Glycated hemoglobin in diagnosis of diabetes mellitus and pre-diabetes; validation by oral glucose tolerance test. The Tromsø OGTT Study.

BACKGROUND: Glycated hemoglobin (HbA(1c)) 6.5% has recently been recommended by the World Health Organization (WHO) and the American Diabetes Association (ADA) as an alternative diagnostic criterion for diabetes mellitus (DM). AIM: To evaluate HbA(1c) as an alternative to oral glucose tolerance test (OGTT) for diagnosis of DM and pre-diabetes and to find the optimal HbA(1c) cut-off points for DM and pre-diabetes in our population. SUBJECTS AND METHODS: The subjects were recruited from the Tromsø Study, performed for the 6th time in 2007-2008 with 12,984 participants. All subjects with HbA(1c) in the range 5.8-6.9% and a random sample of subjects with levels 5.3-5.7% were invited to an OGTT. RESULTS: Among 3476 subjects who completed the OGTT, 199 were diagnosed with DM. The best sensitivity (69.8%) and specificity (81.8%) were found at HbA(1c) 6.2%. For HbA(1c) 6.5% we found a sensitivity of 34.7% and specificity 97.1%. The best cut-off points for impaired fasting glucose (no.=314) and impaired glucose tolerance (no.=404) were found at HbA(1c) 5.9% and 6.0%, respectively. Pre-diabetes detected only by OGTT was associated with worse metabolic characteristics than pre-diabetes detected only by HbA(1c). CONCLUSIONS: The optimum HbA(1c) cutoff point for DM in our population was lower than that proposed by WHO and ADA. To establish more precisely the HbA(1c) levels predictive of micro- and macro-vascular complications, long-term prospective studies are needed. Population- specific optimum cut-off points may be necessary.