**Background:** The US Army replaced the Army Physical Fitness Test with the Army Combat Fitness Test (ACFT), a six-event assessment designed to be sex- and age-neutral. Prior research showed substantial sex-based performance differences, and body size (particularly BMI) was known to correlate with performance—positively for strength/power events and inversely for aerobic events. However, BMI does not capture body shape, which may provide more detailed insights. This study aimed to identify relationships between 3D-imaged body shape, body composition, sex, and ACFT performance.
**Methods:** This cross-sectional study enrolled 265 USMA cadets, of whom 239 (83 females, 156 males) completed the ACFT between February and May 2021. Participants were scanned using a Styku S100 3D body imaging scanner, which generated circumference measurements at 20 body locations (averaged to 13 for analysis). ACFT raw scores were obtained from institutional records. Pearson correlation coefficients were calculated between body measurements and each ACFT event. Multiple linear regression models were developed using the 13 circumferences, ratios, and sex as independent variables. A k-means cluster analysis was performed on normalized body shape measurements (13 circumferences, weight, height), with the number of clusters determined by the elbow of the scree plot. ACFT performance was compared between clusters using t-tests with Holm-Bonferroni correction.
**Key Results:** Five clusters were identified: (1) 'V-shaped' males (n=70, 1% female), (2) larger males (n=27, 0% female), (3) inverted 'V-shaped' males and females (n=50, 64% female), (4) 'V-shaped' smaller males and females (n=39, 8% female), and (5) smallest males and females (n=53, 89% female). Clusters 1 and 2 had the highest performance on all events except the 2-mile run. For example, Cluster 1 achieved mean MDL of 331.3±19.26 lb, SPT of 10.91±1.47 m, HRPU of 51.04±8.10, SDC of 87.33±5.98 sec, LT of 17.63±3.67, and 2-mile run of 854.1±83.74 sec. Cluster 5 had the lowest performance (MDL 178.1±26.24 lb, SPT 5.82±1.07 m, HRPU 32.60±10.21, SDC 123.11±12.30 sec, LT 6.64±6.05, 2-mile run 957.3±113.90 sec). Clusters 3 and 4 showed no statistically significant differences in performance but both outperformed Cluster 5. Correlation analysis revealed that bicep circumference (r=0.82 for MDL, r=0.78 for SPT, r=0.56 for HRPU, r=-0.72 for SDC, r=0.54 for LT), chest-to-waist ratio, and forearm circumference had the highest correlations with performance across events. BMI was correlated with all events except the 2-mile run (r=-0.04, not significant). Regression models achieved adjusted R² values of 0.83 (MDL), 0.77 (SPT), 0.59 (HRPU), 0.72 (SDC), 0.67 (LT), and 0.25 (2-mile run). Adding cluster number as a covariate did not improve model fit.
**Clinical Implications:** Body shape provides more granular performance insights than sex alone, as three of five clusters contained both sexes. The identification of specific body circumferences (bicep, forearm, chest-to-waist ratio) as strong performance predictors suggests that 3D scanning could guide targeted training interventions. However, the authors note that sex-based physiological differences (e.g., females having smaller bicep circumferences even with training) may limit the feasibility of shifting all individuals to the highest-performing clusters. The study is limited by its convenience sample of motivated cadets not representative of the broader Army, the correlational design (causation cannot be inferred), and the subsequent modification of the ACFT (leg tuck replaced with plank, new age/sex-specific scoring). Longitudinal evidence is needed to determine whether altering body shape improves performance.