Fasting cut-points in determining prevalence of diabetes in an Arab population of the Middle East

We used two cross-sectional surveys involving 6356 Omanis aged >= 20 years to estimate the effect of the 1997 American Diabetes Association (ADA) criteria on the prevalence of diabetes mellitus in Oman and develop a validated optimal fasting plasma glucose (FPG) cut-point which best predicts diabetes diagnosed 2-h post oral glucose tolerance test. Applying the 1997 ADA criteria to Oman would underestimate diabetes by 18%. The sensitivity of the ADA criteria was 68.3% (95% CI 64.0-72.4%) and specificity was 98.6% (95% CI 98.2-98.9%). Receiver-operating characteristic (ROC) curve depicted FPG > 5.9 mmol/1 to best predict 2-h post-load glucose >= 11.1 mmol/1. The area under the ROC curve was 0.95 (95% CI 0.94-0.95%) with no significant difference between obese and non-obese individuals. This cut-point had a sensitivity of 87.5% (95% CI 84.3-90.3%), specificity of 90.8% (95% CI 89.9-91.7%) and likelihood ratio of 9.5. On validation in an independent population, the sensitivity and specificity of the depicted cut-point remained high 84.2% (95% Cl 77.0-89.8%) and 80.2% (95% CI 78.0-82.4%) compared to the ADA values 60.4% and 96.6%, respectively. Our study identified a lower cut-point to diagnose diabetes than that suggested by the 1997 ADA criteria. (c) 2006 Elsevier Ireland Ltd. All rights reserved.