Aims/hypothesis: The American Diabetes Association recommended that only a single fasting plasma glucose of greater than or equal to 7.0 mmol/l should be used for diagnosing diabetes in epidemiological studies and did not recommend using a 2-h oral glucose tolerance test. We evaluated the effect of diagnostic changes on the prevalence of diabetes and on the choice of subjects diagnosed with diabetes.
Methods: Existing epidemiological data collected from Asian people between 30 and 89 years of age, was re-analysed separately in 11 population-based studies (n = 17,666), 6 pre-selected hyperglycaemic cohorts (n = 12,221) and one suspected diabetic cohort (n = 8382).
Results: Among the 11 population-based studies, the new fasting glucose criteria resulted in an overall reduction of 1.8% in the prevalence of diabetes, which ranged from a reduction of 4.8% to an increase of 1.7% in the different studies. Of 1215 subjects diagnosed with diabetes by either criteria, only 449 met both criteria, a concordance of 37%. More than half of the diabetic subjects had isolated post-challenge hyperglycaemia and three quarters of the subjects with impaired glucose tolerance, according to the 2-h glucose criteria, were normal according to the fasting glucose criteria. Subjects diagnosed as diabetic based only on the 2-h glucose criteria were, on average, older than those with diabetes according to the fasting criteria.
Conclusion/interpretation: The fasting and the 2-h glucose criteria diagnose different groups of subjects. It would therefore be inappropriate to use the fasting glucose criteria alone for screening diabetes in Asian populations.