Background: Poor glucose control as indicated by elevated levels of hemoglobin A1c (HbA1c) among those with diabetes is a risk factor for stroke; however, whether HbA1c is independently associated with risk of stroke among those without diabetes remains unclear.
Methods: Among 32,826 participants in the Nurses’ Health Study who provided blood samples in 1989-1990, ischemic strokes were identified and confirmed by medical records through 2010. Participants with a history of diabetes or HbA1c ≥ 6.5% at baseline were excluded from the analyses. We conducted a nested case-control analysis of 378 cases, matched to controls (1:1) on age, menopausal status, use of postmenopausal hormones, ancestry, date of blood draw, fasting status and smoking status. The study population was categorized into quartiles based on the baseline HbA1c distribution among the controls Conditional and unconditional logistic regression models were used to estimate the multivariable adjusted (alcohol intake, body mass index [BMI], physical activity, aspirin use, C-reactive protein [CRP] levels, history of hypertension, coronary heart disease [CHD] and total/high density lipoprotein cholesterol) association of HbA1c and the risk of ischemic stroke; results were similar, thus we present results for unconditional analyses.
Results: The mean age at baseline was 60.8 years and the median time to stroke was 9.0 years. In univariate analyses, median HbA1c levels did not differ between stroke cases compared to controls (median of 5.5 for both groups; p-value=0.91). The relative risk (RR) of ischemic stroke was 0.91 (95% confidence interval (CI), 0.57 - 1.45), p-trend = 0.55, among those in the highest quartile compared to the lowest. There was no evidence of effective modification of the association between HbA1c and ischemic stroke by age, BMI, smoking, hypertension or CRP.
Conclusions: In this population of women without diabetes at baseline, HbA1c was not significantly associated with increased risk of total ischemic stroke.