Background/objective: We aimed to develop regression-based prediction equations for estimating total and regional skeletal muscle mass (SMM) from measurements of lean soft tissue mass (LSTM) using dual-energy X-ray absorptiometry (DXA) and investigate the validity of these equations.
Subjects/methods: In total, 144 healthy Japanese prepubertal children aged 6-12 years were divided into 2 groups: the model development group (62 boys and 38 girls) and the validation group (26 boys and 18 girls). Contiguous MRI images with a 1-cm slice thickness were obtained from the first cervical vertebra to the ankle joints as reference data. The SMM was calculated from the summation of the digitized cross-sectional areas. Total and regional LSTM was measured using DXA.
Results: Strong significant correlations were observed between the site-matched SMM (total, arms, trunk and legs) measured by MRI and the LSTM obtained by DXA in the model development group for both boys and girls (R2adj=0.86-0.97, P<0.01, standard error of the estimate (SEE)=0.08-0.44 kg). When these SMM prediction equations were applied to the validation group, the measured total (boys 9.47±2.21 kg; girls 8.18±2.62 kg) and regional SMM were very similar to the predicted values for both boys (total SMM 9.40±2.39 kg) and girls (total SMM 8.17±2.57 kg). The results of the Bland-Altman analysis for the validation group did not indicate any bias for either boys or girls with the exception of the arm region for the girls.
Conclusions: These results suggest that the DXA-derived prediction equations are precise and accurate for the estimation of total and regional SMM in Japanese prepubertal boys and girls.