Glycated hemoglobin and incident type 2 diabetes in singaporean chinese adults: the Singapore Chinese health study

Abstract

Background: The American Diabetes Association recently included glycated hemoglobin in the diagnostic criteria for diabetes, but research on the utility of this biomarker in Southeast Asians is scant. The aim of this study was to evaluate the association between percent HbA1c and incident diabetes in an Asian population of adult men and women without reported diabetes.

Methods: Data analysis of 5,770 men and women enrolled in the Singapore Chinese Health Study who provided a blood sample at the follow-up I visit (1999-2004) and had no cancer and no reported history of diabetes or cardiovascular disease events. Diabetes was defined as self-report of physician diagnosis, identified at the follow-up II visit (2006-2010).

Results: Hazard ratios (and 95% confidence intervals) for incident diabetes by 5 categories of HbA1c were estimated with Cox regression models and continuous HbA1c with cubic spline analysis. Compared to individuals with an HbA1c ≤ 5.7% (≤39 mmol/mol), individuals with HbA1c 5.8-5.9% (40-41 mmol/mol), 6.0-6.1% (42-43 mmol/mol), 6.2-6.4% (44-47 mmol/mol), and ≥ 6.5% (≥48 mmol/mol) had significantly increased risk for incident diabetes during follow-up. In cubic spline analysis, levels below 5.7% HbA1c were not significantly associated with incident diabetes.

Conclusions: Our study found a strong and graded association with HbA1c 5.8% and above with incident diabetes in Chinese men and women.

Fig 1. Timeline and assembly of sub-sample, The Singapore Chinese Health Study.

Fig 2. Cubic spline of the association between glycated hemoglobin (HbA1c %) and self-reported incident diabetes, adjusted for age, gender, and dialect, The Singapore Chinese Health Study.

The hazard ratio (HR) is per each absolute increase of 1 percentage point in the glycated hemoglobin value at baseline. The shaded area is the 95% confidence interval from the restricted-cubic-spline model. The HR scale is logarithmic, the model is centered at the mean/median (5.7%), and the plot was truncated at the 5th and 95th percentiles of glycated hemoglobin (4.9% and 6.5%, respectively). The hazard ratio was adjusted for age, gender, and dialect.