Noncoding RNAs as novel diagnostic biomarkers for diabetic retinopathy: A systematic review and meta-analysis.

Abstract

Background and aims: Diabetic retinopathy (DR), a frequent microvascular complication of diabetes, requires early detection to mitigate progression. Noncoding RNAs (ncRNAs) have emerged as promising diagnostic biomarkers. This meta-analysis evaluates their overall diagnostic accuracy for DR.

Methods: A comprehensive literature search was conducted across multiple electronic databases, including PubMed, Embase, and Web of Science, to identify relevant studies published until August 2025. Data on study characteristics, quality, and diagnostic performance were pooled. Pooled effect sizes were calculated using a random effects model. Heterogeneity was explored via subgroup analysis and meta-regression; publication bias was assessed using Deck's funnel plot.

Results: Twenty-four studies from 17 articles (1,983 patients, 1,437 controls) were included. The pooled analysis demonstrated that ncRNAs achieved a sensitivity of 0.81 and a specificity of 0.84. The area under the curve (AUC) was 0.89, whereas the positive and negative likelihood ratios were 4.96 and 0.23, respectively. LncRNAs demonstrated superior diagnostic performance compared to miRNAs (P = 0.03). Control type, region, and expression trend did not significantly influence outcomes. Significant heterogeneity was observed (I² > 50%), but sensitivity analysis confirmed robustness, and no notable publication bias was detected.

Conclusion: Circulating ncRNAs, particularly lncRNAs, show high diagnostic potential for DR and represent promising noninvasive biomarkers for early screening and differential diagnosis.