Feed

Glycosylated hemoglobin (HbA1c) is a crucial biomarker in the diagnosis and management of dysglycemia. However, its diagnostic accuracy, especially considering its variations by age and sex, remains unclear. This study utilized data from Iran's 2016 Stepwise Approach to Surveillance Survey. A total of 18,729 individuals aged ≥ 25 years old without diabetes were analyzed. The validity of HbA1c for detecting prediabetes and diabetes was assessed using measures of concordance, the Kappa statistic, and the area under the curve (AUC), with fasting plasma glucose (FPG) as the reference. Additionally, optimal HbA1c thresholds for prediabetes and diabetes were determined. The AUCs of HbA1c for detecting diabetes were 0.96 (95% CI 0.95-0.97) in women and 0.94 (95% CI 0.92-0.96) in men. For prediabetes, the AUCs were 0.81 (95% CI 0.79-0.83) in women and 0.76 (95% CI 0.74-0.79) in men. Moreover, an HbA1c threshold of 6.0% was optimal in young (sensitivity: 79.1, specifity: 97.6%) and middle-aged adults (sensitivity: 89.4%, specifity: 88.5%) for detection of diabetes, while an HbA1c of 6.5% was optimal in old adults (sensitivity: 78.2%, specificity: 94.8%). In young women (aged < 40 years old), the HbA1c of 5.7% was optimal for the detection of diabetes (sensitivity = 92.9%, specificity: 91.2%). The clinical performance of HbA1c in the detection of prediabetes was higher when defining prediabetes within the FPG range of 110-125 mg/dL (sensitivity: 71.3%, specifity: 75.8%) instead of 100-125 mg/dL (sensitivity: 52.9%, specifity: 79.3%). The study highlights a higher concordance of HbA1c with FPG in women than in men. We showed that lower HbA1c thresholds are needed in the young and middle-aged adult population, especially in young women.