Abstract 14433: Measures of Adiposity and Future Risk of Ischemic Stroke and Coronary Heart Disease in Older Men and Women: The Cardiovascular Health Study
Background: While obesity is an adverse prognostic factor in middle age, the association between body mass index (BMI) and mortality weakens in older persons. The relationship between measures of general or central adiposity and individual cardiovascular endpoints remains under-studied in older adults.
Methods:We investigated the association of measures of body size and composition with first incident ischemic stroke or coronary heart disease (CHD) in 3,754 community-dwelling adults ages 65-100 participating in the Cardiovascular Health Study. Standardized anthropometric measurements for general (BMI) and regional adiposity (waist circumference, hip circumference, waist-hip ratio [WHR], and waist-height ratio), together with determinations of fat mass and fat-free mass (FFM) by bioelectric impedance, were obtained at baseline. BMI at age 50 (BMI50) was calculated based on recalled weight. Cox proportional hazards models adjusted first for potential confounders (age, sex, race, smoking, alcohol, physical activity, education, estrogen replacement [women], serum creatinine) and then for putative mediators (systolic blood pressure, antihypertensive treatment, pre-diabetes, diabetes, LDL, HDL, triglycerides, aspirin use, CRP).
Results: During median follow-up of 14 years, 490 ischemic strokes and 845 CHD events occurred. Only waist-hip ratio (WHR) was significantly associated with ischemic stroke, with a heightened risk observed for increasing WHR quintiles (Q) in women only (ptrend=0.006) (Q5 vs. Q1, hazard ratio [HR] 1.71, 95% confidence interval [CI] 1.19-2.46). For CHD, there were significant positive associations for all measures except FFM after adjusting for potential confounders, without an interaction with sex. The strongest relationships were for WHR (Q5 vs. Q1: HR 1.56, 95% CI 1.25-1.94, ptrend<0.001) and BMI50 (Q5 vs. Q1: HR 1.71, 95% CI 1.37-2.14, ptrend<0.001), both of which remained significantly associated with CHD after adjusting for potential mediators, for baseline BMI or for each other.
Conclusions: These findings show WHR to be the preferred anthropometric measure for assessing cardiovascular risk in older adults, and demonstrate that mid-life BMI is a stronger predictor of CHD than BMI in later life.
- © 2010 by American Heart Association, Inc.