Rachel Dankner1,2,3, Michael Shanik4,5, Jesse Roth3, Ayala Luski1, Flora Lubin1, Angela Chetrit1. 1. Unit for Cardiovascular Epidemiology, The Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center, Tel Hashomer, Israel. 2. Department for Epidemiology and Prevention, School of Public Health, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. 3. The Feinstein Institute for Medical Research, North Shore-Long Island Jewish Health System, Manhasset, NY, USA. 4. Endocrine Associates of Long Island, P.C, Smithtown, NY, USA. 5. Department of Internal Medicine, Stony Brook University Hospital, Stony Brook, NY, USA.
Abstract
BACKGROUND: Although obesity has been associated with a higher risk for premature death, the sex and ethnic-origin specific body mass index (BMI) levels that are associated with increased mortality are controversial. We investigated the 40-year cumulative all-cause mortality, in relation to the BMI in adult life, among men and women originating from Yemen, Europe/America, Middle East and North Africa, using sex and ethnic-origin specific BMI cut points. METHODS: A random stratified cohort (n = 5710) was sampled from the central population registry and followed since 1969 for vital status. Weight, height and blood pressure were measured, and smoking status was recorded at baseline. BMI was analysed according to conventional categories and according to sex and ethnic-origin specific quintiles. RESULTS: Elevated and significant mortality hazard ratios (HRs) of 1.21 [95% confidence interval (CI) 1.00-1.45] for women and 1.22 (95%CI 1.03-1.44) for men were found for the highest origin-specific BMI quintile. In men, the lowest ethnic-origin specific quintile was also significantly associated with increased mortality (HR of 1.22 95% CI 1.03-1.45), adjusting for age, smoking and blood pressure. Obesity was associated with mortality in non-smokers (HR = 1.29, 95% CI 1.04-1.61 in men and HR = 1.46, 95% CI 1.19-1.79 in women), whereas leanness was associated with mortality only among smoking men (HR = 1.39, 95% CI 1.09-1.77). CONCLUSION: Refinement of BMI categories using country of origin specific quintiles demonstrated significantly increased mortality in the upper quintile in both sexes, while according to the conventional values this association did not prevail in men. We propose the establishment of sex and origin-specific BMI categories when setting goals for disease prevention.
BACKGROUND: Although obesity has been associated with a higher risk for premature death, the sex and ethnic-origin specific body mass index (BMI) levels that are associated with increased mortality are controversial. We investigated the 40-year cumulative all-cause mortality, in relation to the BMI in adult life, among men and women originating from Yemen, Europe/America, Middle East and North Africa, using sex and ethnic-origin specific BMI cut points. METHODS: A random stratified cohort (n = 5710) was sampled from the central population registry and followed since 1969 for vital status. Weight, height and blood pressure were measured, and smoking status was recorded at baseline. BMI was analysed according to conventional categories and according to sex and ethnic-origin specific quintiles. RESULTS: Elevated and significant mortality hazard ratios (HRs) of 1.21 [95% confidence interval (CI) 1.00-1.45] for women and 1.22 (95%CI 1.03-1.44) for men were found for the highest origin-specific BMI quintile. In men, the lowest ethnic-origin specific quintile was also significantly associated with increased mortality (HR of 1.22 95% CI 1.03-1.45), adjusting for age, smoking and blood pressure. Obesity was associated with mortality in non-smokers (HR = 1.29, 95% CI 1.04-1.61 in men and HR = 1.46, 95% CI 1.19-1.79 in women), whereas leanness was associated with mortality only among smoking men (HR = 1.39, 95% CI 1.09-1.77). CONCLUSION: Refinement of BMI categories using country of origin specific quintiles demonstrated significantly increased mortality in the upper quintile in both sexes, while according to the conventional values this association did not prevail in men. We propose the establishment of sex and origin-specific BMI categories when setting goals for disease prevention.