Development of weight-estimation formulae for the bedridden elderly requiring care

AIM: Bedridden elderly persons requiring care need special body-weight measurement implements, and body-weighting assumes more difficult if they live at their own homes. Therefore, we tried to design a new weight-estimation formulae using various anthropometric variables. METHOD: The subjects were 33 male and 132 female elderly inpatients certified to be at care level 4 or 5. The body composition included height, body weight, arm circumference, triceps skinfold thickness, subscapular skinfold thickness, calf circumference, and waist circumference. RESULTS: We studied the correlation between the body weight and each anthropometric variable and age. In men, the highest correlation with body weight was shown by waist circumference (r=0.891, p<0.0001), followed by age (r=0.779, p<0.0001) and calf circumference (r=0.614, p<0.0001). The variables that showed the highest correlation with body weight in women were waist circumference (r=0.806, p<0.0001), followed by triceps skinfold thickness (r=0.723, p<0.0001) and arm circumference (r=0.662, p<0.0001). The weight estimation formulae were obtained by multiple regression analysis. Formulae for men: body weight=0.660×waist circumference (cm)+0.702×calf circumference (cm)+0.096×age (years)-26.917 (R(2)=0.862, p<0.001); formulae for women: body weight=0.315×waist circumference (cm)+0.684×arm circumference (cm)+0.183×height (cm)-28.788 (R(2)=0.836, p<0.001). CONCLUSION: We successfully developed gender-specific weight-estimation formulae with high coefficients of determination. The results suggest that waist circumference, which is an index of visceral fat, is an effective anthropometric variable to estimate the body weight of bedridden elderly patients requiring care.