Diagnostic accuracy of glycated albumin and fructosamine for detecting diabetes and prediabetes in adults compared with the oral glucose tolerance test: a systematic review and meta-analysis

Glycated haemoglobin (HbA1c) is widely used for diagnosis, but is unreliable in conditions affecting erythrocyte turnover, including hemoglobinopathies, iron deficiency anaemia, chronic kidney disease, and pregnancy. Glycated albumin (GA) and fructosamine (FA) reflect shorter-term glycaemic exposure and are independent of red blood cell lifespan. This study aims to evaluate the diagnostic accuracy of GA and FA for detecting diabetes and prediabetes in adults using the oral glucose tolerance test (OGTT) as the reference standard. A systematic review and meta-analysis following PRISMA-DTA guidelines was conducted. PubMed/MEDLINE, Scopus, Cochrane and Web of Science were searched from inception to November 2025. Studies evaluating GA or FA against OGTT or WHO criteria in adults were included. Pooled sensitivity and specificity were estimated using a bivariate random-effects model and hierarchical summary receiver operating characteristic (HSROC) curves. Sensitivity analyses excluding studies at high risk of bias and those using older, non-enzymatic assay methodology were performed, alongside Deeks’ funnel plot testing for small-study effects. Twenty-five unique studies were included, contributing 27 study-marker-outcome datasets. For the diagnosis of diabetes, 14 studies evaluated GA ( n = 14,007) and 10 evaluated FA ( n = 5,701). Pooled sensitivity and specificity were 68.8% (95% CI 59.9–76.5%) and 85.7% (95% CI 77.4–91.3%) for GA (AUC 0.83), and 70.3% (95% CI 65.9–74.3%) and 90.4% (95% CI 83.2–94.7%) for FA (AUC 0.78). No evidence of publication bias was found for either marker (Deeks’ test: GA p = 0.81; FA p = 0.13). Sensitivity analyses excluding studies at high risk of bias and those using older assay methodology showed materially unchanged pooled estimates for both markers. GA and FA demonstrate moderate sensitivity and good specificity for diagnosing diabetes against an OGTT/WHO-based reference standard, supporting their use as adjunctive rather than replacement diagnostic tools. No included study directly compared GA or FA against HbA1c in populations where HbA1c is known to be unreliable; this remains the principal evidence gap motivating further research into their clinical role. Standardised, pre-specified diagnostic thresholds are needed before wider clinical implementation.