Derivation and validation of simple equations to predict total muscle mass from simple anthropometric and demographic data.

Background: Muscle mass reflects and influences health status. Its reliable estimation would be of value for epidemiology.

Objective: The aim of the study was to derive and validate anthropometric prediction equations to quantify whole-body skeletal muscle mass (SM) in adults.

Design: The derivation sample included 423 subjects (227 women) aged 18-81 y with a body mass index (BMI; in kg/m(2)) of 15.9-40.8. The validation sample included 197 subjects (105 women) aged 19-83 y with a BMI of 15.7-36.4. Both samples were of mixed ethnic/racial groups. All underwent whole-body magnetic resonance imaging to quantify SM (dependent variable for multiple regressions) and anthropometric variables (independent variables).

Results: Two prediction equations with high practicality and optimal derivation correlations with SM were further investigated to assess agreement and bias by using Bland-Altman plots and validated in separate data sets. Including race as a variable increased R(2) by only 0.1% in men and by 8% in women. For men: SM (kg) = 39.5 + 0.665 body weight (BW; kg) - 0.185 waist circumference (cm) - 0.418 hip circumference (cm) - 0.08 age (y) (derivation: R(2) = 0.76, SEE = 2.7 kg; validation: R(2) = 0.79, SEE = 2.7 kg). Bland-Altman plots showed moderate agreement in both derivation and validation analyses. For women: SM (kg) = 2.89 + 0.255 BW (kg) - 0.175 hip circumference (cm) - 0.038 age (y) + 0.118 height (cm) (derivation: R(2) = 0.58, SEE = 2.2 kg; validation: R(2) = 0.59, SEE = 2.1 kg). Bland-Altman plots had a negative slope, indicating a tendency to overestimate SM among women with smaller muscle mass and to underestimate SM among those with larger muscle mass.

Conclusions: Anthropometry predicts SM better in men than in women. Equations that include hip circumference showed agreement between methods, with predictive power similar to that of BMI to predict fat mass, with the potential for applications in groups, as well as epidemiology and survey settings.