Blood sugar markers glycated albumin and fructosamine show moderate accuracy for diabetes diagnosis

Glycated albumin and fructosamine, two blood markers that reflect shorter-term blood sugar exposure, show moderate sensitivity and good specificity for detecting diabetes in adults when judged against the oral glucose tolerance test, according to a systematic review and meta-analysis published in BMC Endocrine Disorders. The findings support their use as adjunctive rather than replacement diagnostic tools, the authors conclude.

The review, conducted according to PRISMA-DTA guidelines, searched PubMed/MEDLINE, Scopus, Cochrane and Web of Science from inception to November 2025. The researchers included studies that evaluated glycated albumin or fructosamine against the oral glucose tolerance test or World Health Organization criteria in adults. Twenty-five unique studies were included, contributing 27 study-marker-outcome datasets.

The motivation lies in a well-known limitation of the current standard. Glycated haemoglobin, or HbA1c, is widely used for diagnosis but is unreliable in conditions affecting red blood cell turnover, including haemoglobinopathies, iron deficiency anaemia, chronic kidney disease and pregnancy. Glycated albumin and fructosamine reflect shorter-term glycaemic exposure and are independent of red blood cell lifespan, making them candidate alternatives in such settings.

For the diagnosis of diabetes, 14 studies evaluated glycated albumin in a total of 14,007 participants, and 10 studies evaluated fructosamine in 5,701 participants. Pooled sensitivity and specificity were 68.8 percent (95% CI 59.9–76.5) and 85.7 percent (95% CI 77.4–91.3) for glycated albumin, with an area under the curve of 0.83. For fructosamine, the corresponding figures were 70.3 percent (95% CI 65.9–74.3) sensitivity and 90.4 percent (95% CI 83.2–94.7) specificity, with an AUC of 0.78.

Pooled estimates were calculated using a bivariate random-effects model, with hierarchical summary receiver operating characteristic curves used to summarise overall diagnostic performance. Deeks’ funnel plot testing found no evidence of publication bias for either marker, with p values of 0.81 for glycated albumin and 0.13 for fructosamine. Sensitivity analyses that excluded studies at high risk of bias, and those that used older, non-enzymatic assay methodology, showed materially unchanged pooled estimates for both markers.

The authors note an important gap in the evidence base: no included study directly compared glycated albumin or fructosamine against HbA1c in populations where HbA1c is known to be unreliable. This, they state, remains the principal evidence gap motivating further research into the clinical role of the two markers. They also call for standardised, pre-specified diagnostic thresholds before wider clinical implementation.

That balance of moderate sensitivity and good specificity positions them as complements to, rather than substitutes for, existing diagnostic approaches in the populations studied.

The analysis was restricted to adult populations and to studies using OGTT-based or WHO-based reference standards, so the estimates apply to that diagnostic context. As with any meta-analysis of diagnostic accuracy studies, the pooled figures depend on the thresholds and assay methods used in the included studies, which the sensitivity analyses partially addressed by excluding older assay methodology and high-risk-of-bias studies without materially changing the results.

Subject of Research: Diagnostic accuracy of glycated albumin and fructosamine for detecting diabetes and prediabetes in adults compared with the oral glucose tolerance test

Article Title: 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

Article References: Fernando, K., Kottahachchi, D., Pasan, L., Rathnaweera, C., Dayanath, B., Perera, S., & Premawardhena, A. (2026). 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. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02709-w

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02709-w

Keywords: glycated albumin, fructosamine, diabetes diagnosis, prediabetes, oral glucose tolerance test, HbA1c, systematic review, meta-analysis, diagnostic accuracy, BMC Endocrine Disorders, Diagnostic, accuracy