HbA1c and peripheral arterial disease in diabetes: the Atherosclerosis Risk in Communities study.

OBJECTIVE To assess the relation between HbA(1c) (A1C) and incident peripheral arterial disease (PAD) in a community-based cohort of diabetic adults from the Atherosclerosis Risk in Communities (ARIC) study. A second aim was to investigate whether the association was stronger for severe, symptomatic disease compared with PAD assessed by low ankle-brachial index (ABI). RESEARCH DESIGN AND METHODS This was a prospective cohort study of 1,894 individuals with diabetes using ARIC visit 2 as baseline (1990-1992) with follow-up for incident PAD through 2002. We assessed the relation between A1C and incident PAD, defined by intermittent claudication, PAD-related hospitalization, or a low ABI (<0.9). RESULTS During a mean follow-up of 9.8 years, the crude incidence rates were 2.1 per 1,000 person-years for intermittent claudication (n = 41), 2.9 per 1,000 person-years for PAD-related hospitalization (n = 57), and 18.9 per 1,000 person-years for low ABI at visit 3 or 4 (n = 123). The relative risk (RR) (95% CI) of an incident PAD event comparing the second and third tertiles of A1C to the first, respectively, after adjustment for cardiovascular risk factors was strongest for severe, symptomatic forms of disease, e.g., PAD-related hospitalization (RR = 4.56 [1.86-11.18] for the third A1C tertile compared with the first, P trend <0.001) than for low ABI (RR = 1.64 [0.94-2.87], P trend = 0.08). CONCLUSIONS We found a positive, graded, and independent association between A1C and PAD risk in diabetic adults. This association was stronger for clinical (symptomatic) PAD, whose manifestations may be related to microvascular insufficiency, than for low ABI. Our results suggest that efforts to improve glycemic control in persons with diabetes may substantially reduce the risk of PAD.