used in the treatment of PFPS. One strategy used for the management of PFPS is based on a program developed by McConnell. 27 One of the underlying aims of the program is to improve patellar tracking through specific retraining of VMO timing in functional activities. Our research team is currently investigating whether this treatment is effective in restoring the temporal control of the vastimuscles in functional tasks in people with PFPS. An altered latency of VMO activity in a reflex task has been reported as an intrinsic risk factor for the development of PFPS. 28 However, given the cross-sectional design of our study, we cannot determine whether the deficit in timing of VMO EMG activity was present before the onset of pain, and, thus, a causative factor, or, conversely, if the timing deficit occurred as a result of the presence of pain and dysfunction associated with the condition. Knowledge of this temporal sequence has implications for the prevention and treatment of the disorder. Further research is needed to determine whether the temporal changes in the EMG activity of VL compared with VMO predispose individuals to PFPS, and, thus, determine whether individuals could be screened and preventative programs implemented. Fig 5. Percentage of subjects in which the EMG onset of the VMO preceded, followed, or occurred at the same time as the VL in the PFPS and control subjects. In both phases of stair stepping, the onset of the VL was more likely to precede the VMO for subjects in the PFPS group than for the control group. ( ), VMO onset > 10ms before VL; ( ), VMO onset occurs within 10ms of VL; ( ), VL onset occurs > 10ms before VMO.

188 ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan Simultaneous EMG Onset of Vastii in the Asymptomatic Population The coincident recruitment of the VMO and VL in the asymptomatic population is consistent with previous results 11,15 and concurs with the opinion that the onset of activity of the vastimuscles is relatively balanced in people with no history of patellofemoral pain. Although the VMO and VL may have antagonistic actions for the mediolateral control of the patella, ultimately, the recruitment of the VMO and VL must be appropriately timed for efficient biomechanic function of the knee such that they can act synergistically with each other and the rest of the quadriceps in any functional task. 29 It has been postulated, however, that the VMO should be active before the VL to maintain patellar position in the asymptomatic population. 9 This relation was hypothesized because of the larger cross-sectional area 30 and velocity-producing properties 31 of the VL, which are predicted to result in a dominance of laterally directed patellar motion. Yet, the VMO does have a mechanical advantage over the VL by virtue of the obliquity of its fiber orientation. The orientation of the VMO muscle fibers have been extensively reported to be 50 to 55 medial to the shaft of the femur in the frontal plane, whereas the VL is aligned 12 to 15 lateral in the frontal plane. 30,32-38 This mechanical advantage may be sufficient to balance the superior force and velocity-generating capacities of VL. Alternatively, the magnitude of the VMO s activation may be larger than that of the VL. However, this factor was not investigated in the present study. EMG Onset Timing Differences Vary in Both Populations Although a functional task will have inherently larger variation than a reflex task, the wide variation in the EMG onsets of the VMO and VL in both groups is worthy of closer examination. It is possible that in the control subjects, in which the EMG onset of the VMO occurred after VL, there is a risk of developing PFPS in the future. Perhaps owing to the broad range of factors that may contribute to PFPS, not every sufferer presents with onset timing differences between the 2 vastimuscles. For example, PFPS subgroups with a particular biomechanic or anatomic predisposition may present with timing differences. 39 Additionally, in some individuals, early activation of the VMO may be insufficient to maintain patella position, because the magnitude of this activation may be decreased. It also is important to note that the timing of muscle onset is only 1 component of motor control. Other factors, such as the rate of increase of muscle activation and the amount of activation, may also be important. CONCLUSION It is common in clinical practice to encounter patients with severely malaligned patellae who have no pain and others with seemingly less compromised alignment but severe pain. It is apparent that a number of factors must be present for PFPS to occur. Thus, there is a need for further multifactorial research to focus on defining subgroups of PFPS to ensure not only that researchers are investigating similar patient populations but also, more important, to enable treatment programs to be tailored to the individual needs of patients suffering with PFPS. References 1. Brukner P, Khan K. Clinical sports medicine. Sydney: McGraw- Hill; 1993. p 376. 2. DeHaven KE, Dolan WA, Mayer PJ. Chondromalacia in athletes. Am J Sports Med 1979;7:5-11. 3. Holmes SW, Clancy WG. Clinical classification of patellofemoral pain and dysfunction. J Orthop Sports Phys Ther 1998;28:299-306. 4. Fulkerson JP, Shea KP. Current concepts review: disorder of patellofemoral alignment. J Bone Joint Surg Am 1990;72:1424-9. 5. Insall J. Current concepts review: patellar pain. J Bone Joint Surg Am 1982;64:147-52. 6. Huberti HH, Hayes WC, Stone JL, Shybut GT. Force ratios in the quadriceps tendon and ligamentum patellae. J Orthop Res 1984; 2:49-54. 7. Voight M, Weider D. Comparative reflex response times of the vastus medialis and the vastus lateralis in normal subjects and subjects with extensor mechanism dysfunction. Am J Sports Med 1991;10:131-7. 8. Ahmed AM, Burke DL, Hyder A. Force analysis of the patellar mechanism. J Orthop Res 1987;5:69-85. 9. Grabiner MD, Koh TJ, Draganich LF. Neuromechanics of the patellofemoral joint. Med Sci Sports Exerc 1994;26:10-21. 10. Witvrouw E, Sneyers C, Lysens R, Victor J, Bellemans M. Reflex response times of vastus medialis oblique and vastus lateralis in normal subjects with patellofemoral pain syndrome. J Orthop Sports Phys Ther 1996;24:160-5. 11. Karst GM, Willett GM. Onset timing of electromyographic activity in the vastus medialis oblique and vastus lateralis muscles in subjects with and without patellofemoral pain syndrome. Phys Ther 1995;75:813-23. 12. MacIntyre DL, Robertson GE. Quadriceps muscle activity in women runners with and without patellofemoral pain syndrome. Arch Phys Med Rehabil 1992;73:10-4. 13. Sheehy P, Burdett RC, Irrgang JJ, VanSwearingen J. An electromyographic study of vastus medialis oblique and vastus lateralis activity while ascending and descending steps. J Orthop Sports Phys Ther 1998;27:423-9. 14. Morrish GM, Woledge RC. A comparison of the activation of muscles moving the patella in normal subjects and in patients with chronic patellofemoral problems. Scand J Rehabil Med 1997;29: 43-8. 15. Powers CM, Landel RF, Perry J. Timing and intensity of vastus muscle activity during functional activities in subjects with and without patellofemoral pain. Phys Ther 1996;76:946-55. 16. Kowall MG, Kolk G, Nuber GW, Cassisi JE, Stern SH. Patellar taping in the treatment of patellofemoral pain. A prospective randomized study. Am J Sports Med 1996;24:61-6. 17. Larsen B, Adreasen E, Urfer A, Mickelson MR, Newhouse KE. Patellar taping: a radiographic examination of the medial glide technique. Am J Sports Med 1995;23:465-71. 18. Kannus P, Natri A, Nittymaki S, Jarvinen M. Effect of intraarticular glycosaminoglycan polysulfate treatment on patellofemoral pain syndrome. Arthritis Rheum 1992;35:1053-61. 19. McMullen W, Roncarati A, Koval P. Static and isokinetic treatments of chrondromalacia patella: a comparative investigation. J Orthop Sports Phys Ther 1990;12:256-66. 20. Basmajian J, Blumenstein R. Electrode placement in EMG biofeedback. Baltimore: Williams & Wilkins; 1980. 21. Gilleard W, McConnell J, Parsons D. The effect of patellar taping on the onset of vastus medialis obliquus and vastus lateralis muscle activity in persons with patellofemoral pain. Phys Ther 1998;78:25-31. 22. Tully EA, Stillman BL. Computer-aided video analysis of vertebrofemoral motion during toe touching in healthy subjects. Arch Phys Med Rehabil 1997;78:759-66. 23. Cowan SM, Bennell KL, Hodges PW. The test-retest reliability of the onset of concentric and eccentric vastus medialis obliquus and vastus lateralis electromyographic activity in a stair stepping task. Phys Ther Sport. In press. 24. Di Fabio RP, Badke MB, Duncan PW. Adapting human postural reflexes following localized cerebrovascular lesion: analysis of bilateral long tendon responses. Brain Res 1986;363:257-64. 25. 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: 511-9.

ELECTROMYOGRAPHIC ONSET OF VMO RELATIVE TO VL IN PFPS, Cowan 189 26. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spine associated with low back pain. Spine 1996;21: 2640-50. 27. McConnell J. The management of chondromalacia patellae: a long term solution. Aust J Physiother 1986;32:215-23. 28. Witvrouw E, Lysens R, Bellemans J, Peers K. Intrinsic risk factors for the development of anterior knee pain in an athletic population: a 2 year prospective study. Proceedings of the 13th International Congress of the World Confederation for Physical Therapy; May 23-28, 1999. The Japanese Physical Therapy Association, Yokohama (Japan). p 153. 29. Grelsamer RP, McConnell J. The patella: a team approach. Gaithersburg (MD): Aspen; 1988. 30. Wickiewicz TL, Roy RR, Powell PL, Edgerton VR. Muscle architecture of the human lower limb. Clin Orthop 1983;179:275-83. 31. Johnson MA, Polgar J, Weightman D, Appleton D. Data on the distribution of fiber types in thirty-six human muscles. J Neurol Sci 1973;18:111-29. 32. Lieb FJ, Perry J. Quadriceps function. J Bone Joint Surg Am 1968;50:1535-48. 33. Hubbard JK, Sampson HW, Elledge JR. Prevalence and morphology of the vastus medialis oblique muscle in human cadavers. Anat Rec 1997;249:135-42. 34. Scharf W, Weinstable R, Othrner E. Anatomical separation and clinical importance of two different parts of the vastus medialis muscle. Acta Anat (Basel) 1985;123:108-11. 35. Scott SH, Engstrom CM, Loeb GE. Morphometry of human thigh muscles. Determination of fascicle architecture by magnetic resonance imaging. J Anat 1993;182:249-57. 36. Thiranagama R. Nerve supply of the human vastus medialis muscle. J Anat 1990;170:193-8. 37. Nozic M, Mitchell J, de Klerk D. A comparison of the proximal and distal parts of the vastus medialis muscle. Aust J Physiother 1997;42:277-81. 38. Raimondo RA, Ahmad CS, Blankevoort L, April EW, Grelsamer RP, Henry JH. Patellar stabilization: a quantitative evaluation of the vastus medialis obliquus muscle. Orthopedics 1998;21: 791-5. 39. Radin E. Panel discussion: does chondromalacia patella exist? In: Picket J, Radin E, editors. Chondromalacia of the patella. Baltimore: William & Wilkins; 1983. pp 62-82. Suppliers a. Graphics Control Corp, c/o Medical Services Pty, Ltd, 61 Glass St, Richmond, Australia. b. Peak Performance Technology, Inc, 7388 S Revere Pkwy, Ste 603, Englewood, CA 80117. c. Panasonic S-VHS Movie Camera NV-MS5, Rexel, Australia, 366 Boundary Rd, Burwood, Vic, Australia. d. MT 40 Digital Metronome, Allens Music Inc, 133 Bourke St, Melbourne, Vic, Australia. e. Associative Measurement Pty, Ltd., 12 McGill St, North Ryde, NSW, Australia. f. StatView Version 1.04A, SAS Institute Inc, SAS Campus Dr, Cary, NC 27513.