Cut-off values of fasting and post-load plasma glucose and HbA1c for predicting Type 2 diabetes in community-dwelling Japanese subjects: the Hisayama Study.

AIMS: We examined the optimal cut-off values of fasting plasma glucose, 2-h post-load glucose and HbA(1c) for predicting Type 2 diabetes in community-dwelling Japanese subjects.

METHODS: A total of 1982 subjects without diabetes aged 40-79 years who underwent a 75-g oral glucose tolerance test were followed prospectively for 14 years by annual health examination.

RESULTS: During the follow-up, 295 subjects developed Type 2 diabetes. Compared with the first decile, the crude hazard ratio for incident Type 2 diabetes was significantly higher in the fifth fasting plasma glucose decile [5.4-5.4 mmol/l (97-98 mg/dl)] or higher, in the seventh 2-h post-load glucose decile [6.9-7.2 mmol/l (124-131 mg/dl)] or higher, and in the fifth HbA(1c) decile [34-36 mmol/mol (5.3-5.4%)] or higher. These associations remained substantially unchanged even after adjustment for confounding factors. The receiver operating characteristic curve analysis showed that the optimal cut-off values for predicting Type 2 diabetes were 5.6 mmol/l (101 mg/dl) for fasting plasma glucose, 6.9 mmol/l (124 mg/dl) for 2-h post-load glucose and 37 mmol/mol (5.5%) for HbA(1c). In a stratified analysis, the cut-off values were approximately 5.6 mmol/l (101 mg/dl) for fasting plasma glucose and 37 mmol/mol (5.5%) for HbA(1c), and these values were unchanged over BMI quartile levels, whereas the 2-h post-load glucose cut-off values declined with decreasing BMI levels.

CONCLUSIONS: Our findings suggest that the cut-off value for predicting Type 2 diabetes in the Japanese population is 5.6 mmol/l (101 mg/dl) for fasting plasma glucose and 37 mmol/mol (5.5%) for HbA(1c), while the 2-h post-load glucose cut-off value is lower than the diagnostic criterion for impaired glucose tolerance.