Body composition and body hydration in old age estimated by means of skinfold thickness and deuterium dilution

In this study a group of 30 women and 30 men aged 60-90 years was examined. Body density was predicted from the sum of four skinfold thickness (SFT) as described by Durnin and Womersley (1974). Body composition, expressed as a percentage of total body weight as fat mass (FM) and fat free mass (FFM) was calculated by means of Siri's formula (Siri 1961). Total body water (TBW) was measured from the dilution of an oral deuterium dose and FFM calculated assuming a hydration of 73.2%. The FM% predicted from SFT showed a slight but not significant decreasing age trend from 28.3% to 25.4% in men, and from 40.3% to 37.8% in women. Body hydration (about 50% of body weight in men and 44% in women) remained constant throughout the age range. FFM obtained from TBW assuming a constant hydration of FFM, was found to be unchanged with age. FM% predicted from SFT was correlated with values obtained from TBW on an individual basis (R2 = 0.38; p less than 0.0001 in men and R2 = 0.31; p less than 0.002 in women), but the scatter was large. The Bland and Altman statistical analysis (Bland and Altman 1986) suggests that the discrepancy between the two estimates tends to be significantly more positive with increasing fatness of the subject. The assumption of a constant hydration of FFM between individuals and with advancing age was tested. The hydration of FFM was calculated as the ratio between TBW (from deuterium dilution) and FFM (estimated from SFT). The wide range of values (50% to 90%) suggests large inter-individual differences. The mean values were lower than commonly reported, and differed with age and sex. In men, FFM hydration steadily decreases with age from the 7th decade (70.2 +/- 7.7%) to the 10th decade (65.9 +/- 8.2%). Women showed higher average figures than those of men and a lower age-associated decline (72.5% and 68.5% respectively). Our results suggest that FFM hydration decreases with age, and this indicates the need to develop age-related correction factors in order to derive FFM from TBW in an elderly population.