Abstract
BACKGROUND: Development of atrial fibrillation (AF) portends an increased risk of adverse cardiovascular outcomes. Relatively little is known about the longitudinal association between urine albumin excretion and the development of AF, particularly among generally healthy older individuals. METHODS: In a post hoc analysis of the ASPirin in Reducing Events in the Elderly (ASPREE) trial, we utilized a marginal structural model to quantify the relationship between annual urine albumin:creatinine ratio (UACR) measurements and new-onset AF. Two approaches were used for handling missing data: one utilized multiple imputation and the other used only cases in whom complete information at baseline was available coupled with last observation carried forward for missing data after baseline. RESULTS: UACR data were available for 19 114 participants (mean age 75.1 ± 4.5 years, 56.4% female, 91.3% White). The median follow-up was 4.7 years. UACR values were low: mean UACR was 2.1 ± 8.2 mg/mmol at baseline and 2.1 ± 3.3 mg/mmol at year 7. AF developed in 941 individuals. The rate of AF development was 1.11/100 person-years among participants with UACR <3 mg/mmol at baseline and 1.74/100 person-years among participants with UACR ≥3 mg/mmol at baseline. After adjustment for a broad range of factors, the hazard ratio for new-onset AF was 1.16 [95% confidence interval (CI) 1.11-1.21] when multiple imputation was used and 1.15 (95% CI 1.10-1.19) when only cases with complete baseline information were used. CONCLUSIONS: In older individuals who had low levels of albuminuria, doubling of UACR was associated with a 16% increase in the risk of new-onset AF.