Comparison of predictive equations of lean mass in young and postmenopausal women

Abstract

Background and aim: The lean mass (LM) is a predictor of functional capacity, quality of life, and mortality. In this way, the LM should be measured by reliable methods. However, it presents high cost and generally predictive equations are used in clinical practice, but little is known which is the best predictive equation of LM in women. The purpose of the present study was to verify which predictive equation of LM correctly estimates the LM in young and postmenopausal women.

Methods: Eighty-one women aged 19-81 years were evaluated. Body weight, height, waist circumference, and skin folds (bicipital, tricipital, subscapular and suprailiac) were measured. The LM was evaluated by DXA and also estimated using the predictive equations of Hume I, Hume II, Salamat, Kulkarni I, and Kulkarni II. Bland-Altman analysis was performed to evaluate the over/underestimation of the LM by predictive equations.

Results: The equations of Salamat, Kulkarni II, Hume I and Kulkarni I overestimated the LM by 0.0 (7.0; -6.9) kg; 2.3 (7.5; -3.0) kg; 5.1 (9.0; 0.4) kg; and 9.7 (16.3, 3.1) kg, respectively; whereas Hume II equation underestimated the LM by -16.9 (-11.5; -22.2) kg.

Conclusions: The equation that presented a better prediction of LM was Salamat. However, it should be used with caution in clinical practice since this equation showed elevated confidence intervals and limits of agreements, and can lead to significant errors for some individuals.

Keywords: Body composition; DXA; Lean mass.