Background: Diabetic retinopathy (DR) is a predominant microvascular complication in patients with type 2 diabetes mellitus (T2DM) and constitutes one of the leading causes of vision loss and blindness among this population. Serum uric acid (SUA) acts as a critical biomarker for microangiopathy; however, its relationship with DR remains incompletely elucidated. To adjust for inter-individual heterogeneity and confounding effects of baseline renal function, serum creatinine was used to normalize uric acid concentrations. Accordingly, the serum uric acid-to-creatinine ratio (SUA/SCr) was selected as the study index, as this metric can better reflect the net metabolic burden of uric acid as well as the status of oxidative stress. The present study aimed to explore the association between SUA/SCr levels and DR risk in patients with T2DM.
Methods: A total of 1091 patients diagnosed with T2DM who visited the Department of Internal Medicine at Hefei First People’s Hospital between 2022 and 2024 were initially recruited. Following screening according to pre-defined inclusion and exclusion criteria, 427 eligible subjects were finally enrolled and stratified into DR and non-DR groups. Spearman rank-order correlation analysis was performed to evaluate correlations between DR status and measured clinical variables. Univariable binary logistic regression was performed to screen potential risk factors. Variables with P< 0.05 in the univariable analysis were entered into the baseline multivariable logistic regression model. Post-hoc power calculation was performed for the primary multivariable logistic regression. With an assumed odds ratio of 1.22, α = 0.05, and a case-to-noncase ratio of 1:1.3, the sample size of 427 participants provided adequate statistical power to detect the target effect size.Receiver operating characteristic (ROC) curve analysis was utilized to estimate the predictive value of SUA/SCr for DR in T2DM patients. Furthermore, C-statistics were calculated and compared via DeLong’s test to determine whether adding SUA/SCr could enhance the predictive performance of the baseline DR-risk model. Net reclassification improvement (NRI) and continuous integrated discrimination improvement (IDI) were additionally computed to quantify the incremental predictive value contributed by SUA/SCr.
Results: Baseline-characteristic comparisons showed significant between-group differences in age, disease duration, glycated hemoglobin A1c (HbA1c), albumin (ALB), aspartate transaminase (AST), alanine transaminase (ALT), gamma-glutamyl transferase (γ-GT), total bilirubin (TBIL), triglycerides (TG), SUA, and SUA/SCr (all *P<0.05). Spearman correlation analyses demonstrated that DR status was positively correlated with age, disease duration, HbA1c, SUA, and SUA/SCr (all *P<0.05), whereas negative correlations were observed for C-peptide (C-P), ALB, AST, ALT, γ-GT, and TBIL (all *P<0.05). Univariate and multivariable logistic regression indicated that elevated SUA/SCr was associated with higher DR risk among patients with T2DM. ROC-curve analysis revealed that SUA/SCr exhibited favourable discriminative performance for DR, with an AUC of 0.774 (95% CI: 0.720-0.827, *P<0.05). The optimal cut-off value was 3.68, corresponding to a sensitivity of 79.8% and specificity of 67.2%. When SUA/SCr was incorporated into the baseline DR-prediction model, incremental predictive improvement was observed in terms of IDI but not NRI: NRI was 0.140 (95% CI: -0.068-0.348, *P= 0.187), and IDI was 0.021 (95% CI: 0.006-0.036, *P= 0.006).
Conclusion: SUA/SCr was significantly associated with DR in patients with T2DM. Although its incremental predictive improvement was modest, SUA/SCr yielded significant additional discriminative power for DR risk stratification and may serve as a supportive auxiliary indicator for identifying T2DM patients at high DR risk, which may facilitate targeted intervention strategies for diabetic microvascular complications.
https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2026.1977560/full
