Association of Serum 25-Hydroxyvitamin D, Inflammatory, and Metabolic Biomarkers with Non-Proliferative Diabetic Retinopathy in Type 2 Diabetes Mellitus: A Cross-Sectional Study

Background: Non-proliferative diabetic retinopathy (NPDR) is a common microvascular complication of type 2 diabetes mellitus (T2DM) associated with chronic inflammation, metabolic dysregulation, and multiple interacting systemic factors. This study evaluated the associations of inflammatory biomarkers, serum 25-hydroxyvitamin D [25(OH)D], electrolyte homeostasis, and lipid-derived indices with NPDR. Methods: In this cross-sectional study with prospective recruitment, 98 patients with T2DM were classified into NPDR (n = 55) and non-DR (n = 43) groups. Clinical characteristics, hematological inflammatory indices, serum interleukin-6 (IL-6), C-reactive protein (CRP), 25(OH)D, electrolyte concentrations, lipid profile, and derived biomarkers—including the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), systemic immune–inflammation index (SII), LDL-C/HDL-C ratio, and calculated serum osmolarity—were compared. Receiver operating characteristic (ROC) analysis and multivariable logistic regression were performed to assess discriminatory performance and independent associations with NPDR. Results: Patients with NPDR had significantly lower serum 25(OH)D concentrations than the non-DR group (16.6 ± 7.5 vs. 21.0 ± 5.2 ng/mL, p < 0.001). Serum IL-6 levels (4.2 ± 3.1 vs. 2.8 ± 1.4 pg/mL, p = 0.047), NLR (p = 0.043), LDL-C/HDL-C ratio (p = 0.031), and calculated serum osmolarity (p = 0.033) were significantly higher, whereas lymphocyte count and the lymphocyte-to-monocyte ratio were significantly lower. Among individual biomarkers, serum 25(OH)D showed the best discriminatory performance (AUC = 0.712). A multivariable model including insulin therapy, IL-6, serum 25(OH)D, and the LDL-C/HDL-C ratio demonstrated good discrimination (AUC = 0.813), with 81.8% sensitivity and 81.4% specificity. Bootstrap validation supported the stability of the model. Conclusions: NPDR was associated with lower serum 25(OH)D concentrations, systemic inflammation, and a less favorable LDL-C/HDL-C ratio. A multivariable model integrating inflammatory, metabolic, and nutritional biomarkers showed better discriminatory performance than individual biomarkers. The association between insulin therapy and NPDR likely reflects greater diabetes severity rather than a direct effect of insulin. These findings are exploratory, and prospective studies in larger independent cohorts are needed to determine whether these biomarkers improve risk stratification beyond established clinical factors.