Abstract
BACKGROUND: Excessive visceral adipose tissue (VAT) accumulation is strongly associated with numerous metabolic disorders. Laparoscopic sleeve gastrectomy (LSG) reduces VAT, leading to improved metabolic conditions. However, considerable individual variability results in suboptimal metabolic improvements in certain patients post-LSG. Currently, no predictive model for postoperative VAT content exists, and reliance on macroscopic anthropometric or basic metabolic parameters alone fails to accurately predict postoperative metabolic outcomes. OBJECTIVE: This study aims to evaluate the long-term effects of LSG on VAT reduction, identify factors influencing VAT loss, and develop a clinically applicable risk assessment model. METHODS: This study included 177 patients, randomly divided into a modeling group (132 patients) and a validation group (45 patients). Demographic, metabolic, and imaging data were collected, and patients were categorized based on the median ΔVAT change at 12 months post-LSG. Independent predictors were identified via univariate and multivariate logistic regression, and a nomogram model was developed, followed by external validation. RESULTS: In the modeling group, significant differences in gender, waist-to-hip ratio (WHR), VAT, high-density lipoprotein cholesterol (HDL-c), and hypertension were observed between the high-change and low-change groups. Multivariate logistic regression identified preoperative VAT and HDL-c as independent predictors of weight loss outcomes. The nomogram model demonstrated excellent discriminatory power, with an AUC of 0.7 in the training set and 0.88 in the validation group. The calibration curve confirmed high predictive accuracy, and decision curve analysis (DCA) and clinical impact curve (CIC) analyses underscored the model's strong clinical applicability. CONCLUSION: The combination of preoperative HDL-c and VAT serves as an effective predictor of VAT reduction post-LSG, offering a theoretical basis for improving preoperative assessment and facilitating personalized patient management.