Metabolomic integration with machine learning identifies potential biomarkers for type 2 diabetes mellitus and diabetic retinopathy progression

[Objective] To identify aqueous humor metabolic profiles associated with the progression of type 2 diabetes mellitus(T2DM) and diabetic retinopathy (DR), aiming to discover potential biomarkers for DR progression. Ultra-high performance liquid chromatography tandem quadrupole-electrostatic field orbitrap high-resolution mass spectrometry(UHPLC-HRMS) will be utilized in conjunction with machine learning (ML) for comprehensive analysis. [Methods] A total of 78 patients with T2DM and 30 age- and gender-matched healthy controls were included. UHPLC-HRMS was used to identify metabolites in the aqueous humor of all participants. ML was employed to screen for metabolites associated with T2DM and DR progression, and predictive models were established. [Results] After adjusting for covariates, 1,5-anhydroglucitol, tetradecyl sulfate, and n,n,n-trimethyl-5-aminovaleric acid identified as significant indicators for T2DM compared to controls (all P<0.05). N-acetyltryptophan, linoleamide, oleamide, palmitic amide,valeric acid(FFA(5:0), and succinic acid emerged as predictors for DR progression (all P<0.001). Metabolic pathway analysis revealed that "valine, leucine and isoleucine biosynthesis", "arginine biosynthesis," and "cysteine and methionine metabolism" were the most enriched pathways for T2DM. Predictive models achieved R2 values of 81.3%, and 74% for T2DM and DR progression, respectively. [Conclusions] Metabolomic combined with ML effectively uncovered metabolic characteristics associated with T2DM and DR progression. Linoleamide and oleamide represent promising potential biomarkers for DR progression, offering new opportunities for diagnosis and personalized treatment of DR.