Abstract
IMPORTANCE: Bariatric surgery procedures, in particular Roux-en-Y gastric bypass (RYGB), have been associated with subsequent alcohol-related complications. However, previous studies lack data to account for changes in body mass index (BMI) or alcohol use over time, which are key potential confounders. OBJECTIVE: To evaluate the association between RYGB, sleeve gastrectomy, or gastric banding on subsequent alcohol use disorder (AUD)-related hospitalization and all-cause mortality as compared with referral to a weight management program alone. DESIGN, SETTING, AND PARTICIPANTS: This cohort study included 127 Veterans Health Administration health centers in the US. Patients who underwent RYGB, sleeve gastrectomy, or gastric banding or who were referred to MOVE!, a weight management program, and had a BMI (calculated as weight in kilograms divided by height in meters squared) of 30 or greater between January 1, 2008, and December 31, 2021, were included in the study. EXPOSURES: RYGB, sleeve gastrectomy, or gastric banding or referral to the MOVE! program. MAIN OUTCOMES AND MEASURES: The primary outcome was time to AUD-related hospitalization from the time of bariatric surgery or MOVE! referral. The secondary outcome was time to all-cause mortality. Separate propensity scores were created for each pairwise comparison (RYGB vs MOVE! program, RYGB vs sleeve gastrectomy, sleeve gastrectomy vs MOVE!). Sequential Cox regression approaches were used for each pairwise comparison to estimate the relative hazard of the primary outcome in unadjusted, inverse probability treatment weighting (IPTW)-adjusted (generated from the pairwise logistic regression models), and IPTW-adjusted approaches with additional adjustment for time-updating BMI and categorical Alcohol Use Disorders Identification Test-Concise scores. RESULTS: A total of 1854 patients received RYGB (median [IQR] age, 53 [45-60] years; 1294 men [69.8%]), 4211 received sleeve gastrectomy (median [IQR] age, 52 [44-59] years; 2817 men [66.9%]), 265 received gastric banding (median [IQR] age, 55 [46-61] years; 199 men [75.1%]), and 1364 were referred to MOVE! (median [IQR] age, 59 [49-66] years; 1175 men [86.1%]). In IPTW Cox regression analyses accounting for time-updating alcohol use and BMI, RYGB was associated with an increased hazard of AUD-related hospitalization vs MOVE! (hazard ratio [HR], 1.70; 95% CI, 1.20-2.41; P = .003) and vs sleeve gastrectomy (HR, 1.98; 95% CI, 1.55-2.53; P < .001). There was no significant difference between sleeve gastrectomy and MOVE! (HR, 0.76; 95% CI, 0.56-1.03; P = .08). While RYGB was associated with a reduced mortality risk vs MOVE! (HR, 0.63; 95% CI, 0.49-0.81; P < .001), this association was mitigated by increasing alcohol use over time. CONCLUSIONS AND RELEVANCE: This cohort study found that RYGB was associated with an increased risk of AUD-related hospitalizations vs both sleeve gastrectomy and the MOVE! program. The mortality benefit associated with RYGB was diminished by increased alcohol use, highlighting the importance of careful patient selection and alcohol-related counseling for patients undergoing this procedure.