Democracy and Education

BACKGROUND: Reflux esophagitis (RE) is an endoscopy-defined manifestation of gastroesophageal reflux disease and may be related to metabolic disturbance in type 2 diabetes mellitus (T2DM). The clinical-metabolic characteristics associated with endoscopy-confirmed RE in T2DM remain incompletely defined, particularly with regard to glycemic control, 25-hydroxyvitamin D [25(OH)D] status, and insulin-related indicators. This study evaluates the joint associations of these metabolic factors with endoscopy-confirmed RE in a hospital-based T2DM population. METHODS: The analytic cohort comprises 772 hospital-based patients with T2DM and ascertainable RE status from endoscopic diagnostic records. Demographic characteristics, diabetes-related complications, glycemic indices, lipid parameters, renal function, 25(OH)D, insulin-related indicators, and gastrointestinal test findings are assessed. Prespecified multivariable logistic regression models are applied to identify clinical-metabolic correlates of RE. Restricted cubic spline analyses are used to assess potential nonlinear associations, and exploratory discrimination analyses are performed to evaluate risk-stratification performance. RESULTS: Among 772 patients, 223 are classified as having RE, corresponding to an observed prevalence of 28.9%. Compared with the non-RE group, the RE group has higher HbA1c, fasting insulin, triglycerides, HOMA-IR, fasting-state ISI, and TyG index, and lower 25(OH)D levels. In fully adjusted models, HbA1c remains consistently associated with RE (adjusted odds ratio [aOR] per 1-standard-deviation increase, 1.58; 95% confidence interval [CI], 1.33-1.87). The inverse association between 25(OH)D and RE attenuates after adjustment for HbA1c and TyG index (aOR, 0.84; 95% CI, 0.70-1.00). The TyG index provides limited additional information after HbA1c adjustment (aOR, 1.02; 95% CI, 0.85-1.21), whereas HOMA-IR, fasting insulin, and fasting-state ISI show positive associations with RE when evaluated separately. Restricted cubic spline analyses indicate nonlinear associations for HbA1c and insulin-related indicators, with stronger metabolic signals at higher value ranges. Exploratory discrimination analyses indicate limited-to-moderate risk-stratification performance. CONCLUSION: In hospital-based patients with T2DM, endoscopy-confirmed RE is associated with an adverse clinical-metabolic profile, mainly characterized by higher HbA1c and selected insulin-related abnormalities. The inverse association between 25(OH)D and RE attenuates after glycemic adjustment and requires cautious interpretation. These cross-sectional findings are associational and hypothesis-generating, and prospective studies are needed to clarify temporal relationships and clinical implications.