Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jan 5.
Published in final edited form as: J Biomech. 2024 Jan 5;163:111925. doi: 10.1016/j.jbiomech.2024.111925

Unbalanced Medial-to-Lateral Knee Muscle Co-Contractions are Associated with Medial Tibiofemoral Underloading during Gait Three Months after Anterior Cruciate Ligament Reconstruction

Abdulmajeed Alfayyadh 1,2, Jack R Williams 3,4, Kelsey Neal 4, Ashutosh Khandha 2,5, Kurt Manal 6, Lynn Snyder-Mackler 2,5,7, Thomas S Buchanan 2,4,5
PMCID: PMC10922905  NIHMSID: NIHMS1959063  PMID: 38184905

Abstract

Altered medial/lateral knee muscle co-contraction (measure by co-contraction indices, CCI) occurs during gait early after anterior cruciate ligament reconstruction (ACLR). Changes in peak medial compartment forces (pMCF) are also observed early after ACLR and are linked to the development of knee osteoarthritis. We do not know if imbalanced co-contraction is associated with these alterations in knee load. The purpose of this study was to evaluate the association between pMCF and the CCIs of medial/lateral knee muscle pairs during walking three months after ACLR. Bilateral knee gait mechanics and electromyography (EMG) data were collected from 44 participants 3 months following surgery. CCIs of six muscle pairs and medial-to-lateral (M:L) CCIs ratios were calculated during the weight acceptance interval. Bilateral pMCFs were calculated using a subject-based neuromusculoskeletal model. Based on interlimb pMCF symmetry, participants were divided into three groups: symmetric loaders, underloaders, and overloaders. A 2 × 3 (limb × group) ANOVA was used to compare CCIs between limbs in all groups. A partial Spearman’s test was performed to examine the association between CCIs ratios and pMCF. The CCIs of the vastus lateralis-lateral gastrocnemius muscle pair was higher in the involved limb of underloaders (vs. the uninvolved limb and vs. the involved limb of symmetric loaders). The ratio of M:L CCIs was significantly lower (more lateral CCIs) in the involved limb, which was associated with lower pMCF. These results suggest that individuals early after ACLR who walk with higher CCIs of lateral knee musculature (vs. medial), have medial tibiofemoral underloading.

Keywords: ACL reconstruction, Electromyography, Co-contraction indices, Knee loading, Gait

1. Introduction

Anterior cruciate ligament (ACL) tears are prevalent, particularly among young adults (Kaeding et al., 2017). These injuries result in knee instability, leading to reduced levels of activity and lower functional confidence (Mall et al., 2014; Sanders et al., 2016). Thus, many choose to undergo ACL reconstruction (ACLR) to restore functional knee stability and to return to their pre-injury level of activity (Buller et al., 2015; Herzog et al., 2018). Despite these beneficial outcomes, up to 70% of individuals develop radiographic medial tibiofemoral osteoarthritis (OA) within 10-15 years of surgery (Barenius et al., 2014; Cinque et al., 2018; Øiestad et al., 2011).

One mechanism contributing to the development of post-traumatic knee OA after ACLR is alterations in medial tibiofemoral loading (Wellsandt et al., 2016). These alterations are attributed to altered knee mechanical and neuromuscular strategies. Using a validated EMG-based neuromusculoskeletal model (Manal and Buchanan, 2013), our lab found that those who walk with medial tibiofemoral underloading 6 months after ACLR develop radiographic knee OA 5 years post-surgery (Wellsandt et al., 2016). Additionally, asymmetric peak medial compartment forces (pMCFs, involved less than uninvolved) have been observed as early as 3 months after ACLR (Neal et al., 2021). These asymmetrical loading patterns are associated with early, pre-osteoarthritic, alterations in the biochemical composition of knee cartilage (Williams et al., 2021). Thus, there is a critical need to investigate the potential mechanisms leading to altered knee loading in the early months after ACLR.

While altered knee gait mechanics and their influence on knee load following ACLR have been extensively studied (Khandha et al., 2016; Wellsandt et al., 2017, 2016), little is known regarding the association between neuromuscular alterations and knee loading after surgery. Altered co-contraction of the muscles crossing the knee joint (quadriceps, hamstrings, and gastrocnemii) (Alfayyadh et al., 2022; Blackburn et al., 2019; Sherman et al., 2021) are common following ACLR. Elevated lateral knee muscle co-contractions (quadriceps and hamstrings) have been reported in the involved limb (vs. uninvolved) during gait 3 months after surgery (Alfayyadh et al., 2022) and are found to shift more medially with time (Blackburn et al., 2019). These alterations in knee muscle co-contraction indices (CCIs) (Rudolph et al., 2001), may contribute to asymmetric knee loading between limbs during walking following ACLR.

Altered co-contraction of knee musculature after ACLR is associated with surrogate measures of knee load such as frontal plane moments (Lloyd and Buchanan, 2001) and pMCFs (Wellsandt et al., 2017). Few, however, have examined the association between CCIs of medial/lateral knee muscle pairs and measures of knee load. Prior work among healthy individuals suggests that medial and lateral knee musculature CCIs provide resistance against knee abduction and adduction loading, respectively (Palmieri-Smith et al., 2009; Zhang and Wang, 2001). Thus, alterations to the normal balance of medial-to-lateral co-contraction may be related to abnormal loading following ACLR. Acquiring insights into the relationships between CCIs of medial/lateral knee muscle pairs and pMCF early after ACLR may provide valuable knowledge concerning the underlying mechanisms leading to the development of post-traumatic OA.

The purpose of this study was to evaluate the association between the CCIs of medial and lateral knee muscle pairs and peak medial compartment force (pMCF) during gait 3 months after ACLR. We hypothesized that (1) medial compartment underloaders (involved loading less than uninvolved loading) would show greater CCIs of lateral knee muscle pairs compared to symmetric loaders (involved and uninvolved limb loads equivalent) and overloaders (involved loading greater than uninvolved) in the involved limb, (2) the involved limb would show a more lateral-dominated imbalance in medial-to-lateral [M:L] CCI ratios compared to the uninvolved limb, and (3) lower M:L CCIs ratios would be associated with lower pMCFs in the involved limb during gait 3 months after ACLR.

2. Methods

2.1. Participants

Forty-four participants, from a larger longitudinal cohort study, were included in this study following approval by the Institutional Review Board of the University of Delaware (Table 1). Criteria for participant enrollment included: 16 to 45 years old for age, primary unilateral ACLR, no prior lower-limb injuries or surgeries, no concurrent grade III ligament tears, and no repairable meniscal tears. Prior to participating, all participants signed a written informed consent. For individuals under the age of 18, both minor assessment and parental consent were obtained before enrollment.

Table 1.

Demographic characteristics for each loading group (n = 44).

Variables Symmetric Loaders
(n = 15)		 Underloaders
(n = 22)		 Overloaders
(n = 7)		 p
Sex 6 women, 9 men 9 women, 13 men 3 women, 4 men 0.99
Graft type 6 BPTP, 7 hamstring, 9 BPTP, 6 hamstring, 5 BPTP, 1 hamstring, 0.53
2 allograft 6 allograft, 1 hybrid 1 allograft
Age (years) 23 (6) 23 (7) 23 (10) 0.94
Mass (kg) 77.05 (12.76) 75.52 (14.96) 71.34 (13.42) 0.67
Height (m) 1.74 (0.13) 1.72 (0.08) 1.74 (0.09) 0.69
BMI (kg/m2) 25.20 (1.94) 25.54 (4.35) 23.35 (2.71) 0.35
Walking speed (m/s) 1.51 (0.17) 1.59 (0.17) 1.59 (0.13) 0.37
Open in a new tab

Mean (± standard deviation). BMI = body mass index; BPTP = bone-patellar tendon-bone graft; kg = kilogram; m = meter; s = seconds.

2.2. Gait Analysis

All participants underwent gait analysis during overground walking 3 months (3.1 ± 0.5) following ACLR using previously described methodologies (Gardinier et al., 2012; Manal and Buchanan, 2013). Briefly, prior to walking, surface electromyography (EMG) electrodes were placed bilaterally over knee extensors [vastus medialis (VM), vastus lateralis (VL), rectus femoris (RF)] and flexors [semitendinosus (MH), long head of the biceps femoris (LH), medial gastrocnemius (MG), and lateral gastrocnemius (LG)] after shaving and cleaning the regions over each respective muscle belly. All electrode placements followed SENIAM guidelines (Hermens et al., 2000). Following EMG electrode placement, all participants performed bilateral maximal voluntary isometric contractions (MVICs) for each of the assessed muscle groups (quadriceps, hamstrings, and gastrocnemii) (Gardinier et al., 2012). After MVIC testing, each participant was outfitted with retro-reflective markers which were placed bilaterally on bony landmarks of interest (iliac crests of the pelvis, greater trochanters of the hips, femoral condyles, malleoli, and 1st/5th metatarsal heads) (Gardinier et al., 2012). For motion tracking, rigid shell clusters with four markers were placed on the pelvis, thighs, and shanks. Additionally, two separate markers were placed on the top and bottom of each heel (Gardinier et al., 2012).

Participants completed 16 walking trials (8 for each leg) on a 6-meter walkway at self-selected speed which was maintained throughout the experiment (± 5 %). Participants were instructed to walk at a normal pace, which was monitored using two photoelectric beams (Brower Timing Systems). Following five to eight walking trials, the participants’ average walking speed was calculated. Subsequently, they were asked to maintain their speed within a range of ± 5 % of this average. Trials in which participants exceeded this range were repeated. During walking, kinematic data were collected using an 8-camera Vicon system (Oxford Metrics Limited, London, England) with a sampling rate of 120 Hz and kinetic data were collected at a sampling rate of 1080 Hz using a force platform (Bertec Corp., Worthington, OH, USA). Simultaneously, EMG data were recorded at 1080 Hz using a MA-300 EMG system (Motion Lab Systems, Baton Rouge, LA).

Kinematic, kinetic, and EMG data were processed in Visual3D (C-Motion, Germantown, MD) and used as inputs into a validated EMG-based neuromusculoskeletal model to calculate joint contact forces (Buchanan et al., 2004; Manal and Buchanan, 2013). Prior to analysis, Kinematic and kinetic data were low-pass filtered at 6 and 25 Hz, respectively, using fourth-order Butterworth filter. Also, EMG data were high-pass filtered (cutoff 30 Hz), rectified, low-pass filtered (cutoff 6 Hz), and then normalized to MVICs to create linear envelopes. First, a SIMM (SIMM 4.0.2, Musculographics, Inc., Chicago, IL) musculoskeletal model was scaled for each participant based on their anthropometric measurements. This model has four segments (foot, shank, thigh, and pelvis) driven by the action of 10 muscles crossing the knee joint. The EMG signals of the mentioned seven muscles, in addition to the vastus intermedius (estimated as the average of vastus medialis and lateralis), the semimembranosus (set equivalent to the semitendinosus) and the short head of the biceps femoris (set equivalent to long head of the biceps femoris,), were transformed into muscle activations (Buchanan et al., 2004; Manal and Buchanan, 2013). Model parameters involved in muscle force estimation were then tuned so that the internal frontal plane knee moments were in close agreement with external moments computed via inverse dynamics (Buchanan et al., 2004; Manal and Buchanan, 2013). Finally MCFs were calculated using the estimated muscle forces via a previously described frontal plane moment balance approach (Manal and Buchanan, 2013; Winby et al., 2009). From these, we found the pMCF in each limb during stance phase and subsequently normalized these to an individual’s body weight (BW).

Bilateral co-contraction indices (CCIs) for six muscle pairs (VM-MH, VM-MG, VM-LG, VL-LH, VL-LG, and VL-MG) were calculated as the lowest EMG signal divided by the highest EMG signal and multiplied by the sum of these signals (Rudolph et al., 2001).

CCI=LowestEMCvalueHighestEMGvalue×(LowestEMCvalue+HighestEMGvalue)

The CCI for each muscle pair was calculated for each walking trial then averaged across all walking trials for that limb. The average CCIs were calculated during the weight acceptance interval (100 ms before heel strike to maximal knee flexion angle) of stance phase (Rudolph et al., 2001). Additionally, M:L CCIs ratios were calculated by dividing the medial muscle pair index by the lateral muscle pair index for quadriceps, hamstrings, and gastrocnemii muscles, resulting in the following ratios: VM-MH / VL-LH and VM-MG / VL-LG (Palmieri-Smith et al., 2009).

Participants were divided into three groups based on interlimb medial compartment loading symmetry (involved limb – uninvolved limb): symmetric loaders, underloaders, and overloaders. To do this, we used a previously determined meaningful interlimb difference (MILD) for pMCF (0.4 BW) (Khandha et al., 2016). Specifically, those who had a pMCF interlimb difference (ILD) between ± 0.4 BW were considered symmetric loaders, while those with pMCF ILD greater than 0.4 BW or less than - 0.4 BW were considered overloaders and underloaders, respectively.

2.3. Statistical Analyses

SPSS software (version 26.0, SPSS Inc., Chicago, IL) was utilized to perform all statistical analyses. Shapiro-Wilk tests were performed to check data normality, skewness, and kurtosis. Demographic characteristics were compared between groups (symmetric loaders, underloaders, and overloaders) using one-way ANOVAs and chi-squared tests. A 2X3 [Limb x Loading Group] repeated measures ANOVA was performed to examine the effects of limb and loading classification, and their interactions, on muscle CCIs while controlling for sex, walking speed, body mass index (BMI), and age. If significant effects were observed, two-tailed post hoc Wilcoxon test and Kruskal-Wallis test were used to compare CCIs between limbs and between groups, respectively. The effect size of the findings were evaluated using Partial eta squared values (small size = 0.01, medium size = 0.06, and large size = 0.14) (Richardson, 2011), and Cohen’s D values (small size = 0.2, medium size = 0.5, and large size = 0.8) (Sawilowsky, 2009). M:L CCIs ratios were compared between limbs using a Wilcoxon test. The associations between the M:L CCIs ratios and pMCFs were evaluated using a two-tailed partial Spearman’s correlation, while using gender, walking speed, BMI, and age as covariates. Alpha was set to 0.05 for all analyses. In this study, a total of 24 statistical tests were performed to evaluate the main effects and interactions across six CCIs between loading groups for both limbs (18 tests), to compare the CCIs ratios between limbs (2 tests), and to examine the association between these ratios and medial knee loading (4 tests). No correction for multiple comparisons was applied as the analyses were exploratory in nature, aiming to generate hypotheses rather than to test them definitively (Althouse, 2016).

3. Results

There were no significant differences in demographic characteristics between the loading groups (symmetric loaders, underloaders, and overloaders; Table 1).

The ANOVA analysis for the VL-LG muscle pair displayed a significant limb × loading group interaction (p = 0.02) with large effect size (ηp2 = 0.17; Table 2). For VM-LG muscle pair, the interaction between limb and loading group approached significance (p = 0.07) with medium effect size (ηp2 = 0.13; Table 2). No other interactions or main effects were present for the other muscle pairs (Table 2).

Table 2.

Knee muscle co-contraction indices of the involved and uninvolved limbs during gait 3 months after ACLR.

Variables Symmetric Loaders Underloaders Overloaders Limb Loading Interaction
IN UN IN UN IN UN p ηp 2 p ηp 2 p ηρ 2
VM-MH 0.13
(0.11)		 0.13
(0.08)		 0.12
(0.07)		 0.14
(0.09)		 0.12
(0.04)		 0.11
(0.05)		 0.52 0.01 0.77 0.01 0.55 0.03
VM-MG 0.07
(0.04)		 0.07
(0.05)		 0.07
(0.03)		 0.07
(0.04)		 0.08
(0.05)		 0.07
(0.06)		 0.78 <0.01 0.83 0.01 0.94 <0.01
VM-LG 0.08
(0.04)		 0.11
(0.04)		 0.12
(0.07)		 0.09
(0.06)		 0.13
(0.06)		 0.09
(0.06)		 0.09 0.07 0.49 0.04 0.07 0.13
VL-LH 0.17
(0.16)		 0.12
(0.05)		 0.19
(0.09)		 0.14
(0.06)		 0.19
(0.11)		 0.13
(0.07)		 0.82 <0.01 0.99 <0.01 0.96 <0.01
VL-LG 0.09
(0.05)		 0.11
(0.04)		 0.15
(0.09)		 0.09
(0.05)		 0.14
(0.06)		 0.10
(0.06)		 0.55 0.01 0.41 0.05 0.03 0.17
VL-MG 0.08
(0.04)		 0.07
(0.05)		 0.08
(0.05)		 0.07
(0.04)		 0.10
(0.09)		 0.07
(0.08)		 0.45 0.02 0.62 0.03 0.77 0.01
Open in a new tab

Mean (± standard deviation). Statistically significance (p < 0.05) and large effect size ηp2> 0.14 are indicated by bold values.

IN = involved limb; UN = uninvolved limb; VM-MH = vastus medialis/semitendinosus; VM-MG = vastus medialis/medial gastrocnemius; VM-LG = vastus medialis/lateral gastrocnemius; VL-LH = vastus lateralis/the long head of biceps femoris; VL-LG = vastus lateralis/lateral gastrocnemius; VL-MG = vastus lateralis/medial gastrocnemius.

Post hoc investigations revealed that underloaders walked with significantly higher CCIs in the VL-LG muscle pair of the involved limb (vs. uninvolved) (p < 0.01, d = 0.89). Additionally, underloaders displayed significantly higher involved limb CCIs in the VL-LG muscle pair compared to symmetric loaders (p = 0.02, ηp2 = 0.18).

The M:L CCIs ratio of the VM-MH / VL-LH was significantly lower in the involved limb compared to the uninvolved limb (p < 0.01, d = 0.98; Figure 1) indicating more co-contraction in the lateral muscle pair than medial within the involved limb. There was a significant positive association between the M:L CCIs ratio of VM-MH / VL-LH and pMCF of the involved limb (p = 0.02, r = 0.40; Figure 2), indicating that more laterally dominated co-contraction (vs. medial) was associated with lower medial compartment loading. No other associations were present between the ratio of CCIs and pMCF (Table 3).

Figure 1.

Figure 1.

Open in a new tab

Ratio of medial-to-lateral knee muscle pairs co-contraction indices for participants 3 months after ACLR during walking (Mean [± standard deviation]). The ratio of medial-to-lateral quadriceps/hamstrings CCIs was lower in the involved limb than the uninvolved limb. VM-MH = vastus medialis/semitendinosus; VL-LH = vastus lateralis/long head of the biceps femoris; VM-MG = vastus medialis/medial gastrocnemius; VL-LG = vastus lateralis/lateral gastrocnemius. ✦ significant difference (p < 0.05).

Figure 2.

Figure 2.

Open in a new tab

A positive association between the ratio of medial-to-lateral quadriceps/hamstrings CCIs and pMCF in the involved limb. VM-MH = vastus medialis/semitendinosus; VL-LH = vastus lateralis/long head of the biceps femoris; pMCF = peak medial compartment force; BW = body weight

Table 3.

pMCF versus the ratio of Medial-to-lateral CCIs of knee muscle pairs.

Limb Knee loading Ratio of Medial-to-Lateral CCIs r p
IN limb pMCF VM-MH / VL-LH 0.36 0.02
VM-MG / VL-LG 0.17 0.28
UN limb pMCF VM-MH / VL-LH 0.18 0.28
VM-MG / VL-LG 0.08 0.65
Open in a new tab

Statistically significance association is indicated by bold values (p < 0.05).

IN = involved limb; UN = uninvolved limb; CCIs = co-contraction indices.

4. Discussion

To our knowledge, this is the first study to investigate the association between medial compartment force and the co-contraction of medial and lateral knee muscle pairs following ACLR. Our findings partially supported our hypotheses. We found that that underloaders walked with greater lateral CCIs in the involved limb compared to symmetric loaders for the VL-LG muscle pair. Additionally, we found that individuals walk with more lateral-dominant CCIs (vs. medial) in the involved limb following ACLR, and that this imbalance is associated with lower medial compartment forces. These neuromuscular strategies may contribute to altered knee loading early after ACLR, which, in turn, may negatively impact long-term knee cartilage health.

Regarding our first hypothesis, only the CCIs of VL-LG muscle pair exhibited significant differences between loading groups. Specifically, underloaders walked with higher VL-LG CCIs in the involved limb (vs. the uninvolved limb and vs. the involved limb of symmetric loaders). This suggests that individuals with lower pMCF in the involved limb (vs. uninvolved) early after ACLR may be more likely to walk with elevated co-activation of lateral quadriceps and gastrocnemius. This contradicts the work of Blackburn et al. (2019) which reported higher CCIs of medial knee muscle pairs in the involved limb (vs. healthy control), rather than lateral knee muscle pairs (Blackburn et al., 2019). That study, however, did not include the activity of gastrocnemii in CCIs analysis despite their impact of tibiofemoral loading (Manal and Buchanan, 2013). Additionally, their participants were from a later and broader time range following ACLR (6 months to 5 years after surgery). This broader time range may have obscured important time-dependent alterations. Future studies need to investigate these indices across multiple later time points, such as 6 and 24 months after surgery, and examine how these indices evolve with time.

A recent study found that individuals walking with medial tibiofemoral underloading in the involved limb 6 months after ACLR develop radiographic knee OA 5 years following surgery (Wellsandt et al., 2016). Those individuals showed an inverse association between medial compartment loading and the CCIs of all knee muscle pairs (quadriceps, hamstrings, and gastrocnemii together) in both limbs (Wellsandt et al., 2017). This association, however, could be representative of a normal behavior as it was seen in both limbs. Regarding medial/lateral CCIs of knee muscles, however, our previous work reported that individuals walk with higher CCIs of knee lateral musculature in the involved limb (vs. uninvolved) 3 months after ACLR (Alfayyadh et al., 2022). Similarly, several OA studies demonstrated that OA patients walk with elevated CCIs in lateral knee muscle pairs during the stance phase, and the co-contraction of lateral muscle pairs increases while OA severity increases (Heiden et al., 2009; Hubley-Kozey et al., 2009; Mills et al., 2013). This may suggest that individuals early after ACLR walk with increased lateral CCIs due to pain and/or neuromuscular deficits such as impaired proprioception and excitability of the nervous system, while OA patients use the same technique as protective strategy to reduce compression and pain on the medial compartment. Future work is needed to examine the potential impact of elevated lateral CCIs on knee cartilage early after ACLR, and whether these indices can be altered by rehabilitation interventions after surgery.

Our second and third hypotheses were also partially supported. We found a significantly lower M:L CCI ratio in the involved limb (vs. uninvolved) for the VL-LH / VM-MH pairs. We also saw a trend towards a lower M:L CCI ratio for the quadriceps and gastrocnemii pairs within the involved limb, but statistical significance was not achieved. This suggests that individuals tend to co-contract their involved limb’s knee muscle pairs with a lateral-leaning imbalance during walking early after ACLR. Although it is not fully understood how these neuromuscular patterns develop and how these can impact knee cartilage, it is possible that they could cause alterations in tibiofemoral loading. A recent study showed that healthy people walk with balanced co-contraction between medial and lateral knee muscle pairs (Khandha et al., 2018). Thus, walking with unbalanced M:L CCIs of knee musculature following ACLR might induce aberrant tibiofemoral loading, which may eventually lead to detrimental effects to cartilage.

When examining the influence of unbalanced M:L CCIs on medial knee load, we found a positive association between the M:L CCIs ratio of quadriceps/hamstrings and pMCF in the involved limb. This suggests that walking with lower CCIs ratios (lateral CCIs > medial CCIs) is associated with lower medial tibiofemoral loading in the involved limb 3 months following ACLR. This is not necessarily surprising as medial and lateral knee muscle pairs have varus/valgus moment arms that work oppositely in terms of frontal knee loading resistance (Palmieri-Smith et al., 2009; Zhang and Wang, 2001). Thus, altered medial-to-lateral co-contraction patterns may alter load magnitude and distribution, which may ultimately be harmful to cartilage (Andriacchi et al., 2015, 2009; Kaiser et al., 2016; Palmieri-Smith et al., 2009; Wellsandt et al., 2016; Zhang and Wang, 2001). If load is being shifted, the newly loaded region of cartilage may not have the capacity to withstand the applied loading, which may result in the breakdown of cartilage components (Andriacchi et al., 2015, 2009). If load remains in the same location but changes magnitude, it could be that, given cartilage’s mechanosensitive nature, the cartilage may adapt to the new amount of loading (Arokoski et al., 2000; Van Rossom et al., 2017). This, in turn, could be problematic when knee loading returns to its pre-ACLR level, as it’s possible that the cartilage may be incapable of withstanding such loading, which might cause degenerative changes in cartilage (Vincent and Wann, 2019). Future research should examine whether or not unbalanced CCIs of medial and lateral knee muscle pairs early after ACLR contribute to long-term cartilage degradation.

Estimating tibiofemoral loading using neuromusculoskeletal modeling or via surrogate measures (i.e. external knee adduction moment (Wellsandt et al., 2017)) is difficult and unfeasible for a clinical environment. Given that the magnitude of muscle co-contraction can be measured easily using an EMG device and can be modified by implementing neuromuscular training protocols (Nawasreh et al., 2019), the findings of the current study might be important for clinicians to consider early after ACLR. Future work should explore novel methods for EMG-based biofeedback and gait training aimed at altering co-contraction patterns post-ACL reconstruction and investigate how this alters knee compartmental loads, and ultimately, long-term cartilage health.

Several limitations should be considered in this study. First, the sample sizes of the three groups (symmetric loaders, overloaders, and underloaders) were relatively small which might limit the generalization of our findings to the entire ACLR population. Secondly, we had limited knowledge about the rehabilitation protocols applied to our participants, so discrepancies in neuromuscular strategies may exist because of differing strategies. Additionally, the use of surface EMG electrodes could potentially affect the magnitudes of EMG signals due to skin artifacts (motion). Each electrode was secured to the participant using both tape and wraps to minimize these potential motion artifacts. We designated this study as exploratory in nature due to the novel examination of specific co-contraction indices of medial and lateral knee muscle pairs separately at a defined post-operative time point, a focus not commonly addressed in the literature. The absence of a correction for multiple comparisons aligns with the exploratory aim to identify potential patterns and generate hypotheses in a developing research area, which may limit the statistical conclusiveness of the findings. Finally, as the data of the current study is part of a larger cohort study, we did not conduct a power analysis for the specific questions asked in this study as doing so is considered an inappropriate practice (Hoenig and Heisey, 2001). Thus, there is a chance that other variables may have been deemed not significant due to lack of statistical power rather than true statistical non-significance.

In summary, individuals early after ACLR exhibited higher CCIs of lateral knee musculature during walking. This imbalance between the CCIs of medial and lateral knee muscle pairs in the involved limb was associated with medial compartment underloading of the tibiofemoral joint. These gait neuromuscular strategies early after surgery may contribute to aberrant knee loading in this population and may be a mechanism for eventual post-traumatic OA development.

Acknowledgements

Funding was provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development: R01-HD087459. During the course of this project, Abdulmajeed Alfayyadh was funded by Jouf University through the Saudi Arabian Cultural Mission (SACM), Jack R. Williams was funded by the University of Delaware’s Mechanical Engineering Helwig Fellowship and the Delaware Space Grant College and Fellowship Program (NASA 80NSSC20M0045), and Kelsey Neal was funded by the University of Delaware’s Dissertation Fellowship and the University of Delaware’s Mechanical Engineering Helwig Fellowship. The authors thank Martha Callahan, Jennifer Marmon, and Delaware Rehabilitation Institute Research Core for assistance with subject recruitment and retention.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest

The authors have no conflict of interest with the material to be published in the paper.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Declarations of Interest: None

References

  1. Alfayyadh A, Neal K, Williams JR, Khandha A, Manal K, Snyder-Mackler L, Buchanan TS, 2022. Limb and sex-related differences in knee muscle co-contraction exist 3 months after anterior cruciate ligament reconstruction. J. Electromyogr. Kinesiol. TA - TT - 66. 10.1016/j.jelekin.2022.102693 LK [DOI] [PubMed] [Google Scholar]
  2. Althouse AD, 2016. Adjust for Multiple Comparisons? It’s Not That Simple. Ann. Thorac. Surg 101, 1644–1645. 10.1016/j.athoracsur.2015.11.024 [DOI] [PubMed] [Google Scholar]
  3. Andriacchi TP, Favre J, Erhart-Hledik JC, Chu CR, 2015. A systems view of risk factors for knee osteoarthritis reveals insights into the pathogenesis of the disease. Ann. Biomed. Eng. TA - TT - 43, 376–387. 10.1007/s10439-014-1117-2 LK [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Andriacchi TP, Koo S, Scanlan SF, 2009. Gait mechanics influence healthy cartilage morphology and osteoarthritis of the knee. J. bone Jt. surgery. Am. Vol. TA - TT - 91 Suppl 1, 95–101. 10.2106/JBJS.H.01408 LK [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Arokoski JP, Jurvelin JS, Väätäinen U, Helminen HJ, 2000. Normal and pathological adaptations of articular cartilage to joint loading. Scand. J. Med. Sci. Sport. TA - TT - 10, 186–198. [DOI] [PubMed] [Google Scholar]
  6. Barenius B, Ponzer S, Shalabi A, Bujak R, Norlén L, Eriksson K, 2014. Increased risk of osteoarthritis after anterior cruciate ligament reconstruction: a 14-year follow-up study of a randomized controlled trial. Am. J. Sport. Med. TA - TT - 42, 1049–1057. 10.1177/0363546514526139 LK [DOI] [PubMed] [Google Scholar]
  7. Blackburn T, Pietrosimone B, Goodwin JS, Johnston C, Spang JT, 2019. Co-activation during gait following anterior cruciate ligament reconstruction. Clin. Biomech 67, 153–159. 10.1016/j.clinbiomech.2019.05.010 [DOI] [PubMed] [Google Scholar]
  8. Buchanan TS, Lloyd DG, Manal K, Besier TF, 2004. Neuromusculoskeletal Modeling: Estimation of Muscle Forces and Joint Moments and Movements From Measurements of Neural Command. J. Appl. Biomech 20, 367–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Buller LT, Best MJ, Baraga MG, Kaplan LD, 2015. Trends in Anterior Cruciate Ligament Reconstruction in the United States. Orthop. J. Sport. Med 3, 2325967114563664. 10.1177/2325967114563664 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cinque ME, Dornan GJ, Chahla J, Moatshe G, LaPrade RF, 2018. High Rates of Osteoarthritis Develop After Anterior Cruciate Ligament Surgery: An Analysis of 4108 Patients. Am. J. Sports Med 46, 2011–2019. 10.1177/0363546517730072 [DOI] [PubMed] [Google Scholar]
  11. Gardinier ES, Manal K, Buchanan TS, Snyder-Mackler L, 2012. Gait and neuromuscular asymmetries after acute anterior cruciate ligament rupture. Med. Sci. Sports Exerc 44, 1490–1496. 10.1249/MSS.0b013e31824d2783 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Heiden TL, Lloyd DG, Ackland TR, 2009. Knee joint kinematics, kinetics and muscle co-contraction in knee osteoarthritis patient gait. Clin. Biomech. TA - TT - 24, 833–841. 10.1016/j.clinbiomech.2009.08.005 LK - https://delcat.on.worldcat.org/oclc/4929538666 [DOI] [PubMed] [Google Scholar]
  13. Hermens HJ, Freriks B, Disselhorst-klug C, Rau C, 2000. Development of recommendations for SEMG sensors and sensor placement procedures. J. Electromyogr. Kinesiol 10, 361–374. 10.1007/s10750-015-2551-3 [DOI] [PubMed] [Google Scholar]
  14. Herzog MM, Marshall SW, Lund JL, Pate V, Mack CD, Spang JT, 2018. Trends in Incidence of ACL Reconstruction and Concomitant Procedures Among Commercially Insured Individuals in the United States, 2002-2014. Sports Health 10, 523–531. 10.1177/1941738118803616 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hoenig JM, Heisey DM, 2001. The abuse of power: The pervasive fallacy of power calculations for data analysis. Am. Stat 55, 19–24. 10.1198/000313001300339897 [DOI] [Google Scholar]
  16. Hubley-Kozey CL, Hill NA, Rutherford DJ, Dunbar MJ, Stanish WD, 2009. Co-activation differences in lower limb muscles between asymptomatic controls and those with varying degrees of knee osteoarthritis during walking. Clin. Biomech. (Bristol, Avon) TA - TT - 24, 407–414. 10.1016/j.clinbiomech.2009.02.005 LK [DOI] [PubMed] [Google Scholar]
  17. Kaeding CC, Léger-St-Jean B, Magnussen RA, 2017. Epidemiology and Diagnosis of Anterior Cruciate Ligament Injuries. Clin. Sport. Med. TA - TT - 36, 1–8. 10.1016/j.csm.2016.08.001 LK [DOI] [PubMed] [Google Scholar]
  18. Kaiser J, Vignos MF, Liu F, Kijowski R, Thelen DG, 2016. American Society of Biomechanics Clinical Biomechanics Award 2015: MRI assessments of cartilage mechanics, morphology and composition following reconstruction of the anterior cruciate ligament. Clin. Biomech. (Bristol, Avon) TA - TT - 34, 38–44. 10.1016/j.clinbiomech.2016.03.007 LK [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Khandha A, Manal K, Capin J, Wellsandt E, Marmon A, Snyder-Mackler L, Buchanan TS, 2018. High muscle co-contraction does not result in high joint forces during gait in anterior cruciate ligament deficient knees. J. Orthop. Res 37, 104–112. 10.1002/jor.24141 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Khandha A, Manal K, Wellsandt E, Capin J, Snyder-Mackler L, Buchanan TS, 2016. Gait mechanics in those with/without medial compartment knee osteoarthritis 5 years after anterior cruciate ligament reconstruction. J. Orthop. Res 35, 625–633. 10.1002/jor.23261 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lloyd DG, Buchanan TS, 2001. Strategies of muscular support of varus and valgus isometric loads at the human knee. J. Biomech. TA - TT - 34, 1257–1267. 10.1016/S0021-9290(01)00095-1 LK [DOI] [PubMed] [Google Scholar]
  22. Mall NA, Chalmers PN, Moric M, Tanaka MJ, Cole BJ, Bach BR, Paletta GA, 2014. Incidence and trends of anterior cruciate ligament reconstruction in the United States. Am. J. Sports Med 42, 2363–2370. 10.1177/0363546514542796 [DOI] [PubMed] [Google Scholar]
  23. Manal K, Buchanan TS, 2013. An Electromyogram-Driven Musculoskeletal Model of the Knee to Predict in Vivo Joint Contact Forces During Normal and Novel Gait Patterns. J. Biomech. Eng 135. 10.1115/1.4023457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mills K, Hunt MA, Leigh R, Ferber R, 2013. A systematic review and meta-analysis of lower limb neuromuscular alterations associated with knee osteoarthritis during level walking. Clin. Biomech 28, 713–724. 10.1016/j.clinbiomech.2013.07.008 [DOI] [PubMed] [Google Scholar]
  25. Nawasreh ZH, Marmon AR, Logerstedt D, Snyder-Mackler L, 2019. The effect of training on a compliant surface on muscle activation and co-contraction after anterior cruciate ligament injury. Int. J. Sport. Phys. Ther. TA - TT - 14, 3554–3563. [PMC free article] [PubMed] [Google Scholar]
  26. Neal K, Williams JR, Alfayyadh A, Capin JJ, Khandha A, Manal K, Snyder-Mackler L, Buchanan TS, 2021. Knee joint biomechanics during gait improve from 3 to 6 months after anterior cruciate ligament reconstruction. J. Orthop. Res. TA - TT -. 10.1002/jor.25250 LK [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Øiestad BE, Holm I, Engebretsen L, Risberg MA, 2011. The association between radiographic knee osteoarthritis and knee symptoms, function and quality of life 10-15 years after anterior cruciate ligament reconstruction. Br. J. Sports Med 45, 583–588. 10.1136/bjsm.2010.073130 [DOI] [PubMed] [Google Scholar]
  28. Palmieri-Smith RM, McLean SG, Ashton-Miller JA, Wojtys EM, 2009. Association of quadriceps and hamstrings cocontraction patterns with knee joint loading. J. Athl. Train 44, 256–263. 10.4085/1062-6050-44.3.256 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Richardson JTE, 2011. Eta squared and partial eta squared as measures of effect size in educational research. Educ. Res. Rev 6, 135–147. 10.1016/j.edurev.2010.12.001 [DOI] [Google Scholar]
  30. Rudolph KS, Axe MJ, Buchanan TS, Scholz JP, Snyder-Mackler L, 2001. Dynamic stability in the anterior cruciate ligament deficient knee. Knee Surgery, Sport. Traumatol. Arthrosc 9, 62–71. 10.1007/s001670000166 [DOI] [PubMed] [Google Scholar]
  31. Sanders TL, Kremers HM, Bryan AJ, Larson DR, Dahm DL, Levy BA, Stuart MJ, Krych AJ, 2016. Incidence of Anterior Cruciate Ligament Tears and Reconstruction: A 21-Year Population-Based Study. Am. J. Sport. Med. TA - TT - 44, 1502–1507. 10.1177/0363546516629944 LK [DOI] [PubMed] [Google Scholar]
  32. Sawilowsky SS, 2009. New effect size rules of thumb. J. Mod. Appl. Stat. Methods 8, 597–599. 10.22237/jmasm/1257035100 [DOI] [Google Scholar]
  33. Sherman DA, Glaviano NR, Norte GE, 2021. Hamstrings Neuromuscular Function After Anterior Cruciate Ligament Reconstruction: A Systematic Review and Meta-Analysis. Sports Med. 10.1007/s40279-021-01433-w [DOI] [PubMed] [Google Scholar]
  34. Van Rossom S, Smith CR, Zevenbergen L, Thelen DG, Vanwanseele B, Van Assche D, Jonkers I 2017. Knee Cartilage Thickness, T1ρ and T2 Relaxation Time Are Related to Articular Cartilage Loading in Healthy Adults. PloS one TA - TT - 12, e0170002. 10.1371/journal.pone.0170002 LK [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Vincent TL, Wann AKT, 2019. Mechanoadaptation: articular cartilage through thick and thin. J. Physiol 597, 1271–1281. 10.1113/JP275451 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wellsandt E, Gardinier ES, Manal K, Axe MJ, Buchanan TS, Snyder-Mackler L, 2016. Decreased Knee Joint Loading Associated with Early Knee Osteoarthritis after Anterior Cruciate Ligament Injury. Am. J. Sports Med 44, 143–151. 10.1177/0363546515608475 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wellsandt E, Khandha A, Manal K, Axe MJ, Buchanan TS, Snyder-Mackler L, 2017. Predictors of knee joint loading after anterior cruciate ligament reconstruction. J. Orthop. Res 35, 651–656. 10.1002/jor.23408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Williams JR, Neal K, Alfayyadh A, Lennon K, Capin JJ, Khandha A, Manal K, Potter HG, Snyder-Mackler L, Buchanan TS, 2021. Knee cartilage T(2) relaxation times 3 months after ACL reconstruction are associated with knee gait variables linked to knee osteoarthritis. J. Orthop. Res 10.1002/jor.25043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Winby CR, Lloyd DG, Besier TF, Kirk TB, 2009. Muscle and external load contribution to knee joint contact loads during normal gait. J. Biomech 42, 2294–2300. 10.1016/j.jbiomech.2009.06.019 [DOI] [PubMed] [Google Scholar]
  40. Zhang L-Q, Wang G, 2001. Dynamic and static control of the human knee joint in abduction-adduction. J. Biomech 34, 1107–1115. 10.1016/s0021-9290(01)00080-x [DOI] [PubMed] [Google Scholar]

RESOURCES