HbA1cin relation to incident diabetes and diabetes-related complications in non-diabetic adults at baseline

Aims We compared the utility of glycated hemoglobin (HbA1c) and oral glucose tolerance (oGTT) in non-diabetic patients for identifying incident diabetes; all-cause mortality; cardiovascular disease (CVD) mortality; CVD, coronary heart disease (CHD), and ischemic stroke events; and diabetes microvascular complications. Methods Data from a New Zealand community setting were prospectively linked to hospitalization, mortality, pharmaceutical and laboratory test results data. After applying exclusion criteria (prior laboratory diagnosis or history of drug treatment for diabetes or hospitalization for diabetes or CVD event), there were 31,148 adults who had an HbA1cand 2-h 75g oGTT. HbA1cwas measured by ion-exchange high-performance liquid chromatography, and glucose using a commercial enzymatic method. We compared glycemic measures and outcomes using multivariable Cox proportional hazards regression. Results The median follow-up time was 4years (range 0 to 13). The mean age was 57·6years and 53·0% were male. After adjusting for other glycemic measures (fasting glucose, 2-h glucose and/or HbA1cwhere relevant) in addition to age, sex, ethnicity and smoking habit, the hazard ratios for incident diabetes and diabetes complications of retinopathy and nephropathy were highest for 2-h glucose levels, followed by HbA1cand lastly by fasting glucose. However, all-cause mortality and CHD were significantly associated with HbA1cconcentrations only, and ischemic stroke and CVD events with 2-h glucose only. Circulatory complications showed a stronger association with HbA1c. Conclusion Apart from neuropathy, HbA1cshowed stronger associations with outcomes compared to fasting glucose and provides a convenient alternative to an oGTT.