Association between omega-6 fatty acids and diabetic retinopathy risk: a systematic review and meta-analysis.

Abstract

Background: Diabetic retinopathy (DR) is a progressive microvascular complication of diabetes and a leading cause of vision loss worldwide. Despite advances in glycemic and ophthalmologic management, additional modifiable factors influencing DR onset and progression require further exploration. Emerging evidence suggests a pivotal role of omega-6 polyunsaturated fatty acids in retinal vascular health.

Objectives: To systematically synthesize and quantify associations between omega-6 fatty acids, including linoleic acid (LA), arachidonic acid (AA), and related metabolites, and the incidence, presence, progression, predictive performance, and reported adverse or non-retinal clinical outcomes of diabetic retinopathy (DR) among individuals with diabetes.

Methods: A comprehensive systematic search was conducted across PubMed, Embase, Web of Science, Scopus, Cochrane Library, ScienceDirect, Wiley Online Library, EBSCO Open Research, and Google Scholar from database inception to August 15, 2025. Eligible studies included randomized trials, cohort studies, case-control studies, cross-sectional studies, and Mendelian randomization analyses that evaluated omega-6 fatty acid exposure, circulating omega-6 biomarkers, or omega-6-related metabolites in relation to DR. Quantitative synthesis was performed only for outcomes with sufficient methodological and clinical comparability. Predictive-model studies and Mendelian randomization evidence were synthesized separately. Publication bias assessments were treated as exploratory because of the small number of studies per outcome.

Results: Ten studies were included (n = 1306 clinical participants plus one Mendelian randomization dataset >400,000 individuals). Higher total omega-6 levels were associated with reduced DR risk, with a random-effects odds ratio (OR) of 0.76 (95% confidence interval [CI], 0.72-0.81; I² = 53%). LA demonstrated a consistent inverse association with DR (OR = 0.78; 95% CI, 0.71-0.85; I² = 31%). In contrast, AA was associated with increased DR risk (OR = 1.15; 95% CI, 1.08-1.22; I² = 0%). Mendelian randomization supported an inverse association between genetically predicted omega-6 levels and DR risk in type 2 diabetes, with an OR of 0.75 per standard deviation (SD) increase (95% CI, 0.62-0.92). Predictive models incorporating omega-6 lipid profiles reported area under the receiver operating characteristic curve (AUC) values of 0.79-0.92; exploratory logit-transformed synthesis suggested good discrimination but should be interpreted cautiously because of methodological heterogeneity. Safety data were limited and heterogeneous, and any quantitative synthesis should be considered exploratory.

Conclusion: Current evidence suggests that higher LA and related omega-6 measures may be associated with lower odds of DR, whereas higher AA and certain downstream metabolites may be associated with a higher risk. Because the evidence base remains heterogeneous in design, exposure definition, and outcome ascertainment, omega-6-related biomarkers should be regarded as promising risk-stratification candidates rather than clinically actionable markers at present.