Introduction: Despite the recognized linkage between retinal microvascular abnormalities and cerebrovascular ischemic diseases (CVD), the retinal findings’ standardization and the degree of association remain unclear. We aimed to investigate the quantitative association between various retinal microvascular signs and CVD.
Hypothesis: retinal microvascular abnormalities best predictive of CVD could be identified by synthesizing estimates of CVD risk from population-based studies.
Methods: We followed MOOSE guidelines and systematically searched 6 databases through September 2016 for studies evaluating the association between retinal microvascular abnormalities and white matter hyperintense lesions (WMHI), lacunar infarcts (LI), and cerebral infarctions (CI). Adjusted and unadjusted odds ratios (ORs) and confidence intervals were pooled into the analyses using the DerSimonian and Laird random effects model. Study quality and dissemination biases were assessed and integrated.
Results: Thirty-four prospective studies encompassing 84,144 patients were eligible for meta-analysis. After adjustment for vascular risk factors, focal arteriolar narrowing was associated with WMHI (OR, 1.24 [1.01-1.79]), LI (OR, 1.77 [1.14-2.74]), and CI (OR, 1.75 [1.14-2.69]). Venular dilation was significantly associated with LI (OR, 1.46 [1.10-1.93]), whereas retinal hemorrhage was associated with WMHI (OR, 2.23 [1.34-3.70]). Arteriovenous nicking was associated with WMHI (OR, 1.51 [1.22-1.88]) and LI (OR, 1.70 [1.05-2.76]). Any retinopathy exhibited significant association with WMHI, LI, and CI. Heterogeneity was significant (I2>50%) for all syntheses except retinal hemorrhages and WMHI, retinopathy and CI (I2=0.0%) and respectively LI (I2=45.80%).
Conclusions: Retinal hemorrhages were most highly associated with WMHI. Focal arteriolar narrowing and retinopathy were associated with WMHI, LI, and CI. Our findings may assist in the development of a noninvasive retinal microvascular tool for routine use in CVD research and clinical practice.