Statistical and Machine Learning Modeling of Psychological, Sociodemographic, and Physical Activity Factors Associated with Weight Regain After Bariatric Surgery

Objective: This study aimed to investigate the associations between weight regain and psychological, sociodemographic, and physical activity factors in post-bariatric surgery patients using statistical modeling. Methods: A cross-sectional, single-center quantitative study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) and Checklist for Reporting Results of Internet E-Surveys (CHERRIES) guidelines. A total of 124 patients who underwent bariatric surgery at least 18 months prior were included. Psychological variables were assessed using validated instruments. Results: In the adjusted Poisson regression model, the most significant predictors of weight regain were binge eating severity (prevalence ratio [PR] = 2.41; 95% confidence interval [CI]: 1.75–3.26; p < 0.001), stress levels (PR = 1.92; 95% CI: 1.45–2.79; p = 0.002), negative affectivity (PR = 1.67; 95% CI: 1.29–2.24; p = 0.004), and physical inactivity (PR = 1.58; 95% CI: 1.21–2.41; p = 0.005). In contrast, anxiety (PR = 1.19; 95% CI: 0.87–1.63; p = 0.214) and psychoticism (PR = 1.12; 95% CI: 0.83–1.57; p = 0.278) showed no statistically significant associations. Conclusion: The results highlight the importance of binge eating severity, stress, negative affectivity, and physical inactivity as major contributors to weight regain after bariatric surgery.