Retinal Vasculature, Plasma Metabolites and Cardiovascular Disease Risk: A Prospective Cohort Study from the Canadian Longitudinal Study on Aging
Indzhova, S. I.; Britz-McKibbin, P.; Rafiq, T.; Joshi, D.; Trucco, E.; Mannan, F.; Azab, S. M.; Chong, M.; Pigeyre, M.
Show abstract
BackgroundThis study aimed to investigate whether retinal fractal dimension (Df), a measure of microvascular branching complexity from fundus images, together with plasma metabolites, can help identify pathways linking microvascular changes to cardiovascular disease (CVD) events. MethodsWe analyzed longitudinal data from a subset of 4,781 participants from the Comprehensive cohort of the Canadian Longitudinal Study on Aging (CLSA), free of CVD at baseline (mean age 58.74 {+/-} 8.39; 47% male), and with 811 plasma metabolites measured at baseline and retinal imaging. Total, arterial and venular Df were derived from fundus photographs using the Vessel Assessment and Measurement Platform for Images of the Retina. Incident CVD was defined as one or more of self-reported physician-diagnosed myocardial infarction, angina, coronary heart disease, stroke, transient ischemic attack, or peripheral vascular disease during follow-up. Regression models tested associations among Df, plasma metabolites and incident CVD. ResultsOver a median follow-up of 5.75 years (IQR 5.48-6.04), 546 participants (11.42%) developed CVD. Higher total, arterial and venular Df were associated with lower CVD risk in unadjusted analyses (Odds Ratio (OR)=0.62, 95% CI:0.53-0.72 for total Df; OR=0.78, 95% CI:0.70-0.86 for arterial Df; OR=0.74, 95% CI:0.67-0.82 for venular Df). Total Df demonstrated the strongest predictive value for incident CVD but not independently of established CVD risk factors. Nine plasma metabolites, including amino acids, lipids, and xenobiotics, were associated with both incident CVD and one Df measure (p < 0.05), with cotinine and hydroxycotinine satisfying a false discovery rate adjustment (q < 0.05). Consistent with these findings the Total Nicotine Equivalent (TNE-3) was also associated with both lower arterial Df and increased CVD risk. ConclusionsRetinal microvascular complexity is associated with incident CVD. Nicotine metabolism from tobacco smoking exposures emerged as the strongest association linking microvascular changes and CVD events. Clinical PerspectiveO_ST_ABSWhat is New?C_ST_ABSO_LILower retinal branching complexity, as measured through Df, is significantly associated with incident cardiovascular disease (CVD) in unadjusted analyses. C_LIO_LIIn a population-based sample, we found nine plasma metabolites associated with both retinal vascular complexity and incident CVD, but only nicotine metabolites were significant after multiple testing correction. C_LIO_LINicotine metabolites, particularly cotinine and hydroxycotinine, remained significantly associated with incident CVD even after adjustment for self-reported smoking status. C_LI What Are the Clinical Implications?O_LIRetinal Df measures could be used as a predictor of CVD event, although not independently of age and traditional cardiovascular risk factors. C_LIO_LIMicrovascular changes may lie on the pathway linking nicotine metabolites to CVD events. C_LIO_LIPlasma nicotine metabolites may provide additional cardiovascular risk information beyond self-reported smoking, reflecting individual exposure, metabolism and passive smoke exposure. C_LI
Matching journals
The top 15 journals account for 50% of the predicted probability mass.