Abstract
PURPOSE: To investigate the interaction between metabolic and behavioral factors in non-proliferative diabetic retinopathy (NPDR) among elderly patients with type 2 diabetes mellitus (T2DM). PATIENTS AND METHODS: A cross-sectional study was conducted from October 2023 to May 2024 using cluster random sampling to recruit elderly patients with T2DM from two tertiary hospitals in Nantong. A self-designed questionnaire was utilized to collect lifestyle and behavioral data. Non-mydriatic fundus photography was used for DR screening, and biochemical indicators were extracted from electronic medical records. Descriptive statistics were applied to analyze demographic characteristics, while univariate and multivariable logistic regression analyses were performed to identify factors influencing NPDR. Interaction models evaluated multiplicative and additive effects between metabolic indicators (low-density lipoprotein cholesterol [LDL-C], urine albumin-to-creatinine ratio [UACR]) and weekly fish consumption frequency on NPDR. Receiver operating characteristic (ROC) curves assessed the diagnostic value of elevated UACR, low fish consumption, and their interaction for NPDR. RESULTS: Among 978 eligible elderly patients with T2DM, 794 were included, with 306 (38.5%) diagnosed with DR, including 246 (31.0%) with NPDR. Multivariable logistic regression identified diabetes duration, LDL-C, UACR, and weekly fish consumption as independent risk factors for NPDR. Interaction analysis demonstrated significant positive multiplicative (OR=8.89, 95% CI=2.87-9.52, P<0.001) and additive interactions (RERI=2.77, API=0.31, SI=1.54) were observed between low fish consumption (<2 times/week) and severe UACR (>33.9 mg/mmol). No interactions were found with moderate UACR (3.4-33.9 mg/mmol) or elevated LDL-C (>1.8 mmol/L). ROC analysis demonstrated AUC values of 0.840, 0.816, and 0.893 for severe UACR, low fish consumption, and their combined effect, respectively, indicating enhanced diagnostic utility for NPDR. CONCLUSION: The synergistic interaction between severe UACR and infrequent fish consumption significantly elevates NPDR risk in elderly patients with T2DM, highlighting its potential as a predictive marker for early intervention.