Introduction: 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, all-CVD events, coronary heart disease (CHD) events, ischaemic stroke events and diabetes microvascular complications.
Hypothesis: HbA1c will provide the similar information to an oGTT.
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 hospitalisation for diabetes or CVD event), there were 31,148 adults who had an HbA1c and 2-hr 75g oGTT. HbA1c was measured by ion-exchange high-performance liquid chromatography, and glucose using a commercial enzymatic method. We compared glycaemic measures and outcomes using multivariable Cox proportional hazards regression.
Results: The median follow-up time was 4 years (range 0 to 13). The mean age was 57.6 years and 53.0% were male. After adjusting for other glycaemic measures (fasting glucose, 2-hr glucose and/or HbA1c where relevant) in addition to age, sex and ethnicity, the hazard ratios for incident diabetes, and diabetes complications of retinopathy, nephropathy and neuropathy were highest for 2-hr postchallenge glucose levels, followed by HbA1c and lastly by fasting plasma glucose. However, all-cause mortality was significantly associated with HbA1c concentrations only. CHD events were most strongly associated with HbA1c, followed by 2-hr glucose and circulatory complications showed stronger associations with HbA1c.
Conclusion: HbA1c showed stronger associations with adverse outcomes compared to fasting glucose and provides a convenient alternative to an oGTT.