Properties and clinical implications of body mass indices

The properties of body mass indices were evaluated in a cross sectional study of the weights and heights of 5016 Chinese boys and girls aged between 3 and 18 years. Of the indices examined (weight/height (W/H), weight/height2 (W/H2), weight/height3 (W/H3) and weight/heightp (W/Hp], W/Hp was the only one that consistently showed least correlation with height, and so could be regarded as the optimal body mass index by the criterion of independence of the index from height. The exponent 'p' of W/Hp is, however, highly dependent on age; the value increases steadily between the age of 3 and 7-9 years, and then varies considerably around puberty. Only the age specific exponent ensures a lack of correlation between body mass index (W/Hp) and height. Age specific W/Hp should therefore be used in intrapopulation studies of weight or problems associated with obesity in children. Interpopulation comparison of weight and adiposity by W/H, W/H2, or W/H3 may give misleading results because of their dependence on height. Our results also suggest that the conventional weight for height charts may not be accurate enough for clinical use.