Wong TY, Klein R, Couper DJ, et al. Retinal microvascular abnormalities and incident stroke: the Atherosclerosis Risk in Communities Study. Lancet 2001;358:1134–40. Wong et al. obtained undilated retinal photographs in 10,358 men and women living in North Carolina, Mississippi, Minnesota, and Maryland. They followed the patients for an average of 3.5 years (range 3–5 years), and observed the frequency of stroke (ischemic or hemorrhagic), stratifying it with respect to retinal vessel abnormalities, predisposing systemic diseases (diabetes, hypertension, coronary artery disease), and other risk factors (smoking, alcohol use, cholesterol). There were 110 patients with stroke (96 ischemic, 13 hemorrhagic, and one ischemic/hemorrhagic). After adjusting for age, gender, race, and locality, each of the retinal microvascular lesions observed increased the relative risk of stroke above baseline (microaneurysm: RR 5.0 [95% CI 3.0–8.2], soft exudates: 5.6 [2.8–89], blot hemorrhage: 4.9 [2.7–8.8], flame hemorrhage: 4.6 [2.1–9.9], arteriovenous nicking: 1.9 [1.3–3.0]). Much of this could be expected because these lesions are commonly associated with hypertension, diabetes, and other vascular diseases, and these diseases are all risk factors for stroke. Surprisingly, when a multivariate analysis was done, adjusting for blood pressure, the use of antihypertensive medications, the presence of diabetes, smoking status, fasting glucose, and blood lipid measurements, each of the retinal abnormalities was still a significant risk factor for stroke, although to a lesser degree. A stratified analysis using the presence of diabetes and hypertension also showed that the retinal abnormalities were independent predictors of stroke, although the small number of patients in some of the groups meant that the confidence limits were wide. The biggest potential flaw of this study is that the confounding variables (presence of hypertension and other vascular diseases may not have been completely eliminated in the multivariate analyses. Nonetheless, an important implication of this study is that observation of a retinal microvascular abnormality in a patient should not only prompt a work-up for vascular disease, but even if no disease is present, suggest that the patient is at an increased risk of stroke. The retinal abnormalities may be markers of small vessel disease and may be indicative of vascular pathologic processes occurring in the cerebral circulation.