BackgroundVitamin D status has been hypothesized to protect against development of diabetic retinopathy via its anti-inflammatory and anti-angiogenic properties. Additionally, in vitro and in vivo studies suggest vitamin D favorably influences blood pressure and blood glucose control, strong risk factors for diabetic retinopathy. We examined the association between vitamin D status and prevalent diabetic retinopathy in participants with diabetes from a population-based cohort.MethodsAmong participants in the Atherosclerosis Risk in Communities (ARIC) study with diabetes at visit 3 (1993–1995), 1339 (906 Caucasians, 433 African Americans) had serum 25-hydroxyvitamin (25[OH]D) concentrations assessed at visit 2 (1989–1992) and nonmydriatic retinal photographs taken at visit 3. Dietary intake of vitamin D was assessed at visit 1 (1987–1989). Logistic regression was used to estimate odds ratios (ORs) and 95 % confidence intervals (CIs) for diabetic retinopathy by categories of season-adjusted 25(OH)D (<30 [referent], 30–<50, 50–<75 and ≥75 nmol/L), by quartile of vitamin D intake (IU/day), and use of vitamin D or fish oil supplements (yes/no). P for trend was estimated using continuous 25(OH)D or vitamin D intake. ORs were adjusted for race, and duration of diabetes. We further adjusted for HBA1c and hypertension to examine if 25(OH)D influenced diabetic retinopathy via its effects on either glycemic control or blood pressure.ResultsORs (95 % CIs) for retinopathy, adjusted for race and duration, were 0.77 (0.45–1.32), 0.64 (0.37–1.10), and 0.39 (0.20–0.75), p for trend = 0.001, for participants with 25(OH)D of 30–<50, 50–<75, and ≥75 nmol/L, respectively. Further adjustment for hypertension minimally influenced results (data not show), but adjustment for HBA1c attenuated the OR among those with 25(OH)D ≥75 (0.47 [0.23–0.96], p for trend = 0.030). No statistically significant association was observed between vitamin D intake from foods or supplements and retinopathy.Conclusions25(OH)D concentrations ≥75 nmol/L were associated with lower odds of any retinopathy assessed 3 years later. We speculate this may be due in part to vitamin D’s influence on blood glucose control.Electronic supplementary materialThe online version of this article (doi:10.1186/s12933-016-0434-1) contains supplementary material, which is available to authorized users.