Long-term weight changes in obese young adult men and subsequent all-cause mortality

BACKGROUND: Although the expectation is that weight gain increases mortality and weight loss among those overweight reduces mortality, results on weight gain and mortality in young adults are conflicting, and weight loss is less explored. We investigated the association between long-term weight change and all-cause mortality in a broad range of body mass index (BMI) in young men. METHODS: Among 362200 Danish draftees, examined between 1943 and 1977, all obese (BMI >= 31.0 kg m(-2); n = 1930), and a random 1% sample of the others (n = 3601) were identified at a mean age of 20 years (range: 18-25 years). All the obese and half the controls were re-examined between 4 and 40 years later (mean age 35 years). Weight changes were defined as: weight loss < -0.1 kg m(-2) per year, weight stability within +/- 0.1 kg m(-2) per year and weight gain >0.1 kg m(-2) per year. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated using Cox regression. RESULTS: Among the 908 obese and 1073 controls followed for 30 years after re-examination 220 and 232 died. HR of the weight stable obese was 2.32 (CI: 1.56-3.44) compared with the weight stable controls. In the obese cohort there was no association between weight loss, adjusted for initial BMI, and mortality (HR: 0.99; CI: 0.68-1.45) compared with weight stable obese. Too few controls lost weight to allow assessment of weight loss. Weight gain was associated with increased mortality in the obese (HR: 1.50; CI: 1.07-2.10) and controls (HR: 1.54; CI: 1.14-2.09) compared with weight stable obese and controls, respectively. Neither the time between the two examinations, life-style factors nor exclusion of diseased individuals influenced the results. CONCLUSIONS: Although there were increased mortality of the weight-stable obese compared with controls, there was no association between weight loss and mortality in the obese. Weight gain increased mortality regardless of the initial weight.