Load-velocity profiles in machine-based exercises: comparing trained college students with recreationally active and sedentary middle-aged and older adults

This study aimed to examine the reliability, validity and efficacy of using group or individualized load-velocity profiles (LVPs) in four machine-based exercises, involving trained college students and middle-aged and older adults (recreationally active and sedentary). Thirty students and fifty middle-aged/older adults completed two sessions. Mean concentric velocity (MV) of leg press, lateral pull down, knee extension, and chest press was assessed at 40–80% of estimated one-repetition maximum (1RM). Both sessions’ data assessed test-retest reliability, while first-session MV data validated predicted velocities using individual/group regression models. Reliability and validity were evaluated via intraclass correlation coefficient (ICC), standard error of measurement (SEM), standard error of the estimate (SEE), and Coefficient of Variation of the RMSE (CV-RMSE). MV showed good test-retest reliability across all exercises from 40%1RM to 80%1RM (ICC > 0.77, SEM ≤ 0.05 m/s). Validity of individual LVPs varied from moderate to nearly perfect, with acceptable validity for all groups (R2 > 0.81, ICC > 0.90, SEE: 0.01–0.07 m/s, CV-RMSE: 1.56–14.89%). Validity of group averaged LVPs (linear/nonlinear) was mostly unacceptable for recreationally active middle-aged/older adults (except chest press) and trained college students (ICC ≤ 0.74, SEE: 0.08–0.20 m/s, CV-RMSE: 15.09–33.33%), yet showed moderate predictive validity for sedentary middle-aged/older adults in most exercises except knee extension (ICC > 0.75 for leg press, lateral pull-down, and chest press; SEE: 0.07–0.13 m/s, CV-RMSE: 14.58–23.21%). For high precision, individualized LVPs are recommended. For sedentary middle-aged and older adults with lower precision needs, group-averaged values may be considered as a less optimal alternative for screening purposes only, with the exception of knee extension where individualized profiling is essential.