Content uploaded by John Hubbard
Author content
All content in this area was uploaded by John Hubbard on Jan 22, 2016
Content may be subject to copyright.
Journal of Musculoskeletal Research, Vol. 6, No. 2 (2002) 107–118
©World Scientific Publishing Company
107
ORIGINAL ARTICLES
EMG ANALYSIS OF TWO PORTIONS OF THE VASTUS MEDIALIS MUSCLE
DURING SELECTED KNEE REHABILITATION EXERCISES
John K. Hubbard*, and Steven Opersteny
*Department of Physical Therapy, Angelo State University, San Angelo, TX, USA
Physical Medicine and Rehabilitation Associates, College Station, TX, USA
John.Hubbard@Angelo.edu
Received February 6, 2002; Accepted May 3, 2002
ABSTRACT
Despite a prevalence of literature describing the function and rehabilitation of the vastus medialis
oblique (VMO) muscle, no literature exists which substantiates the existence of the VMO as a
separate muscle entity from the vastus medialis longus (VML) muscle. The purpose of this study
was to determine if the VMO and VML muscles exhibit any differences in muscle activity during
performance of commonly performed knee rehabilitation exercises. Electromyographic (EMG)
action potential amplitude from both the VMO and VML of normal adults were obtained
simultaneously via indwelling electrodes during performance of ten different knee rehabilitation
exercises. A two-way repeated measures analysis of variance was performed with the two portions
of the vastus medialis (VM) muscle as one independent variable, the knee exercises as the second
independent variable, and percent maximal voluntary contraction (%MVC) data as the dependent
variable. Data was analyzed to determine whether any differences existed between the VMO and
VML during performance of any of the selected exercises. Analysis of the data revealed that
although significant differences existed in %MVCs generated between exercises, no differences
were seen between the VMO and VML during any of the individual exercises or the group of
exercises as a whole. Exercises utilized to strengthen one portion of the VM muscle will
simultaneously strengthen the other portion. These findings show that no functional differences
exist between the VMO and VML, and lessen the credibility of the theory that they are separate
muscle entities.
Keywords: Vastus medialis; EMG; In vivo; Knee rehabilitation exercise.
ORIGINAL ARTICLES
Correspondence to: Dr. John K. Hubbard, Department of Physical Therapy, Box 10923, ASU Station, San Angelo, TX 76909, USA.
00081.p65 06/13/2002, 4:47 PM107
108 J. K. Hubbard & S. Opersteny
INTRODUCTION
The vastus medialis (VM) muscle is one of the
four components of the quadriceps femoris (QF)
muscle group. In 1968, Lieb and Perry
22
described
the VM as being subdivided into two components:
a proximal portion referred to as the vastus
medialis longus (VML), and a distal portion
referred to as the vastus medialis oblique (VMO).
The VML was reported to be a direct contributor
to knee extension, while the VMO was reported
to exert its influence as a medial stabilizer of the
patella during knee extension. In a subsequent
study, Lieb and Perry
23
reported on the relative
myoelectric activity of the QF muscle components.
The action potential frequencies and knee extensor
forces measured showed that all of the portions of
the QF were active throughout the range of knee
extension exercises, however, the distal portion of
the VM demonstrated an action potential count
that was consistently twice that of the other QF
components.
In 1990, Thiranagama
32
studied 30 lower limbs
from 15 human cadavers and stated that in all
specimens, two nerves arising from the posterior
division of the femoral nerve innervated the VM.
Although the presence of a fascial plane dividing
the muscle into two portions was not noted, it
was reported that because of differing spinal
segment innervation patterns that the two por-
tions should be inherently functionally different.
Gunal et al.,
13
performed a study in 1992 of the
innervation of the VMO. Their findings agreed
with those of Lieb and Perry
22, 23
and Thirana-
gama
32
in describing an independent innervation
for the VMO, as well as including a statement
that the innervation to the VMO came from the
saphenous nerve.
As a result of these studies, a general acceptance
of the VM as being anatomically and functionally
divided into two portions, the VML and the VMO,
was developed. Each portion was felt to have its
own nerve supply arising from different spinal
segments, with each portion described as being
able to perform distinctly different functions.
Acceptance of this terminology is now widespread
as reflected by the amount of literature describing
the VMO and its hypothesized role in pathology,
stabilization and rehabilitation of the knee.
Despite the amount of literature describing the
role of the VMO, there remains no clear evidence
of effective interventions, and no substantiated
evidence that the VMO is an anatomically
separate structure from the VML.
Recently, a large-scale cadaver study by
Hubbard et al.
15
cast doubts as to the credibility
of the theory that the VMO and VML are separate
anatomical structures. Generating data from 364
cadavers, Hubbard found agreement with the
variations in fiber angle changes of the VM
muscle, but found no sheet of epimysium
separating the VM into two distinct portions. A
fibrofascial plane composed of areolar fascial
tissue as described by Leib and Perry
22
was
selected to represent the division between the
VMO and the VML in contrast to a sheet of
epimysium as dividing muscles from each other
as described by Basmajian
3
and Gray.
11
When
found and dissected out, this fibrofascial plane
was seen to contain a neurovascular bundle con-
sisting of branches of the descending genicular
artery and infrapatellar nerve. These structures
traverse through the distal portion of the VM
muscle on their way to supplying the medial
joint capsule of the knee.
3, 11
In no instances
were there findings where the VM muscle was
clearly separable into distinct VMO and VML
muscles by a sheet of epimysium. There is no
debate that the VM has varied fiber directions
with multiple sites of origin converging to a
common insertion.
5, 13, 22, 23, 28, 31–33
However,
cadaveric evidence does not support the theory
that the VM muscle is separable into two portions.
Multiple studies have been published comparing
EMG activity of the VMO to the other QF, thigh
00081.p65 06/13/2002, 4:47 PM108
EMG Analysis of the VMO and VML during Exercise 109
and leg muscles,
1, 2, 8–10, 12, 14, 16–18, 23, 25–27, 29, 30
how-
ever no in vivo evidence exists to show that the
two portions of the VM can contract differently
from each other. Of the studies previously pub-
lished, only five of the studies
1, 8, 16, 25, 27
attempted
to gather concurrent data from both the VMO and
VML. Unfortunately, none of the studies that did
capture this data performed any statistical tests
for differences between the VMO and VML in
the timing or intensity of EMG activity during
knee extension. Additionally, a review of available
literature revealed no descriptions of exercises or
treatments capable of producing an isolated con-
traction of the VMO from the VML.
The use of cadavers is unquestioned in its
efficacy in studying and learning the structures of
the human body, but cadaveric examination does
not accurately represent the functionality of the
structures being studied. Grays Anatomy
11
states:
It is not necessarily true, however, that the action
in the living body is the same as that deduced
from observing its attachments, nor even from
pulling on it in a dead subject, because incidental
actions may not be utilized or may even be
suppressed in the living body. Koh et al.,
19
further emphasized the need for studies utilizing
in vivo techniques to study muscle function by
stating that in vivo research must be performed
to verify the data collected from cadaveric
specimens. In addition, only in vivo experiments
can be used to investigate the influence of an
in vivo muscle force, and of muscle training
exercises.
EMG studies have been utilized extensively in
determining the activation of muscle function
in vivo. Lippold
24
demonstrated a linear relation-
ship between the action potential amplitude of
the EMG signal and the force of isometric muscle
contractions. Komi
20
wrote of the linear relation-
ship of the EMG action potential amplitude in
both concentric and eccentric work. These results
support Winters
35
theory that the EMG action
potential amplitude indicates the state of activa-
tion of the contractile elements, and is a reflection
of the force being produced by a muscle.
The purpose of this study is to determine
whether the VMO can be selectively activated
in vivo, independently from the VML. Data collec-
tion and statistical analyses were performed to
test the null hypotheses that no differences will be
seen in EMG action potential amplitude between
the VML and VMO portions of the VM muscle
while performing selected rehabilitation activities.
This is the first study in which the in vivo function
of the VMO and VML will be compared and sta-
tistically analyzed with each other to determine
the validity of the concept of the VMO function-
ing as an individual muscle entity, separate from
the VML.
METHODS
Twenty-four healthy human subjects (21 males
and 3 females) who met selection criteria were re-
cruited to participate in this study to examine the
in vivo function of the VM muscle. Characteristics
of the human subjects are listed in Table 1. The
use of human subjects in this study was approved
by the Texas A&M University Institutional Review
Board. Prior to being included in this study, each
subject read, signed and received a copy of an in-
formed consent form as well as a consent form to
be photographed and/or video taped. Study sub-
jects were selected from either gender, between
the ages of 18 and 28, with no history of anterior
knee pain or patellofemoral dysfunction. Inclusion
criteria was the ability of the subject to actively
adduct the patella, or goniometric measurement
of the most distal portion of the VM muscle at
or greater than 60° from the shaft of the femur.
Subjects with a history of cellulitis, lymphedema,
or bleeding disorders, or who were on any blood
thinning medications, including aspirin, were dis-
qualified from participation in this study.
00081.p65 06/13/2002, 4:47 PM109
110 J. K. Hubbard & S. Opersteny
The dominant lower extremity of each subject
was utilized for data collection. The dominant
extremity was selected by having the subjects kick
a ball lying on the floor. The subjects were all
instructed in performing each of the study
exercises with a 5-second maximal voluntary
contraction, and given an opportunity to perform
each of them until they were comfortable with the
exercises, positions and equipment. The subjects
were also instructed in the EMG insertion and
data collection technique, and instructed to inform
the investigators if any discomfort was present, or
if they wished to discontinue their study parti-
cipation. The exercises selected for performance
in this study (Table 2) are all commonly used
clinically in efforts to rehabilitate patellofemoral
disorders, and have been studied in prior research
efforts.
6, 7, 9, 10, 14, 16, 18, 30, 36
The skin overlying the VM muscle was cleansed
using standard alcohol wipes. Each subject had
two bipolar 37 mm × 26 gauge sterile disposable
wire electrodes inserted by a physician, following
aseptic techniques, into the two portions of the
VM muscle. Due to the fact that the VMO and
VML are very close to one another, the indwelling
EMG electrode technique was chosen to avoid
recording the electrical activity from motor units
in close proximity to the muscle being studied.
No Exercise Abbreviation
1 Isotonic knee extension with the ankle plantarflexed.
2 Isotonic knee extension with the ankle dorsiflexed.
3 Isometric knee extension with the knee fixed at 90° of knee flexion.
4 Isometric knee extension with the knee fixed at 60° of knee flexion.
5 Isometric knee extension with the knee fixed at 30° of knee flexion.
6 Isometric knee extension with the knee fixed at 0° of knee flexion.
7 Isotonic hip flexion to 60° with the subject lying supine and the knee fully extended.
8 Isotonic hip adduction to 30° with the subject sidelying on their dominant side, and the nondominant
extremity placed on a bench at a standard height of 46 cm.
9 Isotonic hip adduction in the standing position, with the knee fully extended while attached to the
hip adduction cable of the Icarian Multi-station® exercise machine, with resistance equaling 10% of
the body weight of the subject.
10 Isotonic hip and knee extension to be performed while the subject is seated on a Cybex Eagle Leg
Press® machine with resistance equaling 50% of the body weight of the subject.
Table 2 Knee Rehabilitation Exercises.
KEPF
KEDF
ISO90
ISO60
ISO30
QS
SLR
HADSL
HADST
LP
Table 1 Human Subject Descriptive Data.
Range xSD
Age (years) 20–27 23.4 2.47
Weight (kg) 65.9–111.4 90.6 13.84
Distal VM fiber angle (degrees) 60 –80 69.3 6.28
Male Female Total
Number of subjects 21 3 24
00081.p65 06/13/2002, 4:47 PM110
EMG Analysis of the VMO and VML during Exercise 111
This crosstalk effect is frequently observed and
cited as a potential source of error in surface EMG
electrode studies of the QF muscles.
9
The elec-
trodes were placed according to an established
protocol,
36
with the VMO placement 4 cm supe-
rior and 3 cm medial to the superomedial aspect
of the patella, and the VML placement 15 cm
superior and 3 cm medial to the superomedial
aspect of the patella (Fig. 1). Having each subject
actively contract the quadriceps muscle group,
and viewing the resultant action potentials on
the oscilloscope screen ensured insertion of the
electrodes into both portions of the muscle. Stabi-
lization of the electrodes into the muscle bellies
was also performed according to an established
protocol.
34
The lead wires were secured to the
thigh of each subject by hypoallergenic tape,
and connected to a Cadwell 5200A (Cadwell
Instruments, Kennewick, WA) dual channel EMG
recorder. A preamplifier was used to provide
initial amplification of the EMG signal, thus re-
ducing background noise. The EMG signal was
processed and filtered within the Cadwell 5200A
system signal acquisition and processing compo-
nents. The amplifier bandwidth was preset from
30–10,000 Hertz (HZ). A common mode rejection
ratio was set at 100 decibels. Input impedance
was negligible, as is normal for fine wire electrode
studies, and was constant throughout the study.
All subjects performed three trials of each
exercise. Contractions were timed, started and
stopped on verbal command by the same investi-
gator for all subjects. Between each contraction,
subject comfort and electrode integrity were
checked, and the subjects were instructed to relax
until muscle EMG activity returned to zero on the
oscilloscope. Following completion of all of the
maneuvers, the wire electrodes were removed, the
insertion areas cleaned with alcohol, and small
adhesive bandage strips
applied to the insertion
sites if requested by the subjects, or advised by
the attending physician.
The simultaneous EMG amplitudes for both
portions of the VM muscle were viewed on the
EMG oscilloscope screen during each trial of each
exercise (Fig. 2) and measured in µV with the
Fig. 1 Electrode placements into the VM muscle. VMO placement is 3 cm medial and 4 cm superior to the superomedial
patellar pole. VML placement is 3 cm medial and 15 cm superior to the superomedial patellar pole.
00081.p65 06/13/2002, 4:47 PM111
112 J. K. Hubbard & S. Opersteny
Cadwell 5200s internal measuring features. The
maximum peak-to-peak EMG (PEMG) action
potential amplitude produced during the middle
3-seconds of each 5-second contraction was
utilized for data analysis. The amplitude was
measured from the most positive to the most
negative peak as suggested by the Nomenclature
Committee of the American Association of Electro-
diagnostic Medicine, and calculated within the
Cadwell Sierras internal system. The data were
then recorded and stored on paper data collection
sheets as well as in the memory of the EMG
device. The PEMG amplitude for each of the three
trials was averaged, then converted to a percen-
tage of maximal voluntary contraction (%MVC)
for statistical analysis. For each portion of the VM
muscle, the exercise that produced the greatest
averaged PEMG amplitude was designated as
that muscles maximum volitional contraction.
The %MVC for each of the ten exercises was
calculated by dividing the PEMG obtained during
each of the exercises into that portions max-
imal PEMG. The %MVC was then utilized for
statistical analysis to determine if any statistical
difference exists between the VMO and the VML
during performance of any of the selected study
exercises.
A randomized experimental research design
was incorporated in this study, with randomiza-
tion introduced by the order of performance of
the selected rehabilitation exercises. Limiting the
data collection session to a single event
25
con-
trolled internal validity. Thus, there was no effect
from history, selection, maturation, testing effects,
performance, or statistical regression. Carryover
effects were limited by giving each subject
sufficient rest periods between exercises to allow
the EMG action potentials to return to their
resting levels. This was indicated by absence of
electrical activity in the muscles at rest. Instru-
mentation reliability was controlled by utilizing
only one person, a board certified medical prac-
titioner, to calibrate and collect all of the EMG
data. External validity was limited to the portion
of the population possessing the characteristics
utilized for inclusion in the study.
Fig. 2 Simultaneous recording of VMO and VML action potentials during exercises viewed on Cadwell 5200 oscilloscope
screen.
00081.p65 06/13/2002, 4:47 PM112
EMG Analysis of the VMO and VML during Exercise 113
Analysis of the data was performed through
a Two-Factor (exercise × muscle) Repeated Mea-
sures Analysis of Variance design. An alpha level
of 0.05 was used as evidence of statistical signifi-
cance in all tests. The ten rehabilitation exercises
were designated as independent variables for the
first factor, and the VMO and VML portions of the
VM muscle were each designated as independent
variables for the second factor. The dependent
variables were the %MVC from both the VMO
and VML for each exercise condition. If significant
interactions or main effects were noted, post hoc
tests were performed to determine specific differ-
ences utilizing the Student Newman Keuls (SNK)
method for Pairwise Multiple Comparisons. All
statistical analyses were performed using Sigma-
Stat, version 2.0 (Jandel Scientific, San Rafael, CA).
RESULTS
The mean %MVC, with standard deviations and
difference between the means of both the VMO
and VML for each of the exercises preformed is
presented in Table 3. Both portions of the VM
muscle developed measurable EMG activity dur-
ing performance of each of the selected exercises.
Throughout this study, there were no main effects
seen between either level of muscle. No statisti-
cally significant difference in the %MVC was
recorded between the VMO and the VML during
any single exercise, or over the entire set of exer-
cises. The difference in %MVC between the two
portions of the muscle was only 0.018 µV over
the entire set of exercises, which was also statisti-
cally nonsignificant. SNK post hoc tests showed a
significant main effect between levels of exercise.
Multiple pairwise comparisons showed that the
QS exercise developed significantly more %MVC
than all of the other test exercises. Additional sig-
nificant differences existed between a few of the
other test exercises as indicated in the legend of
Fig. 3. Results of the Two-Factor Repeated Mea-
sures Analysis of Variance test (Table 4) showed
a significant interaction of the two main effects,
with a statistical power between the interaction
terms of 0.59. As depicted in Fig. 3, the interac-
tion occurred during the QS, HADSL, and
HADST exercises where the VMO produced a
greater, although statistically non-significant,
%MVC than did the VML. For all of the other ex-
ercises, the VML produced a greater, but statisti-
cally non-significant %MVC than did the VMO.
Despite the interaction shown, and significant
main effect for levels of exercise, results from
the Two-Way Repeated Measures Analysis of
Variance indicated that there were no significant
No Exercise VMO VML
X SD X SD Mean Difference
1. KEPF 33.8% 15.5% 39.5% 23.3% 5.7%
2. KEDF 34.6% 15.3% 44.0% 28.7% 9.4%
3. ISO90 59.2% 15.2% 60.5% 23.0% 1.3%
4. ISO60 45.0% 15.7% 49.9% 19.5% 4.9%
5. ISO30 53.7% 26.2% 56.4% 22.9% 2.7%
6. QS 91.5% 14.6% 87.3% 19.0% 4.2%
7. SLR 36.4% 19.7% 37.8% 20.6% 1.4%
8. HADSL 26.7% 15.4% 25.4% 10.9% 1.3%
9. HADST 42.4% 32.3% 38.9% 27.0% 3.5%
10. LP 51.1% 23.0% 52.3% 28.8% 1.2%
Table 3 Average %MVC Produced during Knee Rehabilitation Exercises with %SD.
00081.p65 06/13/2002, 4:47 PM113
114 J. K. Hubbard & S. Opersteny
Source of Variance DF SS MS F p
Subject 23 3.1749 0.13804
Exercise 9 13.3244 1.48049 20.901 < 0.0001
Exercise × Subject 207 14.6625 0.07083
Muscle 1 0.0376 0.03761 0.575 0.4559
Muscle × Subject 23 1.5038 0.06538
Exercise × Muscle 9 0.1923 0.02136 2.224 0.0218*
Residual 207 1.9883 0.00961
Total 479 34.8837 0.07283
Table 4 Summary Table for Two-Factor Repeated Measures Analysis
of Variance Procedure.
Dependent Variable: %MVC
Power for Exercise × Muscle: (p<0.05) = 0.59
*Post hoc pairwise comparisons for interaction were not significant.
Fig. 3 SNK pairwise multiple comparison results. A: Significantly different from all other exercises. B: Significantly
different from KEPF, KEDF, SLR, HADSL, HADST. C: Significantly different from KEPF and HADSL. D: Significantly
different from HADSL. Asterisk (*) indicated exercises where VMO produced greater %MVC than VML.
(Key: KEPF = knee extension in plantar flexion; KEDF = knee extension in dorsiflexion; IOS90 = isometric knee extension at
90 degrees; ISO60 = isometric knee extension at 60 degrees; ISO30 = isometric knee extension at 30 degrees; QS = quadriceps
set; SLR = straight leg raise; HADSL = hip adduction in sidelying; HADST = hip adduction in standing; LP = leg press).
00081.p65 06/13/2002, 4:47 PM114
EMG Analysis of the VMO and VML during Exercise 115
differences between the two portions of the VM
muscle during performance of any of the exer-
cises. Examination of the data and statistical
analysis failed to reject the stated null hypoth-
eses, therefore, it continues to remain tenable.
DISCUSSION
Grants Method of Anatomy
3
describes skeletal
muscles as having features in common which
distinguish them from one another. They all have
a distinct origin and insertion where the muscle
attaches to the bones, and an aponeurotic plane of
fascial tissue (epimysium) that separates muscles
from one another. Each muscle will have a distinct
nerve innervation that allows it to be stimulated
and contracted independently of other muscles. A
vascular network of arteries and veins supplies
each muscle with oxygen and nutrients and
cleanses it of waste products. These nerves
and vessels frequently travel together in a neuro-
vascular bundle in the fascial planes before
entering the muscle belly and distributing their
terminal branches to the muscle fibers.
Motor nerve fibers frequently branch as they
spread throughout a muscle in order to supply in-
nervation to each motor unit.
3, 11
Although most
muscles are innervated from their deep side, it
is not unusual for a nerve to travel in a fascial
plane superficially prior to entering a muscle
belly. The nerve to the VM is described
11
as hav-
ing a short branch entering the muscle proximally,
and a longer nerve fiber traveling in the adductor
canal prior to entering the muscle belly in its
midportion. The femoral nerve, which innervates
the entire QF muscle complex, is composed of
nerve roots from the L2–4 spinal segments. No
descriptions of motor nerve innervation patterns
were found to validate Thiranagamas
32
claim of
the femoral nerve containing fibers from the L1
spinal segment. Gunal et al.,
13
stated that an addi-
tional nerve branch existed to the VMO, arising
from the saphenous nerve. The saphenous nerve
is the terminal branch of the femoral nerve, con-
taining sensory fibers only. One of the terminal fi-
bers of the saphenous nerve is the descending
genicular nerve, a branch of which innervates the
medial knee joint capsule.
3, 11
This nerve runs its
normal course through the fibers of the distal VM
muscle, on its way to the medial capsule of the
knee joint. Since it is contained in a neurovascular
bundle, this nerve may have been thought to be
an additional motor nerve providing additional
innervation the distal VM muscle. Contrary to
Thiranagama
32
and Gunal et al.s
13
claims, all of
the distal VM muscles examined in Hubbard
et al.s
15
large scale cadaveric study were seen
as being innervated by nerve fibers originating
directly from the femoral nerve.
An intact muscle contracts to produce move-
ment at a joint when it receives sufficient internal
or external stimulation to depolarize the cell
membrane and produce an action potential. Direct
external stimulation of innervated muscle tissue,
such as is performed routinely in physical therapy
clinics, and affects only those terminal motor
nerve branches and the motor fibers they inner-
vate through which the electrical current passes.
This generally is limited to the portion of a muscle
that lies between the stimulating electrodes. If a
portion of a muscle is unable to contract through
injury or disease, or only a portion of a muscle is
able to receive sufficient stimulation to initiate a
contraction, then the effects of the muscle contrac-
tion on its insertion will be altered or diminished
from what would normally be expected in an
intact muscle. The ability to selectively contract a
portion of a muscle with electrical stimulation, or
relax a portion of a muscle through biofeedback
training does not reflect the function of the muscle
as an intact structure in vivo. Studies that have
advocated the use of these modalities for selective
training and strengthening of the VMO
4, 21, 36
also
stated that the effects were not seen without
00081.p65 06/13/2002, 4:47 PM115
116 J. K. Hubbard & S. Opersteny
the assistance of the modality. Additionally, these
studies noted that no carryover effects were
demonstrated either following the treatment
sessions or by the conclusion of the studies. This
indicates that true training of the muscle through
use of these procedures did not occur.
None of the subjects of this study stated that
they were unable to perform maximal voluntary
contractions or elicit an EMG signal from both
parts of the VM muscle due to perceived or real
discomfort from the indwelling electrodes. Visual
observation of the human subjects who partici-
pated in this study uniformly demonstrated an
inability of any of the subjects to contract the
VMO independently from other portions of the
QF muscle group. None of the subjects of either
gender who participated in this study were able to
actively adduct the patella, despite the presence of
a hypertrophied distal portion of the VM muscle
and insertion of the distal portion of the VM well
past midline of the patella. In all subjects tested,
both portions of the VM muscle were electrically
active during performance of all of the selected
exercises.
Throughout this study, there were no statis-
tically significant differences observed in the
%MVC measured between the two portions of the
VM muscle. All differences observed were due to
differences in the exercises performed and not to
differing amounts of muscle EMG activity. There
were no instances during the testing procedures
where either part of the VM muscle was able to
contract without electrical activity being present
and recordable in the other portion of the muscle.
The differences present in %MVC between the
two portions of the VM muscle were not great
enough to exclude the possibility that the differ-
ences were due to random sampling variability.
The effect of the different exercises on %MVC
produced did not depend on which portion of the
VM muscle was involved. There was no evidence
seen in this study, or in any of the other refer-
enced studies that the VMO can be selectively
activated or strengthened independently from
the VML. Therefore, any activity or exercises
that elicits an increase in EMG action potential
amplitude in either portion of the VM muscle will
be accompanied by an increase in activity in the
other portion of the muscle.
Clinicians attempting to rehabilitate intact
muscle units must be able to rely on clinical
research that has demonstrated the effectiveness
of techniques that utilize the entire muscle under
question in their research methodologies. Based
on the findings of this and other studies, there
is not sufficient evidence to support the concept
that the VMO is an anatomically or functionally
separate muscle from the VML in the QF muscle
group. In the original anatomical study by Lieb
and Perry,
22
the VMO was identified as being a
separate muscle from the VML due to a change
in the VM fiber angles, and the presence of an
areolar fascial plane dividing the VM into two
portions. Subsequent anatomical studies
5, 28, 33
indicated that the VM was divisible into two
portions in all specimens examined, based on the
change in muscle fiber angle, and variable nerve
innervation patterns. A large-scale cadaver dissec-
tion study
15
showed that although the VM muscle
did vary in its fiber orientation from more vertical
proximally to more horizontal distally, there were
no distinct fascial planes of epimysium separating
the two portions of the VM muscle from each
other and no innervation to the distal portion
from the saphenous nerve. Additionally, there
continues to be no instances of EMG recording of
VMO muscle activity without concurrent VML
muscle activity of similar amplitude.
CONCLUSION
The results of this study do not support published
claims that the VMO is a separate muscle from
the VML. Statistical analysis of in vivo VM muscle
00081.p65 06/13/2002, 4:47 PM116
EMG Analysis of the VMO and VML during Exercise 117
activity in human subjects indicated that no differ-
ences existed in %MVC between the VMO and
VML portions of the VM muscle during perfor-
mance of selected exercises. During performance
of the exercises included in this study, there were
no instances where the VMO was able to contract
separately from the VML. There were also no
instances observed in which either the VMO or
VML were able to generate measurable electrical
activity during any of the selected exercises with-
out simultaneous generation of measurable electri-
cal activity in the other portion of the muscle. The
VMO does not possess a unique origin and inser-
tion, a distinct fascial plane of epimysium, a sepa-
rate nerve innervation, or specific agonistic action
with an antagonistic muscle. It does not possess
the requisite characteristics of a distinct muscle
entity, and should not be considered as a separate
muscle entity or strengthened separately from
the VML.
References
1. Adler N, Perry J, Kent B, Robertson K. Electromyogra-
phy of the vastus medialis oblique and vasti in normal
subjects during gait. Electromyogr Clin Neurophysio 233:
643–649, 1983.
2. Basmajian JV, Hardin TP, Regenos EM. Integrated
actions of the four heads of the quadriceps femoris: An
electromyographic study. Anat Rec 172:15–20, 1972.
3. Basmajian JV. Grants Method of Anatomy, Williams &
Wilkins, Baltimore, MD, 1980, pp. 20–26.
4. Bohannon R. Effect of electrical stimulation to the
vastus medialis muscle in a patient with chronically
dislocating patella. Phys Ther 63:1445–1447, 1983.
5. Bose K, Kanagasuntheram R, Osman MBH. Vastus
medialis oblique: An anatomic and physiologic study.
Orthopedics 3:880–883, 1980.
6. Cerny K. Vastus medialis oblique/vastus lateralis mus-
cle activity ratios for selected exercises in persons with
and without patellofemoral pain syndrome. Phys Ther
75:672–683, 1995.
7. Douchette SA, Child DD. The effect of open and closed
chain exercise and knee joint position on patellar track-
ing in lateral patellar compression syndrome. J Orthop
Sports Phys Ther 23:104–110, 1996.
8. Grabiner MD, Campbell KR, Hawthorne DL, Hawkins
DA. Electromyographic study of the anterior cruciate
ligament-hamstrings synergy during isometric knee
extension. J Orthop Res 7:152–155, 1989.
9. Grabiner MD, Koh TJ, Miller GF. Fatigue rates of
vastus medialis oblique and vastus lateralis during
static and dynamic knee extension. J Orthop Res 9:
391–397, 1991.
10. Grabiner MD, vonHaeffen L, Kohn TJ. Concurrent hip
joint adduction and knee joint extension moments do
not influence quadriceps femoris excitation (abstract).
Med Sci Sports Exerc 24(Suppl): S128, 1992.
11. Gray H. Grays Anatomy, 30th Am edn., edited by CM
Goss, Lea & Febiger, Philadelphia, 1973, p. 377.
12. Gryzlo SM, Patek RM, Pink M, Perry J. Electromyogra-
phic analysis of knee rehabilitation exercises. J Orthop
Sports Phys Ther 20:36–43, 1994.
13. Gunal I, Arac S, Sahinoglu K, Birvar K. The innervation
of the vastus medialis obliquus. J Bone Joint Surg 74B:
624, 1992.
14. Hanten WP, Schulthies SS. Exercise effect on electro-
myographic activity of the vastus medialis oblique and
vastus lateralis muscles. Phys Ther 70:561–565, 1990.
15. Hubbard JK, Sampson HW, Elledge JR. Prevalence and
morphology of the vastus medialis oblique muscle in
human cadavers. Anat Rec 249:135–142, 1997.
16. Karst GM, Jewett PD. Electromyographic analysis of
exercises proposed for differential activation of medial
and lateral quadriceps femoris muscle components.
Phys Ther 73:286–299, 1992.
17. Karst GM, Willett GM. Onset timing of electromyogra-
phic activity in the vastus medialis oblique and vastus
lateralis muscles in subjects with and without patello-
femoral pain syndrome. Phys Ther 75:813–823, 1995.
18. King, MA, Boucher JP, Cyr A, Lefebvre R. Evaluation
of VMO/VL ratio with alteration of patellofemoral me-
chanics and hip adduction in patello-femoral pain syn-
drome (PFPS) (abstract). Med Sci Sports Exerc 24(Suppl):
S148, 1992.
19. Koh TJ, Grabiner MD, DeSwart RJ. In vivo tracking of
the human patella. J Biomech 25:637–643, 1992.
20. Komi PV. Relationship between muscle tension, EMG,
and velocity of contraction under concentric and eccen-
tric work. In: New Developments in Electromyography and
Clinical Neurophysiology, edited by JE Desmedt, Karger,
Basel, Switzerland, 1973, pp. 596–606.
21. LeVeau BF, Rogers C. Selective training of the vastus
medialis muscle using EMG biofeedback. Phys Ther 60:
1410–1415, 1980.
22. Lieb FJ, Perry J. Quadriceps function. An anatomical
and mechanical study using amputated limbs. J Bone
Joint Surg [Am] 50:1535–1548, 1968.
00081.p65 06/13/2002, 4:47 PM117
118 J. K. Hubbard & S. Opersteny
23. Lieb FJ, Perry J. Quadriceps function. An electromyo-
graphic study under isometric conditions. J Bone Joint
Surg [Am] 53:749–758, 1971.
24. Lippold OCJ. The relationship between integrated
action potentials in a human muscle and its isometric
tension. J Physiol 177:492–499, 1952.
25. Powers CM. The reliability of obtaining VL:VMO and
VL:VML EMG ratios using indwelling wire electrodes
(abstract). Phys Ther 76(Suppl): S87, 1996.
26. Reynolds L, Levin TA, Medeiros JM, Adler NS, Hallum
MS. EMG activity of the vastus medialis oblique and
the vastus lateralis in their role in patellar alignment.
Am J Phys Med 62:61–70, 1983.
27. Salzman A, Torburn L, Perry J. Contribution of rectus
femoris and vasti to knee extension. An electromyogra-
phic study. Clin Orthop 290:236–243, 1993.
28. Scharf W, Weinstabl R, Orthner E. Anatomical separa-
tion and clinical importance of two different parts of
the vastus medialis muscle. Acta Anat 123:108–111,
1985.
29. Soderberg GL, Cook TM. An electromyographic analy-
sis of quadriceps femoris muscle setting and straight
leg raising. Phys Ther 63:1434–1438, 1983.
30. Souza DR, Gross MT. Comparison of vastus medialis
obliquis: Vastus lateralis muscle integrated electromyo-
graphic ratios between healthy subjects and patients
with patellofemoral pain. Phys Ther 71:310–316, 1991.
31. Speakman HG, Weisberg J. The vastus medialis contro-
versy. Physiotherapy 63:249–254, 1977.
32. Thiranagama R. Nerve supply of the human vastus
medialis muscle. J Anat 170:193–198, 1990.
33. Weinstabl R, Scharf W, Firbas W. The extensor appara-
tus of the knee joint and its peripheral vasti: Anatomic
investigation and clinical relevance. Surg Radiol Anat
11:17–22, 1989.
34. Westfall DC, Worrell TW. Anterior knee pain syndrome:
Role of the vastus medialis oblique. J Sport Rehabil
1:317–325, 1992.
35. Winter DA. Biomechanics and Motor Control of Human
Movement, 2nd edn., John Wiley & Sons, Inc., New York,
1990, pp. 207–209.
36. Wise HH, Fiebert IM, Kates JL. EMG biofeedback as
treatment for patellofemoral pain syndrome. J Orthop
Sports Phys Ther 6:95–103, 1984.
00081.p65 06/13/2002, 4:47 PM118
