Sleeve surgery ranked first among six obesity treatments, on evidence the authors graded very low certainty

A network meta-analysis published on 26 August 2026 in the Journal of Clinical Medicine Research lined up six ways of treating obesity and ranked them by how much weight people lost in a year. Laparoscopic sleeve gastrectomy, the most common weight-loss operation, came out on top. The same paper graded the certainty of its own evidence as very low on every outcome it measured.

A network meta-analysis compares treatments that have often never been tested against each other, by linking them through the comparisons that do exist. It is only as reliable as the studies feeding into it.

Sleeve gastrectomy led, and the gap to the endoscopic version was 11.3 points

The researchers searched for studies published between January 2010 and April 2026 and found 17 that qualified. Using sleeve gastrectomy as the reference, the difference in percentage of total body weight lost at 12 months came out as follows, with the 95% confidence interval in brackets.

• Laparoscopic greater curve plication: 2.18 points less (4.93 less to 0.56 more)
• Endoscopic sleeve gastroplasty: 11.33 points less (13.15 to 9.51 less)
• Laparoscopic adjustable gastric banding: 15.81 points less (18.76 to 12.87 less)
• High-intensity diet and lifestyle therapy: 17.63 points less (21.54 to 13.73 less)
• Intragastric balloon: 19.00 points less (23.37 to 14.63 less)

One correction is worth making. The paper's abstract states that all of these came out significantly lower than sleeve gastrectomy. The plication figure does not: its confidence interval crosses zero and its p value is 0.118, which means the analysis did not detect a difference for that procedure.

The serious complication comparison did not favor the endoscopic procedure

Comparing endoscopic sleeve gastroplasty directly with sleeve gastrectomy, serious complications were 30% more common after the endoscopic procedure, with a range running from 3% fewer to 73% more (risk ratio 1.30, 95% confidence interval 0.97 to 1.73). Because that range crosses the point of no difference, the result does not settle the question either way.

The authors are direct about it. They write that the finding should be read as inconclusive rather than as evidence of equivalent safety, and that it challenges the prevailing assumption that endoscopic therapies are universally safer than conventional surgery.

One large registry study inside the analysis shows what that looks like in practice. Major adverse events were similar, at 1.4% after the endoscopic procedure and 1.1% after surgery. But the endoscopic group returned to hospital more often: 30-day readmission 3.8% against 2.6%, reoperation 1.4% against 0.8%, and reintervention 2.8% against 0.7%.

Reflux ran the other way. In one of the smaller cohorts, new heartburn and acid reflux appeared in 4.3% of endoscopic patients and 20% of surgical patients. Type 2 diabetes remission across four cohorts was 64% after the endoscopic procedure and 82% after surgery.

Seventeen studies, none of them randomized

Every study in this analysis was observational. Patients chose, or were selected for, their treatment, which is the single biggest problem with the numbers above. In the studies where patients were not matched, the people sent for endoscopic treatment started noticeably lighter: average body mass index of 38.6 against 47.2 in one study, and 39.6 against 44.9 in another. People at different starting weights are not directly comparable.

The weight-loss comparison also varied a lot from study to study, with a heterogeneity measure of 84.3%. Roughly speaking, that means the individual studies were not all measuring the same thing, so the pooled average stands in for a wide spread of different results.

Under GRADE, the standard system for rating how much confidence to place in a body of evidence, the authors rated all three of their outcomes at the lowest of four levels, very low certainty. That rating is the authors' own assessment, not an outside criticism.

Roux-en-Y gastric bypass, one of the two most common weight-loss operations, was left out of the network entirely.