Introduction: African origin coding variants in APOL1, encoding apolipoprotein L1, are strongly associated with various kidney diseases in African Americans, with odds ratio ranging from 7 to 89 for two renal risk allele carriers. APOL1 renal risk genotypes may influence risk for cardiovascular disease and mortality, but findings have been inconsistent.
Hypothesis: Given that African-American aging women are at high risk for cardiovascular disease and stroke, we asked whether APOL1 risk genotypes affect these outcomes.
Methods: We used data from 11,137 African-American postmenopausal women who participated in the Women’s Health Initiative prospective cohort, followed from enrollment (1993-1998) to 2014. APOL1 genotypes were obtained directly or by imputation from whole exome sequencing data. Adjudicated outcomes were incident coronary heart disease, stroke and subtypes, heart failure and subtypes, and overall and cause-specific mortality obtained from hospital records and death certificates. End-stage renal disease status (ESRD) was obtained from the US Renal Data System. We estimated incident rates for each outcome and used Cox proportional hazard to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the associations of APOL1 high and low risk groups with outcomes.
Results: Mean age was 61.7 years. APOL1 high risk carriers (12.3% of participants) had higher prevalence of hypertension, use of cholesterol lowering medication, and reduced estimated glomerular filtration rate (eGFR), defined as <60 mL/min/1.73 m2. After 11 years mean follow-up APOL1 high risk subjects had a higher incident rate of ESRD and hospitalized heart failure with preserved ejection fraction (HFpEF) than low risk carriers, but showed no difference in the incident rates of other outcomes. In adjusted models, there were a significant 43% increased hazard of ESRD (95% CI 1.01, 2.02), and a 58% increased hazard of hospitalized HFpEF (95% CI 1.03, 2.41) among high risk compared to low risk APOL1 carriers. In sensitivity analyses restricted to the genome-wide association sample and accounting for population stratification (n=7,797), the associations remained significant for HFpEF. These associations were no longer significant after adjusting for eGFR.
Conclusions: We identified novel associations of APOL1 high risk status with hospitalization for HFpEF among postmenopausal women, which are likely mediated by APOL1-induced chronic kidney disease. We also showed lack of association of APOL1 with incident coronary heart disease, stroke and mortality.