Development and validation of prediction models for one-repetition maximum based on bioelectrical impedance analysis in adults of all ages

[Purpose] This study aimed to develop and internally validate one-repetition maximum prediction models based on skeletal muscle mass or skeletal muscle mass index measured using bioelectrical impedance analysis in adults across all ages and both sexes. [Participants and Methods] Eighty-two participants (40 males, 42 females, aged 20–84 years) completed measurement of dominant leg skeletal muscle mass and skeletal muscle mass index using multifrequency bioelectrical impedance analysis. Unilateral leg press one-repetition maximum was measured. Two multiple linear regressions that incorporated age and sex were fitted and internally validated through bootstrap resampling. [Results] The model using skeletal muscle mass produced an adjusted R2 of 0.701 and a standard error of estimate of 16.01 kg. The model using skeletal muscle mass index demonstrated higher accuracy (adjusted R2=0.737; standard error of estimate=15.0 kg). Internal validation indicated minimal overfitting, with optimism-corrected adjusted R2 values of 0.677 and 0.715 and standard error of estimate values of 16.82 and 15.81 kg, respectively. [Conclusion] The developed and internally validated models can reasonably predict unilateral leg press one-repetition maximum in adults across a broad age range. However, caution is warranted owing to the magnitude of the prediction error.