Measurement of body fat in healthy elderly men: a comparison of methods

BACKGROUND: Nutritional evaluation of elderly people is of great importance. Two-component methods for body composition assessment, such as anthropometry and bioelectrical impedance (BIA), are widely used in clinical practice, but their fundamental assumptions may be invalid in older people. Dual-energy X-ray absorptiometry (DXA) is a relatively new method for reliable and direct measurements of body mass in its three basic components: total body bone mineral content (TBBMC), mineral free lean tissue mass (LTM), and fat. In this study, percent body fat (%BF) estimates from anthropometry and BIA in men of various ages were compared with corresponding measurements by DXA. METHODS: Body fat percentage was estimated in 67 men aged 20-95 by anthropometric measurements (skinfold thickness, body mass index, or BMI), BIA, and DXA. Age-specific equations were used for anthropometry and BIA. Limits of agreement were calculated between DXA and the other methods. RESULTS: The equations based on BMI and BIA systematically overestimated %BF with respect to %BF measured by DXA in people of all ages. Intermethod difference between DXA and skinfold thicknesses was less marked, but in over-80-year-olds %BF predicted by skinfold measurements underestimated %BF measured by DXA. Interindividual and age-related variation in TBBMC and in fat-free mass mineralization could partly explain the intermethod differences found between DXA and the other methods. CONCLUSIONS: Because of practical constraints, anthropometry and BIA are often the only available options for body composition assessment in clinical routine; therefore, further research on the validity and improvement of these methods in older people is indicated.