Purpose In patients with type 2 diabetes mellitus (T2DM), bariatric surgery appears to be more effective than medical treatment (MT) at improving glycaemic control and decreasing cardiovascular risk. However, long-term effectiveness has not been systematically assessed using randomised controlled trials. In this study, we aimed to systematically assess randomised controlled trials, with at least 5 years of follow-up, on bariatric surgery in patients with T2DM and BMI ≥ 25 kg/m 2 , as well as to compare different bariatric procedures. Material and Methods PubMed, SCOPUS and Web of Science were searched. We performed a network meta-analysis to evaluate the comparative effectiveness of the different procedures and MT in terms of full T2DM remission, weight loss, complications and cardiometabolic biomarkers. The quality of evidence was assessed using the Cochrane Risk of Bias Tool and CINeMA. ResultsWe included 11 primary studies. Laparoscopic one-anastomosis gastric bypass (LOAGB) was found to be one of the most effective treatments for full remission of T2DM (I 2 = 0, inconsistency p value = 0.9223). Biliopancreatic diversion without duodenal switch was found to be more effective than laparoscopic Roux-en-Y gastric bypass and laparoscopic adjustable gastric banding regarding percent total cholesterol and low-density lipoprotein change. Regarding percent high-density lipoprotein and weight change, all analysed surgical treatments were shown to be more effective than MT. The confidence rating in our results was overall moderate, but most studies had high risk of performance and detection bias. Conclusion Bariatric surgery seems to be effective for T2DM remission. LOAGB appears to be a good option in this context, and a possible alternative to laparoscopic duodenal switch, but the included primary studies in our review are not sufficiently powered to establish a more definitive conclusion. More studies with longer follow-up times are needed to comprehensively assess bariatric surgery in T2DM.

y Some of the formatting mistakes identified during the proofing phase were not corrected which impairs comprehension of the information provided in the tables of the main article.Regarding Table 1, the last ten columns of the table were moved to the beginning of the table in the published version. These columns should be in the end (right side) of the table. The correct version of Table 1 is provided here.Table 2 included in the published version of the article is not from the correct version (the "clean" version) of the manuscript; it is from the annotated version, with all the changes made during the revision phase written in red, instead of black. The correct version is displayed here.Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.