Risk factor analysis on body mass rebound after laparoscopic sleeve gastrectomy and establishment of a nomogram prediction model

Abstract

Objective: To investigate the potential independent risk factors of body mass rebound following laparoscopic sleeve gastrectomy (LSG) and construct a nomogram prediction model based on these factors. Methods: In this retrospective observational study, patients with obesity who had undergone LSG at the Department of Gastrointestinal Surgery of the Affiliated Changzhou No. 2 People's Hospital of Nanjing Medical University between January 2015 and July 2017 were retrospectively enrolled. These patients were divided according to their status of postoperative body mass rebound. The inclusion criteria were patients aged between 16 and 65 years who had undergone LSG bariatric surgery with surgical indications according to the 2014 Chinese Guidelines for the Surgical Management of Obesity and Type 2 Diabetes Mellitus. The exclusion criteria were patients who had undergone other bariatric surgeries, who were taking weight-loss drugs or drugs that affected their body weight, who had severe gastroesophageal reflux and hiatal hernia, who were pregnant, who had incomplete clinical data, and who were lost to follow-up or were followed up for <3 years. In total, 241 patients with obesity (69 males and 172 females) who had undergone LSG surgery were enrolled. The mean age and body mass index (BMI) were (29.9±5.8) years and (40.8±4.8) kg/m2, respectively. The patients were followed up till July 2022, with a focus on their body weight. Postoperative body mass rebound was defined as a percentage increase of ≥10% from the nadir body mass, which was the lowest body mass during the 3-year follow-up period. The body weight rebound following LSG and its influencing factors were observed, based on which a nomogram model was constructed and evaluated. The relationships between the patients' basic data, clinical indicators, preoperative hematological indicators, postoperative indicators, and body weight rebound following LSG were analyzed via univariate analysis. Independent risk factors were further screened by multivariate logistic regression analysis. Factors with a statistically significant difference were included into the nomogram prediction model. Moreover, the model was internally (modeling set) and externally (validation set, 80 baseline data-matched patients with obesity from our center) validated using receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) via R software. ROC curve analysis was used to analyze the predictive and cutoff values of the measurement data for body mass rebound. Results: Overall, 90 patients (37.3%) exhibited postoperative body weight rebound, with the lowest BMI of (29.5±2.6) kg/m2 and time to reach the lowest BMI of (15.4±2.3) months; 151 patients (62.7%) did not exhibit body weight rebound, with the lowest BMI of (29.8±2.3) kg/m2 and time to reach the lowest BMI of (14.7±2.1) months. The results of univariate analysis showed that BMI, depression, anxiety, C-reactive protein (CRP) levels, systemic immune inflammatory index (SII), prognostic nutritional index (PNI), and albumin/fibrinogen ratio (AFR) were associated with body weight rebound following LSG with statistically significant differences (all P<0.05). The results of multivariate regression analyses suggested that depression [odds ration (OR) = 1.31, 95% confidence interval (CI): 1.08-1.62, P=0.010], preoperative CRP levels of ≥8 mg/L (OR = 1.34, 95% CI: 1.09-1.69, P=0.007), SII (OR = 0.58, 95% CI: 0.41-0.86, P=0.013), PNI (OR = 2.06, 95% CI: 1.03-4.21, P=0.007), and AFR (OR: 0.49, 95% CI: 0.33-0.69, P=0.011) were five independent risk factors for body mass rebound. A nomogram prediction model was constructed based on the multivariate analysis results. The scores of PNI, SII, AFR, CRP, and depression were 92.5, 100, 72.5, 25, and 27.5, respectively. The total score was calculated by adding the individual scores of each risk factor, which was used to calculate the probability of body mass rebound following LSG. The evaluation results of the nomogram model showed a C-index of 0.713 and 0.762, sensitivity of 0.656 and 0.594, and specificity of 0.715 and 0.909 in the modeling and validation sets, respectively. The calibration curve analysis and DCA indicated that the nomogram model has a good predictive value for body mass rebound after LSG. Conclusion: Preoperative depression, CRP of ≥8 mg/L, SII, PNI, and AFR were independent risk factors for body mass rebound following LSG. Hence, the nomogram prediction model based on these factors can effectively predict body mass rebound in patients undergoing LSG.