Time impact on the antidiabetic effects of key bariatric surgeries: a network meta-analysis of randomized controlled trials with meta-regression.

Background: Few studies compare the efficacy of the key bariatric procedures in type 2 diabetes management over the long term. None offer a reliable comparison of their respective efficacy loss over time.

Objectives: To analyze and compare the time evolution of the antidiabetic effects of the key bariatric procedures.

Setting: Obesity surgery departments in America, Europe, and Asia.

Methods: All the randomized clinical trials assessing the efficacy of bariatric surgery in type 2 diabetes management with 1-5 years of follow-up were reviewed. A network meta-analysis with meta-regression was performed to compare the effectiveness of each technique and its respective efficacy loss temporal dynamics.

Results: Thirty-one trials involving 1906 patients were included. In comparison to Roux-en-Y gastric bypass, the 5-year complete or partial diabetes remission rates were inferior with medical treatment (odds ratio [OR] = .05; 95% credible interval [CrI]: .02-.13) and gastric banding (OR = .38; 95% CrI: .16-.87), equivalent with sleeve gastrectomy (OR = 1.08; 95% CrI: .59-1.97), and superior with 1 anastomosis gastric bypass (OR = 3.00; 95% CrI: 1.12-8.33) and biliopancreatic diversion and its affiliated techniques (OR = 3.71; 95% CrI: 1.16-12.55). However, remission rates and glycemic control progressively decreased whatever the treatment option evaluated. Moreover, this loss of efficacy followed a statistically comparable temporal dynamic to those of Roux-en-Y gastric bypass regardless of the therapeutic strategy implemented.

Conclusions: No therapeutic modality offered stable antidiabetic effects. The gap observed between the techniques after a 5-year follow up concerning remission rates and glycemic control could depend essentially on the magnitude of the effects initially obtained. However, these results need to be confirmed over longer follow-up periods.