RESUMO
PURPOSE: Weight regain (WR) compromises the effectiveness of bariatric surgery. The objective of this study was to determine differences in long-term WR prevalence using different definitions and analyze possible preoperative predictors involved. METHODS: Single-center retrospective cohort study including 445 adults who underwent 3 modalities of bariatric surgery between 2009 and 2014. EXPOSURE: age, gender, ethnicity, body mass index (BMI), type 2 diabetes (T2D), hypertension (HTN), and type of surgery. MAIN OUTCOMES: WR at year 6 assessed by 4 definitions and 6 multivariate models based on common thresholds. RESULTS: Our cohort (71.1% female) had a mean age of 44.78 ± 11.94 years, and mean presurgery BMI of 44.94 ± 6.88 kg/m2, with a median follow-up of 6 years (IQR=5-8). The prevalences of T2D and HTN were 36.0% and 46.7% respectively. WR rates over thresholds ranged from 25.4 to 68.1%, with significant differences between groups in the WR measured as the percentage of maximum weight loss (MWL) and the increase in excess weight loss (EWL). Presurgery BMI was a significant predictor in 3 models; restrictive techniques were associated with WR in all the models except for those considering WR over 10 kg and WR over 15% from nadir as dependent variables. CONCLUSIONS: In this long-term study, WR defined as percentage of MWL and increase in EWL from nadir had the greatest significance in logistic regression models with preoperative BMI and type of surgery as independent variables. These findings could serve to establish a standardized outcome reporting WR in other longitudinal studies. KEY POINTS: ⢠Lack of standardized outcome to measure weight regain after bariatric surgery. ⢠Lowest rates of weight regain in malabsorptive techniques in all definitions applied. ⢠Weight regain measured as percentage of maximum weight lost.