The Knee 10 (2003) 277–281

Synergy of medial and lateral hamstrings at three positions of tibial

rotation during maximum isometric knee flexion夞
Olfat Mohameda,*, Jacquelin Perryb,c, Helen Hislopd
a
Department of Physical Therapy, California State University, Long Beach, CA, USA
b
Pathokinesiology Laboratory, Rancho Los Amigos National Rehabilitation Center, Downey, CA, USA
c
Department of Orthopedic, University of Southern California, Los Angeles, CA, USA
d
Department of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, CA, USA

Accepted 24 September 2002

Abstract

Rotation of the knee has been used to isolate the strength of the medial and lateral hamstrings during manual testing of the
knee flexors. The purpose of this study was to determine if medial and lateral rotation of the knee during manual knee flexor
strength testing increased the electromyographic activity of the respective hamstrings. Twenty-three women between 22 and 36
years old with no history of lower extremity injury or disease participated in the study. Indwelling fine wire electrodes were used
to record EMG activity of the medial (semitendinosus and semimembranosus) and lateral (long and short heads of the biceps
femoris) hamstring muscles during maximally resisted knee flexion with neutral, medial, and lateral rotation of the knee. Repeated
measures analysis of variance with post hoc Bonferroni adjustments were used to compare EMG activity across the three tests.
EMG activity increased significantly for the target hamstrings during ipsilateral rotation (P-0.05). The semitendinosus had a
mean activity of 109% Max. during medial rotation as opposed to 95% Max. during lateral rotation. The semimembranosus
averaged 107 and 89% Max. in medial and lateral rotation respectively. Conversely, both the long and short head of the biceps
muscle showed significantly higher activity (P-0.05) during lateral compared to medial rotation (110 and 108% compared to 93
and 97%, respectively). Even though the differences are statistically significant they ranged from 2 to 13% only of maximum
activity, the clinical importance of this small change in EMG activity is questionable.
䊚 2002 Elsevier Science B.V. All rights reserved.

Keywords: EMG; Electromyography; Hamstrings; Hamstring muscles; Manual muscle testing

1. Introduction strings separately which is not possible using isokinetic

Hamstrings strength training is an integral part of
rehabilitation protocols after knee injuries. Clinicians
have especially concentrated on increasing hamstrings
strength in ACL-deficient patients as well as those who
had ACL reconstructive or repair surgeries to provide
anterior tibial stability. To assess the strength of the
hamstring muscles, clinicians commonly use manual
muscle testing (MMT) because of its ease and conven-
ience. In addition, the manual technique allows for
testing the medial and lateral components of the ham-
夞 The Institutional Review Board of Rancho Los Amigos Medical
Center.
*Corresponding author. Tel.: q1-562-985-7139; fax: q1-562-985-
4069.
E-mail address: osm@csulb.edu (O. Mohamed).
equipment. The purpose of this study was to document
the ability of MMT to ‘bias’ the participation of the
medial and lateral components of the hamstrings through
tibial rotation using indwelling fine wire electrode
technique.
2. Review of the literature
Several studies have demonstrated the role of the
hamstring muscles in preventing anterior tibial displace-
ment w1–3x. Knee rehabilitation protocols designed for
ACL-deficient patients or for those who had ACL repair
or reconstruction regularly include a hamstring-strength-
ening component w4,5x. Testing hamstrings’ strength by
applying resistance to knee flexion challenges the func-
tion of the muscles as knee flexors. Modification of this

0968-0160/03/$ - see front matter 䊚 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 6 8 - 0 1 6 0 Ž 0 2 . 0 0 1 4 0 - 0
278 O. Mohamed et al. / The Knee 10 (2003) 277–281
test by placing the tibia in medial and lateral rotation
during knee flexion is suggested to better isolate the
medial and lateral hamstrings from each other w6,7x.
Similar modifications of other MMT techniques are used
to isolate other muscles, such as knee flexion during toe
standing differentiating the action of the soleus muscle
from that of the gastrocnemius. These modifications
however, are mostly based on clinical experience and
anatomical information rather than research data.
Perry et al. using wire electrode EMG investigated
the specificity of MMT of the calf muscles w8x. Two
variations of coming to toe standing in the upright
position were used to isolate the muscles of the triceps
surae; the knee was extended for the gastrocnemius test
and flexed approximately 408 for the soleus test. Results
showed that the standard test for the gastrocnemius
induced maximum activity in this muscle but, at the
same time, elicited strong soleus activity, approximately
87% of its maximum. Standing with the knee flexed
provided better soleus isolation as it markedly reduced
gastrocnemius activity to only 21% of its maximum
activity while inducing maximum soleus action. The
tibialis posterior participated in both tests at approxi-
mately 34% of its maximum activity w8x.
Using surface electrodes, Fiebert et al. studied the
effect of tibial rotation on the root mean square (RMS)
of the medial and lateral hamstrings at 458 of knee
flexion against a resistance equivalent to 5% of the
subject’s body weight w9x. They found that the medial
hamstrings generated 40% more activity in medial rota-
tion compared to lateral rotation while the lateral ham-
strings had a 59% increase with lateral tibial rotation
compared to medial rotation. Fiebert et al. acknowledged
the several limitations of the surface electrode technique
including the major problem of cross talk. Another
problem arises from lack of normalization of EMG data,
which attempts to reduce the sampling differences
between electrodes and allows for better comparison
between EMG activities of different muscles.
In strength training and MMT, resistance is usually
applied at a much higher level than the 5% of body
weight used in Fiebert’s study, the question of partici-
pation of the different components of the hamstrings
during maximum effort remains unanswered.
3. Methods
3.1. Subjects
Twenty-three adult women volunteers between the
ages of 22 and 36 years participated in the study.
Participants reported no history or signs of musculo-
skeletal or neurological dysfunction that would affect
motor performance in either lower extremities.
3.2. Instrumentation
Myoelectric activity (EMG) was sensed by fine wire
electrodes (50-mm) of nickel–chromium alloy with a
bipolar configuration. The wire electrodes were insulated
except for their 2 mm tips and passed through the
cannula of a 25-guage hypodermic with the distal ends
staggered and folded over the needle tip. After insertion
into the muscle, the distal ends of the wires were
connected to the corresponding ground plate. The EMG
signal from each muscle was transmitted through an
eight channels FMyFM1 telemetry system to a DEC 11y
23 computer.2 The computer sampled each channel at
2500 Hz. The modulation of each channel was adjusted
to insure 100–1000 Hz frequency response, with an
overall gain of 1000. Calibration was achieved by
passing a 1 V, 400 Hz signal through the ground
electrodes; a baseline noise level above 50 mV was not
accepted. The instrumentation also included a panel of
seven oscilloscopes, which allowed continuous, on line,
monitoring of the signal.
Digitization of the data was accomplished with the
aid of a DEC 11y23 minicomputer with an A–D
converter. All EMG data were digitized at 2500 samples
per second.
3.3. Procedure
Prior to testing, all procedures were explained to each
subject. All subjects provided informed consent prior to
participation and were informed about their right to end
participation at any time. All testing was done at the
Pathokinesiology Laboratory, Ranch Los Amigos
National Rehabilitation Center. The sterilized fine wire
electrodes were inserted using Basmajian’s single needle
technique into the mid-belly of the of the following four
muscles; semitendinosus, semimembranosus, biceps
femoris (long head) and biceps femoris (short head). A
mild electrical stimulation was delivered through the
wires to confirm electrode placement. Placement was
confirmed with a palpable andyor visible contraction of
the muscle stimulated.
Subjects were positioned prone and stabilized with a
belt over the buttocks area. The knee of the tested
extremity was flexed to 708 and placed in maximum
medial tibial rotation to test the medial hamstrings
‘medial test’. The subject was asked to hold that position
(knee flexed to 708 and the tibia maximally rotated
medially) while the principal investigator applied max-
imal manual resistance to knee flexion just above the
ankle joint. The subject was verbally encouraged to give
their best effort during contraction. The ‘lateral test’
1
Model 2600S, BioSentry Telemetry INC., 20720 Earl Street,
Torrance, CA 90503, USA.
2
Digital Equipment Co., 146 Main Street, Maynard, MA 01754,
USA.
 O. Mohamed et al. / The Knee 10 (2003) 277–281 279

Table 1
EMG during MMT (% Max.) means and SEM

Muscle Medial test (M) Neutral test (N) Lateral test (L) ANOVA Significant post hoc
P value
Mean SEM Mean SEM Mean SEM
STEN 109 2 96 3 95 2 0.003 M vs. L, N
SMEM 107 3 105 3 89 4 0.003 M vs. L
BFLH 93 5 97 4 110 3 0.025 L, N vs. M
BFSH 97 3 92 2 108 3 0.003 L vs. M, N
Abbreviations: BFLH, long head of biceps femoris; BFSH, short head of biceps femoris; SEM, standard error of the mean; SMEM, semimem-
branosus; STEN, semitendinosus.

was adapted for the lateral hamstring muscles by posi-
tioning the leg in maximum lateral rotation and repeating
the procedure. Another manual resistance test, the ‘neu-
tral test’ was carried out by applying resistance to knee
flexion while the knee was in neutral rotation. The order
of the test was randomized for each subject.
To control for the variability of electrode placement
and allow for comparisons of EMG intensity between
muscles and subjects, EMG data were normalized to the
average of the three tests: ‘medial’, ‘lateral’ and ‘neu-
tral’, for each subject. EMG data from each test is
expressed as a ratio of that average and cited as %
maximum (% Max.).
3.4. Data analysis
All statistical procedures were performed with BMDP
statistical software.3 Repeated measure analysis of vari-
ance (ANOVA) was conducted to compare the EMG
3
SPSS Inc, 444 N Michigan Avenue, Chicago, IL 60611, USA.
activity across the three MMT positions for all tested
muscles. Bonferroni adjustment of the alpha level was
used for post hoc comparisons when ANOVA results
were significant. An overall significance level of 0.05
was accepted.
4. Results
A statistically significant increase of EMG activity
(P-0.05) was found in the medial hamstrings with
medial rotation (Table 1 and Fig. 1) over lateral rotation.
The semitendinosus had a mean activity of 109% Max.
during medial rotation as opposed to 95% Max. during
lateral rotation. The semimembranosus averaged 107
and 89% Max. in medial and lateral rotation, respective-
ly. Conversely, both the long and short head of the
biceps muscle showed significantly higher activity (P-
0.05) during lateral compared to medial rotation.
Modification of the hamstring manual muscle test to
place the leg in a neutral position slightly reduced the
activity of the medial hamstrings compared to the medial

Fig. 1. EMG activity of medial and lateral hamstrings during maximum resistance to knee flexion; with the knee in medial, lateral or neutral
rotation. BFLH, biceps femoris long head; BFSH, biceps femoris, short head. STEN, semitendinosus; SMEM, semimembranosus.
280 O. Mohamed et al. / The Knee 10 (2003) 277–281
test, but only the semitendinosus had significantly lower
activity. The mean EMG value for the semitendinosus
was 109% Max. in the medial hamstring test vs. 96%
with the leg in neutral (P-0.05). When the ‘lateral
test’ and the ‘neutral test’ were compared for the lateral
hamstrings, there was a significantly higher activity
during ‘lateral testing’ for both heads of the biceps
femoris muscle (P-0.05) (Table 1 and Fig. 1).
5. Discussion
The textbooks on MMT recommend using knee rota-
tion to differentiate the flexor strength of the medial
and lateral hamstrings. The leg is laterally rotated to
isolate the action of the biceps femoris, conversely,
medial rotation is assumed to isolate the medial ham-
strings w6,7,10x. The ability to isolate the strength medial
versus lateral hamstrings via MMT was not supported
by the data in this study.
With the leg medially rotated the lateral hamstring
muscles, which theoretically should show low activity,
displayed an EMG level that was 95% of their average
maximum value for all testing positions. Likewise,
during the ‘lateral test’, the semitendinosus and semi-
membranosus muscles had an EMG activity level of
92% of their average maximum. The neutral position
produced a level of EMG in all the muscles that was
between 92 and 105% Max. These data, therefore,
contradict the claim for hamstring muscle testing in
which rotation of the leg is used to isolate the strength
of the medial vs. lateral groups. It is clear that the three
test positions challenged all three muscles, which was
evidenced by the high EMG activity levels recorded in
all positions.
The participation of the hamstrings as a group in all
three varieties of the muscle test is in agreement with
their anatomical capabilities. The biarticular hamstring
muscles all arise from the ischial tuberosity, they all
pass behind the knee joint to be inserted in the tibia and
the fibula. All, therefore, are capable of flexing the knee
w11x. When referring to the hamstring muscles as a
group, the short head of biceps femoris is usually not
included because it is a monoarticular knee flexor unlike
the rest of the hamstrings. Its alignment however, would
contribute to both knee flexion and tibial lateral rotation
w12x which should not be overlooked when evaluating
knee flexion torque.
Our results are in agreement with those of Fiebert et
al. who documented with surface EMG that both medial
and lateral hamstrings participated in knee flexion with
medial and lateral tibial rotation when their subjects
resisted a weight that is equivalent to 5% body weight
w9x. Fiebert et al. however, reported much lower level
of participation of the medial hamstrings in the lateral
test and the lateral hamstrings in the medial test, they
found that the medial hamstrings created 40% (indicated
by RMS values) more activity with medial tibial rotation
while the lateral hamstrings had a 59% increase with
lateral tibial rotation. This difference between the Fiebert
and our data confirms that the medial and lateral
hamstrings differ in their rotatory capability but have a
common flexor function and this dominates when max-
imum contraction is elicited. In our study the medial
and lateral tests by altering leg position differed by only
13% of average maximum activity. Despite the fact that
this difference was statistically significant, its clinical
relevance is questionable.
Our findings are in parallel with those of Perry et al.,
who were unable to isolate the gastrocnemius muscle
using the gastrocnemius muscle test w8x. They demon-
strated, using wire electrode EMG, that the gastrocne-
mius standing heel rise test with an extended knee did
not exclude the soleus. To the contrary, the standing
heel rise test induced soleus activity at approximately
87% of its maximum w8x. Likewise, Katz et al., in a
similar procedure, showed that the scapular rotation test,
which had been claimed to isolate the serratus anterior,
did not fulfill that task. Resistance to scapular rotation
was accompanied by considerable activity of the levator
scapula, and all three portions of the trapezius muscle
w13x. It follows that any weakness in the latter muscles
would affect the strength grade assigned to the serratus
anterior.
Some studies showed atypical hamstring electromyo-
graphic activity of either the medial or lateral hamstrings
during gait and other athletic activities in ACL-deficient
patients w14,15x. Boerboom et al. showed that this
atypical pattern was not consistent among patients; some
had increased EMG activity of the medial and others of
the lateral hamstrings w14x. While these recordings
indicate rotatory correction of knee position, these dif-
ferences at submaximal levels are not likely to be
identified with a maximal manual resistance test.
Weakness of any of the four knee flexor muscles that
were tested in this study would affect the grade assigned
to the knee flexors whether the test was done with or
without rotation of the leg. It follows that isometric
training of the knee flexors will elicit activity in all four
muscles. We disagree with the recommendation of Fie-
bert et al. of performing two separate tests of medial
and lateral hamstrings.
6. Conclusion
Based on the data of this study, selective separation
of the strength of the medial vs. lateral hamstring
muscles is not feasible with published manual muscle
tests. Despite a statistical difference in the intensity of
EMG activity of the medial and lateral hamstring mus-
cles by modifying the position of leg rotation, the
differences averaged only 13%. The clinical significance
of this difference is questionable. Until a method of
 O. Mohamed et al. / The Knee 10 (2003) 277–281
isolating the activity of the medial and lateral hamstrings
is found, clinicians should use only one test for the
hamstring muscles. Further investigation is needed to
document the participation of medial and lateral ham-
strings in different dynamic exercises.
References
w1x Roberts CS, Rash GS, Honaker JT, Wachowiak MP, Shaw JC.
A deficient anterior cruciate ligament does not lead to quadri-
ceps avoidance gait. Gait Posture 1999;10(3):189 –199.
w2x Pandy MG, Shelburne KB. Dependence of cruciate-ligament
loading on muscle forces and external load. J Biomech
1997;30(10):1015 –1024.
w3x Li G, Rudy TW, Sakane M, Kanamori A, Ma CB, Woo SL.
The importance of quadriceps and hamstring muscle loading
on knee kinematics and in situ forces in the ACL. J Biomech
1999;32(4):395 –400.
w4x Reid D. Current concepts in rehabilitation of the anterior
cruciate deficient knee. Current Orthop 1993;7:101 –105.
w5x Shelbourne KD, Nitz P. Accelerated rehabilitation after anterior
cruciate ligament reconstruction. Am J Sports Med
1990;18(3):292 –299.
w6x Hislop JH, Montgomery J. Daniels and Worthingham’s muscle
testing. 7th ed. Philadelphia: WB Saunders, 2002.
281
w7x Kendall F, McCreary E. Muscle testing and function. Baltimore,
MD: Williams & Wilkins, 1993.
w8x Perry J, Easterday CS, Antonelli DJ. Surface versus intramus-
cular electrodes for electromyography of superficial and deep
muscles. Phys Ther 1981;61(1):7 –15.
w9x Fiebert IM, Roach KE, Fingerhut B, Levy J, Schumacher A.
EMG activity of medial and lateral hamstrings at three posi-
tions of tibial rotation during low-force isometric knee flexion
contractions. J Back Musculoskeletal Rehab 1997;8:215 –222.
w10x Clarkson HM. Musculoskeletal assessment (joint range of
motion and manual muscle strength). 2nd ed.Lippincot Wil-
liams & Wilkins, 2000.
w11x Perry J, Bekey GA. EMG-force relationships in skeletal muscle.
Crit Rev Biomed Eng 1981;7(1):1 –22.
w12x Wickiewicz TL, Roy RR, Powell PL, Edgerton VR. Muscle
architecture of the human lower limb. Clin Orthop
1983;179:275 –283.
w13x Katz D, Gronley J, Perry J, et al. EMG comparison of two
manual muscle tests for the serratus anterior muscle. Abstract
Phys Ther 1983;63:765.
w14x Boerboom AL, Hof AL, Halbertsma JP, et al. Atypical ham-
strings electromyographic activity as a compensatory mecha-
nism in anterior cruciate ligament deficiency. Knee Surg Sports
Traumatol Arthrosc 2001;9(4):211 –216.
w15x Branch TP, Hunter R, Donath M. Dynamic EMG analysis of
anterior cruciate deficient legs with and without bracing during
cutting. Am J Sports Med 1989;17(1):35 –41.