Validity of ultrasound prediction equations for total and regional muscularity in middle-aged and older men and women.

To test the validity of published equations, 79 Caucasian adults (40 men and 39 women) aged 50-78 y had muscle thickness (MT) measured by ultrasound at nine sites of the body. Fat-free mass (FFM), lean soft tissue mass (LM) and total muscle mass (TMM) were estimated from MT using equations previously published in the literature. Appendicular LM (aLM) was estimated using dual-energy X-ray absorptiometry (DXA) and this method served as the reference criterion. There were strong correlations (range r = 0.85-0.94) between DXA-derived aLM and estimated FFM, leg LM or TMM. Total error between DXA-derived aLM and TMM (∼2 kg) was lower compared with the three other selected equations (6-10 kg). A Bland-Altman plot revealed that there was no systematic bias between aLM and TMM; however, the other three equations included systematic error. Our results suggest that an ultrasound equation for TMM is appropriate and useful for evaluating skeletal muscle mass in the body.