Body mass index, body composition and the chronic energy deficiency classification of rural adult populations in Guatemala.

The present study tested the hypotheses that: (a) individual body composition estimates obtained with the Durnin-Womersley (D-W) equations have low validity in certain populations in developing countries; (b) there exists a poor relationship between the body mass index (BMI) and body composition estimates (fat mass (BFM) and fat-free mass (FFM)), and (c) BMI cut-off estimates (fat mass (BFM) and fat-free mass (FFM)), and (c) BMI cut-off points provide an invalid classification of chronic energy deficiency (CED) in adults. The study involved four samples of rural men and women in Guatemala, who had mean BMI of approximately 21 kg/m2. Body composition estimates were obtained by densitometry in three of the samples. Mean body fat (%) and mean FFM (kg) were: men: 11.6 (+/- 4.7) and 47.7 (+/- 4.9); and women: 21.6 (+/- 5.3) and 35.8 (+/- 3.5), respectively. The D-W equations based on various combinations of skinfold measurements consistently overestimated body fat content with low precision and validity. The BMI was more related to BFM and FFM than to fat proportion, but explained little of the variation in both body components, particularly at low BMI levels. A small number of men and women had BMI values below 18.5 kg/m2, and only one woman fell below 16 kg/m2. The power coefficients of height in the weight/height ratio which provided the strongest correlations with BFM and FFM were: BFM: women: 1.0; men: 1.5; FFM: 0.5 for both women and men. We conclude that the Quetelet index should not be recommended as a universally valid indicator to classify CED in adult groups similar to the study population.