A Retrospective Study of Foot Biomechanics and Injury History in Varsity Football Athletes at the U.S. Naval Academy

Abstract

Introduction: The six-item Foot Posture Index (FPI-6) was previously developed as an assessment tool to measure the posture of the foot across multiple segments and planes. It was derived from a criterion-based observational assessment of six components of each foot during static standing. The association between abnormal foot posture and musculoskeletal injuries remains unclear and is in of need further exploration.

Hypothesis/purpose: The purpose of this study was to assess the association between foot biomechanics and self-reported history of musculoskeletal pain or injury.

Study design: Retrospective, cross-sectional study of collegiate football players at the U.S. Naval Academy.

Materials and methods: For each athlete, data were recorded on height, weight, self-reported history of pain or injury, and foot posture, which was measured using a FPI-6 with each item measuring the degree of pronation/supination. The primary outcome was each athlete's maximum deviation from neutral posture across the six-item index (FPImax). The prespecified primary analysis used generalized linear models to measure the association between FPImax and self-report history of pain or injury. Exploratory analyses measured the association using penalized regression (L1-norm) and a type of tree-based ensemble known as extreme gradient boosting (XGBoost).

Results: Data were collected on 101 athletes, 99 of whom had sufficient body mass index (BMI) data to be included for analysis. Among the 99 athletes, higher FPImax was associated with a prior history of musculoskeletal pain (odds ratio [OR] 1.15, 95% confidence interval [CI] 0.97 to 1.35), although the sample size was too small for the association to be significant with 95% CI (P = .107). FPImax was not associated with a history of knee injury/pain (OR 0.98, 95% CI 0.83 to 1.15, P = .792), nor with a history of ankle/foot injury or pain (OR 1.04, 95% CI 0.90 to 1.21, P = .599). From the L1-penalized model, the FPI components with the strongest linear associations were the L6, R2, R1-squared, and FPImax. From the XGBoost model, the most important variables were FPItotal, BMI, R1, and R2.

Conclusions: The U.S. Naval Academy football players whose foot postures deviated from neutral were more likely to have reported a previous history of musculoskeletal pain. However, this deviation from normal was not strongly associated with a specific history of pain or injury to the knee, ankle, or foot.

Clinical relevance: The information ascertained from this study could be used to better inform clinicians about the value of the FPI in predicting or mitigating injuries for varsity football athletes.