Abstract
BACKGROUND AND AIMS: Body adiposity is known to affect mortality risk in patients with coronary artery disease (CAD). We examined associations of body mass index (BMI) and waist circumference (WC) with long term mortality in Dutch CAD patients, and potential and effect modification of these associations by lifestyle and health determinants. METHODS: 10,370 CAD patients (mean age ∼65 y; 20% female; >80% on cardiovascular drugs) from the prospective Alpha Omega Cohort and Utrecht Cardiovascular Cohort-Secondary Manifestations of ARTerial disease study were included. Cox models were used to estimate categorical and continuous associations (using restricted cubic splines) of measured BMI and WC with all-cause and cardiovascular mortality risk, adjusting for age, sex, smoking, alcohol, physical activity and educational level. Analyses were repeated in subgroups of lifestyle factors (smoking, physical activity, diet quality), education and health determinants (diabetes, self-rated health). RESULTS: During ∼10 years of follow-up (91,947 person-years), 3,553 deaths occurred, including 1,620 from cardiovascular disease. U-shaped relationships were found for BMI and mortality risk, with the lowest risk for overweight patients (BMI ∼27 kg/m2). For obesity (BMI ≥30), the HR for all-cause mortality was 1.31 (95% CI: 1.11, 1.41) in male patients and 1.10 (95% CI: 0.92, 1.30) in female patients, compared to BMI 25-30 kg/m2. WC was also non-linearly associated with mortality, and HRs were 1.18 (95%CI:1.06, 1.30) in males and 1.31 (95%CI:1.05, 1.64) in females for the highest vs. middle category of WC. Results for cardiovascular mortality were mostly in line with the results for all-cause mortality. U-shaped associations were found in most subgroups, associations were moderately modified by physical activity, smoking and educational level. CONCLUSIONS: CAD patients with obesity and a large WC were at increased risk of long-term CVD and all-cause mortality, while mildly overweight patients had the lowest risk. These associations were consistent across subgroups of patients with different lifestyles and health status.