Predicting muscular strength using demographics, skeletal dimensions, and body composition measures

The purpose of this study was to develop an equation to predict strength for seven common resistance training exercises using anthropometric and demographic measures. One-hundred forty-seven healthy adults (74 males, 73 females, 35 ± 12 yr, 174 ± 10 cm, 88 ± 19 kg) volunteered to participate. Body composition values (regional/total) and body dimensions were assessed using dual-energy x-ray absorptiometry (DEXA). Subjects underwent the following maximal strength assessments: Leg Press, Chest Press, Leg Curl, Lat Pulldown, Leg Extension, Triceps Pushdown, and Biceps Curl. Multiple linear regression with stepwise removal was used to determine the best model to predict maximal strength for each exercise. Independent predictor variables identified (p < 0.05) were height (cm); weight (kg); BMI; age; sex (0 = F,1 = M); regional lean masses (LM,kg); fat mass (FM,kg); fat free mass (FFM,kg); percent fat (%BF); arm, leg, and trunk lengths (AL, LL, TL; cm); and shoulder width (SW,cm). Analyses were performed with and without regional measures to accommodate scenarios where DEXA is unavailable. All models presented were significant (p < 0.05, R ² = 0.68-0.83), with regional models producing the greatest accuracy. Results indicate that maximal strength for individual resistance exercises can be reasonably estimated in adults.