The albumin-to-globulin ratio and diabetic retinopathy: an analysis of NHANES data (1999–2020)

Diabetic retinopathy (DR) is a major microvascular complication of diabetes. The albumin-to-globulin ratio (AGR) reflects a systemic inflammatory state, and its role in the course of DR is unclear. This analysis aimed to examine the association between AGR and DR, using data from the National Health and Nutrition Examination Survey. This cross-sectional study employed data from ten NHANES cycles (1999–2020) to explore the association between AGR and DR using weighted logistic regression models. Restricted cubic splines (RCS) were used to investigate possible nonlinear associations. Subgroup analysis was used to identify susceptible populations, and sensitivity analysis was performed to evaluate the robustness of the study results. Among 6,250 eligible participants, 1,392 (22.3%, crude unweighted sample proportion) had DR; the nationally weighted prevalence of DR was 20.42%. Weighted logistic regression showed a significant negative association between AGR and DR (odds ratio = 0.576, 95% confidence interval: 0.420–0.791) in the fully adjusted Model 3. RCS analysis confirmed a significant linear negative association between AGR and DR risk (overall P < 0.05, non-linear P = 0.268). Subgroup analysis indicated this inverse association was stronger among participants aged < 60 years (interaction P < 0.05). Our cross-sectional study found a significant linear negative association between AGR and DR. This association was more evident in patients aged < 60. As a routine, low-cost biochemical indicator, AGR may provide additional information for the clinical assessment of inflammatory and nutritional status in patients with diabetes. Its potential clinical or screening value remains to be validated in prospective studies and formal predictive models. Owing to the limitations of the cross-sectional study design, this study could not establish the longitudinal predictive value of AGR or causal relationships. Not applicable.