Abstract
BACKGROUND: Type 2 diabetes mellitus (T2DM) is associated with diabetic retinopathy (DR). The complement system maintains the normal physiologic microenvironment of the retina. The relationship between serum complement levels and clinical features of DR remains unclear. METHODS: Clinical characteristics of 252 patients with T2DM including 101 with non-DR (NDR), 79 with nonproliferative DR (NPDR), and 72 with proliferative DR (PDR) were prospectively analyzed. Serum complement levels were compared between NDR and DR patients. The correlation between clinical characteristics and complement levels in DR patients was analyzed. A multifactorial logistic analysis was constructed to predict the risk of developing DR in T2DM. RESULTS: Serum C4, CFB, CFI, C3 and C5 levels were higher in DR patients than in NDR patients (all P < 0.05). In T2DM patients, C3 and C4 levels were higher in PDR patients than in DR patients (all P < 0.05), and MBL levels were not statistically different between the two cohorts (P > 0.05). These complement components or fragments were positively correlated with the duration of diabetes, glycosylated hemoglobin (HbA1c), and triglycerides (TG) (all P < 0.05). C3, C5, the duration of diabetes, HbA1c, and TG were the independent risk factors for DR in T2DM patients. The ROC model showed good value for predicting the risk of developing DR in T2DM with an area under the curve of 0.887. CONCLUSION: Serum complements C3 and C5 are predictive factors for DR in patients with T2DM. The prediction model constructed by the clinical characteristics of patients with T2DM and complement can better distinguish between NDR and DR, and can be used as a potential biomarker for assessing the risk of developing DR.