Upper extremity and trunk body segment parameters are affected by BMI and sex

Although body mass index (BMI) relates to body segment parameters (BSPs), unknowns persist over whether: 1) BSPs relate to BMI group classifications, 2) sex influences BMI/BSP relationships, and 3) simple anthropometric measures sufficiently predict BSPs. Dual energy X-ray absorptiometry (DXA) scans and anthropometric measures were obtained from 76 participants (33M, 43F) of varying body composition. Trunk, neck and head (TNH), arm, forearm and hand masses were obtained from DXA scans and center of mass locations (COM) estimated from geometric models. Groups with larger BMIs had lesser hand and forearm mass (%total body mass; p<0.001) and greater TNH mass (p=0.014). Males had greater hand, forearm, and arm masses (p<0.05). TNH COM was inferior in groups with larger BMI (p<0.01) and in males (p=0.006). In females, arm COM was distal in Obese II/III versus Normal (p=0.024). Two sets of linear models were created to predict BSPs; a simplified set with only BMI, height, weight, sex and mass potential predictors and a complex set with additional anthropometric measures. Complex arm and TNH mass models (arm R²=0.43, TNH R²=0.61) explained more variance than simplified models (arm R²=0.1, TNH R²=0.33). Complex hand mass, forearm mass and TNH COM models had smaller R² increases versus simplified models (hand=0.05, forearm=0.06, TNH=0.08). Explained variance in forearm COM (R²=0.2) and arm COM (R²=0.27) complex models was low, suggesting a constant may provide reasonable estimates. Certain BSPs can be estimated using simplified measures, whereas prediction of other BSPs markedly improves if additional anthropometric measures are included.