Objective: In 1997, the American Diabetes Association (ADA) recommended a new diagnostic category, impaired fasting glucose (IFG), to describe individuals with borderline glucose tolerance. On the other hand, the World Health Organization (WHO) suggested retaining the category of impaired glucose tolerance (IGT). We studied the prevalence of IFG and IGT in a multiethnic society and compared the cardiovascular risk profiles of subjects with IFG, IGT, or both IFG and IGT.
Research design and methods: A total of 3,568 subjects were examined from the 1992 National Health Survey of Singapore, which involved a combination of disproportionately stratified sampling and systematic sampling. Anthropometric, blood pressure, insulin, lipid profile, and uric acid measurements were taken, and a standard 75-g oral glucose tolerance test was performed after a 10-h overnight fast.
Results: The prevalence rates of IFG only, IGT only, and both IFT and IGT were 3.45, 10.2, and 3.4%, respectively. The degree of agreement (kappa) between the two diagnostic criteria (the ADA IFG and the WHO IGT) was only 0.25. A fasting glucose level of 5.5 mmol/l was the optimal cutoff for predicting a 2-h postload glucose level of > or =7.8 mmol/l. The following cardiovascular risk factors were higher in subjects with both IFG and IGT compared with those with either IFG or IGT alone: systolic blood pressure (131 +/- 20 vs. 125 +/- 21 and 125 +/- 19 mmHg, respectively; P < 0.05 and P < 0.001, respectively); diastolic blood pressure (77 +/- 12 vs. 73 +/- 12 and 74 +/- 12 mmHg, respectively; P < 0.05); BMI (26.2 +/- 4.2 vs. 24.4 +/- 4.0 and 24.6 +/- 4.4 kg/m2, respectively; P < 0.01 and P < 0.001, respectively); waist circumference (84.1 +/- 10.3 vs. 79.3 +/- 10.7 and 79.3 +/- 10.6 cm, respectively; P < 0.001); waist-to-hip ratio (0.84 +/- 0.08 vs. 0.82 +/- 0.09 and 0.81 +/- 0.08, respectively; P < 0.05 and P < 0.001, respectively); fasting insulin (12.1 +/- 9.7 vs. 9.2 +/- 5.3 and 9.9 +/- 7.7 mU/l; P < 0.01); insulin resistance (by homeostasis model assessment [HOMA]) (3.41 +/- 2.77 vs. 2.58 +/- 1.50 and 2.43 +/- 1.83, respectively; P < 0.01 and P < 0.001, respectively); total cholesterol (5.81 +/- 1.1 vs. 5.51 +/- 1.1 and 5.53 +/- 1.1 mmol/l, respectively; P < 0.05) and apolipoprotein(B) [apo(B)] (1.5 +/- 0.38 vs. 1.40 +/- 0.34 and 1.39 +/- 0.35 mmol/l, respectively; P < 0.01). The pattern of difference remained significant only for fasting insulin, insulin resistance (HOMA), and apo(B) (borderline) after adjustment for age, sex, and ethnic differences.
Conclusions: Obvious discordance was evident in the classification of glycemic status when applying the criteria proposed by the ADA (IFG) or WHO (IGT) in a multiethnic society like Singapore. However, subjects with either IFG or IGT had similar cardiovascular risk profiles. Therefore, both criteria identified individuals at high risk for cardiovascular disease. Individuals with both IFG and IGT had a greater incidence of the cardiovascular dysmetabolic syndrome.