Waaler revisited: The anthropometrics of mortality.

Although many studies have been written about the relationship between BMI and human height on the one hand and mortality on the other, the issue of socio-economic status as confounding variable has been at times less emphasized. This study analyzes the influence of education and income on the relationship between BMI and mortality and between height and mortality. It is based on data collected between 1963 and 1975 by the Norwegian National Health Screening Service. 1.7 million subjects were recorded. The Norwegian statistics bureau linked these data to the national death records and to socio-economic information. We apply Cox proportional hazards regressions in order to determine whether adding income and education as covariates affects the relations among BMI, height, and mortality. Previous findings and insights are either not present or ambiguous. We conclude that the omission of SES does not significantly bias the effect of BMI on most causes of death, with one exception: type 2 diabetes mellitus, where the effect of BMI is substantially lower for both adults and adolescents when adjusted for education.