Anterior cruciate ligament research has resulted in more
than 2000 scientific articles published outlining injury inci-
dence, mechanism, surgical repair techniques, and rehabili-
tation of this important stabilizing knee ligament.10
However, despite the many scientific advances in the treat-
ment of ACL injury, osteoarthritis occurs at a 10 times
greater rate in ACL-injured patients,regardless of the treat-
ment (nonsurgical management vs surgical treatment).8
Epidemiologic research has demonstrated that female
athletes have a 4- to 6-fold increased risk of ACL injury
compared with their male counterparts playing at similar
levels in the same sports.1,30,34The increased ACL injury
risk coupled with increased sports participation by young
women over the past 30 years (9-fold increase in high
school37and 5-fold increase in collegiate sports36) have
increased public awareness and fueled many gender-specific
mechanistic and interventional investigations. This signifi-
cant problem is associated with a large health care cost, in
the range of $625 million annually, in addition to increased
potential for loss of entire seasons of sports participation,
loss of possible scholarship funding, lowered academic per-
formance, long-term disability, and up to 100 times greater
risk of radiographically diagnosed osteoarthritis.5,9,11
Efforts to prevent ACL injury in female athletes should
focus on the factors that make women more susceptible to
injury and to develop interventions to aid in the prevention
of these injuries. The following meta-analysis attempts to
quantitatively combine the results of 6 independent studies
Anterior Cruciate Ligament Injuries
in Female Athletes
Part 2, A Meta-analysis of Neuromuscular Interventions
Aimed at Injury Prevention
Timothy E. Hewett,*†‡PhD, Kevin R. Ford,†MS, and Gregory D. Myer,†MS, CSCS
From the †Cincinnati Children’s Hospital Research Foundation, Sports Medicine Biodynamics
Center and Human Performance Laboratory, Cincinnati, Ohio, and the ‡University of Cincinnati
College of Medicine, Departments of Pediatrics, Orthopaedic Surgery and Rehabilitation
Sciences, College of Allied Health Sciences, Cincinnati, Ohio
Female athletes have a 4 to 6 times higher incidence of anterior cruciate ligament injury than do male athletes participating in
the same landing and pivoting sports. This greater risk of anterior cruciate ligament injury, coupled with a geometric increase in
participation (doubling each decade), has led to a significant rise in anterior cruciate ligament injuries in female athletes. The
gender gap in anterior cruciate ligament injury, combined with evidence that the underpinnings of this serious health problem are
neuromuscular in nature, leads to the development of neuromuscular interventions designed to prevent injury. A systematic
review of the published literature yielded 6 published interventions targeted toward anterior cruciate ligament injury prevention
in female athletes. Four of 6 significantly reduced knee injury incidence, and 3 of 6 significantly reduced anterior cruciate liga-
ment injury incidence in female athletes. A meta-analysis of these 6 studies demonstrates a significant effect of neuromuscular
training programs on anterior cruciate ligament injury incidence in female athletes (test for overall effect, Z = 4.31, P < .0001).
Examination of the similarities and differences between the training regimens gives insight into the development of more effec-
tive and efficient interventions. The purpose of this “Current Concepts” review is to highlight the relative effectiveness of these
interventions in reducing anterior cruciate ligament injury rates and to evaluate the common training components between the
training studies. In addition, the level of rigor of these interventions, the costs and the difficulty of implementation, the compli-
ance with these interventions, and the performance benefits are discussed. This review summarizes conclusions based on evi-
dence from the common components of the various interventions to discuss their potential to reduce anterior cruciate ligament
injury risk and assess their potential for combined use in more effective and efficient intervention protocols.
Keywords: neuromuscular training; balance training; strength training; plyometrics; knee injury; anterior cruciate ligament (ACL)
injury; injury prevention; gender differences
*Address correspondence to Timothy E. Hewett, PhD, Cincinnati
Children’s Hospital, 3333 Burnet Avenue, MLC 10001, Cincinnati, OH
45229 (e-mail: firstname.lastname@example.org).
No potential conflict of interest declared.
The American Journal of Sports Medicine, Vol. 34, No. 3
© 2006 American Orthopaedic Society for Sports Medicine
Clinical Sports Medicine Update
Vol. 34, No. 3, 2006 Prevention Programs
drawn from a systematic review of the published literature
regarding ACL injury interventions in female athletes.This
analysis summarizes and synthesizes the findings of all
6 studies to draw generalized conclusions on the effective-
ness of neuromuscular training interventions in reducing
ACL injuries during sports competition in female athletes.
This meta-analysis was designed to identify the effectiveness
of training interventions to prevent ACL injuries during ath-
letics. We searched electronic databases, Medline (1966-
2004) and CINAHL (1982-2004), with the subject terms knee
injury and sports injury.44The results were further limited to
the terms intervention and control. Articles were included in
the meta-analysis if they were a randomized controlled trial
(RCT) or prospective cohort study and investigated a neuro-
muscular training intervention used for prevention of ACL
injury in female athletes. Abstracts were excluded from this
review. Six articles were identified that met the systematic
review criteria (Table 1).The total number of ACL injuries in
the training and control groups was entered into a statistical
package to calculate the overall effect of training.4The
analysis used weighted sample sizes for each study.
The results of the meta-analysis are outlined in Figure 1.
The meta-analysis shows the total ACL injuries in the train-
ing group (n = 29) versus the control group (n = 110). These
6 studies significantly favor injury prevention training
programs for reducing ACL injuries (test for overall effect,
Z = 4.31, P < .0001). A power analysis yielded a required
minimum number of 344 athletes in both the trained and
untrained groups for 80% power in the reviewed studies.
The 6 studies are described in the following section by order
of publication date.
The Effects of Intervention Training
on ACL Injury Incidence
Hewett et al.Hewett et al17(Table 2) conducted a prospective
cohort study monitoring high school–aged female soccer, bas-
ketball,and volleyball players;15 female teams (n = 366) were
included in the neuromuscular training intervention, and an
additional 15 female teams (n = 463) were used as a control
group.Thirteen male teams (n = 434) were also included as an
additional control group. A limitation of this study was that
there were more volleyball players in the trained group than
Estimation of Injury Incidence per 1000 Exposures
ACL Knee Injuries,
Hewett et al17
14-18Trained (n = 366)17 2220.12a
Untrained (n = 463)
Trained (n = 42)
Untrained (n = 258)
Trained (n = 62)
Untrained (n = 78)
Trained (n = 263)d
Heidt et al15
Soderman et al43
Soccer20.4 ± 4.6
20.5 ± 5.4
Myklebust et al34
Untrained (n = 645)d
Trained (n = 850)e
Untrained (n = 942)f
Trained (n = 1885)
Untrained (n = 3818)
Untrained (n = 134)
Mandelbaum et al31
Petersen et al40
Untrained (n = 142)—c
aSignificant decrease in injuries through training intervention.
bInjury exposures estimated from Hewett et al17soccer exposures times 2 seasons.
cInjury exposures estimated as 2 hours = 1 practice or game exposure.
dScreening year 2000-2001, completed intervention versus dropped out.
eScreening year 2000-2001, intervention, all athletes.
fScreening year 1998-1999, no intervention.
Hewett et alThe American Journal of Sports Medicine
in the untrained group.The intervention consisted of a 6-week
neuromuscular training intervention19performed 3 times a
week (60-90 min/session) before their competitive season.
Noncontact ACL injury risk was significantly reduced in
the trained female athletes (P ≤ .05).The rate of noncontact
ACL injury was decreased 72% in those athletes who
underwent preseason neuromuscular training compared
with the untrained group. Five untrained female athletes
(3 basketball, 2 soccer) sustained a noncontact ACL injury
compared with none of the trained female athletes (P ≤ .05).
This was the first study to demonstrate the effects of neu-
romuscular training on reducing ACL injury rates in female
athletes. It is a well-designed prospective study but would
be stronger as an RCT.
Heidt et al. Heidt et al15performed a neuromuscular
training intervention on high school female soccer players
(Table 1). The study consisted of a control group (n = 258)
and an intervention group (n = 42) trained before the start
of their competitive seasons.The intervention group partic-
ipated in 13 treadmill speed-training sessions (2 times/wk)
and 7 foot agility sessions (line jumps that progress from
unidirectional to multidirectional to 2-in incremented
barrier hops) completed within a 7-week period.
The trained group had significantly fewer (14%) over-
all injuries than did the control group (33.7%, P < .01).
However, there were no differences in the occurrence of
ACL injuries between the groups. Anterior cruciate liga-
ment rupture occurred in 2.4% of the trained group com-
pared with 3.1% of the controls. The occurrence of medial
collateral ligament sprain/tear was 2.4% compared with
2.3% in the intervention and control groups, respectively.
The lack of significant difference in ACL injury rates may
possibly be attributed to the fact that only minimal low-
intensity plyometrics were incorporated into the training
protocol (footwork and agility drills). In addition, the study
was underpowered (the N was too low) to demonstrate dif-
ferences in ACL injury rates. Furthermore, the definitions
of injuries were vague and nonobjective (eg, bursitis was
defined as a knee injury in this study).
Soderman et al. A randomized control trial on profes-
sional female soccer players from Sweden was conducted
by Soderman et al43(Table 1). Seven teams (n = 62) were
randomized into the intervention group,and 6 teams (n = 78)
served as controls. There were no significant differences in
traumatic injuries or ACL injuries between groups. The
intervention group had 4.45 injuries per 1000 hours of prac-
tices and games compared to 3.83 in the control group. The
intervention group sustained significantly more major
injuries, 8 major injuries, than did the controls (1 major
injury, P = .02).
This study was conducted with a very low number of sub-
jects and inappropriate statistical power. The treatment did
not appear to be effective in reducing ACL injury incidence.
The lack of significant difference in ACL injury rates may
possibly be attributed to the fact that only minimal balance
training was incorporated into the protocol. Furthermore,
the injury reporting mechanisms were poorly described. In
addition, the balance board training was performed by the
athlete at home.This training approach may have poor com-
pliance.The balance training program performed 3 times per
week did not reduce lower extremity sprains.
Myklebust et al. Myklebust et al34performed an ACL
intervention study in female team handball players. This
prospective cohort study monitored ACL injury incidence for
3 consecutive seasons in 3 divisions of Norwegian female
handball. An intervention designed to prevent ACL injuries
was instituted during the second (58 teams, n = 855) and
third seasons (52 teams, n = 850) of play.There were 29 ACL
injuries in the initial control season compared with 23 and
17 in the next 2 intervention seasons, respectively (P = .62
and .15, respectively). There was a significant reduction in
the number of noncontact injuries from the control season to
the second intervention year (18 control year, 7 intervention
year 2; P = .04). When separated by division and training
intervention compliance, the elite division that performed
the intervention had a significant reduction in ACL injuries
(2.3%) compared with the athletes who did not complete the
intervention (8.9%, P = .01). When normalized to injury
Figure 1. Effect of injury prevention training programs on the odds of an ACL injury occurring. Total ACL injuries in the training
group (n = 29) versus the control group (n = 110). Meta-analysis favors injury prevention training on ACL injury outcome. OR, odds
ratio; CI, confidence interval.
Vol. 34, No. 3, 2006 Prevention Programs
Details on Compliance of the 6 Intervention Programs
Hewett et al17
(3 per week
Olympic Gym ($4000);
for 6 weeks)
($500); videos ($10) (15 teams)
Heidt et al15
Fee for service
in 7 weeks)
Soderman et al43
108 in season
(30 in first month; 3
times/wk for 26 weeks)
Myklebust et al34
62: 21 preseason
12 balance boards and
(3 times/wk for 7 weeks)
mats ($50.00 each);
and 41 in season
(1 time/wk for 41 weeks)
62: 21 preseason
boards and mats
for 7 weeks) and 41
in season (1 time/wk
for 41 weeks)
Mandelbaum et al31
36 in season (assumes
that the warm-up was
performed for all
Petersen et al40
65: 24 preseason
Coach or PT
boards and mats
for 8 weeks); assumes
41 in-season weeks
aPer athlete per season. Based on $28.80/h for physical therapist (PT) salary and $15.30/h for certified athletic trainer (ATC) salary (national averages from the American
Physical Therapy Association and the National Athletic Trainers’ Association).
bPer athlete per season or per team if equipment was not given to every individual. Prices estimated and standardized between studies.
cPer athlete per season.
dThe coach was given the questionnaire on the last week of the season and was asked to observe if the athletes on the team performed the prescribed protocol. If the question-
naire was returned, then all athletes on the team were considered compliant.
Hewett et alThe American Journal of Sports Medicine
exposures (player-hours),ACL injury risk decreased 36%. In
the elite division among those who met the compliance
criteria (at least 15 training sessions of 15-21 possible ses-
sions), there was a significant drop in injury rates (P = .01).
Half of the ACL injuries were noncontact injuries.When the
authors separated out the contact injuries, they observed 18
noncontact ACL injuries in the control season and 7 in the
second intervention season (P = .04).
Although there was a trend toward a reduction in ACL
injuries for the entire cohort (P = .15),it was not statistically
significant except for the elite division (P = .06). One expla-
nation for this phenomenon is the fact that the elite players
participated in 5 to 10 practice sessions per week (in con-
trast to the minimum 15 sessions) over the 5- to 7-week
training period. Therefore, the Division I athletes may have
had more training opportunities to gain ACL injury preven-
tion protective effects through training.
Mandelbaum et al. This controlled cohort study31enrolled
soccer players between the ages of 14 and 18 years over a
2-year period. During the first year, 52 teams (n = 1041)
were enrolled in the intervention group, and 95 age- and
skill-matched teams (n = 1905) that were untrained served
as controls. The second-year intervention group consisted of
45 teams (n = 844), and 112 teams (n = 1913) served as the
During the first season, there were 2 noncontact ACL
injuries resulting from 37476 athlete exposures in the
intervention group (0.05 incidences per 1000 exposures),
which was significantly fewer (P < .001) than the injuries
of the control group—32 ACL injuries resulting from
68 580 athlete exposures (0.47 incidences per 1000 expo-
sures). Similar results were found in the second year, with
0.13 and 0.51 incidences per 1000 exposures in the inter-
vention and control groups, respectively (P < .01). Combined
over the 2 years of the study, a total of 6 ACL ruptures
occurred in the training group in comparison with 67 in
the control group.
This was a prospective study that would be stronger as
a randomized control trial. Furthermore, the mechanism of
injury to differentiate contact from noncontact injuries was
not clearly defined. This study employed end-of-season
injury reporting, which can be problematic. The ACL
injuries may have been underreported in this setting. In
addition, the mechanism (contact vs noncontact) may have
Petersen et al. Petersen et al40performed a controlled,
prospective case control study of ACL injury prevention in
German female team handball players. An intervention
designed to prevent ACL injuries was instituted with
10 teams, with a total of 134 players, 10 other teams
(142 players) followed their normal training routines. The
ACL injury prevention intervention was based primarily
on the work of Myklebust et al34and consisted of 3 exercise
components: balance board exercises, jump exercises, and
balance mat exercises. Each component was progressed in
6 phases from easy to more difficult.
There were 5 ACL injuries in the control group compared
with 1 in the trained group (odds ratio,0.17;95% confidence
interval, 0.02-1.5). There was not a significant reduction in
the number of ACL injuries in the intervention compared
with the control group, although ACL injury risk was 80%
lower in the intervention group.
Although the findings of the study were not statistically
significant, the results looked potentially promising.A prob-
lem with this study was that it was underpowered;the num-
ber of subjects was likely too low to accept the null
hypothesis of no difference. Our power analysis predicted
that 134 players in the intervention group and 142 players
in the control group would not give sufficient power to detect
differences between groups. However, in this case, the treat-
ment did not appear to be effective in bringing about the
desirable change (ACL injury reduction).
Common Components of the Effective Interventions
The effects of 3 of the 6 interventions that reduced ACL
injury rates appear to be relatively similar, arising from a
common rationale derived from performance enhancement
training and physical rehabilitation for athletes.15,17,31,34,40,43
A comprehensive review of all 6 interventions reviewed
suggests that multiple neuromuscular training compo-
nents may provide some level of ACL injury risk reduction.
Neuromuscular training likely alters active knee joint sta-
bilization and appears to aid in decreasing ACL injury
rates in female athletes.
An examination of the data extracted from the interven-
tion studies leads one to a few potentially valuable general-
izations. Plyometric training combined with biomechanical
analysis and technique training were common components
of all 3 studies that effectively reduced ACL injury rates.
Balance training alone is probably not as effective for injury
prevention as when it is combined with other types of train-
ing. One needs to consider whether the teams’ or athletes’
primary goal is injury prevention, performance enhance-
ment, or both. In-season training alone is probably the most
cost-effective and efficient method for achieving beneficial
injury prevention effects,although the lack of high-intensity
overload in these programs likely precludes measurable per-
formance enhancement effects. Recent studies employing
the in-season training program of Mandelbaum et al31demon-
strated that the ACL injury reduction is not observed until
later in the season in collegiate soccer, and this in-season
program did not appear to change biomechanical risk fac-
tors.12,41Finally, the most effective and efficient programs
appear to require a combination of components, and the
effects of these components are potentially additive.
Plyometric Component of the Effective Interventions. The
studies by Hewett et al,17Myklebust et al,34Mandelbaum
et al,31and Petersen et al40incorporated high-intensity
jumping plyometric movements that progressed beyond
footwork and agility into their intervention designs. The
studies by Heidt et al15and Soderman et al43did not. All 4
studies that incorporated plyometrics reduced ACL risk,
whereas the 2 studies that did not incorporate plyometrics
did not reduce ACL injury risk. The plyometric component
of an exercise intervention, which trains the muscles,
connective tissue,and nervous system to effectively carry out
Vol. 34, No. 3, 2006 Prevention Programs
the stretch-shortening cycle and focuses on proper technique
and body mechanics, appears to reduce serious ligamentous
injuries, specifically ACL injuries. Training interventions
that incorporate plyometrics with safe levels of varus or val-
gus stress may induce more muscle-dominant neuromuscu-
lar adaptations to correct for neuromuscular imbalances in
female athletes.18,28Such adaptations may better prepare an
athlete for more multidirectional sport activities and may
reduce positioning that puts high loads on the ACL.27
Biomechanics Technique Feedback Effects.The 3 programs
that significantly reduced ACL injury risk also used analy-
sis of the movement biomechanics and feedback to the
athlete regarding proper body position and technique.
The studies by Hewett et al,17Myklebust et al34and
Mandelbaum et al31all incorporated critical technique
analysis and feedback during training into their inter-
vention designs. The studies by Heidt et al15and Soderman
et al43did not. Of these 2 studies, neither reduced ACL
injury risk. Education and enforced awareness of dangerous
positions and mechanisms of ACL injury have also been
shown to decrease ACL injuries.21Ski instructors viewed
videotapes of ACL injuries and were encouraged to formu-
late their own preventive strategies. Anterior cruciate liga-
ment injuries were decreased by more than 50% with this
technique. Olsen et al38reported that a video-based injury
awareness program did not decrease injury rates in soccer.
Awareness programs alone without training may not be
effective in landing and cutting sports. However, elements
from this ski study may be applicable to other sports. It is
important to teach athletes to avoid biomechanically disad-
vantageous and dangerous positions in any sport. Griffin13
identified 3 potentially dangerous maneuvers in basketball
that she proposed should be modified through training to
prevent ACL injury. She suggested that athletes land in a
more bent knee position and decelerate before a cutting
maneuver. Preliminary work implementing the different
techniques on a small sample of athletes showed a trend
toward a decrease in injury rates between the trained versus
untrained study groups.13
Hewett et al17expanded this concept and used a trainer to
provide feedback and awareness to an athlete during train-
ing. Verbalization and visualizations such as “on your toes,”
“straight as an arrow,” “light as a feather,” “shock absorber,”
and “recoil like a spring” were used by the trainer as verbal
and visualization cues for each phase of the jump. Athletes
were required to perform jumps using only proper technique.
As the athletes became fatigued,they were required to stop if
they could not execute each jump with correct biomechanics.
Myklebust et al34used partner training to provide the criti-
cal feedback. Partners encouraged each other to focus on the
quality of their movements, specifically on the knee-over-
the-toe position. Mandelbaum et al31used a training video
to emphasize proper body position and movement mechanics
during running and landing. Three studies17,31,34specifically
cited critical analysis and feedback as contributors to the
reduction of ACL injuries in their respective studies.
Balance and Core Stability Training. Balance training
alone may not be sufficient to produce significant ACL
injury prevention effects. Soderman et al43focused on
balance training, primarily using unstable wobble boards.
However, the intervention in this study was not effective in
reducing ACL injuries. Caraffa et al3prospectively evalu-
ated the effect of balance board exercises on noncontact
ACL injury rates in male soccer players. Soderman et al43
attempted to replicate the Caraffa et al3study performed
in male athletes with female athletes, although the
impressive effect on ACL injury was not replicated in the
female soccer players. The training consisted of approxi-
mately 20 minutes of balance board exercises divided into
5 phases. They compared athletes who participated in
proprioceptive training before their competitive seasons
versus controls and found a significantly decreased rate of
ACL injuries in the trained group.3
The studies by Hewett et al17and Mandelbaum et al31
incorporated single-leg core stability (functional balance)
training, primarily using hold positions from a decelerated
landing, into their intervention designs. Myklebust et al34
examined the effects of a relatively comprehensive functional
balance training intervention. Their intervention elaborated
on the balance board protocol of Caraffa et al3by adding
a focus to improve awareness and knee control during
standing, cutting, jumping, and landing.They demonstrated
a reduction in the incidence of ACL injury in women’s elite
handball division over 2 competitive seasons.34Others have
shown that this type of proprioceptive and balance training
can improve postural control and that lack of postural con-
trol and stability was also related to increased risk of ankle
injury.20,39,45,46Likewise, improvement in single-leg stability
can be gained with a neuromuscular training intervention
that incorporates perturbations into balance training on
unstable surfaces.39Balance training has also been shown to
improve maximum lower extremity strength and decrease
side-to-side imbalances in stabilometric measures.16Side-to-
side imbalances in lower extremity measures have been
shown to be a risk factor for ACL injury.22The above findings
support the integration of proprioceptive stability and bal-
ance training in ACL injury interventions. However, it
appears that balance drills using unstable platforms alone
may not be sufficient to reduce ACL injury risk.
Strength Training. The studies by Hewett et al17and
Mandelbaum et al31incorporated strength training in their
intervention protocols. Myklebust et al,34Heidt et al,15
Soderman et al,43and Petersen et al40did not include
strength training in their interventions. The designs that
incorporated strength training were among the most effec-
tive at decreasing ACL injury rates, but strength training
may not be a prerequisite for prevention, as the Myklebust
et al34study was effective and it did not incorporate strength
training. Strength training may be optional for injury pre-
vention; however, the biomechanical and strength changes
observed may have been owing in part to the strength train-
ing component.19Resistance training alone has not been
shown to reduce ACL injuries. However, there is inferential
evidence that resistance training may reduce injury based
on the beneficial adaptations that occur in bones, ligaments,
and tendons after training.7,23Lehnhard et al26were able
to significantly reduce injury rates with the addition of a
strength training regimen to a men’s soccer team. They
monitored injuries for 2 years without training and 2 years
with the strength training treatment added. Although they
Hewett et al The American Journal of Sports Medicine
did not observe a reduction specifically in ACL injuries, they
did report a decrease in percentage of injuries that were lig-
ament sprains.The significant reduction of ligament sprains
may have been related to reduced knee injury (43%)
reported in the second year of posttrained competition.26
Resistance training may aid in the reduction of ACL injuries
when combined with other training components; however,
the efficacy of a single-faceted resistance training protocol
on ACL injury prevention has yet to be determined.
Efficacy and Efficiency of These Interventions
Relative Difficulty, Intensity, and Time Cost. The relative
difficulty, cost, and efficiency of these interventions need to
be addressed. The details of each program are shown in
Table 2. The relative training volume and intensity could
be ranked as follows: Hewett et al17> Heidt et al15>>
Myklebust et al34= Petersen et al40> Mandelbaum et al31>>
Soderman et al.43The studies by Hewett et al,17Heidt
et al,15Mandelbaum et al31and Myklebust et al34incor-
porated more high-intensity movements into their inter-
vention designs. The Hewett et al17and Heidt et al15study
involved preseason high-intensity neuromuscular train-
ing, Myklebust et al34and Petersen et al40involved both
preseason and in-season medium-intensity (prepractice
warm-up) training, Mandelbaum et al31involved in-season
medium-intensity (prepractice warm-up) training, and
Soderman et al43involved in-season low-intensity balance
board neuromuscular training. The training time per ath-
lete is initially high for the preseason programs, but this
time per athlete tends to balance out over the length of the
season. These programs also save time during the season.
However, a combination preseason andin-seasonprogram,
similar to that of Myklebust et al,34may prove the most
Compliance. Reported compliance rates and definition of
compliance vary greatly between studies (Table 2). Hewett
et al17reported 70% compliance rates, Soderman et al43
reported 63%, Heidt et al15reported 100%, Mandelbaum
et al31reported 98%, and Myklebust et al34reported rates as
low as 28%. Petersen et al40did not report compliance rates.
Myklebust et al34examined compliance with the greatest
rigor and in the greatest detail,which may account for their
relatively low percentage compliance. They separated their
data by training intervention compliance. In the elite divi-
sion among those who met the compliance criteria (at least
15 training sessions of 15-21 possible sessions), there was a
significant drop in injury rates. Myklebust et al34should be
commended for their more detailed analysis of compliance,
which was not well addressed in the other studies. They
also had the most stringent inclusion criteria. Their data
demonstrate what a challenge compliance is,even with ath-
letes of high caliber, as only 26% of the teams were judged
compliant the first year and 29% the second year of inter-
vention.Table 2 indicates that when compliance is an impor-
tant consideration, the team coaches are likely to direct the
most compliant programs, and they are the most inexpen-
sive source of training assistance.
Effects of the Interventions
The high-intensity neuromuscular overload associated with
strength training and plyometrics likely enhances both mus-
cular power and performance and the injury prevention
effects of neuromuscular training. Strength training may
be optional for injury prevention; however, this type of train-
ing is prerequisite to overloading the muscle and gaining opti-
mal performance-enhancement effects. Overload is required
for measurable muscular adaptation. Hewett et al17and
Mandelbaum et al31both incorporated a strength training
component into their interventions.The difficulty in assessing
the relative effects on performance is that only one study has
assessed effects on performance.19These authors demon-
strated an approximately 10% increase in vertical jump height
with 6 weeks of training. However, numerous neuromuscular
training programs designed for young women can be effective
at improving performance measures of speed, strength, and
power.19,24,25,32Female athletes may especially benefit from
neuromuscular training,as they often display decreased base-
line levels of strength and power compared with their male
counterparts.19,29Dynamic neuromuscular training has also
been demonstrated to reduce gender-related differences in
force absorption, active joint stabilization, muscle imbalances,
and functional biomechanics while increasing strength of
structural tissues (bones, ligaments, and tendons).6,7,19,32,33,42
Neuromuscular training may reduce the risk of injury
in female athletes; however, without the performance-
enhancement effects, athletes may not be motivated to par-
ticipate in a neuromuscular training program. Prevention
training that is oriented toward reducing ACL injuries in
female athletes may have compliance rates as low as 28%.34
However, training for performance enhancement can have
better compliance ranging from 80% to 90%.2,14,23,24,47
Hence, if protocols are designed for both performance
enhancement and ACL injury prevention techniques, neu-
romuscular training may be instituted on a widespread
basis with potentially higher athlete compliance.
Limitations of These Intervention Studies
There are several limitations to this body of studies as a
whole and to each of the 6 individual studies. The number
of participants in each study is low for epidemiologic stud-
ies.Therefore, the number of both injuries and exposures is
relatively low, and statistically, these studies are under-
powered. In addition, the measure of exposures was not
consistent. Four of these studies are not randomized con-
trolled trails. Randomized controlled trials are needed to
better discern the effects of these interventions. There was
no mention made of a power analysis in these studies.
Teams in the Hewett et al17study were not cluster ran-
domized, and compliance was not rigorously assessed. The
Soderman et al43and Petersen et al40studies were under-
powered to attempt to examine ACL injury epidemiology.
Soderman et al43had only 62 trained and 78 untrained
athletes, and Petersen et al40had only 137 and 142 athletes
Vol. 34, No. 3, 2006Prevention Programs
in their intervention and control groups, respectively. The
training protocol was not well documented in the Soderman
et al43study. Compliance was also not well documented by
either of these studies.
The Heidt et al15study also had low numbers for attempt-
ing to examine injury epidemiology. The lack of a random-
ized design was also a distinct weakness. The lack of
significant difference in ACL injury rates in the Heidt et al15
study may possibly be attributed to the fact that only mini-
mal low-intensity plyometrics were incorporated into the
training protocol (footwork and agility drills). Furthermore,
the injury reporting methods and the injury definitions were
poorly described. In addition, although the study was pur-
ported to be randomized in the methods, it was not an RCT.
Subjects volunteered to either participate in the training or
not (personal communication from author).
The Myklebust et al34study was a well-done prospective
study.However,there were methodological limitations to this
study.They include a lack of sufficient documentation of ACL
injury status. In addition, the authors make several refer-
ences to nonsignificant results without performing a power
analysis for the study. A power analysis should have been
performed to determine if the lack of significance was likely
owing to an insufficient number of subjects. This inaction is
a potential weakness of all the reported studies. The
Mandelbaum et al31study also had poor compliance reporting.
Exposures were not closely surveyed, as the reported num-
bers appear to be gross estimates of exposure.In addition,the
incidence of injury in the untrained athletes was high for that
age group compared to other values reported in the literature.
Limitations of the Meta-analysis
There are several limitations to our meta-analysis of the
literature. One potential limitation is publication bias. Only
positive findings tend to get published in the literature.This
factor could potentially bias our analysis toward positive
results. An important potential limitation of this analysis is
general heterogeneity between studies. Another potential
problem, which is a subset of the previous limitation, is the
mixed designs between studies. It may not be appropriate to
compare studies with different study designs. Yet another
potential limitation of this analysis related to study hetero-
geneity in the different treatments. It is not clear whether
one should expect the same effect from the 5 different treat-
ments we analyzed.However,each of the authors did hypoth-
esize that the intervention would reduce ACL injury. Finally,
there is the problem of different follow-up times in the vari-
ous investigations,implying that the odds ratio is potentially
not the best statistic to meta-analyze. However, it would be
very difficult to calculate relative risks (adjusted to follow-up
times and expressed in person-years), so all that can be done
is to list these problems as limitations of the analysis.
CONCLUSION AND FUTURE DIRECTIONS
There is evidence that neuromuscular training decreases
potential biomechanical risk factors for ACL injury and
decreases ACL injury incidence in female athletes. Three
of the 6 interventions in this meta-analysis demonstrated
significant effects on ACL injury rates.
Five of 6 demonstrated positive trends and reduction in
odds ratios. However, we do not yet know which of these
components is most effective or whether their effects are
combinatorial. Future directions will be to assess the rela-
tive efficacy of these interventions alone and in combination
to achieve the optimal effect in the most efficient manner
possible. Final conclusions from this examination of these 6
studies are that neuromuscular training may assist in the
reduction of ACL injuries in females athletes if
1. plyometrics,balance,and strengthening exercises are
incorporated into a comprehensive training protocol;
2. the training sessions are performed more than
1 time per week; and
3. the duration of the training program is a minimum
of 6 weeks in length.
The studies by Hewett et al,17Myklebust et al34and
Mandelbaum et al31incorporated high-intensity plyomet-
ric movements that progressed beyond footwork and
agility in the intervention. The studies by Heidt et al15
and Soderman et al43did not. All 3 studies that incorpo-
rated high-intensity plyometrics reduced ACL risk,
whereas the studies that did not incorporate high-inten-
sity plyometrics did not reduce ACL injury risk. The plyo-
metric component of these interventions, which trains
the muscles, connective tissue, and nervous system to
effectively carry out the stretch-shortening cycle and that
focuses on proper technique and body mechanics, appears
to reduce ACL injuries.
The authors acknowledge funding support from National
Institutes of Health grant R01-AR049735-01A1 (T.E.H.).
The authors thank the authors of all 5 reviewed studies for
their excellent work in the field of ACL injury prevention.
The authors also acknowledge Dr Bert Mandelbaum for
inspiring this analysis and thank Tiffany Evans for her
assistance with preparation of the article.We acknowledge
Dr Paul Succop from the Department of Biostatistics at the
University of Cincinnati College of Medicine for his statis-
tical expertise and input regarding the meta-analysis.
1. Arendt E, Dick R. Knee injury patterns among men and women in
collegiate basketball and soccer: NCAA data and review of literature.
Am J Sports Med. 1995;23:694-701.
2. Ben-Sira D, Ayalon A, Tavi M. The effect of different types of strength
training on concentric strength in women. J Strength Cond Res. 1995;
3. Caraffa A, Cerulli G, Projetti M, Aisa G, Rizzo A. Prevention of anterior
cruciate ligament injuries in soccer: a prospective controlled study of
proprioceptive training. Knee Surg Sports Traumatol Arthrosc. 1996;
498 Download full-text
Hewett et al The American Journal of Sports Medicine
4. Cochrane Collaboration. RevMan Analyses. Oxford, England:
Cochrane Collaboration; 2002.
5. Deacon A, Bennell K, Kiss ZS, Crossley K, Brukner P. Osteoarthritis of
the knee in retired, elite Australian Rules footballers. Med J Aust. 1997;
6. Faigenbaum AD, Kraemer WJ, Cahill B, et al. Youth resistance training:
position statement paper and literature review. Strength Cond. 1996;
7. Fleck SJ, Falkel JE. Value of resistance training for the reduction of
sports injuries. Sports Med. 1986;3:61-68.
8. Fleming BC. Biomechanics of the anterior cruciate ligament. J Orthop
Sports Phys Ther. 2003;33:A13-A15.
9. Ford KR, Myer GD, Toms HE, Hewett TE. Gender differences in the
kinematics of unanticipated cutting in young athletes. Med Sci Sports
10. Frank CB, Jackson DW. The science of reconstruction of the anterior
cruciate ligament. J Bone Joint Surg Am. 1997;79:1556-1576.
11. Freedman KB, Glasgow MT, Glasgow SG, Bernstein J. Anterior cruciate
ligament injury and reconstruction among university students. Clin
Orthop Relat Res. 1998;356:208-212.
12. Gilchrist JR, Mandelbaum BR, Melancon H, et al. A randomized con-
trolled trial to prevent non-contact ACL injury in female collegiate
soccer players. Paper presented at: American Orthopaedic Society
for Sports Medicine; March 10-14, 2004; San Francisco, Calif.
13. Griffin LY. The Henning program. In: Prevention of Noncontact ACL
Injuries. Rosemont, Ill: American Academy of Orthopaedic Surgeons;
14. Hakkinen K, Alen M, Kraemer WJ, et al. Neuromuscular adaptations
during concurrent strength and endurance training versus strength
training. Eur J Appl Physiol. 2003;89:42-52.
15. Heidt RS Jr, Sweeterman LM, Carlonas RL, Traub JA, Tekulve FX.
Avoidance of soccer injuries with preseason conditioning. Am J Sports
16. Heitkamp HC, Horstmann T, Mayer F, Weller J, Dickhuth HH. Gain in
strength and muscular balance after balance training. Int J Sports
17. Hewett TE, Lindenfeld TN, Riccobene JV, Noyes FR. The effect of
neuromuscular training on the incidence of knee injury in female ath-
letes: a prospective study. Am J Sports Med. 1999;27:699-706.
18. Hewett TE, Paterno MV, Myer GD. Strategies for enhancing proprio-
ception and neuromuscular control of the knee. Clin Orthop Relat Res.
19. Hewett TE, Stroupe AL, Nance TA, Noyes FR. Plyometric training in
female athletes: decreased impact forces and increased hamstring
torques. Am J Sports Med. 1996;24:765-773.
20. Holm I, Fosdahl MA, Friis A, Risberg MA, Myklebust G, Steen H. Effect
of neuromuscular training on proprioception, balance, muscle strength,
and lower limb function in female team handball players. Clin J Sport
21. Johnson RJ. The ACL injury in female skiers. In: Griffin LY, ed. Prevention
of Noncontact ACL Injuries. Rosemont, Ill: American Academy of
Orthopaedic Surgeons; 2001:107-111.
22. Knapik JJ, Bauman CL, Jones BH, Harris JM, Vaughan L. Preseason
strength and flexibility imbalances associated with athletic injuries in
female collegiate athletes. Am J Sports Med. 1991;19:76-81.
23. Kraemer WJ, Duncan ND, Volek JS. Resistance training and elite ath-
letes: adaptations and program considerations. J Orthop Sports Phys
24. Kraemer WJ, Hakkinen K, Triplett-Mcbride NT, et al. Physiological
changes with periodized resistance training in women tennis players.
Med Sci Sports Exerc. 2003;35:157-168.
25. Kraemer WJ, Mazzetti SA, Nindl BC, et al. Effect of resistance train-
ing on women’s strength/power and occupational performances.
Med Sci Sports Exerc. 2001;33:1011-1025.
26. Lehnhard RA, Lehnhard HR, Young R, et al. Monitoring injuries on a
college soccer team: the effect of strength training. J Strength Cond
27. Lloyd DG. Rationale for training programs to reduce anterior cruciate
ligament injuries in Australian football. J Orthop Sports Phys Ther.
2001;31:645-654, discussion 661.
28. Lloyd DG, Buchanan TS. Strategies of muscular support of varus and val-
gus isometric loads at the human knee. J Biomech. 2001;34:1257-1267.
29. Malina RM, Bouchard C. Growth, Maturation, and Physical Activity.
Champaign, Ill: Human Kinetics; 1991.
30. Malone TR, Hardaker WT, Garrett WE, et al. Relationship of gender to
anterior cruciate ligament injuries in intercollegiate basketball players.
J South Orthop Assoc. 1993;2:36-39.
31. Mandelbaum BR, Silvers HJ, Watanabe D, et al. Effectiveness of a
neuromuscular and proprioceptive training program in preventing
anterior cruciate ligament injuries in female athletes: two-year follow
up. Am J Sports Med. 2005;33:1003-1010.
32. Myer GD, Ford KR, Palumbo JP, Hewett TE. Neuromuscular training
improves performance and lower-extremity biomechanics in female
athletes. J Strength Cond Res. 2005;19:51-60.
33. Myer GD, Hewett TE, Noyes FR. The use of video analysis to identify
athletes with increased valgus knee excursion: effects of gender and
training. Med Sci Sports Exerc. 2000;32:S298.
34. Myklebust G, Engebretsen L, Braekken IH, Skjolberg A, Olsen OE,
Bahr R. Prevention of anterior cruciate ligament injuries in female
team handball players: a prospective intervention study over three
seasons. Clin J Sport Med. 2003;13:71-78.
35. Myklebust G, Maehlum S, Holm I, Bahr R. A prospective cohort study
of anterior cruciate ligament injuries in elite Norwegian team handball.
Scand J Med Sci Sports. 1998;8:149-153.
36. National Collegiate Athletic Association. NCAA Injury Surveillance
System Summary. Indianapolis, Ind: National Collegiate Athletic
37. National Federation of State High School Associations. 2002 High
School Participation Survey. Indianapolis, Ind: National Federation of
State High School Associations; 2002.
38. Olsen OE, Myklebust G, Engebretsen L, et al. Injury mechanisms for
anterior cruciate ligament injuries in team handball: a systematic
video analysis. Am J Sports Med. 2004;32:1002-1012.
39. Paterno MV, Myer GD, Ford KR, Hewett TE. Neuromuscular training
improves single-limb stability in young female athletes. J Orthop
Sports Phys Ther. 2004;34:305-316.
40. Petersen W, Braun C, Bock W, et al. A controlled prospective case
control study of a prevention training program in female team handball
players: the German experience. Arch Orthop Trauma Surg. 10.1007
41. Powers CM, Sigward SM, Ota S, Pelley K. The influence of an ACL
injury training program on knee mechanics during a side-step cutting
maneuver. J Athl Train. 2004;39:S27.
42. Rooks DS, Micheli LJ. Musculoskeletal assessment and training: the
young athlete. Clin Sports Med. 1988;7:641-677.
43. Soderman K, Werner S, Pietila T, Engstrom B, Alfredson H. Balance
board training: prevention of traumatic injuries of the lower extremi-
ties in female soccer players? A prospective randomized intervention
study. Knee Surg Sports Traumatol Arthrosc. 2000;8:356-363.
44. Thacker SB, Stroup DF, Branche CM, Gilchrist J, Goodman RA, Porter
Kelling E. Prevention of knee injuries in sports: a systematic review of
the literature. J Sports Med Phys Fitness. 2003;43:165-179.
45. Tropp H, Ekstrand J, Gillquist J. Stabilometry in functional instability
of the ankle and its value in predicting injury. Med Sci Sports Exerc.
46. Tropp H, Odenrick P. Postural control in single-limb stance. J Orthop
47. Wroble RR, Moxley DR. The effect of winter sports participation on high
school football players: strength, power, agility, and body composition.
J Strength Cond Res. 2001;15:132-135.