Aim: To determine the incidence of Type 2 diabetes and to examine the effect of different cut-points for impaired fasting glucose (IFG) on diabetes incidence.
Methods: Population-based longitudinal study (1990-2000) with clinical, anthropometric and biochemical measurements, including an oral glucose tolerance test (OGTT), in 1040 non-diabetic adults aged 40-69 years at baseline. Baseline glucose status was defined as normoglycaemia < 5.6, IFG-lower 5.6-6.0 and IFG-original 6.1-6.9 mmol/l. The all-IFG group included fasting glucose values of 5.6-6.9 mmol/l.
Results: The 10-year cumulative incidence of diabetes was 7.3 per 1000 person-years. Diabetes incidence was 2.4 [95% confidence interval (CI) 1.2, 4.8], 6.2 (4.0, 9.8) and 17.5 (12.5, 24.5) per 1000 person-years in those with normoglycaemia, IFG-lower and IFG-original, respectively. Compared with normoglycaemia, the age/sex-adjusted risk [hazard ratio (HR) and 95% CI] for incident diabetes was greatest in the IFG-original category (HR 6.9; 3.1, 15.2) and increased to a lesser degree in the IFG-lower (HR 2.5; 1.1, 5.7) and all-IFG categories (HR 4.1; 1.9, 8.7). When adjusted for confounding factors, the magnitude and direction of associations persisted, with HR 1.9, 4.4 and 2.9, for the categories IFG-lower, IFG-original and all-IFG, respectively.
Conclusions: Diabetes incidence is more strongly related to IFG defined as fasting glucose between 6.1 and 6.9 mmol/l than to the lower category of 5.6-6.0 mmol/l, or entire range of 5.6-6.9 mmol/l. Future studies should examine the association of IFG with cardiovascular outcomes, but for diabetes risk our study supports the use of the IFG cut-point at 6.1 mmol/l.