Epidemiology of obesity in relation to health hazards.

Examination of total mortality in relation to Metropolitan Relative Weight among 2223 men followed for 30 years in the Framingham Heart Study shows that the duration of follow-up influences the nature of the relationship. During the first 6 years, among nonsmokers, mortality and weight had an inverse relation, but as time goes on the relation becomes U-shaped with a minimum mortality at a relative weight of 100% to 109%. Similar relations are found among men who smoke. Time trends in national mortality rates, however, do not show a direct relation to national trends in mean body mass index. Caution is necessary in relating to the effects of changes in obesity levels in populations to indices of health and longevity.