Duration of Diabetes and Incident Heart Failure: The ARIC (Atherosclerosis Risk In Communities) Study

Abstract

ObjectivesThis study assessed the association of diabetes duration with incident heart failure (HF). BackgroundDiabetes increases HF risk. However, the independent effect of diabetes duration on incident HF is unknown. MethodsWe included 9,734 participants (mean age 63 years, 58% women, 22% Black) at ARIC (Atherosclerosis Risk In Communities) Visit 4 (1996-1998) without HF or coronary heart disease. We calculated diabetes duration at Visit 4 (baseline), utilizing diabetes status at the first 4 ARIC visits spaced 3 years apart, and self-reported diagnosis date for those with diabetes diagnosed before Visit 1. We used Cox regression to estimate associations of diabetes duration with incident HF, accounting for intercurrent coronary heart disease and other risk factors. We performed analyses stratified by age (<65 years or ≥65 years), race, sex, and glycemic control (hemoglobin A1C [HbA1C] consistently <7%, vs HbA1C ≥7%), with tests for interaction. ResultsOver 22.5 years of follow-up, there were 1,968 HF events. Compared to those without diabetes, HF risk rose with longer diabetes duration, with the highest risk among those with ≥15 y diabetes duration (HR: 2.82; 95% CI: 2.25-3.63). Each 5-year increase in diabetes duration was associated with a 17% (95% CI: 11-22) relative increase in HF risk. Similar results were observed across HF subtypes. The HF and diabetes duration associations were stronger among those aged <65 years, those with HbA1C ≥7%, those with a body mass index ≥30 kg/m2, women, and Blacks (all P interactions <0.05). ConclusionsDelaying diabetes onset may augment HF prevention efforts, and therapies to improve HF outcomes might target those with long diabetes duration.