Comparative Diagnostic Performance of Short-Term Glycemic Markers and HbA1c for Diabetes and Prediabetes: A Systematic Review

Abstract

Background: Glycated hemoglobin (HbA1c) is widely used for the diagnosis and monitoring of diabetes, but its diagnostic performance may be affected by conditions such as altered erythrocyte lifespan, hemoglobinopathies, chronic kidney disease, and pregnancy. Short-term glycemic markers, including glycated albumin, fructosamine, and 1,5-anhydroglucitol (1,5-AG), have been proposed as alternative or complementary biomarkers, particularly in settings where HbA1c may be less reliable. This systematic review synthesized evidence on the comparative diagnostic performance of HbA1c and short-term glycemic markers for the detection of diabetes and prediabetes. Methods: A systematic review was conducted in accordance with PRISMA 2020 guidelines. Electronic databases searched included PubMed/MEDLINE, Scopus, Web of Science, Embase, the Cochrane Library, and Google Scholar for studies published in English since January 2011. Eligible studies evaluated HbA1c and/or at least one short-term glycemic marker against recognized glucose-based reference standards such as fasting plasma glucose or oral glucose tolerance testing. Data on study characteristics, index tests, reference standards, and reported diagnostic performance measures were extracted and synthesized narratively because of heterogeneity in study populations, assay methods, and diagnostic thresholds. Results: Nineteen studies met the inclusion criteria. HbA1c was the most extensively evaluated biomarker and generally demonstrated high specificity but variable sensitivity for diabetes detection across populations. Reported HbA1c sensitivity ranged from 45.5% to 89.4%, specificity from 80.0% to 100%, and area under the curve (AUC) from 0.73 to 0.970. Glycated albumin showed moderate to good diagnostic performance, with AUC values ranging from 0.550 to 0.951, sensitivity from 27.3% to 77.5%, and specificity from 65.5% to 99.2%. 1,5-AG demonstrated variable performance depending on population and clinical context, with AUC values up to 0.850. Fructosamine was less frequently studied but showed promising performance in selected populations, particularly where HbA1c interpretation may be limited, such as in children with β-thalassemia major. Several studies also suggested that combining biomarkers improved diagnostic sensitivity and case detection compared with single-test approaches. Conclusion: HbA1c remains the most established and clinically useful biomarker for diabetes diagnosis because of its convenience, standardization, and generally high specificity. However, its sensitivity is inconsistent across populations and clinical settings. Short-term glycemic markers, particularly glycated albumin and fructosamine, may provide complementary diagnostic value, especially in conditions where HbA1c is less reliable or where recent glycemic exposure is of interest. A multimarker approach may improve the detection of diabetes and prediabetes in selected clinical contexts.