Long-term effects of gastric bypass and sleeve gastrectomy in type 2 diabetes: a matched retrospective cohort study from Sweden.

Background: Long-term data on the efficacy and safety of Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG) in people with type 2 diabetes mellitus (T2DM) are still limited. Using a matched cohort design, we aimed to evaluate the long-term effects of RYGB and SG on individuals with T2DM, focussing on obesity- and surgery-related outcomes over a follow-up period of up to 14 years. Methods: A nationwide, matched, longitudinal study was conducted using data from the Swedish National Diabetes Register (NDR) and the Swedish Obesity Surgery Registry (SOReg). Between 2007 and 2020, all individuals with T2DM who underwent primary surgery (RYGB = 7294 and SG = 1105) were identified through SOReg and matched by age, sex, and BMI to a control group of individuals with T2DM from NDR who had not undergone surgery (n = 8399). Data on all-cause mortality and obesity- and surgery-related outcomes after RYGB and SG were retrieved from national registers with almost complete coverage. Risks were expressed as incidence rates per 10,000 person-years and analysed using adjusted Cox regression models, which included duration of diabetes, yielding adjusted hazard ratios (HR) with 95% confidence intervals (CI). Findings: During follow-up, the percentage total weight loss and reductions in HbA1c levels were significantly greater after RYGB and SG than in unexposed individuals (%TWL: RYGB 23·2 vs. 3·6 and SG 17·1 vs. 3·1 at two years, smd > 0·1) and (mean HbA1c: RYGB 46 (SD 14) vs. 58 (SD 17) and SG 46 (SD 13) vs. 55 (SD 15) at two years, smd > 0·1). RYGB was associated with sustainable reductions in all-cause mortality (adjusted HR of 0·62 (95% CI [0·51-0·71])) and obesity-related comorbidities, with risks as much as 45% lower compared to unexposed individuals (p < 0·001). However, individuals after RYGB face as much as a twofold increased risk of malabsorption and micronutrient deficiency (adjusted HR of 2·00 (95% CI [1·76-2·28])) and alcohol use disorder (adjusted HR of 2·82 (95% CI [2·37-3·36])), p < 0·001. The risk of other psychiatric disorders, such as depression (adjusted HR of 1·28 (95% CI [1·14-1·43])), and surgical complications, such as bowel obstruction (adjusted HR of 3·96 (95% CI [3·15-4·98])), was also higher after RYGB (p < 0·001). In contrast, the SG cohort showed no significant effects on obesity-related conditions and risk of surgical complications, despite similar weight reduction in both surgery groups. Interpretation: The study highlights the advantages and limitations of RYGB and SG, providing insights to guide an individualised approach. The limited efficacy of SG in lowering obesity-related disease risks should be a key consideration when selecting individuals with T2DM for surgery. Funding: A grant from the Swedish state under the agreement with the county councils.