When is sitting height a better measure of adult body size than total height, and why? The contrasting examples of body mass, waist circumference, and lung volume.

OBJECTIVES: We aimed to establish which of sitting height (SH) and total height (Ht) is most appropriately used in the scaling of adult body mass (BM), waist circumference (WC), and forced vital capacity (FVC), considering likely explanations and proposing a suitable index for each. METHODS: Data were from the US Third National Health and Nutrition Survey for white and black American women and men aged 20 to 50 years. Statistical analysis involved mainly correlation coefficients, the multiple regression of BM, WC, or FVC on SH and leg length (LL), and fitting of allometric regression equations relating each of BM, WC, and FVC to SH or Ht. RESULTS: BM and WC correlated more strongly with SH than with Ht, and FVC correlated more strongly with Ht. Associations with LL were negative for WC, negative or nonsignificant for BM, and positive for FVC. Using round-number exponents for Ht and SH, the allometric relationships indicated that appropriate indices are BM/SH3 , WC/SH, and FVC/Ht2 . CONCLUSIONS: Contrary to usual practice, BM and WC are better scaled in relation to SH than to Ht. FVC is slightly better scaled in relation to Ht, as is conventional. Interpretations involve the small influence of LL on BM and the influence both of gluteo-femoral fatness on measured SH and of childhood health and nutrition on adult LL, WC, and FVC. It is evident that SH should be measured more often for research purposes.

The waist-hip ratio: a flawed index.

The waist-hip ratio, namely waist circumference (WC) divided by hip circumference (HC), has been referred to in thousands of articles, generally as a correlate and predictor either of health conditions such as cardiovascular disease and diabetes, or of amounts of visceral and subcutaneous abdominal fat. It has been argued that combining WC and HC as a ratio is inappropriate, and yet their individual roles can only be fully elucidated if considered jointly. Whereas WC is positively associated with cardiovascular disease, diabetes and premature mortality, the opposite is true of HC. With health-related measures taken as dependent variables, the present novel approach establishes that WC and HC are far better treated as separate independent variables in multiple regression equations than as their ratio. This necessarily produces closer fits to data. One should then allow for variations in height, or some other such measure of general body size, by including this in the regression equations. The widespread concern with the ratio seems to have distracted attention from HC, for this is discussed notably less often than WC. Given that other body parts, such as the thighs, may share relevant properties with the hips, measurements of these could perhaps replace HC.