Repeated measures of body mass index and risk of health related outcomes.

Most studies examining the association between body mass index (BMI) and mortality neglected changes in weight over time, which may have led to underestimation of the true association. The aim of this study is to examine the relationship between BMI and health related outcomes while accounting for variations of BMI over time. The association between BMI and both mortality and occupational disability was examined in a follow-up of 5,554 male construction workers in Württemberg/Germany, who participated at least two times in routine occupational health examinations between 1986 and 2005. Using Cox proportional hazards model with time dependent variables, hazard ratios were calculated with normal weight (<25 kg/m²) as reference after adjustment for potential confounding factors. Overall, an U-shaped association between baseline BMI and mortality (370 events) as well as occupational disability (658 events) was observed, with lowest risk at BMI levels between 25 and 30 kg/m². Men with a baseline BMI ≥ 30 kg/m² experienced a 10% higher mortality and disability risk than normal weight men. The association between BMI and occupational disability became stronger after accounting for temporal variability of BMI with a significant increased risk of 1.26 (95% confidence interval: 1.01-1.56) among obese men. In contrast, the association between BMI and mortality did not materially change after accounting for time dependent effects. Stable obesity as defined by a BMI of 30 kg/m² and above increases risk of disability in male construction workers. Accounting for changes of BMI over time is crucial for disclosing full impact of obesity.