To study the main influencing factors of anxiety disorders in patients with functional dyspepsia using logistic regression and decision tree models and to understand their psychological characteristics, in order to take targeted measures to improve the mental health status of them.

A questionnaire survey was conducted on 537 patients with functional dyspepsia from two tertiary hospitals in Xinjiang, using a general demographic data survey, Generalized Anxiety Scale (GAD-7), Gastrointestinal Symptom Rating Scale (GSRS), Multidimensional Health Control Source Scale (MHLC), and Simplified Niping Digestive Dysfunction Index (SF-NDI). Binary logistic regression model and CHAID algorithm based classification decision tree to analyze the influencing factors of anxiety disorders in patients with functional dyspepsia. Compare the differences between the two models.

Logistic regression results showed that the leaders and staffs of state institutions (OR=5.13, 95% CI:1.031-25.529), freelances, the unemployed(OR=6.3, 95% CI:1.291-30.749), the worse the emotional states (OR=1.167, 95% CI:1.03-1.323), the worse the gastrointestinal symptoms(OR=1.065, 95% CI:1.037-1.095), and health authority control (OR=1.032, 95% CI:1.006-1.06) were associated with a higher risk of anxiety disorders in patients with functional dyspepsia (P<0.05). The results of decision tree modeling showed that gastrointestinal symptoms, health authority figure control and emotional state were the influencing factors of anxiety disorders in patients with functional dyspepsia, among which gastrointestinal symptoms were the most important influencing factors.

Both models have classification and prediction value, but each has its own advantages and disadvantages. The two can complement each other, making the analysis results more practical.