Metabolic Predictors and Diagnostic Performance of Fundus Autofluorescence for Diabetic Macular Edema: A...

To evaluate the diagnostic performance of fundus autofluorescence (FAF) imaging for detecting diabetic macular edema (DME) using swept-source optical coherence tomography (SS-OCT) as the reference standard, and to identify metabolic factors associated with FAF detection accuracy in patients with type 2 diabetes mellitus.

This cross-sectional diagnostic accuracy study enrolled 120 eyes from 68 patients with type 2 diabetes mellitus at the Vitreoretina Subdivision, Department of Ophthalmology, RSUP Dr. Mohammad Hoesin, Palembang, Indonesia, between January 2024 and June 2025. All participants underwent comprehensive ophthalmic examination including FAF imaging and en face SS-OCT (DRI OCT Triton Plus, Topcon). Metabolic parameters including glycated hemoglobin (HbA1c), fasting lipid profile, body mass index (BMI), estimated glomerular filtration rate (eGFR), and diabetes duration were recorded. Diagnostic accuracy indices were calculated using 2×2 contingency tables. Multivariate logistic regression identified metabolic predictors of FAF detection concordance with SS-OCT.

DME was present in 78 eyes (65.0%) by SS-OCT. FAF demonstrated sensitivity of 73.1% (95% CI: 61.8-82.5%), specificity of 76.2% (95% CI: 61.5-87.2%), positive predictive value of 85.1%, negative predictive value of 60.4%, positive likelihood ratio of 3.07, negative likelihood ratio of 0.35, and diagnostic odds ratio of 8.68. HbA1c ≥8.5% (OR 3.42, 95% CI: 1.28-9.14, p=0.014), diabetes duration ≥10 years (OR 2.87, 95% CI: 1.15-7.16, p=0.024), and BMI ≥30 kg/m² (OR 2.31, 95% CI: 0.94-5.68, p=0.068) were independently associated with FAF-OCT concordance. Center-involved DME showed higher FAF sensitivity (82.4%) compared to non-center-involved DME (55.6%, p=0.012).

FAF imaging demonstrates moderate diagnostic accuracy for DME detection, with performance significantly influenced by metabolic status and DME subtype. Integration of metabolic risk stratification may optimize FAF-based screening in diabetic populations.