Purpose: Retinal microvascular abnormalities reflect persistent arteriolar damage from hypertension and independently predict stroke. We examined their associations with long-term cardiovascular mortality.
Design: Population-based, nested, case-control study.
Population: Cases were Beaver Dam Eye Study participants (age range, 43-84 years) who died of coronary heart disease or stroke between the baseline examination in 1988 to 1990 and 1999 (n = 413). Nearly 3 controls per case were selected from the baseline cohort, frequency-matched on 5-year age intervals and gender (n = 1198).
Methods: Retinal photographs of cases and controls at baseline were evaluated for retinopathy, focal arteriolar narrowing, and arteriovenous nicking by graders masked to case-control status using standardized protocols. To obtain an estimate of generalized arteriolar narrowing, photographs were digitized and diameters of individual retinal vessels were measured and summarized by a computer program.
Main outcome measure: Ten-year cardiovascular mortality.
Results: After controlling for systolic blood pressure, diabetes, glycosylated hemoglobin levels, and other risk factors, retinopathy was associated with increased cardiovascular mortality, with odds ratios of 1.8 (95% confidence interval [CI], 1.2, 2.7). For other retinal abnormalities, associations with cardiovascular mortality were present only in younger people, with odds ratios of 2.7 (95% CI, 1.0, 7.4) for focal arteriolar narrowing, 1.8 (95% CI, 0.8, 4.5) for arteriovenous nicking, and 1.9 (95% CI, 1.2, 2.9) for generalized arteriolar narrowing in persons 43 to 74 years of age but odds ratios of 1.1, 0.4, and 1.0 for the corresponding retinal abnormalities in persons 75 years and older.
Conclusions: Retinopathy is independently associated with cardiovascular mortality. Associations for other retinal abnormalities were only observed in middle-aged persons. These data support recent studies that suggest retinal microvascular abnormalities provide independent information regarding cardiovascular risk.