Background: To clarify the effect of age on the association between body mass index (BMI) and all-cause mortality.
Methods: We followed 43 972 Japanese participants aged 40 to 79 years for 12 years. Cox proportional hazards regression analysis was used to estimate hazard ratios (HRs), using the following BMI categories: <18.5 (underweight), 18.5-20.9, 21.0-22.9, 23.0-24.9 (reference), 25.0-27.4, 27.5-29.9, and ≥30.0 kg/m(2) (obese). Analyses were stratified by age group: middle-aged (40-64 years) vs elderly (65-79 years).
Results: We observed a significantly increased risk of mortality in underweight elderly men: the multivariate HR was 1.26 (0.92-1.73) in middle-aged men and 1.49 (1.26-1.76) in elderly men. In addition, we observed a significantly increased risk of mortality in obese middle-aged men: the multivariate HR was 1.71 (1.17-2.50) in middle-aged men and 1.25 (0.87-1.80) in elderly men. In women, there was an increased risk of mortality irrespective of age group in the underweight: the multivariate HR was 1.46 (0.96-2.22) in middle-aged women and 1.47 (1.19-1.82) in elderly women. There was no excess risk of mortality with age in obese women: the multivariate HR was 1.47 (0.94-2.27) in middle-aged women and 1.26 (0.95-1.68) in elderly women.
Conclusions: As compared with the reference category, obesity was associated with a high mortality risk in middle-aged men, whereas underweight, rather than obesity, was associated with a high mortality risk in elderly men. In women, obesity was associated with a high mortality risk during middle age; underweight was associated with a high mortality risk irrespective of age. The mortality risk due to underweight and obesity may be related to sex and age.