Article Text

Download PDFPDF

Gender differences in the knee adduction moment after anterior cruciate ligament reconstruction surgery
  1. Kate E Webster,
  2. Jodie A McClelland,
  3. Simon E Palazzolo,
  4. Luke J Santamaria,
  5. Julian A Feller
  1. Musculoskeletal Research Centre, La Trobe University, Melbourne, Australia
  1. Correspondence to Kate E Webster, Musculoskeletal Research Centre, La Trobe University, Melbourne, Victoria 3086, Australia; k.webster{at}


Background The external knee adduction moment during gait has previously been associated with knee pain and osteoarthritis (OA). Recently, the knee adduction moment has been shown to be increased following anterior cruciate ligament (ACL) reconstruction surgery and has been suggested as a potential mechanism for the progression of early onset knee OA in this population. No study has investigated the gender differences in gait biomechanics following ACL reconstruction.

Aim To examine gender differences in gait biomechanics following ACL reconstruction surgery.

Methods 36 subjects (18 females, 18 males) who had previously undergone ACL reconstruction surgery (mean time since surgery 20 months) underwent gait analysis at a self-selected walking speed. Males and females were well matched for age, time since surgery and walking speed. Maximum flexion and adduction angles and moments were recorded during the stance phase of level walking and compared between the male and female groups.

Results The knee adduction moment was 23% greater in the female compared with the male ACL group. No gender differences were seen in the sagittal plane. No differences were seen between the reconstructed and contralateral limb.

Conclusion The higher knee adduction moment seen in females compared with males may suggest an increased risk for the development of OA in ACL-reconstructed females.

Statistics from


Osteoarthritis (OA) of the knee is one of the most common and debilitating forms of OA. The pain and stiffness associated with knee OA often interferes with activities of daily living, compromises quality of life and is a significant public health issue. The biomechanical variable most researched for its association with knee OA is the knee adduction moment. Research has shown that people with knee OA have greater knee adduction moments during gait than controls.1 Furthermore, greater knee adduction moments have been associated with more severe OA2 and were the best predictor of radiographic progression of knee OA over a 6-year period3 as well as the development of chronic knee pain over a 3–4-year period.4 Although this research suggests that the knee adduction moment may influence OA advancement, a causal link between the knee adduction moment and OA development remains to be established.

A proposed mechanism by which ambulatory joint mechanics may lead to knee cartilage degeneration involves gait patterns which increase the load on articular surfaces. Increased knee adduction moments are indicative of increased loads on the medial compartment relative to the lateral compartment of the knee.5 They have been associated with greater bone mineral distribution on the proximal medial aspect of the tibia relative to the lateral aspect in both normal subjects and subjects with knee OA.6 7 It is hypothesised that this increase in bone density may alter the mechanical load distribution on the cartilage. Over time, this abnormal cartilage loading may change the mechanical properties of the tissue resulting in the degradation of the cartilage matrix, leading to the development of OA.8 As studying this hypothesis in the general population is difficult due to the large number of subjects that would be required, researchers have turned to populations at high risk for the development of knee OA to better understand the link between knee adduction moments and knee OA.9

Early onset knee OA following both anterior cruciate ligament (ACL) injury and reconstruction has been reported in a growing number of studies. In soccer players, 51% of females10 and 41% of males11 who had sustained an ACL injury developed radiographic evidence of knee OA within 12–14 years. A recent study which followed patients who had undergone ACL reconstruction surgery found that 62% of patients who sustained an isolated ACL injury and 80% of patients with combined injuries (ACL injury combined with meniscal injury, medial collateral injury or chondral lesion) developed radiographic evidence of OA within 10–15 years after the surgical procedure.12 Asano et al13 observed articular cartilage degeneration as early as 15 months in a cohort of 105 patients who had undergone ACL reconstruction surgery. In their study, women and older patients (>30 years) had significantly greater degrees of cartilage damage and it was suggested that the cartilage of women may be more vulnerable than that of men.

Although the mechanisms behind the development of knee OA following ACL reconstruction surgery remain to be determined, Butler et al14 suggested that one potential mechanism may be an increased knee adduction moment. In their study, which has been the only study to date to have investigated the knee adduction moment following ACL reconstruction, they reported significantly increased knee adduction moments during level gait in a group of patients after ACL reconstruction who had returned to their preinjury level of activity compared with a gender-, age- and activity level–matched control group.

The findings of Butler et al14 and their suggestion that the knee adduction moment may be related to early onset OA following ACL reconstruction are both interesting and potentially important and also highlight the limited information available regarding the knee adduction moment in this patient population. For instance, the type of graft, which has been shown to influence gait biomechanics,15 was not reported and the gender distribution was also relatively unequal in the Butler study (13 women, 4 men) meaning that gender differences were not able to be explored. Although the incidence of OA development has not been specifically studied between males and females following ACL reconstruction surgery, the risk of sustaining an ACL injury has been shown to be increased in females compared with their male counterparts.16 Altered movement patterns during landing and pivoting, particularly in the coronal plane, have been suggested as a potential explanation for the gender difference in ACL injury risk.17,,21 As such, it should not be assumed that gait biomechanics are the same for males as they are for females after ACL reconstruction surgery. Therefore, the purpose of this study was to compare knee angles and moments during level gait in both the sagittal and coronal planes for both females and males who had undergone ACL reconstruction surgery in order to provide more information about these biomechanical variables in this patient population.



Eighteen female and 18 male participants were compared. All participants had undergone uncomplicated primary ACL reconstruction with a four-strand hamstring autograft. The females were matched to the males for age (±2 years) and the amount of time postsurgery (±2 months). A pool of 52 males who had previously undergone gait analysis was available for this matching. To be eligible to participate all patients met the following inclusion criteria: aged from 18 to 40 years, no previous cruciate ligament damage to either knee, time from injury to reconstruction between 3 weeks and 12 months, no collateral ligament injury, no chondral disruption greater than Noyes grade IIA,22 no meniscal pathology treated by repair and no other significant injuries or surgery to any lower extremity joint. The demographic characteristics of the participants are shown in table 1.

Table 1

Descriptive data for the subjects with ACL reconstructions

Twenty-seven subjects had an isolated tear of the ACL and nine (four females, five males) had an additional meniscal tear. For the females, the meniscal tears were treated by partial meniscectomy in three subjects (one medial, two lateral) and left untreated (stable tear) in the remaining female subject. All five males had partial meniscectomies performed (one medial, four lateral). For all subjects, the ACL reconstruction was performed arthroscopically by an experienced knee surgeon. The gracilis and semitendinosus tendons were harvested through an oblique 3–4-cm incision. The tendons were doubled and attached to a 15-mm closed loop Endobutton (Smith & Nephew Endoscopy, Mansfield, Massachusetts, USA). The femoral tunnel was drilled using a transtibial technique. The surgeon aimed to start the femoral tunnel between the 10 o'clock or 10.30 positions, as visualised with the knee at 80° flexion. Before distal fixation, the knee was taken through 10 cycles of extension and flexion with traction applied to the graft. The distal end of the graft was fixed with interference screw with the knee in 70° of flexion and 15 lbs of traction applied to the graft. Patients were discharged from hospital on the first postoperative day.

Postoperatively, all subjects underwent the same rehabilitation protocol. This protocol encouraged immediate full knee extension and the restoration of quadriceps function as soon as possible. Particular emphasis was placed on the restoration of vastus medialis function. Weight bearing was allowed on an as tolerated basis from the first postoperative day. No braces or splints were used. Progression was guided by the presence and degree of pain and swelling. No emphasis was placed on gait retraining. Subjects completed the IKDC Subjective Knee Evaluation Form23 on the same day as gait analysis. At 12-month follow-up, all patients underwent radiographic review. Standing orthogonal x-rays of the operative leg included an posterior–anterior view with 15° of knee flexion, and a standard lateral view in extension.


Subjects were informed of the nature of the experiment and gave written consent which was approved by University Ethics Committee. Analysis of each participant's gait was conducted over a single test session. Measurements of each subject's pelvis and lower limbs were obtained and 15 reflective markers were attached to the anterior superior iliac spines, sacrum, lateral aspects of the thigh and leg, knee joint axis, lateral malleolus, heel and forefoot of each subject.24 25 The thigh and leg markers were attached to 5-cm-long wands. All markers were attached directly to the skin surface using double-sided tape and secured with adhesive elastic tape.

Data were captured in the central portion of a 10-m linoleum covered walkway using a three-dimensional motion analysis system (Vicon; Oxford Metrics, Oxford, UK). Ground reaction forces were recorded from two force plates (AMTI; Kistler, Winterthur, Switzerland) set in the floor of the laboratory. The system was calibrated according to the manufacturer's instructions before each data collection session. This calibration defined a common coordinate system for the three-dimensional video data. To obtain a reference point for the markers, a static trial was obtained with the subject in quiet standing. For this trial, knee alignment devices (KADs; Motion Lab Systems, Baton Rouge, LA, USA) were fitted over each knee. The KAD is a low tension spring-loaded metal jig on which three equidistant markers are mounted. The design of the KAD enables the Vicon software to establish a virtual knee marker at the central joint of the KAD and this, combined with the KAD wand markers, enables the knee flexion axis to be measured. The KAD markers were also used to calculate the offset of thigh rotation and leg rotation. The KAD was replaced with knee joint axis markers for the gait trials.

Subjects were asked to walk up and down the walkway several times at their own speed until they were relaxed and accustomed to the markers. This also enabled a starting point to be identified so that the subject would contact the force plates in normal stride. Subjects were then asked to complete a number of walks at their self-selected comfortable speed while data were collected. They were not aware of the presence of the force plates until data collection was completed. Data collection continued until a minimum of four trials with good force plate contact was recorded for both left and right limbs.

Data analysis

Lower limb segment trajectories were filtered by fitting Woltring's quintic spline with a mean squared setting of 20 to the data before running the biomechanical model (Vicon Plug-in-Gait; Oxford Metrics). Vicon Plug-in-Gait (Oxford Metrics) biomechanical modelling software was used to process and output the data as kinematic and kinetic profiles.

During walking, the timing of the kinematic and kinetic profiles were normalised as a percentage of a single complete (100%) gait cycle, with initial contact of the foot at 0% and the following contact of the same foot at 100%. Joint moment data were normalised to body mass and height (Nm/kg.m) to allow for between-group comparisons, and the moments reported are external moments. Data for all four walk trials were averaged and dependent variables measured included maximum flexion and adduction angles and moments at the knee during midstance.

Independent samples t tests were used to assess for differences between the female and male participants for each dependent variable. Statistical significance was set at p<0.05. This strategy was adopted because we were interested in looking at the individual effect of each dependent variable rather than the effect of taking all variables together. Despite performing a number of t tests and potentially increasing type I error, we chose to maintain p<0.05 for the level of significance so that potentially important findings were not missed due to an overly conservative α level.26 We have therefore reported the exact p values for each comparison so that the reader can select a more conservative level of significance.


There were no significant differences in age, interval between injury and surgery, time since surgery, subjective knee evaluation or walking speed between the male and female subjects (tables 1 and 2). The males were significantly taller and heavier than the females (p<0.01). None of the subjects had radiographic changes of OA on lateral and flexion weight-bearing posteroanterior views at 12 months.

The group of females who had undergone ACL reconstruction surgery exhibited a 23% greater peak knee adduction moment about the operated knee than the group of male ACL reconstruction patients, which was a statistically significant difference (p=0.025, figure 1). No significant between-group differences were seen for the peak adduction angle or the peak flexion angle and moment (table 2). There were also no significant differences between the operated and contralateral limbs for any of the dependent variables for either patient group.

Table 2

Descriptive statistics for knee kinematic and kinetic variables for the operated knee recorded during the stance phase of gait between the male and female ACL reconstruction subjects

Figure 1

Mean (SD) knee adduction moment for the operated knee in males and females during the stance phase of level walking. Error bars are shown in one direction for clarity.


The results of this study showed that females who had undergone ACL reconstruction surgery had a 23% greater knee adduction moment about the operated knee during level walking compared with males who had undergone the same ACL reconstruction procedure. Out of the number of gait variables measured, the knee adduction moment was the only one to show a significant gender difference.

The current results can be compared with the only other study to have reported on the knee adduction moment during level walking following ACL reconstruction. Butler et al14 reported an adduction moment of 0.36 Nm/kg.m for their ACL group, which was significantly greater than a control group which had an adduction moment of 0.30 Nm/kg.m. Both groups comprised mostly of females. This corresponds to the present study in which the adduction moment in our female ACL patients was 0.38 Nm/kg.m.

ACL reconstruction surgery aims to allow the patient to return to sport participation with normal knee function that does not lead to symptomatic or radiographic evidence of OA in later life.27 A growing number of studies are reporting radiographic evidence of early onset knee OA following both ACL injury and reconstruction.10,,13 The most recent of these studies showed a high prevalence of radiographic knee OA in a cohort of 181 patients 10–15 years following patellar tendon ACL reconstruction. The incidence was 62% in patients who had an isolated rupture but increased to 80% in patients with a combined (meniscal or collateral ligament injury or chondral damage) injury.12 Although knee OA has not been shown to be more prevalent in females compared with males following ACL reconstruction surgery, previous reports have shown the prevalence and severity of knee OA of all types to be greater in women compared with men.28,,32 As such, the higher knee adduction moment seen in females compared with males in the current study may suggest an increased risk for the development of OA in ACL-reconstructed females and this warrants further exploration.

A potential limitation of the present study is that the male and female ACL patients were not compared with an uninjured control group. Therefore, it is possible that the greater adduction moment seen in the female ACL patients was not due to the reconstruction procedure but instead reflects an intrinsic gender difference. This, however, seems unlikely and is not supported by previous research. In a large gait study of 110 healthy men and women, there were no gender differences for a number of gait measures, including the knee adduction moment, during barefoot level walking.33 Nonetheless, the premise that the knee adduction moment is increased following ACL reconstruction surgery was based largely on the findings of Butler et al14 and although we refer to the knee adduction moment as being greater in females compared with males, we do not provide evidence in this study that it is greater than ‘normal’. In fact, the adduction moment was not significantly greater than the contralateral side. However, this may simply mean that the contralateral side is not the best comparator, as there is some suggestion of bilateral biomechanical changes after ACL injury and reconstruction.34 That said, a clear difference between the ACL-reconstructed and control groups was seen in the Butler et al study in which the control group was well matched for activity level.

Although increased knee adduction moments have been related to both the increased severity of OA as well as the progression of knee OA in other populations,2 3 35 36 there is no evidence to suggest that high adduction moments per se cause OA to develop in joints without cartilage damage. As such, it is relevant to note that abnormal biomechanics in other movement planes may equally contribute to the development or progression of knee OA. A number of studies have reported abnormal tibial rotation movements during the stance phase of both walking37,,39 and running gait40 as well as during lunging41 and pivoting42 43 in patients who have undergone ACL reconstruction surgery. However, it is currently not clear what amount of rotational change may be clinically important for the development of degenerative changes in the knee joint. It is possible that degenerative changes may develop from the combined changes in a number of biomechanical variables.

A number of potential load-modifying interventions have been put forward to reduce the knee adduction moment during walking. Footwear interventions, such as shoes with a lateral wedge, have been shown to reduce the knee adduction moment in some studies.44,,46 The use of a brace as well as various walking strategies, such as an increased toe-out angle, increased mediolateral trunk sway and reduced walking speed, have also been shown to be effective.47,,50 Some of these strategies may have limited clinical use due to the difficulty in sustaining such modifications over time. Recent research which used a real-time active feedback system to encourage participants to shift pressure to the medial side of the foot during walking and resulted in significant reductions in the knee adduction moment is also promising.51 However, no research has evaluated the effectiveness of any of these interventions in patients with ACL reconstruction.

Strengths of the present study include the close matching of the male and female participants in terms of age, time between injury and surgery, follow-up period and walking speed. However, the males were taller and heavier than the females, so joint moments were normalised for weight and height to account for this. A single surgeon performed the reconstruction procedures and the rehabilitation protocol was also similar for all patients. This, however, means that the population was homogenous and this may limit the ability to generalise the findings to other populations. Although previous studies have shown strong associations between the magnitude of the knee adduction moment and OA incidence and severity, it is also important to note that it was beyond the scope of this study to investigate the actual cartilage contact forces.

In summary, the results of this study showed that a group of females who underwent ACL reconstruction surgery exhibited a significantly larger knee adduction moment about the operated knee during walking than a group of males who had undergone the same reconstructive procedure. It is currently not clear why females have this greater knee adduction moment. Future research should aim to explore the relationship between greater knee adduction moments and the risk for development of OA in this patient group. Long-term studies are also required to determine whether the adduction moment changes as participants continue to engage in strenuous activities and whether load-modifying interventions are warranted.

What is already known on this topic

  • The external knee adduction moment has been shown to be greater in individuals who have undergone ACL reconstruction surgery compared with control participants.

  • A higher knee adduction moment may reflect a potential mechanism for the early progression of knee osteoarthritis in this population.

What this study adds

The knee adduction moment was 23% greater in the female participants who had undergone ACL reconstruction surgery compared with the male participants. It should, therefore, not be assumed that biomechanical variables are the same between genders, and future studies should be conducted to determine the relationship between a greater knee adduction moment and osteoarthritis development in this population.


Supplementary materials

Related Data


  • Funding The authors would like to acknowledge the Australian Orthopaedic Association Research Foundation for providing funding for this project.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Faculty of Health Sciences, La Trobe University.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.