The association of retinal vessel calibres with heart failure and long-term alterations in cardiac structure and function: the Atherosclerosis Risk in Communities (ARIC) Study.

Abstract

Narrower retinal arterioles and wider retinal venules have been associated with macrovascular forms of cardiovascular disease (CVD). However, whether they are predictive of the development of heart failure (HF) independent of atherosclerotic CVD is unclear. We aimed to describe long-term associations of retinal vessel calibres with incident HF, in those with and without prevalent macrovascular disease, and how they relate to cardiac structure and function. This analysis included Atherosclerosis Risk in Communities Study participants who underwent retinal photography between 1993 and 1995. HF outcomes were followed in these participants until the end of 2013. Returning participants underwent echocardiography between 2011 and 2013. Participants with retinal vessel measurements who were free of CVD at baseline (n=10692) were followed for a mean of 16 years (baseline mean age 60 ± 6 years). Wider central retinal venular equivalent (CRVE) and narrower central retinal arteriolar equivalent (CRAE), adjusted for age, gender, and race, were significantly linearly associated with incident HF; however, a non-linear association was detected with CRVE and incident HF (P-value for overall trend < 0.001; P-value for non-linearity=0.002). After adjustment with clinical risk factors, CRVE association with incident HF remained significant (P-value for overall trend=0.025). Adjusted for age, gender, and race, CRVE widening and CRAE narrowing were associated with larger left ventricular size, higher prevalence of left ventricular hypertrophy, and worse measures of diastolic and systolic function. Retinal vessel calibre imaging, which characterizes retinal microvasculature, is a simple, non-invasive test that predicts incident HF and adverse cardiac structure/function 18 years in the future, thereby providing insight into systemic cardiovascular health.