Sex and age differences in the effect of obesity on incidence of hypertension in the Japanese population: A large historical cohort study.

Elevated risk of death from any cause and stroke associated with high body mass index (BMI) values decline with aging. However, it is not clear whether the effect of obesity on the incidence of hypertension varies by sex or age. Cox regression analyses were performed using a large historical cohort composed of 6803 men and 22,800 women. In the final model, which included the main effect (sex, age, and BMI), all two-way interactions (sex*age, sex*BMI, and age*BMI), a three-way interaction (sex*age*BMI), and adjusted variables determined by backward elimination, the three-way interaction and two of the two-way interactions (age*sex and age*BMI) were not significant; the remaining two-way interaction between sex and BMI was significant (P = .016). In the next step, the effect of BMI on the incidence of hypertension stratified by sex was evaluated. Hazard ratios (HRs) were significantly higher for BMI 27.0 to <29.0 and ≥29.0 kg/m(2) than for the HR for BMI 21.0 to <23.0 kg/m(2) in women (HR, 1.45; 95% confidence interval [CI], 1.32-1.59 and HR, 1.46; 95% CI, 1.29-1.65, respectively), but not in men (HR, 1.14; 95% CI, 0.98-1.33 and HR, 1.01; 95% CI, 0.77-1.32, respectively). The effect of obesity on the incidence of hypertension is stronger in women than in men.