Original Research

Hip Strength and Knee Pain in High School Runners:

A Prospective Study
Jonathan T. Finnoff, DO, Mederic M. Hall, MD, Kelli Kyle, ATC,
David A. Krause, PT, DSc, Jim Lai, Jay Smith, MD

Objective: To determine whether pre-injury hip muscle weakness is associated with the
development of patellofemoral pain (PFP) in high school running athletes.
Design: Prospective cohort study.
Setting: Academic institution sports medicine center.
Participants: High school running athletes.
Methods: Baseline hip strength of high school running athletes was assessed at the
beginning of the running season. Strength testing was repeated in athletes who developed
PFP. Peak hip muscle strengths and strength ratios were compared between the injured and
non-injured groups.
Results: Six injuries occurred in 5 of the 98 subjects who completed the study. The
baseline hip external-to-internal strength ratio was lower in injured than in uninjured
subjects (P ⫽ .008). In the injured group, hip abduction and external rotation strengths
decreased from pre-injury to post-injury (P ⫽ .002 and P ⫽ .01, respectively). Logistic
regression analysis demonstrated that a greater baseline hip abduction strength (odds ratio ⫽
5.35, 95% confidence interval [CI] 1.46-19.53; P ⬍ .01) and abduction-to-adduction
strength ratio increased the risk of injury (odds ratio ⫽ 14.14, 95% CI 0.90-221.06; P ⫽
.05), and a greater pre-injury hip external-to-internal rotation strength ratio decreased the
risk of injury (odds ratio ⬍ 0.01, 95% CI ⱕ .01, 0.44; P ⫽ .02).
Conclusions: The findings of the current study suggest that stronger pre-injury hip
abductors (particularly in relation to their hip adductors) and weaker pre-injury hip
J.T.F. Department of Physical Medicine and
external rotators (particularly in relation to their hip internal rotators) are associated with Rehabilitation, Mayo Clinic College of Med-
the development of PFP. In addition, persons in whom PFP develops appear to lose hip icine, Mayo Clinic Sports Medicine Center,
abduction and external rotation strength when compared with their pre-injury strength. 200 1st St SW, Rochester, MN. Address
correspondence to J.T.F.; e-mail: Finnoff.
Finally, a higher hip external-to-internal rotation strength ratio may protect against the jonathan@mayo.edu
development of PFP. Disclosure: nothing to disclose
PM R 2011;3:792-801 M.M.H. Department of Physical Medicine and
Rehabilitation, Mayo Clinic College of Medi-
cine, Mayo Clinic Sports Medicine Center,
Rochester, MN
INTRODUCTION Disclosure: nothing to disclose

K.K. Mayo Clinic Sports Medicine Center,

Running is a popular sport in the United States, with an estimated 30-34 million people Rochester, MN
participating in some form of running exercise on a regular basis [1,2]. Unfortunately, running Disclosure: nothing to disclose
is not always a benign activity. The incidence of lower extremity injuries in runners ranges from D.A.K. Department of Physical Medicine and
19%-79% [3]. The knee is the most frequently injured lower extremity joint, and the most Rehabilitation, Mayo Clinic College of Medi-
cine, Rochester, MN
common diagnoses are patellofemoral pain (PFP) and iliotibial band syndrome (ITBS) [3-6]. The Disclosure: nothing to disclose
incidences of PFP and ITBS in runners are 25%-40% and 1.6%-12%, respectively [3-6].
J.L. Mayo Clinic Sports Medicine Center,
Although the etiologies of PFP and ITBS are likely multifactorial, one contributing factor Rochester, MN
may be a lack of proximal femoral control because of weak hip musculature. Several Disclosure: nothing to disclose
investigators have found hip girdle weakness in subjects with PFP or ITBS [2,4,6-15], and J.S. Department of Physical Medicine and
several authors have suggested that hip girdle strengthening should be an integral part of the Rehabilitation, Mayo Clinic College of Medi-
cine, Mayo Clinic Sports Medicine Center,
rehabilitation program for patients with PFP and ITBS [4,11,16]. In fact, in a recent Rochester, MN
randomized controlled trial in which the authors compared standard patellofemoral reha- Disclosure: nothing to disclose
bilitation focusing on the thigh musculature with a rehabilitation program incorporating hip Submitted for publication November 18,
abduction and external rotation strengthening, they found a more significant reduction in 2010; accepted April 8, 2011.

PM&R © 2011 by the American Academy of Physical Medicine and Rehabilitation

1934-1482/11/$36.00 Vol. 3, 792-801, September 2011
792
Printed in U.S.A. DOI: 10.1016/j.pmrj.2011.04.007
PM&R
PFP in the group that received hip girdle strengthening than
in the standard rehabilitation group [17].
It has been hypothesized that weak hip musculature reduces
the ability to control femoral adduction and internal rotation
during the stance phase of running, thus leading to increased
femoral adduction and internal rotation, abnormal patellofemo-
ral joint reaction forces, and eventually the development of PFP
[2,4,6-14,18-20]. The authors of 2 recent kinematic studies
have corroborated this theory by demonstrating an association
between hip muscle strength deficits and increased femoral
adduction [15,21]. Furthermore, excessive femoral adduction
and internal rotation during the stance phase of running in-
creases iliotibial band tension, leading to ITBS [4,22-25]. There-
fore similar mechanisms may predispose a person to the devel-
opment of both ITBS and PFP.
To our knowledge, only one prospective study has been
conducted to investigate the association between hip girdle
strength and PFP [26], and no prospective studies have been
performed to evaluate the association between hip girdle
strength and ITBS. In a study by Boling et al [26], subjects in
whom PFP developed had greater pre-injury hip external
rotation strength and a larger hip internal rotation angle
during a jump-landing task compared with subjects in whom
PFP did not develop. Although strong hip external rotation
muscles were contrary to what the authors expected to find in
the subjects with PFP, they hypothesized that the larger
degree of hip internal rotation that occurred during jump
landing tasks placed significant eccentric demands on the
subject’s external rotators, thus leading to an increase in
external rotation strength. However, the investigators con-
cluded that further research was required to confirm their
findings. Because of the paucity of prospective studies inves-
tigating the association between hip girdle strength and PFP
or ITBS, the primary objective of this study was to determine
whether hip muscle weakness predisposes to PFP or ITBS in
high school running athletes.
MATERIALS AND METHODS
Study Design
The study used a prospective cohort design and was con-
ducted from the fall of 2007 through the fall of 2009.
Throughout this period, 98 running athletes (53 male and 45
female) from 5 local high schools were recruited to partici-
pate in the study by the investigators and by the certified
athletic trainers (ATCs) who worked at their schools. Ap-
proximately 1500 athletes had the opportunity to enroll in
the study. Recruitment involved handing out a flyer describ-
ing the study to all running athletes during preseason regis-
tration and during the first day of practice. An assent and
consent form was attached to the flyer. Athletes who were
interested in participating in the study were instructed to
contact their ATC to schedule a meeting time and determine
Vol. 3, Iss. 9, 2011 793
whether they met inclusion/exclusion criteria for study par-
ticipation and to obtain baseline data.
Subjects were continuously enrolled in the study at the
beginning of the track and/or cross-country running seasons.
The beginning of the season was defined as the first 2 weeks
of organized practice. Each subject was enrolled in the study
for the duration of one track or cross-country running sea-
son. However, upon completion of the season, they were
allowed to re-enroll in the study at the beginning of the next
track or cross-country running season if they continued to
meet inclusion and exclusion criteria and had not sustained
knee pain during their previous participation in the study.
Re-enrollment was accounted for in the statistical analysis
(discussed in the section “Data Analysis”), and a new set of
intake data were acquired upon re-enrollment in the study.
However, every effort was made to enroll new subjects, and thus
only 4 subjects re-enrolled in the study. Subjects participated in
the study on a voluntary basis and signed an informed consent.
Parents were required to co-sign the informed consent form for
subjects who were younger than 18 years. The study was ap-
proved by the Mayo Clinic Institutional Review Board. Because
the subjects were recruited from local high schools, permission
from the school board, athletic directors, and track and cross-
country running coaches from the participating schools also was
obtained before we initiated the study.
Inclusion criteria included male or female running athletes,
14 to 18 years of age, who were members of a high school track
or cross-country running team. Exclusion criteria included cur-
rent lower extremity or back pain, acute illness, pregnancy, a
history of previous knee or hip surgery, knee pain precluding
testing, inflammatory arthritis, or any medical condition that
precluded their participation in running sports.
Upon enrollment in the study at the beginning of the track
or cross-country running season, each subject was assigned a
unique subject number for identification and data collection
purposes. The subject’s age, gender, height, weight, year in
school, sport (track or cross-country running), running
event(s), dominant leg (determined by which leg they kick
with), bilateral full leg, upper and lower leg lengths, and hip
girdle strengths were recorded.
The weight measurement was taken with the athlete
dressed in a shirt and shorts. Their bilateral leg lengths were
determined by measuring from their anterior superior iliac
spine (ASIS) to a point 2 cm proximal to the apex of their
medial malleolus [27]. The subjects’ bilateral upper leg
lengths were determined by measuring from the ASIS to the
midpatellar region of each leg with the knee in full extension,
whereas their lower leg lengths were assessed by measuring
from the midpatellar region to a point 2 cm proximal to the
highest point of their medial malleolus with the knee in full
extension. Their hip abduction, adduction, flexion, exten-
sion, internal rotation, and external rotation strengths were
determined with use of a hand-held dynamometer (HHD).
Strength testing via use of an HHD has been shown to be
794 Finnoff et al
valid and reliable in the assessment of muscle strength.
[2,4,7,28-32]. Testing took place before the subjects’ daily
exercise to control for fatigue.
The subjects were monitored weekly for knee pain by the
ATCs who worked at their school. An e-mail with the study
subjects’ names were sent to the ATCs weekly to remind them
to ask study subjects whether they had developed knee pain.
When a subject reported knee pain, he or she was assessed by
the ATC. If it was determined that the pain was likely PFP or
ITBS, the ATC contacted the physician investigators (J.T.F.,
M.M.H.), and the subject was evaluated by one of the physi-
cian investigators within 2 days to confirm the presence or
absence of PFP or ITBS.
The diagnosis of ITBS was suspected in athletes who re-
ported lateral knee pain and was confirmed by the presence of
local tenderness over the lateral knee where the ITB crossed the
lateral femoral condyle, exacerbated by flexing and extending
the knee while pressing over the lateral femoral condyle, with
the greatest pain occurring at approximately 30° of knee flexion
(Noble compression test). The diagnosis of isolated ITBS was
excluded if there was joint line tenderness other than where the
ITB crosses the joint line, popliteus tendonitis, knee ligamen-
tous injury, or a knee effusion [4].
PFP was suspected in athletes who reported anterior knee
pain that was exacerbated by deep knee bending and/or
climbing stairs and was confirmed by the reproduction of the
pain by at least one of the following maneuvers: (1) pressure
over the subject’s distal quadriceps tendon combined with
active contraction of his or her quadriceps muscle (patellar
grind test) or (2) direct palpation of the medial or lateral
patellar facets, along with the absence of exclusion criteria
[8]. The diagnosis of PFP was excluded if tenderness was
present in areas other than the medial and lateral patellar
facets. The subject was asked to quantify the maximum
severity of his or her pain using a 10-cm visual analog scale
(VAS), with a score of 0 corresponding to no pain and 10
being the worst imaginable pain. A VAS score was obtained
for each diagnosis (PFP or ITBS) and side (right, left, bilat-
eral) of their symptoms.
When it was determined that the subject had PFP or ITBS,
the subject’s bilateral hip girdle strength was re-assessed. The
patient’s diagnoses, side of pathology (right, left, or bilateral),
VAS, and hip girdle strength were recorded for analysis.
At the end of the season, the ATCs contacted all of the
study subjects to determine whether they completed the
entire competitive season. Those who did not complete the
season for a reason other than the development of PFP or
ITBS were removed from data analysis.
Strength Testing of the Hip Musculature
Hip muscle strength was determined with use of 1 of 3
MicroFET 2 HHDs (Hoggan Health Industries, West Jordan,
UT). The MicroFET 2 HHD has a manufacturer’s reported
HIP STRENGTH AND KNEE PAIN IN RUNNERS
accuracy of 97%. The 3 HHDs were calibrated before each
testing session. Five testers (J.T.F., M.M.H., K.K., D.A.K., and
J.L.) performed the testing.
Hip strength assessments included hip flexion, extension,
abduction, adduction, external rotation, and internal rota-
tion. The tests were performed in agonist-antagonist pairs (ie,
hip flexion with extension, abduction with adduction, and
external rotation with internal rotation). The order of leg
tested and of agonist-antagonist pairs tested for each leg
were randomized to minimize threats to internal validity.
All strength testing occurred on an examination table. A
“break-test” was used in which the subject’s maximum
isometric contraction was overcome by a force applied by
the examiner.
The following procedure was used for all strength tests.
After the investigator explained and demonstrated the
strength testing technique, the subject was allowed one prac-
tice bout before each individual test. After 30 seconds of rest,
the subject performed 3 maximal bouts of each test. Subjects
rested a minimum of 15 seconds between bouts [4]. Al-
though the subjects were instructed to provide maximal
effort, no encouragement was given while testing to maintain
uniformity of the testing procedure. The presence or absence
of pain during testing was recorded. Although previous re-
search has established that this method of strength testing is
valid and reliable [2,4,7,28-32], we attempted to perform a
reliability study to strengthen the study’s methodology.
However, because each subject had to perform 3 maximal
bouts of each test with each of the 5 examiners, the subjects
experienced significant fatigue and discomfort during data
acquisition, which prevented them from completing the re-
liability study. Therefore we were unable to determine the
reliability of the strength testing method used in this study,
and it is possible that the use of 5 different strength testers
introduced error into our data.
The HHD values for the 3 trials of each test were recorded.
Torque was calculated with the following equation: torque ⫽
force (N) ⫻ lever arm (m). The lever arm length used for the
hip abduction, adduction, and extension torque calculations
was the subject’s full leg length measurement. The hip flexion
torque calculation used the upper leg length measurement,
whereas the hip internal and external rotation torque calcu-
lations used the lower leg length measurement. Torque val-
ues were normalized for the subject’s body weight (BW)
reported in Newtons (N) and height (h) reported in meters.
The full equation for normalized torque percent was as
follows [4]:
%(BW ⫻ h) ⫽ Torque(N ⫻ m) ⫻ 100 ⁄ [BW(N) ⫻ h(m)]
Hip abduction strength was assessed with the subject lying
on his or her side. The subject grasped the table with his or
her upper arm, and flexed the lower leg, hip, and knee for
stabilization. The subject abducted the upper leg approxi-
PM&R Vol. 3, Iss. 9, 2011 795

Figure 1. HHD hip strength testing. (A) Hip abduction, (B) hip adduction, (C) hip flexion, (D) hip extension, (E) hip external rotation,
and (F) hip internal rotation.

mately 30° while maintaining neutral hip flexion, extension, Hip adduction strength was assessed with the subject
and rotation. A break test was performed with the HHD lying on his or her side. The upper, non-test extremity was
positioned against the lateral aspect of the lower leg, 2 cm supported by a bench placed anterior to the test leg [30]. The
proximal to the apex of the lateral malleolus (Figure 1A). subject adducted his or her lower leg while maintaining
796 Finnoff et al
neutral hip flexion, extension, and rotation. A break test was
performed with the HHD placed 2 cm proximal to the
subject’s medial malleolus (Figure 1B).
Hip flexion strength was assessed with the subject seated
and the trunk-to-thigh angle approximately 90°. The subject
was allowed to grasp the table for stability. The subject flexed
his or her hip to 120°, and a break test was performed with
the HHD positioned on the distal aspect of the subject’s thigh
(Figure 1C).
Hip extension strength was assessed with the subject
prone with an 8-inch diameter foam roll placed at the level of
their ASIS. The subject grasped the table for trunk stability
and was asked to extend the test hip to a neutral position with
the knee extended while maintaining neutral hip rotation. A
break test was performed with the HHD against the subject’s
posterior calcaneus (Figure 1D).
Hip external rotation strength was assessed with the sub-
ject seated with a trunk-to-thigh angle of approximately 90°.
The subject’s knees were also flexed 90° with the hip in
neutral rotation. The subject was allowed to grasp the table
for stability, and a break test was performed with the HHD
positioned 2 cm proximal to the apex of the medial malleolus
(Figure 1E).
Hip internal rotation strength was assessed with the sub-
ject in a position identical to the one used for hip external
rotation strength testing. The break test was performed with
the HHD positioned 2 cm proximal to the lateral malleolus
(Figure 1F). During each of the strength tests, the examiner
monitored the subject to ensure that proper positioning was
maintained throughout the hip strength testing procedure.
Data Analysis
The primary focus of the analysis was to evaluate the associ-
ation between baseline hip strength measures (ie, individual
hip muscle strengths and agonist-antagonist ratios) and the
occurrence of PFP and/or ITBS during the study period. This
evaluation was accomplished in 2 ways. First, the mean
baseline hip muscle strengths and strength ratios were com-
pared between the injured and non-injured groups. Second,
baseline hip muscle strengths and strength ratios were eval-
uated in a logistic regression model, with the occurrence of
PFP and/or ITBS during the study period as the outcome.
Because some subjects were enrolled in the study more than
once (eg, cross-country in the fall and track in the spring, or
subsequent school years), all analyses used generalized esti-
mating equations [33] in a generalized linear models [34]
framework to properly account for the within-subject corre-
lation and to generate the correct variance estimate. As stated
earlier, only 4 subjects enrolled in the study more than once.
Analyses were conducted for the entire study sample. All
statistical tests were 2-sided, and the threshold of statistical
significance was set at ␣ ⫽ 0.05.
HIP STRENGTH AND KNEE PAIN IN RUNNERS
RESULTS
All 98 subjects completed the study. Knee pain developed in
6% of the study subjects (n ⫽ 6) during the study. One
subject had bilateral PFP, whereas the remaining four cases of
PFP were unilateral (2 on the left side and 2 on the right), as
was the single case of ITBS (left side). Therefore 7 injuries
occurred in 6 subjects. The injured subjects reported a mean
VAS pain level of 4.95 (range, 2.30-8.00). All of the subjects
who had knee pain were right-leg dominant, but no associa-
tion existed between the side of leg dominance and the side of
knee pain development. Sixty percent of the injured subjects
with patellofemoral pain, or 3 of 5, were female, whereas
40% of subjects, or 2 of 5, were male. PFP did not develop in
any of the 4 subjects who participated in the study more than
once. Subject demographic and anthropomorphic character-
istics are listed in Table 1.
Because ITBS developed in only one subject, which pre-
cluded our ability to generalize this information, the data for
this subject were withdrawn from the statistical analysis. No
statistically significant differences in demographic or anthro-
pomorphic characteristics were found between the uninjured
and injured groups, but the difference in body mass index
(BMI) and weight between injured and uninjured subjects
approached significance, with injured subjects demonstrat-
ing a higher BMI and weight (23.69N and 699N, respec-
tively) than uninjured subjects (21.11N and 612N, respec-
tively; P ⫽ .09 and P ⫽ .08, respectively).
During our preliminary analyses of the hip strength data,
both the peak normalized torque percent of the 3 trials and the
average normalized torque percent of the 3 trials were analyzed.
However, no differences between the analyses were found when
the peak normalized torque percent value was used compared
with when the average normalized torque percent value was
used. Therefore only the peak normalized torque percent values
were reported in this article (Tables 2 and 3).
The baseline hip external-to-internal rotation normalized
torque percent ratio was significantly lower in injured sub-
jects (mean ⫽ 0.74% BW ⫻ h, range ⫽ 0.53-0.88% BW ⫻ h)
than in uninjured subjects (mean ⫽ 0.87% BW ⫻ h, range ⫽
0.28-1.44% BW ⫻ h; P ⫽ .008). A trend toward higher
baseline hip abduction-to-adduction normalized torque per-
cent ratio was seen in injured subjects (mean ⫽ 1.12% BW ⫻
h, range ⫽ 0.75-1.46% BW ⫻ h) compared with uninjured
subjects (mean ⫽ 0.94% BW ⫻ h, range ⫽ 0.56-2.20% BW ⫻ h;
P ⫽ .09). In the injured group, hip abduction normalized
torque percent significantly decreased from before injury
(mean ⫽ 3.14% BW ⫻ h, range ⫽ 2.58-4.26% BW ⫻ h) to
after injury (mean ⫽ 2.80% BW ⫻ h, range ⫽ 2.26-3.77%
BW ⫻ h; P ⫽ .002), as did hip external rotation normalized
torque percent (pre-injury mean ⫽ 1.34% BW ⫻ h, range ⫽
1.13-1.72% BW ⫻ h; post-injury mean ⫽ 1.20% BW ⫻ h,
range ⫽ 0.98-1.49% BW ⫻ h; P ⫽ .01). No significant
differences in hip normalized torque percent or hip normal-
PM&R Vol. 3, Iss. 9, 2011 797

Table 1. Demographic and anthropomorphic characteristics of uninjured and injured subjects

Age, Dominant Height [SD], Weight [SD], BMI [SD], Leg Length [SD],
Range Gender Sport Event Leg, Range Range Range Range
Subject P ⴝ .58 P ⴝ .66 P ⴝ 1.00 P ⴝ 1.00 P ⴝ 1.00 P ⴝ .65 P ⴝ .08 P ⴝ .09 P ⴝ .72
Uninjured 15.97, Male ⫽51 CC ⫽54 ⬍ 400 m ⫽ 27 Right⫽85 1.72 m [0.1], 612 N [119], 21.11 [3.1], 0.87 m [0.06],
14-18 Female⫽41 Track⫽38 400-800 m ⫽ 6 Left ⫽ 7 1.30-1.98 423-1228 15.54-33.7 0.73-1.03
⬎800 m ⫽ 59
Injured 15.6, Male ⫽ 2 CC ⫽ 3 ⬍400 m ⫽ 2 Right⫽ 6 1.73 m [0.1], 699 N [109], 23.69 [3.8], 0.86 m [0.04],
14-17 Female⫽ 3 Track⫽ 2 400-800 m ⫽ 0 Left ⫽ 0 1.63-1.88 551-809 19.99-29.44 0.80-0.90
⬎800m ⫽ 3

BMI ⫽ body mass index; CC ⫽ cross-country.

ized torque percent ratios were found between the injured
and uninjured sides in injured subjects.
Logistic regression analysis demonstrated several interest-
ing findings (Table 4). A higher pre-injury hip external-to-
internal rotation normalized torque percent ratio decreased the
risk of injury (odds ratio ⬍ 0.01, 95% CI ⱕ 0.01, 0.44; P ⫽ .02),
whereas higher baseline hip abduction normalized torque per-
cent (odds ratio ⫽ 5.35, 95% CI ⫽ 1.46, 19.53; P ⬍ .01) and
abduction-to-adduction normalized torque percent ratios sig-
nificantly increased the risk of injury (odds ratio ⫽ 14.14,
95% CI ⫽ 0.90, 221.06; P ⫽ .05).
DISCUSSION
The primary aim of this study was to determine whether hip
muscle weakness predisposed high school running athletes
to PFP or ITBS. Because ITBS developed in only one subject,
that subject’s data were not included in the analysis because
no significant conclusions could be determined from a single
case. Our findings suggest that a lower pre-injury hip exter-
nal-to-internal rotator strength ratio may be associated with
the development of PFP in high school running athletes.
Athletes with an injury also had a significant reduction in hip
abduction and external rotation strength from before injury
to after injury. Predictors of future PFP development in-
cluded higher pre-injury hip abduction strength and abduc-
tion-to-adduction strength ratios, whereas a higher pre-in-
jury hip external-to internal rotation strength ratio decreased
the risk of developing PFP. Trends suggested an association
between the development of PFP and a greater baseline hip
abduction-to-adduction strength ratio, body weight, or BMI.
The findings of our study are intriguing, particularly when
compared with those of Boling et al [26].
The incidence of PFP was 5% in this study, which was
twice the incidence of PFP in the study by Boling et al [26]
but was lower than the incidence reported in several other
studies [3-6]. The difference in injury incidence between our
study and that of Boling et al [26] may have been attributable
to the younger age of the subjects in our study (mean, 16
years; range, 14-18 years) compared with those who partic-
ipated in the study by Boling et al [26]. Although Boling and
colleagues [26] did not report the age range in their study,
because their participants were midshipmen at the United
States Naval Academy, subjects in their study were likely
older than the population in our study. However, further
research is required to determine whether this hypothesis is
correct.
In our study, 60% of PFP injuries occurred in female
subjects and 40% occurred in male subjects. These findings
are consistent with those of previous investigators [5,26,35].
In our study, a higher external-to-internal rotation strength
ratio protected runners from developing PFP, whereas a low
hip external-to-internal rotation strength ratio appeared to
predispose subjects to the development of PFP. The hip
external rotation strength in subjects with PFP was even
lower at their post-injury assessment than at baseline. These
findings contradict those of Boling et al [26] but are sup-
ported by the findings of multiple previous case-control
studies in which the authors investigated hip strength in
subjects with PFP [7-11,13-15,36].
In the study by Boling et al [26], kinematic analysis of their
study subjects demonstrated increased femoral internal rota-
tion during a jump landing task in the subjects who had PFP.
The researchers theorized that the increased femoral internal
rotation may have placed significant eccentric demands upon

Table 2. Preinjury hip normalized torque percent of injured and uninjured subjects (%BW ⫻ h)
External-Internal
Internal External Adduction Abduction Flexion-Extension Rotation Ratio Abduction-Adduction
Flexion [SD], Extension [SD], Rotation [SD], Rotation [SD], [SD], [SD], Ratio [SD], [SD], Ratio [SD],
Range Range Range Range Range Range Range Range Range
Subject P ⴝ .36 P ⴝ .41 P ⴝ .46 P ⴝ .38 P ⴝ .67 P ⴝ .67 P ⴝ .22 P ⴝ .008 P ⴝ .09

Uninjured 2.84 [0.61], 3.15 [0.79], 1.68 [0.40], 1.44 [0.31], 2.79 [0.61], 2.57 [0.53], 0.94 [0.24], 0.87 [0.17], 0.94 [0.20],
1.42-4.32 1.27-5.20 0.84-3.37 0.40-2.50 1.30-5.22 1.53-4.28 0.59-2.43 0.28-1.44 0.56-2.20
Injured 2.49 [0.92], 2.87 [0.79], 1.88 [0.68], 1.34 [0.26], 2.87 [0.45], 3.14 [0.63], 0.86 [0.15], 0.74 [0.13], 1.12 [0.28],
1.69-4.17 2.19-4.01 1.40-3.21 1.13-1.72 2.32-3.46 2.58-4.26 0.70-1.12 0.53-0.88 0.75-1.46
798 Finnoff et al HIP STRENGTH AND KNEE PAIN IN RUNNERS

Table 3. Comparison of pre-injury to post-injury hip normalized torque percent in injured subjects (%BW ⫻ h)
External-Internal
Internal External Adduction Abduction Rotation Ratio
Flexion [SD], Extension [SD], Rotation [SD], Rotation [SD], [SD], [SD], Flexion-Extension [SD], Abduction-Adduction
Range Range Range Range Range Range Ratio [SD], Range Range Ratio [SD], Range
Subject P ⴝ .31 P ⴝ .78 P ⴝ .71 P ⴝ .01 P ⴝ .01 P ⴝ .002 P ⴝ 0.20 P ⴝ .47 P ⴝ .34

Preinjury 2.49 [0.92], 2.87 [0.79], 1.88 [0.68], 1.34 [0.26], 2.82 [0.45], 3.14 [0.63], 0.86 [0.15], 0.74 [0.13], 1.12 [0.28],
1.69-4.17 2.19-4.01 1.40-3.21 1.13-1.72 2.32-3.46 2.58-4.26 0.70-1.12 0.53-0.88 0.75-1.46
Post-injury 2.73 [0.75], 2.81 [0.42], 1.81 [0.49], 1.20 [0.19], 2.69 [0.47], 2.80 [0.57], 0.96 [0.13], 0.69 [0.15], 1.06 [0.25],
1.87-3.88 2.25-3.27 1.22-2.39 0.98-1.49 2.13-3.44 2.26-3.77 0.83-1.18 0.57-0.98 0.80-1.50

their subject’s hip external rotation musculature, thus lead-
ing to an increase in external rotation strength. This theory
may explain the discrepancy between their findings and
those of other studies.
It has been suggested that weak hip external rotators
decrease the dynamic control of femoral internal rotation
during weight-bearing activities, thus leading to increased
femoral internal rotation. In support of this hypothesis,
Souza and Powers [15] found weak hip external rotators and
increased hip internal rotation during weight-bearing tasks
in subjects with PFP when compared with control subjects.
Femoral internal rotation during weight-bearing activities
results in lateral patellar subluxation [20] and increased
patellofemoral joint stress [18], which may predispose a
person to the development of PFP.
Several other researchers also have demonstrated in-
creased femoral internal rotation and/or adduction in sub-
jects with weak hip external rotators [37,38]. However, con-
trol of dynamic femoral internal rotation and/or adduction
during weight-bearing activity probably is multifactorial,
with an important role played by appropriate neuromuscular
control [39]. The significant decrease in hip external rotation
strength that occurred in our PFP subjects from the time of
their baseline strength testing to the time of their injury may
have resulted in an increase in femoral internal rotation, thus
contributing to the development of PFP. Future researchers
may wish to perform baseline and post-injury kinematic
analyses to determine whether changes in lower extremity
kinematics occur after injury and whether these changes
correlate with changes in hip strength.
We were surprised that subjects in our study with strong
hip abductors and a higher hip abduction-to-adduction
strength ratio were predisposed to the development of PFP.
These findings appear to be in direct contrast to those of
Boling et al [26] as well as previous case-control studies in
which the authors have demonstrated hip abduction weak-
ness in subjects with PFP [2,4,6-9,11-16,21,36,40]. There
are several possible explanations for this discrepancy. First,
the injured subjects in our group had a higher weight and
BMI than did the uninjured subjects. Because the single most
influential factor on the hip adduction moment magnitude
during the early and mid-stance phases of gait is weight, it is
possible that the injured subjects in this study developed
stronger hip abductors to compensate for their larger hip
adduction moment. Because Boling et al [26] did not report
the weight or BMI of their subjects, the influence of weight or
BMI on their subject’s hip abduction strength could not be
evaluated.
In addition, the gluteus medius can act not only as a hip
abductor but also as a hip external rotator via its middle and
posterior fibers [41,42]. Because the injured subjects in our
study had weak hip external rotators, their gluteus medius
muscles may have been recruited to serve as accessory exter-
nal rotators to assist with the eccentric control of femoral
internal rotation during the stance phase of gait, leading to
increased hip abduction strength in the injured subjects.
Interestingly, the hip abduction strength in our study sub-
jects decreased significantly from the time of their baseline
strength testing to the time of their injury. None of the
subjects experienced pain during repeat hip strength testing.
Therefore their decrease in hip abduction strength was not
inhibited by pain provocation. As suggested previously, it is
possible that this decrease in strength resulted in uncompen-
sated femoral adduction and/or internal rotation, thus con-
tributing to the development of PFP. However, future kine-
matic research is required to determine if this phenomenon
indeed occurs. Finally, because the subjects in our study
were high school running athletes, direct comparison with

Table 4. Logistic regression analysis of the pre-injury hip normalized torque percent’s ability to predict future injury via odds ratios
(95% confidence interval)

Internal External Flexion-Extension External-Internal Abduction-Adduction

Hip Flexion Hip Extension Rotation Rotation Adduction Abduction Ratio Rotation Ratio Ratio
P ⴝ .38 P ⴝ .42 P ⴝ .35 P ⴝ .42 P ⴝ .67 P < .01 P ⴝ .32 P ⴝ .02 P ⴝ .05
0.40, 0.64, 2.75, 0.35, 1.23, 5.35, 0.17, 0.01, 14.14,
0.05-3.09 0.21-1.90 0.33-23.17 0.03-4.48 0.48-3.17 1.46-19.53 0.021 5.61 ⬍0.01-0.44 0.90-221.06
PM&R
previous studies is difficult because of the differences in age
ranges, study designs, and sports participation of the subjects
in the other studies. Therefore it is possible that despite the
findings of previous researchers, in high school running
athletes, greater hip abduction strength or abduction-to-
adduction strength ratio may predispose to the development
of PFP.
The possible association between increased weight and
BMI and the development of PFP found in our study has not
been reported in the few prospective studies in which the
authors have investigated the intrinsic risk factors of PFP
[26,43,44]. Because a major determinant of joint reaction
forces and joint kinetics is weight [45], it stands to reason that
an increase in weight or BMI may increase patellofemoral
joint reaction forces during running, thus increasing the risk
of developing PFP. In fact, Rutherford and Hubley-Kozey
[45] found that body mass was the most influential factor
associated with increased femoral adduction moments dur-
ing the initial and mid-stance phases of gait.
Moreover, if abnormal femoral kinematics are a risk factor
for the development of PFP and a combination of muscular
strength, endurance, and coordination is required to control
femoral kinematics, an increase in weight or BMI may lead to
more difficulties in controlling femoral kinematics, thus ex-
aggerating the kinematic abnormalities that predispose one
to the development of PFP. Although this finding was inter-
esting, because the difference in weight or BMI between the
injured and uninjured athletes was not statistically significant
in our study (P ⫽ .08 and P ⫽ .09, respectively), future
research is required to determine whether weight or BMI is
associated with the development of PFP.
Our study has several limitations. Although previous re-
search has established that the strength testing method used
in this study is valid and reliable [2,4,7,28-32], our attempt
to perform a reliability study failed because the study subjects
developed significant fatigue and discomfort when trying to
perform repetitive bouts of maximal muscle contraction, thus
precluding their ability to complete the data collection por-
tion of the reliability study. Therefore it is possible that the
use of 5 different strength testers introduced error into our
data. The age range and sport participation was narrow, thus
minimizing the ability to generalize the results to other pop-
ulations. However, because PFP frequently develops in high
school runners, the findings of our study certainly have
clinical applicability.
Multiple variables that may predispose individuals to PFP
were not controlled or accounted for in this study, including
anatomic variants, biologic age, footwear, trunk and lower
extremity neuromuscular control/running gait pattern, expe-
rience, training volume, and intensity, among others. Be-
cause these factors were not tracked, we were unable to
determine whether the hip strength deficits demonstrated in
this study caused PFP; rather, we were only able to conclude
that a relationship existed between hip girdle strength deficits
Vol. 3, Iss. 9, 2011 799
and the development of PFP. Furthermore, because of the
limited number of subjects in our study, our conclusions are
susceptible to type 2 errors. Therefore the findings of this
study should be interpreted with caution, and further re-
search should be conducted to determine whether our con-
clusions are accurate.
However, despite the inherent difficulties of conducting a
prospective study of this type, they are the preferred method
to determine associated or causative factors associated with
injuries and are therefore important. Moreover, the authors
of several prospective studies with similar limitations have
provided significant contributions to the sports medicine
literature and have led to future research that increased our
knowledge regarding sports injuries [22,46-48]. Thus de-
spite this limitation, our findings provide significant addi-
tional information related to factors associated with the de-
velopment of PFP.
Although the athletes in this study were monitored for
symptoms by the ATCs at their schools, the study relied upon
self-reporting of symptoms by the athletes. Because many
athletes minimize symptoms or are concerned that they will
be restricted from activity if they report their symptoms to a
medical provider such as an ATC, it is possible that this study
underreported the incidence of PFP. Because only one case of
ITBS occurred in this study, further analysis could not be
performed to differentiate between factors associated with
ITBS from those associated with PFP, and the subject with
ITBS was eliminated from the data analysis. Although previ-
ous research suggests that similar mechanisms predispose
athletes to both injuries, further study is required to deter-
mine if the findings of our study are applicable to both ITBS
and PFP. Despite these limitations, several statistically and
clinically significant findings were present in this study, and
because of the lack of prospective studies in which the
authors evaluated the association between hip girdle strength
and the development of PFP, we believe our findings provide
an important contribution to the available literature.
CONCLUSIONS
In conclusion, PFP commonly occurs in running athletes.
Several investigators have detected hip strength deficits in
athletes with PFP, but only one previous prospective study
has investigated the association of hip strength and the de-
velopment of PFP. The findings of the current study suggest
that stronger pre-injury hip abductors (particularly in rela-
tion to their hip adductors) and weaker pre-injury hip exter-
nal rotators (particularly in relation to their hip internal
rotators) are associated with the development of PFP. Sub-
jects in whom PFP develops appear to lose hip abduction and
external rotation strength when compared with their pre-
injury strength. Finally, a higher hip external-to-internal
rotation strength ratio may protect against the development
of PFP.
800 Finnoff et al
REFERENCES
1. McCrory J, Martin DF, Lowery RB, et al. Etiologic factors associated
with Achilles tendinitis in runners. Med Sci Sports Exerc 1999;31:
1374-1381.
2. Neimuth P, Johnson RJ, Myers MJ, Thieman TJ. Hip muscle weakness and
overuse injuries in recreational runners. Clin J Sport Med 2005;15:14-21.
3. Van Gent R, Van Middelkoop M, Van Os AG, Bierma-Zeinstra SMA,
Koes BW. Incidence and determinants of lower extremity running
injuries in long distance runners: A systematic review. Br J Sports Med
2007;41:469-480.
4. Fredericson M, Cookingham CL, Chaudhari AM, Dowdell BC, Oest-
reicher N, Sahrmann SA. Hip abductor weakness in distance runners
with ilitibial band syndrome. Clin J Sport Med 2000;10:169-175.
5. Taunton J, Ryan M, Clement D, McKenzie D, Lloyd-Smith D, Zumbo B.
A retrospective case-control analysis of 2002 running injuries. Br J
Sports Med 2002;36:95-101.
6. Piva S, Goodnite EA, Childs JD. Strength around the hip and flexibility
of soft tissues in individuals with patellofemoral pain syndrome. J Or-
thop Sports Phys Ther 2005;35:793-801.
7. Robinson R, Nee RJ. Analysis of hip strength in females seeking physical
therapy treatment for unilateral patellofemoral pain syndrome. J Or-
thop Sports Phys Ther 2007;37:232-238.
8. Cichanowski H, Schmitt JS, Johnson RJ, Niemuth PE. Hip strength in
collegiate female athletes with patellofemoral pain. Med Sci Sports
Exerc 2007;39:1227-1232.
9. Ireland M, Willson JD, Ballantyne BT, Davis IM. Hip strength in females
with and without patellofemoral pain. J Orthop Sports Phys Ther
2003;33:671-676.
10. Leetun D, Ireland ML, Willson JD, Ballantyne BT, Davis IM. Core
stability measures as risk factors for lower extremity injury in athletes.
Med Sci Sports Exerc 2004;36:926-934.
11. Mascal C, Landel R, Powers C. Management of patellofemoral pain
targeting hip, pelvis, and trunk muscle function: 2 case reports. J Or-
thop Sports Phys Ther 2003;33:647-660.
12. Tyler T, Nicholas SJ, Mullaney MJ, McHugh MP. Hip muscle function
in the treatment of patellofemoral pain syndrome. Am J Sports Med
2006;34:630-636.
13. Bolgla L, Lori A. Hip strength and hip and knee kinematics during stair
descent in females with and without patellofemoral pain syndrome.
J Orthop Sports Phys Ther 2008;38:12-18.
14. Willson J, Davis IS. Lower extremity strength and mechanics during
jumping in women with patellofemoral pain. J Sports Rehabil 2009;18:76-90.
15. Souza R, Powers CM. Predictors of hip internal rotation during running.
An evaluation of hip strength and femoral structure in women with and
without patellofemoral pain. Am J Sports Med 2009;37:579-587.
16. Nakagawa TH, Muniz TB, de Marche Baldon R, Maciel CD, de Menezes
Reiff RB, Serrao FV. The effect of additional strengthening of hip abductor
and lateral rotator muscles in patellofemoral pain syndrome: A random-
ized controlled pilot study. Clin Rehabil 2008;22:1051-1060.
17. Fukuda T, Rossetto FM, Magalhaes E, Bryk FF, Lucareli PRG, Carvalho
NADA. Short-term effects of hip abductors and lateral rotators strengthen-
ing in females with patellofemoral pain syndrome: A randomized con-
trolled clinical trial. J Orthop Sports Phys Ther 2010;40:736-742.
18. Lee T, Morris G, Csintalan RP. The influence of tibial and femoral
rotation on patellofemoral contact area and pressure. J Orthop Sports
Phys Ther 2003;33:686-693.
19. Powers C, Ward SR, Fredericson M, Guillet M, Shellock FG. Patell-
ofemoral kinematics during weight-bearing and non-weight-bearing
knee extension in persons with lateral subluxation of the patella: A
preliminary study. J Orthop Sports Phys Ther 2003;33:677-685.
20. Powers C, Ward SR, Fredericson M, Guillet M, Shellock FG. Patell-
ofemoral kinematics during weight-bearing and non-weightbearing
knee extension in persons with lateral subluxation of the patella: A
preliminary study. J Orthop Sports Phys Ther 2003;33:677-685.
HIP STRENGTH AND KNEE PAIN IN RUNNERS
21. Dierks T, Manal KT, Hamill J, Davis IS. Proximal and distal influences
on hip and knee kinematics in runners with patellofemoral pain during
a prolonged run. J Orthop Sports Phys Ther 2008;38:448-456.
22. Noehren B, Davis I, Hamill J. ASB clinical biomechanics award winner
2006 prospective study of the biomechanical factors associated with
iliotibial band syndrome. Clin Biomech 2007;22:951-956.
23. Ferber R, Noehren B, Hamill J, Davis I. Competitive female runners
with a history of iliotibial band syndrome demonstrate atypical hip and
knee kinematics. J Orthop Sports Phys Ther 2010;40:52-58.
24. Powers C. The influence of abnormal hip mechanics on knee injury: A
biomechanical perspective. J Orthop Sports Phys Ther 2010;40:42-51.
25. Hamill J, Miller R, Noehren B, Davis I. A prospective study of iliotibial
band strain in runners. Clin Biomech 2008;23:1018-1025.
26. Boling M, Padua DA, Marshall SW, Guskiewicz K, Pyne S, Beutler A. A
prospective investigation of biomechanical risk factors for patellofemo-
ral pain syndrome. Am J Sports Med 2009;37:2108-2116.
27. Magee D. Pelvis. In: Magee DJ, ed. Orthopedic Physical Assessment.
4th ed. Philadelphia, PA: Saunders; 2002, 567-605.
28. Marino M. The efficacy of manual assessment of muscle strength using
a new device. Am J Sports Med 1982;10:360-364.
29. Bohannon R, Saunders N. Hand-held dynamometry: A single trial may
be adequate for measuring muscle strength in healthy individuals.
Physiother Can 1990;42:6-9.
30. Krause D, Schlagel SJ, Stember BM, Zoetewey JE, Hollman JH. Influ-
ence of lever arm and stabilization on measures of hip abduction and
adduction torque obtained by hand-held dynamometry. Arch Phys
Med Rehabil 2007;88:37-42.
31. Hyde S, Scott CM, Goddard CM. The myometer: The development of a
clinical tool. Physiotherapy 1983;69:424-427.
32. Bohannon R, Andrews AW. Interrater reliability of hand-held dyna-
mometry. Phys Ther 1987;67:931-933.
33. Liang K, Zeger SL. Longitudinal data analysis using generalized linear
models. Biometrika 1986;73:13-22.
34. McCullagh P, Nelder JA. Generalized Linear Models. New York, NY:
Chapman and Hall Ltd; 1983.
35. Wilson T. The measurement of patellar alignment in patellofemoral
pain syndrome: Are we confusing assumptions with evidence? J Orthop
Sports Phys Ther 2007;37:330-341.
36. Boling M, Padua DA, Creighton RA. Concentric and eccentric torque of
the hip musculature in individuals with and without patellofemoral
pain. J Athl Train 2009;44:7-13.
37. Willson J, Ireland ML, Davis IM. Core strength and lower extremity
alignment during single leg squats. Med Sci Sports Exerc 2006;38:945-
952.
38. Souza R, Powers CM. Differences in hip kinematics, muscle strength,
and muscle activation between subjects with and without patellofemo-
ral pain. J Orthop Sports Phys Ther 2009;39:12-19.
39. Hollman J, Ginos BE, Kozuchowski J, Vaughn AS, Krause DA, Youdas
JW. Relationships between knee valgus, hip-muscle strength, and hip
muscle recruitment during a single-limb step-down. J Sports Rehabil
2009;18:104-117.
40. de Marche Baldon R, Nakagawa TH, Muniz TB, Amorim CF, Maciel
CD, Serrao FV. Eccentric hip muscle function in females with and
without patellofemoral pain syndrome. J Athl Train 2009;44:490-496.
41. Soderberg G. Hip. In: GL S, ed. Kinesiology: Application to Patholog-
ical Motion. Baltimore, MD: Williams and Wilkins; 1986, 175-199.
42. Neumann D. Kinesiology of the hip: A focus on muscular actions.
J Orthop Sports Phys Ther 2010;40:82-94.
43. Milgrom C, Finestone A, Eldad A, Shlamkovitch N. Patellofemoral pain
caused by overactivity. A prospective study of risk factors in infantry
recruits. J Bone Joint Surg 1991;73A:1041-1043.
44. Witvrouw E, Lysens R, Bellemans J, Cambier D, Vanderstraeten G.
Intrinsic risk factors for the development of anterior knee pain in the
athletic population: A two year prospective study. Am J Sports Med
2000;28:480-489.
PM&R
45. Rutherford D, Hubley-Kozey C. Explaining the hip adduction moment
variability during gait: Implications for hip abductor strengthening.
Clin Biomech 2009;24:267-273.
46. Hewett T, Lindenfeld TN, Riccobene JV, Noyes FR. The effect of
neuromuscular training on the incidence of knee injury in female
athletes: A prospective study. Am J Sports Med 1999;27:699-706.
Vol. 3, Iss. 9, 2011 801
47. Gilchrist J, Mandelbaum BR, Melancon H, et al. A randomized con-
trolled trial to prevent non-contact anterior cruciate ligament injury in
female collegiate soccer players. Am J Sports Med 2008;36:1476-1483.
48. Zebis M, Andersen LL, Bencke J, Kjaer M, Aagaard P. Identification of
athletes at future risk of anterior cruciate ligament ruptures by neuro-
muscular screening. Am J Sports Med 2009;37:1967-1973.