Vastus medialis oblique/vastus lateralis muscle activity ratios for selected exercises in persons with and without patellofemoral pain syndrome.

Background and Purpose. The purpose of this study was to determine which of selected exercises with and without the feet free to move would enhance vastus medialis oblique muscle (VMO) activity over that of the vastus lateralis muscle (VL) and whether the use of taping would increase VMO activity. Subjects. Twenty-one subjects without patellofemoral pain (PFP) syndrome and 10 subjects with PFP syndrome, aged 19 to 43 years (X=26, SD=7), participated. Methods. Subjects were studied for the normalized, integrated electromyographic (IEMG) activity of their VMO, VL, and adductor magnus muscle (subjects without PFP syndrome) and the VMO/VL ratio using wire electrodes. Results. One exercise demonstrated greater activation of the VMO over the VL when compared with similar exercises in subjects without PFP syndrome. The mean VMO/VL activity ratio for terminal knee extension was 1.2 (SD= 0.5) with the hip medially rotated and 1.0 (SD=0.4) with the hip laterally rotated. Although subjects reported that patellar taping decreased pain 94% during the step-down exercise, the VMO/VL ratio was not changed. Conclusion and Discussion. The results suggest that neither exercises purported to selectively activate VMO activity nor patellar taping improve the VMO/VL ratio over similar exercises.

Patellofemoral joint pain (PFP) is common in the general population, occurring more often in women and in athletes, with the joint being the most common site of knee pain in sports medicine clinics.[1-5] Lateral malalignment of the patella has been suggested as one of the major causes of PFP syndrome.[4,6-8] As a result, both surgical and conservative treatments to correct this malalignment have been suggested.[1,4-10]

Many exercise treatments emphasize the importance of the vastus medialis oblique muscle (VMO)[6,8,10] because of its medial pull on the patella.[11-15] Some researchers[16-19] suggest that contraction of the hip adductor and quadriceps femoris muscles simultaneously would preferentially activate the VMO. Other researchers[17,19] report that the vastus medialis muscle (VM) is activated preferentially in response to valgus stress at the knee caused by hip lateral rotation during knee extension exercises and that hip medial rotation, therefore, decreases the activity of the VM. Still others[9,20,21] have reported greater VMO activity during knee extension with the knee relatively flexed than in terminal knee extension. Knee extension exercise with tibial medial rotation has been proposed because the VMO is purported to prevent lateral rotation of the tibia[18,19] and therefore to decrease the quadriceps femoris muscle angle (Q angle) and lateral patellar tracking.[22] Pronation of the subtalar joint and medial rotation of the tibia, however, have also been claimed to increase lateral tracking.[6,23]

McConnell[6,16] trains patients with PFP syndrome in unilateral and bilateral limb flexion exercises in weight bearing (walk stance and wall slide) and step-down exercises because these patients have increased pain with these activities. Although soft bracing has previously been suggested to negate the valgus force of the quadriceps femoris muscles on the patella,[1,5] McConnell[6,16] suggests improving patellar tracking, decreasing pain, and increasing the vastus medialis oblique/vastus lateralis muscle (VMO/VL) activity ratio in persons with PFP by taping the patella or tensor fascia lata muscle medially. Little evidence exists to support the use of the exercises and procedures reviewed.

The primary purpose of this study was to determine whether any of the exercises purported to increase the activity of the VMO over the VL, the VMO/VL activity ratio, did so when compared with similar exercises in subjects with and without PFP. A secondary purpose of this investigation was to determine whether therapeutic medial-glide taping altered the activity of the VMO or the VL or the VMO/VL activity ratio. Another purpose was to determine which of similar exercises increased the activity of the VMO, VL, and adductor magnus muscle (AM). Finally, I wanted to determine whether gender influenced the muscle activity studied.

Method

Subjects

Twenty-one subjects who had no known lower-limb musculoskeletal impairments and who exhibited no signs of neurological impairment and 10 subjects with PFP participated in this study. The subjects' average age, average body weight, and gender are shown in Table 1. Subjects with symptoms had a physician's diagnosis of PFP within 6 years of the testing date and reported retropatellar pain during at least two of the following activities: (1) squatting, (2) ascending and descending stairs, and (3) prolonged sitting.[1,3,4] All subjects with PFP performed a step-down exercise from a 22.9-cm (9-in) stool with and without medial-glide taping of the patella before beginning the study. The subjects rated their pain after step-down exercise on a scale of 1 to 10, with 1 being minimal pain and 10 being the worst pain that they could imagine, and they reported the percentage of change in pain with step down after taping. The subjects with PFP were required to have at least a 50% reduction in pain with patellar taping to participate in the study. All subjects provided informed consent consent with university policy.

[TABULAR DATA OMITTED]

Electromyographic Procedure

Myoelectnc activity was measured by use of indwelling wire electrodes of 50-[mu]m nickel alloy. The electrodes, insulated except for 2 mm at the ends, were inserted into the muscle with a 25-gauge needle. The needle was withdrawn, leaving the barbed ends of the wire electrodes in place. The wire electrodes were inserted into the VMO and VL of all subjects. In addition, a wire electrode was placed in the AM of the subjects without PFP syndrome to ensure that they were contracting the muscle during exercises. The wire for the VMO was placed in the middle of the muscle belly. The wire for the VL was placed in the muscle approximately one third of the distance from the patella to the anterior superior iliac spine. The wire for the AM was inserted into the muscle just anterior to the gracilis muscle, approximately one third of the distance from the medial femoral epicondyle to the symphysis pubis. The locations of the VL and AM insertions were chosen because they proved to be the most distal locations, consistently affording a full electromyographic (EMG) interference pattern upon muscle activation during pilot testing. The distal VL insertion was used to sample the oblique portion of the VL, which is purported to best oppose the action of the VMO.[12] The distal AM insertion was used to sample its activity near the origin of the VMO.[13] The leg each subject said was the dominant leg was tested in all except one of the subjects without PFP syndrome. The nondominant limb was tested in one subject without PFP syndrome because he had a previous knee surgery on the dominant side. The most severely involved leg of the subjects with PFP was tested.

In order to decrease the chance of wire electrode migration during testing, subjects contracted the inserted muscle maximally several times after needle insertion to pull the wire into the muscle. in addition, the investigator moved the limb into full knee flexion and extension and full hip abduction and adduction to allow the wire to slip further into the tissues before taping the external wire to the limb with a stress-relief loop.

Surface ground plates and the telemetry system were attached to the subject, and the electrode wires were connected to attachment posts on the ground plates. Myoelectric signals were differentially amplified, bandpass filtered (50-850 Hz), and transmitted by FM-FM telemetry to a receiver interfaced to a B&L computer (model 286).(*) Placement of electrodes in the vastus muscles rather than rectus femoris muscle was confirmed by noting activity during isometric knee extension and silence during combined isometric hip and knee flexion. Placement in the AM rather than the VM was confirmed by noting activity during isometric hip adduction and silence during isometric knee extension. The EMG activity was then recorded during rest and during maximal manual resistance tests. All recordings during the resistance tests were obtained with the subjects positioned supine and supported on their elbows. For the VMO and VL tests, the subjects' knee extension was resisted while the hip and knee were flexed approximately 30 degrees. For the AM test, the subjects were resisted for hip adduction with the knee extended.

Exercise Procedures

Exercises included what Lehmkuhl and Smith[24] and Soderberg[25] have described as "open-chain" and "closed-chain" activities of the lower limb, performed m random order for 5 seconds each. Lower-limb open-chain exercises were performed with the sole of the foot free to move, whereas closed-chain exercises had the sole of the foot planted on the floor. The terms "open chain" and "closed chain" are used here to denote whether the distal limb segment was free to move when the quadriceps femoris muscles contracted. Open-chain exercises were randomly chosen to be performed prior to or after closed-chain exercises. Subjects practiced dynamic exercises until the investigators were satisfied that movements were smoothly timed with a metronome at 1 beat per second. Starting positions of the knee were monitored with a standard goniometer. Subjects moved and their EMG activity was recorded during exercises, beginning on the beat of a metronome. Movement began with a "go" command from the investigator immediately following a "ready" command at the previous metronome beat. Electromyographic recording during isometric exercises began after the raw EMG level stabilized in a full interference pattern. In consideration of the tolerance of the subjects with PFP syndrome, fewer exercises were performed by the subjects with PFP syndrome than by the subjects without PFP syndrome. Abbreviations and definitions of the exercises used in this study are presented in Table 2.

[TABULAR DATA OMITTED]

Exercise for Subjects Without

PFP Syndrome: Open Chain

Three groups of open-chain activities were used for interexercise comparison: quadriceps femoris muscle setting ("quad sets"), knee extension, and isometric holds. All exercises were performed against the resistance of an ankle cuff weight equal to 5% of each subject's body weight to the nearest pound.

Quad sets (QS). Quad sets were isometric exercises performed in full knee extension with the subjects positioned long sitting and supported on their hands with their heels lifted off of the table to decrease the possibility of substitution by hip extensor activity. The six quad set exercises were done with (1) the hip and ankle positioned in neutral (QS), (2) the hip maximally medially rotated and the ankle positioned in neutral (QSMR), (3) the hip maximally laterally rotated and the ankle positioned in neutral (QSLR), (4) the hip maximally adducted against a pillow bolster with the ankle positioned in neutral (QSA), (5) the hip positioned in neutral and the ankle maximally dorsiflexed (QSDF), and (6) the hip positioned in neutral and the ankle maximally plantar flexed (QSPF).

Knee extension (KE). The knee extension exercises were performed with the subjects in a sitting position with the knee flexed from 30 to 0 degrees and the ankle positioned in neutral. The movement was timed with a metronome for 3 seconds, followed by a 2-second hold at full extension. The three exercises were performed with (1) the hip positioned in neutral (KE), (2) the hip maximally laterally rotated (KELR), and (3) the hip maximally medially rotated (KEMR).

Isometric holds in flexion (IS). Isometric hold exercises were isometric knee extension exercises performed with the subjects in a sitting position with their hip and ankle positioned in neutral. The five exercises were done with (1) the knee flexed 15 degrees and the tibia in neutral rotation (IS15), (2) the knee flexed 60 degrees and the tibia in neutral rotation (IS60), (3) the knee flexed 45 degrees and the tibia in neutral rotation (IS45), (4) the knee flexed 45 degrees and the tibia maximally laterally rotated (IS45LR), and (5) the knee flexed 45 degrees and the tibia maximally medially rotated (IS45MR).

Exercises for Subjects Without PFP Syndrome: Closed Chain

Two groups of closed-chain exercises were used for interexercise comparison: (1) walk-stance and step-down exercises and (2) wall-slide exercises. Movements were performed for 3 seconds, followed by a 2-second hold at the end position.

Walk stance-step down (WS-SD). Walk-stance exercises were unilateral exercises performed to 45 degrees of knee flexion with the hip in neutral rotation and the subject's weight supported on the forward, tested limb. The opposite toe was permitted to remain on the floor for balance only. Balance was also provided by touching the hands of an investigator. The five exercises were performed with (1) the subtalar joint unconstrained (WS); (2) the subtalar joint maximally supinated (WSS); (3) subtalar joint maximally pronated (WSP); (4) the subtalar joint unconstrained after the patella was taped medially with a media frontal-plane tilt and rotation to position the error pole of the patella inferiorly (WSPT); and (5) the subtalar joint unconstrained after tensor fascia lata muscle medial-glide taping (WSTT).

The final exercise was a step down from a 22.9-cm stool, with the subject leading with the contralateral limb while the tested limb was in neutral hip rotation. The hold was with the contralateral foot just off of the floor (SD).

Wall slide (WSl). The wall-slide exercises were bilateral exercises that the subjects performed from an upright standing position to 45 degrees of knee flexion while the trunk maintained contact with the wall. This was done to prevent decreasing the quadriceps femoris muscle demand by shifting the center of gravity of the trunk anteriorly. The subjects' feet were parallel, shoulder width apart, and just far enough from the wall to allow knee flexion to 45 degrees. The hips were in neutral rotation. The two wall-slide exercises were performed as (1) a straight wall slide (WSl) and (2) a wall slide while squeezing a pillow bolster between the knees (WSlA).

Exercises for Subjects With PFP Syndrome

Procedures were identical to those for subjects without PFP syndrome, unless otherwise noted. Groups of exercises for interexercise comparison were open-chain exercises and closed-chain exercises with and without patellar taping.

Open-chain exercises were QS, IS15, IS60, and KE. Closed-chain exercises were done both before and after the patella was taped medially with a medial frontal-plane tilt and rotation to position the inferior pole of the patella inferiorly. The closed-chain exercises were WS, WSl, WSlA, SD, and isometric knee extension and hip adduction in a sitting position with hips and knees flexed to 90 degrees (ISQA). This exercise was considered closed chain because subjects were instructed to contract the vastus muscles by pushing the foot against the floor and to squeeze the bolster maximally.

At the end of testing, wire electrodes were slipped out of the muscle, and the skin was cleaned with alcohol. Total testing time was 1 1/2 to 2 hours per subject.

Data Analysis

The EMG activity was digitized at 2,000 samples per second through an A/D converter(dagger) run by the B&L software version 4.19).(*) The software then rectified the signals and set noise thresholds from the first 2 seconds of activity of the resting EMG record. The threshold was the lowest level of activity recorded below which 95% of the resting EMG signal was found. Only EMG activity greater than threshold was then quantified by integration. Signals were integrated each 1/50 second for each 5-second exercise. The EMG values for each exercise were normalized by the software by division by the EMG value from the maximal exercise tests. All EMG values reported are therefore expressed as a percentage of maximal activity.

Within-day reliability of the integrated, normalized EMG values for the VMO, the VL, and the VMO/VL ratio was previously established in 12 subjects without PFP syndrome by intraclass correlation coefficients (ICC[3, 1]) for two repeated measures of QS, QSA IS60, IS15, WS, WSl, WSlA, SD, an& knee extension from 90 to 0 degrees of flexion using the same EMG processing as in this study. Two insertion sites in locations bordering within 1.27 cm (0.5 in) of those used in this study were sampled for each muscle. Eighty percent of all ICCs were above .90. Reliability averaged .81 for the VMO, ranging from .37 to .98, and averaged .91 for the VL, ranging from .77 to .98. The VMO/VL ratio averaged .93, ranging from .88 to .99, with the exception of .55 for walk stance using the lower insertion sites. The walk-stance ratio for the upper insertion sites was .97.

The average EMG value was calculated for each muscle of each subject for each exercise, and a VMO/VL ratio was calculated from these values. Mean values for VMO, VL, and AM myoelectric activity and the VMO/VL ratio were calculated for each exercise. Exercises in each of the five groups of exercises for subjects without PFP syndrome (QS, KE, IS, WS-SD, WSl) were compared across exercises for muscle activity and VMO/VL ratio by a two-way repeated-measures multivariate analysis of variance (MANOVA), with gender as the grouping factor. Exercises for open-chain activities for patients with PFP syndrome were compared across exercises for muscle activity and VMO/VL ratio by a oneway repeated-measures MANOVA. Exercises for closed-chain activities for subjects with PFP were compared across exercises for muscle activity and VMO/VL ratio by a two-way repeated-measures MANOVA, with taping as the grouping factor. When MANOVA results were significant, a subsequent univariate analysis of variance (ANOVA) was done for each muscle tested. A level of significance of .05 was accepted, and a Bonferroni adjustment was used for post hoc t tests. The BMDP statistical package(double dagger) was used for all analyses.

Results

All reported significant results refer to post hoc significance. For these results, the MANOVA and ANOVA results were also significant.

Subject Without PFP Syndrome

Due to recording difficulties, one male subject's data were lost for the quad set and closed-chain exercises.

Open-chain exercises. Means and standard deviations of data and comparisons that yielded statistically significant results are reported in Tables 3 through 5. No differences in myoelectric activity due to gender were seen. Comparison of the quad set exercise with all its variants showed no differences due to ankle or hip rotation position for the VMO, VL, or VMO/VL ratio (Tab. 3). The EMG activity rugged from 48% to 58% of maximum for the VMO and from 46% to 52% of maximum for the VL. The VMO/VL ratio ranged from 1.1 to 1.2. Although AM activity was higher for QSA (32% [+ or -] 20%) than for QS (12% [+ or -] 22%), the increased AM activity did not affect the VMO, VL, or VMO/VL ratio activity (Tab. 3).

[TABULAR DATA OMITTED]

Knee extension exercises were compared post hoc between KE and both KEMR and KELR and between KEMR and KELR. Comparisons of KE with KEMR and KELR showed higher VMO and VL activity in the KE exercise (34% [+ or -] 18% and 35% [+ or -] 15%, respectively) than in either rotated position of the hip. The VMO activity was 28% [+ or -] 6% for KEMR and 22% [+ or -] 12% for KELR, whereas VL activity was 28% [+ or -] 13% for KEMR and 26% [+ or -] 12% for KELR (Tab. 4). No difference, however, was seen in the VMO/VL ratio between KE and either KEMR or KELR. Comparison between KEMR and KELR showed higher VMO activity and VMO/VL activity ratio in KEMR than in KELR (Tab. 4). The VMO/VL ratio was 1.2 [+ or -] O.5 for KEMR and 1.0 [+ or -] O.4 for KELR.

[TABULAR DATA OMITTED]

Isometric hold exercises were compared post hoc between IS60 and both IS15 and IS45 and between IS45 and both IS45MR and IS45LR. Less myoelectric activity of the VMO and VL was seen in IS60 (5% [+ or -] 3% for both) than in IS15 (18% [+ or -] 12% and 22% [+ or -] 10%, respectively) without a change in the VMO/VL ratio (Tab. 5). Tibial rotation did not affect the VMO or VL activity or the VMO/VL ratio for the 45-degree position.

[TABULAR DATA OMITTED]

Closed-chain exercises. Data were lost while recording the activity during the WSPT in 1 male subject. Analysis was therefore performed on 19 subjects for all six exercises in the WS-SD group and also for all 20 subjects for the remaining five exercises. No differences resulted between these analyses.

Means and standard deviations of data and comparisons that yielded statistically significant results are reported in Tables 6 and 7. Gender/exercise interactions were seen for the VMO and VL for the WS-SD exercise group. Athough muscle activity was similar between men and women for all WS exercises, ranging from 11% to 15% for women and from 11% to 16% for men, women required approximately twice the activity as men for SD (Tab. 6). Average VMO activity was 24 [+ or -] 11% for men and 45% [+ or -] 6% for women, whereas VL activity averaged 19% [+ or -] 7% for men and 41% [+ or -] 3% for women.

[TABULAR DATA OMITTED]

Separate analysis by gender for the VMO and VL comparing WS with WSS, WSP, WSPT, WSTT, and SD tended to show greater activity for both the VMO and VL during the SD exercise than during the WS exercise (Tab. 6). Both muscle values for women but only VL values for men were different between SD and WS. When data from men and women were combined, the difference in VMO activity between SD and WS was significant (P=.0000). No influence of gender or exercise was seen for either AM activity or VMO/VL ratio for the WS-SD exercise comparisons, although a trend toward increased AM activity from WS to SD was found.

No differences in EMG activity due to gender were seen in WSl exercises. Likewise, no difference in VMO/VL ratio was seen between WSl and WSlA, but greater activity was seen in the AM, VMO, and VL during WSlA than in WSl (Tab. 7). The AM activity increased from 2% [+ or -] 3% to 30% [+ or -] 53% when adduction was added to the WSl exercise. The VMO increased from 9% [+ or -] 6% to 17% [+ or -] 7%, whereas the VL increased from 9% [+ or -] 5% to 17% [+ or -] 8% with the addition of adduction to WSl.

Subject With PFP Syndrome

Means and standard deviations of data and comparisons that yielded statistically significant results for open- and closed-chain exercises are given in Tables 8 and 9, respectively. The VMO/VL ratio did not differ in comparisons of QS with IS60, IS15, and KE and of IS60 with IS15 (Tab. 8). Higher VMO and VL activity, however, occurred during the QS activity than in any other open-chain activity. The QS activity was 101% [+ or -] 30% for the VMO and 90% [+ or -] 36% for the VL. The highest activity in other open-chain exercises was 49% [+ or -] 17% for the VMO and 48% [+ or -] 17% for the VL during KE. In addition, IS15 demanded more vastus muscle activity than did IS60 (Tab. 8). The VMO activity increased from 6% [+ or -] 5% for IS60 to 40% [+ or -] 25% for IS15, whereas VL activity increased from 7% [+ or -] 6% to 37% [+ or -] 21%, respectively.

[TABULAR DATA OMITTED]

Patellar taping did not affect closed-chain muscle activity, even though the decrease in pain after patellar taping for the SD exercise averaged 94%. Analyses comparing WS with WSl, SD, and ISQA and WSl with WSlA were not significant for the VMO/VL ratio (Tab. 9). Both the VMO and VL were less active in the WS (31% [+ or -] 23 for VMO and 38% [+ or -] 36% for VL) than in the SD (65% [+ or - ] 22% for VMO and 77% [+ or -] 36% for VL) or ISQA (63% [+ or -] 31% for VMO and 69% [+ or -] 36% for VL). The VMO increased its activity when adduction was added to the wall slide from 13% [+ or -] 7% to 30% [+ or -] 18%. The

increase in the VL activity from 16%[plus or minus]l0% for WSl to 36%[plus or minus]32% with WSlA did not reach the rigor of post hoc significance.

[TABULAR DATA OMITTED]

Discussion

Technique

The amount of EMG signal recorded is dependent on the location and size of the recording electrodes. In order to effectively compare the EMG signal between muscles, each of which may be a different distance from the recording electrode, and between subjects, each of whom may have different-sized muscles and different distances from active muscle to electrodes, a method of expressing the EMG activity of a specific muscle as a ratio of activity to some reference value eliminates the influence of location and size of recording electrodes. Use of the maximum isometric EMG activity as the normalizing factor allows expression of activity in an easily understandable ratio and has been shown to provide better reliability (ICC) than using either dynamic maximal or submaximal EMG activity in the gastrocnemius muscle.[26]

The software used for data collection prevented collection of data for less than 5 seconds per trial. Because movement during dynamic exercise would be much slower than customary in clinical practice if movement was prolonged for 5 seconds, only 3 seconds of movement was used, with a 2-second hold at the end position. As a result, two fifths of each dynamic exercise was actually isometric. Because no differences were found in VMO/VL ratios at different isometric positions, this practice of collecting 2 seconds of isometric data in each dynamic exercise probably had no effect on the overall- ratio. Furthermore, exercises with movement were compared only with other exercises with movement, except that KE was compared with ISs and ISQA was compared with exercises with movement in subjects with PFP syndrome.

I had no method of quantifying my ability to control the speed of movement. Undoubtedly, some error existed in subjects' ability to move at a constant speed with the metronome. Likewise, I had no better method of controlling isometric knee flexion positions than use of a standard goniometer. The error introduced by these shortcomings is unknown. These techniques, however, approximate clinical practice more than would a more elaborate method used to control speed or position.

The VMO/VL ratio was used in this study because it reflects the relative contributions of the VMO and VL. I believe an increase of VMO activity with a specific exercise is meaningless if the relative activity of the VL is unknown, as both muscles may be increasing their activity the same amounts. Normalized EMG data are ratio data. A true absence of activity can exist, and ratios are used in normalization of the data.

Knee flexor activity may have occurred in some of the open-chain exercises and most likely occurred in the closed-chain exercises. The purpose of this study was to compare results across similar exercises, regardless of whether other activity was present. Rather than examine vastus muscle activity when no other activity was present, this study was designed to study exercises widely used in physical therapy.

VMO/VL Ratio

This study did not support the claims that certain commonly used exercises or patellar or tensor fascia lata muscle medial-glide taping enhance VMO activity over VL activity. The only exercise to show a higher VMO/VL ratio in comparison with similar exercises was KEMR in comparison with KELR. This finding is in contrast to the commonly held hypothesis that hip lateral rotation, by creating a knee abduction torque, enhances the activity of the VM[19] but is consistent with the reported lack of preferential activation of the VMO during knee extension with a knee abduction torque.[27,28] Whether the small magnitude of increase in VMO/VL ratio (0.2) found with medial over lateral rotation of the femur is clinically significant is not known. Its significance most likely depends on the magnitude of the lateral trucking of the patella.

Likewise, preferential activation of the VMO over the VL is not consistent with this and other EMG studies of hip adduction[27] or tibial rotation [18,29] with knee extension. Two investigations of open-chain adduction exercises that did not require a simultaneous quadriceps femoris muscle contraction, however, did show preferential activation of the VM or VMO.[18,19] Wheatley and Jahnke[19] reported that VM action potentials occurred during hip adduction, but they did not quantify or statistically compare their data and their results could be in error due to volume conduction of adductor muscle activity to their widely spaced ([greater than or equal to]2.54 cm [[greater than or equal to]1 in]) surface electrodes over the VM. Hanten and Sculthies' adduction exercise, although not requiring quadriceps femoris muscle activity, still elicited high levels of activity in both the VMO and VL.[18] Perhaps conscious activation of the vastus muscles with adduction negates any benefit of adduction exercise to preferentially activate the VMO. Wheatley and Jahnke[19] also reported greater VM activity during QS with leg lateral rotation. In addition, the findings of my study are not consistent with the theories that the VMO is selectively activated in a flexed position of the knee, during tibial medial or lateral rotation, or during subtalar joint pronation.[6,9,18-23]

Results of taping the patella and the tensor fascia lata muscle in my study are in contrast to McConnell's finding of an increased VMO/VL activity ratio in symptomatic subjects with such taping, even though subjects with PFP in my study reported greatly reduced pain during the SD exercise after medial-glide taping.[16] McConnell did not report her EMG recording procedures. Whether her VMO data may have been contaminated by volume conduction of activity of nearby muscles to the VMO electrode, therefore, cannot be judged.

Some caution must be taken in interpreting the findings of no effects of specific exercise or taping on the VMO/VL activity ratio because of the low number of subjects in this study. Ratio differences of 0.7 for ISs, 0.3 for QSs, and 0.6 for KEs, WS-SD, and WSls in subjects without PFP syndrome and of 0.7 for open-chain exercises, 0.5 for closed-chain exercises, and 0.4 for taping in subjects with PFP syndrome were necessary to satisfy a statistical power of .80.[30] True differences less than these ratios, therefore, could have been present without rejection of the null hypothesis. The magnitude of change in the VMO/VL ratio necessary for therapeutic effect is unknown. Certainly, the greater the increase in magnitude, the greater will be the medial pull on the patella.

Lack of preferential activation of the VMO over the VL due to exercise purported to produce such activation or due to patellar taping leads to the question of whether the VMO can be trained to selectively increase its activity. Studies of EMG biofeedback training for the VMO are needed to answer this question.

Because no difference in the VMO/VL ratio was found with patellar taping although pain was decreased an average of 940%, I question that taping decreases pain due to appropriate realignment of the patella. The ability to reliably determine patellar alignment is poor.[31] Therefore, the ability to appropriately realign the patella is questionable. Furthermore, I found no evidence in the literature that patellar taping can maintain the position of the patella during exercise. The effect on PFP of placebo taping of the patella or of taping the patella with randomly chosen direction has not been studied. The positive effect of taping may be due to additional sensory input or the placebo effect. The effect of patellar taping, therefore, should be studied in a large group of subjects with PFP. Two groups of subjects without knowledge of taping theory randomly assigned to receive medial or lateral patellar taping can be studied using a blind research design.

Both subjects with and without PFP appeared to have similar VMO/VL ratios in open-chain exercise, whereas subjects with PFP appeared to have lower ratios in closed-chain exercise than subjects without PFP. Whether this apparent difference is statistically significant or whether it is important is unknown. Statistical comparison in further studies would be beneficial.

VMO, VL, and AM Activity

Subjects in this study averaged higher VMO and VL activity during the QS exercises than in other open-chain exercises. They were at end range of knee extension for 5 seconds during the QS exercises but were either more flexed or moving for at least 3 seconds of all other open-chain activities. As a result, subjects had the opportunity to produce high levels of EMG activity in the QS exercises while maintaining their test position. I suggest that they produced this high level of activity because they were well motivated. Attempts at increasing EMG activity during the other open-chain exercises would have either increased the speed of movement (controlled with a metronome) or moved the leg from the isometric test position. Either of these activities would have resulted in discarding the data and repeating the exercise until the desired velocity or posture was attained. As a result, the EMG activity was lower for these exercises. Because others [27,32] have found higher vastus muscle activity in QS than in straight leg raising, the QS has consistently been shown to be an excellent exercise for recruiting vastus muscle activity.

The increased activity in both the VMO and VL during IS15 over IS60 in both the subjects with and without PFP syndrome is expected due to the increased quadriceps femoris muscle demand at 15 degrees without preferential demand for VMO activity.[14,19,20,33-35] This increased mechanical demand at 15 degrees is due to the combined effects of increased gravitational lever arm and decreased muscle length and lever arm of the quadriceps femoris muscle.

Because the WSlA tended to recruit greater VMO, VL, and AM activity than the WSl and because the SD and ISQA recruited greater VMO, VL, and AM activity than the WS, the suggested benefit of these closed-chain exercises in persons with PFP may be due to a high level of coactivation of the knee extensors and hip extensors (AM) to better control the femur. A comparison of lower-limb extensor muscle activity between closed-chain exercises that mandate multisegment control of the lower limb and open-chain exercises in subjects with PFP syndrome would be beneficial.

Conclusions

Whether in the subjects with PFP or in those without PFP, only one exercise resulted in a higher VMO/VL activity ratio over similar exercises. The KEMR showed a higher VMO/VL ratio than the KELR. Exercises more commonly prescribed to enhance VMO activity over that of the VL, however, failed to selectively activate the VMO. Furthermore, the results of this study indicate that medial-glide taping of the patella or tensor fascia lata muscle does not alter the VMO/VL ratio.

Acknowledgments

I thank the following California State University, Long Beach, physical therapy students for their assistance in this project: Janet Froggatt, Anthony Granger, Cynthia Grauf, Kathy Harbert Greenwood, Michael Greenwood, Gregory R Jue, Mark Klem, Sonja Maul, Nancy Rhoan, Susan Royce, Stacy Samano, Ernie Sanchez, Milan Steijn, David Swink, Carol Whitmire, and Laura Olsen. I thank Michael Monahan for his question in class that inspired this research, and I thank Charles Felder, PT, OCS, instructor for McConnell seminars, who consulted with me on this project and who taught me the McConnell taping technique.

(*) B&L Engineering, 12309 E Florence Ave, Santa Fe Springs, CA 90670. (dagger) Model DT2801-A, Data Translation, 100 Locke Dr, Marlborough, MA 01752-1192. (double dagger) BMDP Statistical Inc, Los Angeles, CA 90086.

References

[1] Levine J. Chondromalacia patellae. The Physician and Sportsmedicine, 1979;7(8):41-49.

[2] Outerbridge RE. Further studies on the aetiology of chondromalacia patellae. J Bone Joint Surg [Br]. 1964;46:179-190.

[3] Puniello MS. Iliotibial band tightness and medial patellar glide in patients with patellofemoral dysfunction. J Orthop Sports Phys Ther. 1993;17:144-148.

[4] Fulkerson JP, Hungerford DS. Disorders of the Patellofemoral Joint. 2nd ed. Baltimore, Md: Williams & Wilkins; 1990.

[5] Malek MM, Mangine RE. Patellofemoral pain syndrome: a comprehensive and conservative approach. J Orthop Sports Phys Ther. 1981;2: l08-l16.

[6] McConnell J. The management of chondromalacia patellae: a long-term solution. Australian Journal of Physiotherapy. 1986;32:215-223.

[7] Schutzer SF, Ramsby GR, Fulkerson JP. Computed tomographic classification of patellofemoral pain patients. Orthop Clin North Am. 1986;17:235-248.

[8] Kettlekamp DB. Management of patellar malalignment. J Bone Joint Surg [Am]. 1981;63: 1344 - 1348.

[9] Mariani PP, Caruso I. An electromyographic investigation of subluxation of the patella. J Bone Joint Surg [Br]. 1979;61:169-171.

[10] Antich TJ, Brewster CE. Modification of quadriceps femoris muscle exercises during knee rehabilitation. Phys Ther. 1986;66:1246-1251.

[11] Lieb FJ, Perry J. Quadriceps function: an anatomical and mechanical study using amputated limbs. J Bone Joint Surg [Am]. 1968;50: 1535-1548.

[12] Weinstabl R, Schaf W, Firas W. The extensor apparatus of the knee joint and its peripheral vasti: anatomic investigation and clinical relevance. Surg Radiol Anat. 1989;11:17-22.

[13] Bose K, Kanagasuntheram R, Osman MBH. Vastus medialis oblique: an anatomic and physiologic study. Orthopedics. 1980;3:880-883.

[14] Hehne HJ. Biomechanics of the patellofemoral joint and its clinical relevance. Clin Orthop. 1990;258:73-85.

[15] Slocum DB, Larson RL. Rotary instability of the knee. J Bone Joint Surg [Am]. 1968;50:211-225.

[16] McConnell J. Training the vastus medialis oblique in the management of patellofemoral pain. In: Proceedings of the 10th International Congress of the World Confederation for Physical Therapy; Sydney, New South Wales, Australia; 1987.

[17] Doucette SA, Goble EM. The effect of exercise on patellar tracking in lateral patellar compression syndrome. Am J Phys Med. 1992; 20:434-440.

[18] Hanten WP, Schulthies SS. Exercise effect on electromyographic activity of the vastus medialis oblique and vastus lateralis muscles. Phys Ther. 1990;70:561-565.

[19] Wheatley MD, Jahnke WD. Electromyographic study of the superficial thigh and hip muscles in normal individuals. Arch Phys Med Rehabil. 1951;32:508-515.

[20] Boucher JP, King MA, Lefebre R, Pepin A. Quadriceps femoris muscle activity in patellofemoral pain syndrome. Am J Phys Med. 1992;20:527-532.

[21] Sczepanski TL, Gross MT, Cuncan PW, Chandler JM. Effect of contraction type, angular velocity and arc of motion on the VMO:VL ratio. J Orthop Sports Phys Ther. 1991;14:256-262.

[22] vanKampen A, Huiskes R. The three-dimensional tracking pattern of the human patella. J Orthop Res. 1990;8:372-382.

[23] McCulloch MU, Brunt D, Vander Linden D. The effect of foot orthotics and gait velocity on lower limb kinematics and temporal events of stance. J Orthop Sports Phys Ther. 1993;17: 2-10.

[24] Lehmkuhl LD, Smith LK. Brunnstrom's Clinical Kinesiology. 4th ed. Philadelphia, Pa: FA Davis Co; 1983:7-8.

[25] Soderberg GL. Kinesiology: Application to Pathological Motion. Baltimore, Md: Williams & Wilkins; 1986:60-61.

[26] Knutson LM, Soderberg GL, Ballantyne BT, Clarke WR. Study of various normalization procedures for within-day electromyographic data. Journal of Electromyography and Kinesiology. 1994;4:47-59.

[27] Karst GM, Jewett PD. Electromyographic analysis of exercises proposed for differential activation of medial and lateral quadriceps femoris muscle components. Phys Ther. 1993; 73:286-299.

[28] Andriacchi TP, Andersson GBJ, Ortengren R, Mikosz RP. A study of factors influencing muscle activity about the knee joint. J Orthop Res. 1984;1:266-275.

[29] Duarte-Cintra AI, Furlani J. Electromyographic study of the quadriceps femoris in man. Electromyogr Clin Neurophysiol. 1981;21: 539-554.

[30] Glantz SA. Primer of Bio-Statistics. 3rd ed. New York, NY: McGraw Hill; 1992:155-187, 311, 401-405.

[31] Fitzgerald GK, McClure PW. Reliability of measurements obtained with four tests of patellofemoral alignment. Phys Ther. 1995;75: 84-90.

[32] Soderberg GL, Cook TM. An electromyographic analysis of quadriceps femoris muscle setting and straight leg raising. Phys Ther. 1983;63:1434-1438.

[33] Basmajian JV. Muscles Alive. 5th ed. Baltimore, Md: Williams & Wilkins; 1985:324-332.

[34] Eloranta V. Patterning of muscle activity in static knee extension. Electromyogr Clin Neurophysiol. 1989;29:369-375.

[35] Moller BN, Krebs B, Tidemand-Dal C, Aaris K. Isometric contractions in the patellofemoral pain syndrome: an electromyographic study. Arch Orthop Trauma Surg. 1986;105:24-27.

K Cerny, PhD, PT, is Professor, Department of Physical Therapy, College of Health and Human Services, California State University, Long Beach, CA 90840-5603 (USA) ([e-mail]).

The study protocol was approved by the Human Subjects Committee, California State University, Long Beach.

Result of this research were previously presented at the 1991 and 1992 Annual Conferences of the California Chapter of the American Physical Therapy Association and at the 1991 and 1922 Annual Conferences of the Long Beach Veterans Administration Hospital-California State University, Long Beach-Memorial Medical Center of Long Beach.