Using Machine Learning to Predict Lower-Extremity Injury in US Special Forces

Machine learning approaches to injury risk prediction in sport: a scoping review with evidence synthesis

OBJECTIVE

This study reviewed the current state of machine learning (ML) research for the prediction of sports-related injuries. It aimed to chart the various approaches used and assess their efficacy, considering factors such as data heterogeneity, model specificity and contextual factors when developing predictive models.

DESIGN

Scoping review.

DATA SOURCES

PubMed, EMBASE, SportDiscus and IEEEXplore.

RESULTS

In total, 1241 studies were identified, 58 full texts were screened, and 38 relevant studies were reviewed and charted. Football (soccer) was the most commonly investigated sport. Area under the curve (AUC) was the most common means of model evaluation; it was reported in 71% of studies. In 60% of studies, tree-based solutions provided the highest statistical predictive performance. Random Forest and Extreme Gradient Boosting (XGBoost) were found to provide the highest performance for injury risk prediction. Logistic regression outperformed ML methods in 4 out of 12 studies. Three studies reported model performance of AUC>0.9, yet the clinical relevance is questionable.

CONCLUSIONS

A variety of different ML models have been applied to the prediction of sports-related injuries. While several studies report strong predictive performance, their clinical utility can be limited, with wide prediction windows or broad definitions of injury. The efficacy of ML is hampered by small datasets and numerous methodological heterogeneities (cohort sizes, definition of injury and dependent variables), which were common across the reviewed studies.

Lower Extremity Musculoskeletal Injuries in United States Active Duty Service Members: Prevalence/Incidence, Health Care Utilization, and Cost Analysis Spanning Fiscal Years 2016-2021

INTRODUCTION

Lower Extremity musculoskeletal injuries (LE MSKIs) represent a significant portion of overuse injuries in active duty service members (ADSMs). However, variations in study methods and research gaps related to LE MSKIs have prevented Department of Defense (DoD) leaders from assessing the burden of these conditions. The purpose of this study was to report the incidence, prevalence, and types of LE MSKIs sustained by ADSMs across four branches of service and describe associated health care utilization and private sector costs.

MATERIALS AND METHODS

This was a retrospective, longitudinal population study, including ADSMs from the Air Force, Army, Marine Corps, and Navy. Prevalence and incidence rates for LE MSKIs, associated health care utilization, and costs were obtained by querying electronic health records (EHR) from military treatment facilities (MTFs), private sector care (PC) claims, and theater medical data from the Military Health System Data Repository (MDR) from October 1, 2015 to September 30, 2021 (FY16-21). Utilization associated with LE MSKIs in both the direct care (DC) and PC settings was classified into mutually exclusive outpatient encounter categories and acute inpatient stays. PC costs related to LE MSKIs were captured for each year.

RESULTS

In FY21, LE MSKIs occurred in 25.3% of ADSMs (n = 378,615). A higher proportion of females sustained an LE MSKI (33.3%), compared to males (23.7%). From FY16-21, the Army had the highest annual prevalence of LE MSKIs (30.9-35.5%), followed by the Air Force (23.8-31.0%), Marine Corps (23.4-27.0%), and Navy (17.2-19.8%). Incidence rate patterns were similar, with the Army sustaining LE MSKIs at 320 to 377 injuries per 1,000 person-years, followed by the Air Force (241-318), Marines (255-288), and Navy (173-203). Overuse/non-specific MSKIs of the knee were the most common injury type and body region affected, respectively. There were 10,675,543 DC and 1,875,307 PC outpatient encounters from FY16-21 with a primary or secondary diagnosis of LE MSKI. The Air Force was most reliant on PC, with 21.5 to 36.8% of LE MSKI-related encounters occurring outside MTFs during FY16-21. Over $99 million was paid by TRICARE on LE MSKI in FY21 alone with Same Day Surgeries accounting for almost half ($48 million) of this total.

CONCLUSIONS

Among U.S. ADSMs, LE MSKIs remain highly prevalent and costly. We observed disparities across the Services in the prevalence and incidence of LE MSKIs, and their respective reliance on the private sector for management of these conditions. Findings from this work may support military leaders and MSKI researchers who seek to reduce the impact of LE MSKIs on the readiness and overall health of the U.S. Military.

A stacked autoencoder-based aid system for severity degree classification of knee ligament rupture

BACKGROUND

Knee ligament rupture is one of the most common injuries, but the diagnosis of its severity tends to require the use of complex methods and analyses that are not always available to patients.

AIM

The objective of this research is the investigation and development of a diagnostic aid system to analyze and determine patterns that characterize the presence of the injury and its degree of severity.

METHODS

Implement a novel proposal of a framework based on stacked auto-encoder (SAE) for ground reaction force (GRF) signals analysis, coming from the GaitRec database. Analysis of the raw data is used to determine the main features that allow us to diagnose the presence of a knee ligament rupture and classify its severity as high, mid or mild.

RESULTS

The process is divided into two stages to determine the presence of the lesion and, if necessary, evaluate variations in features to classify the degree of severity as high, mid, and mild. The framework presents an accuracy of 87 % and a F1-Score of 90 % for detecting ligament rupture and an accuracy of 86.5 % and a F1-Score of 87 % for classifying severity.

CONCLUSION

This new methodology aims to demonstrate the potential of SAE in physiotherapy applications as an evaluation and diagnostic tool, identifying irregularities associated with ligament rupture and its degree of severity, thus providing updated information to the specialist during the rehabilitation process.

Interaction of Medical Conditions and Football Exposures Associated with Premortem Chronic Traumatic Encephalopathy Diagnosis in Former Professional American Football Players

BACKGROUND AND OBJECTIVE

Despite being a postmortem diagnosis, former professional American-style football players report receiving chronic traumatic encephalopathy (CTE) diagnoses from medical care providers. However, many players also report other health conditions that manifest with cognitive and psychological symptoms. The purpose of this study was to identify how medical conditions, psychological disorders, and football exposure combinations are associated with former athletes reporting a premortem CTE diagnosis.

METHODS

This study was a cross-sectional cohort survey from 2015 to 2019 of 4033 former professional American-style football players. Demographics (age, race, domestic status, primary care recipient), football-related factors (position, years of professional play, burden of symptoms following head impacts, performance-enhancing drug use), and comorbidities (sleep apnea, psychological disorder status [depression and anxiety; either depression or anxiety; neither depression nor anxiety], diabetes mellitus, attention-deficit/hyperactivity disorder, hypertension, heart conditions, high cholesterol, stroke, cancer, low testosterone, chronic pain, current and maximum body mass index) were recorded. A Chi-square automatic interaction detection (CHAID) decision tree model identified interactive effects between demographics, health conditions, and football exposures on the CTE diagnosis.

RESULTS

Depression showed the strongest univariate association with premortem CTE diagnoses (odds ratio [OR] = 9.5, 95% confidence interval [CI] 6.0-15.3). CHAID differentiated participants with premortem CTE diagnoses with 98.2% accuracy and area under the curve = 0.81. Participants reporting both depression and anxiety were more likely to have a CTE diagnosis compared with participants who reported no psychological disorders (OR = 12.2; 95% CI 7.3-21.1) or one psychological disorder (OR = 4.5; 95% CI 1.9-13.0). Sleep apnea was also associated with a CTE diagnosis amongst those with both depression and anxiety (OR = 2.7; 95% CI 1.4-5.2).

CONCLUSIONS

Clinical phenotypes including psychological disorders and sleep apnea were strongly associated with an increased likelihood of having received a pre-mortem CTE diagnosis in former professional football players. Depression, anxiety, and sleep apnea produce cognitive symptoms, are treatable conditions, and should be distinguished from neurodegenerative disease.

An Overview of Machine Learning Applications in Sports Injury Prediction

Use injuries, i.e., injuries caused by repetitive strain on the body, represent a serious problem in athletics that has traditionally relied on historic datasets and human experience for prevention. Existing methodologies have been frustratingly slow at developing higher precision prevention practices. Technological advancements have permitted the emergence of artificial intelligence and machine learning (ML) as promising toolsets to enhance both injury mitigation and rehabilitation protocols. This article provides a comprehensive overview of recent advances in ML techniques as they have been applied to sports injury prediction and prevention. A comprehensive literature review was conducted searching PubMed/Medline, Institute of Electrical and Electronics Engineers (IEEE)/Institute of Engineering and Technology (IET), and ScienceDirect. Ovid Discovery and Google Scholar were used to provide additional aggregate results and a grey literature search. A focus was placed on papers published from 2017 to 2022. Algorithms of interest were limited to K-Nearest Neighbor (KNN), K-means, decision tree, random forest, gradient boosting and AdaBoost, and neural networks. A total of 42 original research papers were included, and their results were summarized. We conclude that given the current lack of open source, uniform data sets, as well as a reliance on dated regression models, no strong conclusions about the real-world efficacy of ML as it applies to sports injury prediction can be made. However, it is suggested that addressing these two issues will allow powerful, novel ML architectures to be deployed, thus rapidly advancing the state of this field, and providing validated clinical tools.

Employing body-fixed sensors and machine learning to predict physical activity in military personnel

INTRODUCTION

This was a feasibility pilot study aiming to develop and validate an activity recognition system based on a custom-made body-fixed sensor and driven by an algorithm for recognising basic kinetic movements in military personnel. The findings of this study are deemed essential in informing our development process and contributing to our ultimate aim which is to develop a low-cost and easy-to-use body-fixed sensor for military applications.

METHODS

Fifty military participants performed a series of trials involving walking, running and jumping under laboratory conditions in order to determine the optimal, among five machine learning (ML), classifiers. Thereafter, the accuracy of the classifier was tested towards the prediction of these movements (15 183 measurements) and in relation to participants' gender and fitness level.

RESULTS

Random forest classifier showed the highest training and validation accuracy (98.5% and 92.9%, respectively) and classified participants with differences in type of activity, gender and fitness level with an accuracy level of 83.6%, 70.0% and 62.2%, respectively.

CONCLUSIONS

The study showed that accurate prediction of various dynamic activities can be achieved with high sensitivity using a low-cost easy-to-use sensor and a specific ML model. While this technique is in a development stage, our findings demonstrate that our body-fixed sensor prototype alongside a fully trained validated algorithm can strategically support military operations and offer valuable information to commanders controlling operations remotely. Further stages of our developments include the validation of our refined technique on a larger range of military activities and groups by combining activity data with physiological variables to predict phenomena relating to the onset of fatigue and performance decline.

Development and Validation of a Dynamically Updated Prediction Model for Attrition From Marine Recruit Training

ABSTRACT

Dijksma, I, Hof, MHP, Lucas, C, and Stuiver, MM. Development and validation of a dynamically updated prediction model for attrition from Marine recruit training. J Strength Cond Res 36(9): 2523-2529, 2022-Whether fresh Marine recruits thrive and complete military training programs, or fail to complete, is dependent on numerous interwoven variables. This study aimed to derive a prediction model for dynamically updated estimation of conditional dropout probabilities for Marine recruit training. We undertook a landmarking analysis in a Cox proportional hazard model using longitudinal data from 744 recruits from existing databases of the Marine Training Center in the Netherlands. The model provides personalized estimates of dropout from Marine recruit training given a recruit's baseline characteristics and time-varying mental and physical health status, using 21 predictors. We defined nonoverlapping landmarks at each week and developed a supermodel by stacking the landmark data sets. The final supermodel contained all but one a priori selected baseline variables and time-varying health status to predict the hazard of attrition from Marine recruit training for each landmark as comprehensive as possible. The discriminative ability (c-index) of the prediction model was 0.78, 0.75, and 0.73 in week one, week 4 and week 12, respectively. We used 10-fold cross-validation to train and evaluate the model. We conclude that this prediction model may help to identify recruits at an increased risk of attrition from training throughout the Marine recruit training and warrants further validation and updates for other military settings.

Exploration of Race and Ethnicity, Sex, Sport-Related Concussion, Depression History, and Suicide Attempts in US Youth

IMPORTANCE

Racial, ethnic, and sex disparities for concussion incidence and suicide attempts in youth have been reported, but data on the interaction of these variables in a large national data set are lacking. Understanding how race and ethnicity interact with sex and concussion to influence suicide attempts could yield critical information on the sociocultural impact of brain injury and mental health in US youth.

OBJECTIVE

To examine the associations of concussion history, race and ethnicity, and sex with reported suicide attempts among adolescents.

DESIGN, SETTING, AND PARTICIPANTS

This population-based cross-sectional cohort study used data from US Youth Risk Behavior Surveillance System (YRBSS) survey respondents between 2017 and 2019. Data were analyzed from May 2021 to January 2022.

EXPOSURES

Respondents reported sport- or recreation-related concussion (yes or no), depression (yes or no), and suicide attempt (yes or no) over the previous 12 months, along with race and ethnicity (categorized as American Indian or Alaska Native, Asian, Black, Hispanic/Latino, multiracial, Native Hawaiian or other Pacific Islander, and White), and sex (male or female).

MAIN OUTCOMES AND MEASURES

Two Chi-Square Automatic Interaction Detection (CHAID) decision tree models were built. The first was suicide attempt with depression history (SA-DEP), the second was suicide attempt without depression history (SA-NO DEP). CHAID uses risk factors (eg, number of concussions, race and ethnicity, sex) to divide the study sample into a series of subgroups that are nested within each other. Risk ratios (RRs) and 95% CIs were calculated for each subgroup to provide effect estimates.

RESULTS

A total of 28 442 youths aged up to 18 years (mean [SD] age, 14.6 [3.0] years; 14 411 [50.7%] female) responded to the survey. The CHAID decision trees revealed a complex interaction between race, sex, and concussion history for attempting suicide, which differed by depression history (overall accuracy, 84.4%-97.9%). Overall, depression history was the variable most strongly associated with SA (adjusted odds ratio, 11.24; 95% CI, 10.27-12.29). Concussion was the variable most strongly associated with SA-DEP (RR, 1.31; 95% CI, 1.20-1.51; P < .001). Black, Hispanic/Latino, or multiracial race and ethnicity were associated with increased risk for SA-DEP compared with others (RR, 1.59; 95% CI, 1.38-1.84; P < .001). American Indian or Alaska Native, Black, and Hispanic/Latino race and ethnicity were associated with increased risk for SA-NO DEP (RR, 1.89; 95% CI, 1.54-2.32; P < .001) compared with the remaining population.

CONCLUSIONS AND RELEVANCE

These findings suggest that clinicians should consider race, ethnicity, and sex when evaluating the role of sport- or recreation-related concussion on suicide risk among US youth.

Forecasting football injuries by combining screening, monitoring and machine learning

Identifying players or circumstances associated with an increased risk of injury is fundamental for successful risk management in football. So far, time-constant and volatile risk factors are generally considered separately in either a screening (constant) or a monitoring (volatile) approach each resulting in a restricted set of explanatory variables. Consequently, improvements in predictive accuracy may be expected when screening and monitoring data are combined, especially when analysed with current machine learning (ML) techniques. This trial was designed as a prospective observational cohort study aiming to forecast non-contact time-loss injuries in male professional football (soccer). Injuries were registered according to the Fuller consensus. Gradient boosting with ROSE upsampling within a leave-one-out cross-validation was used for data analysis. The hierarchical data structure was considered throughout. Different splits of the original dataset were used to probe the robustness of results. Data of 88 players from 4 teams and 51 injuries could be analysed. The cross-validated performance of the gradient boosted model (ROC area under the curve 0.61) was promising and higher compared to models without integration of screening data. Importantly, holdout test set performance was similar (ROC area under the curve 0.62) indicating prospect of generalizability to new cases. However, the variation of predictive accuracy and feature importance with different splits of the original dataset reflects the relatively low number of events. It is concluded that ML-based injury forecasting based on the integration of screening and monitoring data is promising. However, external prospective verification and continued model development are required.

Blood sample profile helps to injury forecasting in elite soccer players

Purpose

By analyzing external workloads with machine learning models (ML), it is now possible to predict injuries, but with a moderate accuracy. The increment of the prediction ability is nowadays mandatory to reduce the high number of false positives. The aim of this study was to investigate if players’ blood sample profiles could increase the predictive ability of the models trained only on external training workloads.

Method

Eighteen elite soccer players competing in Italian league (Serie B) during the seasons 2017/2018 and 2018/2019 took part in this study. Players’ blood samples parameters (i.e., Hematocrit, Hemoglobin, number of red blood cells, ferritin, and sideremia) were recorded through the two soccer seasons to group them into two main groups using a non-supervised ML algorithm (k-means). Additionally to external workloads data recorded every training or match day using a GPS device (K-GPS 10 Hz, K-Sport International, Italy), this grouping was used as a predictor for injury risk. The goodness of ML models trained were tested to assess the influence of blood sample profile to injury prediction.

Results

Hematocrit, Hemoglobin, number of red blood cells, testosterone, and ferritin were the most important features that allowed to profile players and to analyze the response to external workloads for each type of player profile. Players’ blood samples’ characteristics permitted to personalize the decision-making rules of the ML models based on external workloads reaching an accuracy of 63%. This approach increased the injury prediction ability of about 15% compared to models that take into consideration only training workloads’ features. The influence of each external workload varied in accordance with the players’ blood sample characteristics and the physiological demands of a specific period of the season.

Conclusion

Field experts should hence not only monitor the external workloads to assess the status of the players, but additional information derived from individuals’ characteristics permits to have a more complete overview of the players well-being. In this way, coaches could better personalize the training program maximizing the training effect and minimizing the injury risk.

Unsupervised Clustering Techniques Identify Movement Strategies in the Countermovement Jump Associated With Musculoskeletal Injury Risk During US Marine Corps Officer Candidates School

Musculoskeletal injuries (MSKI) are a significant burden on the military healthcare system. Movement strategies, genetics, and fitness level have been identified as potential contributors to MSKI risk. Screening measures associated with MSKI risk are emerging, including novel technologies, such as markerless motion capture (mMoCap) and force plates (FP) and allow for field expedient measures in dynamic military settings. The aim of the current study was to evaluate movement strategies (i.e., describe variables) of the countermovement jump (CMJ) in Marine officer candidates (MOCs) via mMoCap and FP technology by clustering variables to create distinct movement strategies associated with MSKI sustained during Officer Candidates School (OCS). 728 MOCs were tested and 668 MOCs (Male MOCs = 547, Female MOCs = 121) were used for analysis. MOCs performed 3 maximal CMJs in a mMoCap space with FP embedded into the system. De-identified MSKI data was acquired from internal OCS reports for those who presented to the OCS Physical Therapy department for MSKI treatment during the 10 weeks of OCS training. Three distinct clusters were formed with variables relating to CMJ kinetics and kinematics from the mMoCap and FPs. Proportions of MOCs with a lower extremity and torso MSKI across clusters were significantly different (p < 0.001), with the high-risk cluster having the highest proportions (30.5%), followed by moderate-risk cluster (22.5%) and low-risk cluster (13.8%). Kinetics, including braking rate of force development (BRFD), braking net impulse and propulsive net impulse, were higher in low-risk cluster compared to the high-risk cluster (p < 0.001). Lesser degrees of flexion and shorter CMJ phase durations (braking phase and propulsive phase) were observed in low-risk cluster compared to both moderate-risk and high-risk clusters. Male MOCs were distributed equally across clusters while female MOCs were primarily distributed in the high-risk cluster. Movement strategies (i.e., clusters), as quantified by mMoCap and FPs, were successfully described with MOCs MSKI risk proportions between clusters. These results provide actionable thresholds of key performance indicators for practitioners to use for screening measures in classifying greater MSKI risk. These tools may add value in creating modifiable strength and conditioning training programs before or during military training.

Data analytics in military human performance: Getting in the game: Summary of a keynote address

OBJECTIVES

The rise of data analytics has been not only central to the digital transformation of many industries and governments, but is now ubiquitous in daily life. But what is it? Researchers in military human performance may very well ask themselves: What is new? After all, aren't they already collecting, analysing, interpreting, and presenting data? Do they need to adapt?

DISCUSSION

Defence and security have often been at the forefront of new technologies, but has lagged other industries with respect to data analytics. Sports science is one of the industries that are on the leading edge and this presents an opportunity that researchers in military human performance must seize.

CONCLUSIONS

Researchers must embrace data analytics and seek opportunities to 'operationalize' their research via data science: responsible analytics respecting scientific development supporting decision making at the necessary speed of relevance.

Much work remains to reach consensus on musculoskeletal injury risk in military service members: A systematic review with meta-analysis

Musculoskeletal injuries are the most common reason military service members cannot perform their military duties. Not only are they costly and associated with long-term disability, often long after completion of military service, but injuries also adversely affect the military readiness of a nation. This can be seen as a threat to national security and part of the impetus behind many efforts to better understand, predict, and mitigate injury risk in the military. A systematic review of the literature published between 1995 and October 31, 2020 was conducted to identify significant risk factors of musculoskeletal injury in military populations across the world. 74 out of 170 eligible studies addressed comprehensive injuries, providing 994 unique risk factors. 46 of these studies provided data that could be included in a meta-analysis, which was possible for 15 predictor variables. Seven predictors were significant in meta-analysis: female sex(RR=1.46;95CI 1.30,1.64), high body mass index(RR=1.36;95CI 1.21,1.53), functional movement screen pain (RR=1.70;95CI 1.55,1.87) or scores ≤ 14(RR=1.42 95CI 1.29,1.56), prior injury(RR=1.54;95CI 1.32,1.80), slower running performance(RR=1.33;95CI 1.18,1.51), and poorer push-up performance(RR=1.15;95CI 1.04,1.27). Low BMI, height, weight, smoking, physical activity scores, and sit-up and jump performance were not significant risk factors in the meta-analysis. Most studies had a high risk of bias. Lack of raw data and large heterogeneity in definitions of predictors and injury outcomes limited comparison across many studies.Highlights Female sex, high body mass index, pain with functional movement screen or a score of ≤ 14, prior injury, slower running performance and poorer push-up performance were all significant predictors of musculoskeletal injury.Low body mass index, height, weight, smoking, physical activity scores, and sit-up and jump performance were not significant predictors of musculoskeletal injury.Many other predictors were present only in single studies, but large heterogeneity in definitions of both outcomes and predictors limited comparison across studies.Overall, studies assessing risk factors to predict musculoskeletal injuries in the military were at high risk for bias, especially in regards to statistical approaches.

Can Injuries Be Predicted by Functional Movement Screen in Adolescents? The Application of Machine Learning

ABSTRACT

Karuc, J, Mišigoj-Duraković, M, Šarlija, M, Marković, G, Hadžić, V, Trošt-Bobić, T, and Sorić, M. Can injuries be predicted by functional movement screen in adolescents? The application of machine learning. J Strength Cond Res 35(4): 910-919, 2021-This study used machine learning (ML) to predict injuries among adolescents by functional movement testing. This research is a part of the CRO-PALS study conducted in a representative sample of adolescents and analyses for this study are based on nonathletic (n = 364) and athletic (n = 192) subgroups of the cohort (16-17 years). Sex, age, body mass index (BMI), body fatness, moderate-to-vigorous physical activity (MVPA), training hours per week, Functional Movement Screen (FMS), and socioeconomic status were assessed at baseline. A year later, data on injury occurrence were collected. The optimal cut-point of the total FMS score for predicting injury was calculated using receiver operating characteristic curve. These predictors were included in ML analyses with calculated metrics: area under the curve (AUC), sensitivity, specificity, and odds ratio (95% confidence interval [CI]). Receiver operating characteristic curve analyses with associated criterium of total FMS score >12 showed AUC of 0.54 (95% CI: 0.48-0.59) and 0.56 (95% CI: 0.47-0.63), for the nonathletic and athletic youth, respectively. However, in the nonathletic subgroup, ML showed that the Naïve Bayes exhibited highest AUC (0.58), whereas in the athletic group, logistic regression was demonstrated as the model with the best predictive accuracy (AUC: 0.62). In both subgroups, with given predictors: sex, age, BMI, body fat percentage, MVPA, training hours per week, socioeconomic status, and total FMS score, ML can give a more accurate prediction then FMS alone. Results indicate that nonathletic boys who have lower-body fat could be more prone to suffer from injury incidence, whereas among athletic subjects, boys who spend more time training are at a higher risk of being injured. Conclusively, total FMS cut-off scores for each subgroup did not successfully discriminate those who suffered from those who did not suffer from injury, and, therefore, our research does not support FMS as an injury prediction tool.

Load carriage magnitude and locomotion strategy alter knee total joint moment during bipedal ambulatory tasks in recruit-aged women

Knee osteoarthritis (OA) is prevalent among female soldiers, resulting in limited duty and long term adverse ambulatory effects. A proposed mechanism to the development of knee OA is the assiduous execution of load carriage tasks. Soldiers are often required to maintain a walking gait with load at velocities beyond their gait transition velocity (GTV) known as forced marching. The primary aim of this investigation is to determine the interactive effects of load magnitude and locomotion pattern on relative knee total joint moment (KTJM) in healthy recruit-aged women. The secondary aims are to determine knee total joint moment limb differences and to determine the interactive effect of load magnitude and locomotion pattern on the percent contributions of each plane of motion moment. Individuals were tasked with running and forced marching at 10% above their GTV at body weight (BW) and with an additional 25% and 45% of their BW. KTJM was analyzed at two specific gait events of heel-strike and mid-stance. At heel-strike, forced marching exhibited greater KTJM compared to run for all load conditions but running had greater KTJM than forced marching at mid-stance. The forced marching pattern exhibited larger KTJM for the dominant limb at both gait events compared to running. Lastly, at mid-stance the knee adduction moment percent (KAM%) contribution was greater for forced marching compared to running. The forced marching pattern demonstrates joint kinetics that may be more deleterious with prolonged exposure. Likewise, forced marching induced KAM% similar to those already suffering from knee OA.