Metabolic predictors of obesity: cross-sectional versus longitudinal data

In prospective studies in Pima Indians, four metabolic parameters, which are known to have a familial component, have been found to predict weight gain. These are: (i) low relative resting metabolic rate (RMR, relative to the average predicted rate for body size), (ii) low level of spontaneous physical activity (SPA), (iii) high 24 h respiratory quotient (RQ) and (iv) high insulin sensitivity (IS). Cross-sectional studies show that all four parameters correlate with body size: RMR vs. fat-free mass (FFM), r = 0.87, P < 0.0001; energy cost of SPA vs. weight, r = 0.69, P < 0.001; RQ vs. body fat, r = -0.23, P < 0.05; IS vs. weight, r = -0.38, P < 0.001. When these parameters are adjusted for differences in body size, then the initial value predicts the rate of change in body weight over the subsequent years: RMR (adjusted for FFM, fat mass, age and sex), r = -0.39, P < 0.001; SPA, r = -0.35, P < 0.005 (males); RQ (adjusted for fat mass), r = 0.24, P < 0.01; IS (adjusted for weight), r = 0.34, P < 0.0001. After gaining weight, the original deviation from the value predicted on the basis of the population (e.g. low relative RMR, high RQ and high IS) tends to diminish, suggesting a progressively decreasing physiological drive for further body weight gain. Thus, the high RMR, high energy cost of SPA, low RQ and low IS seen in obesity may act to limit additional weight gain.