The short answer
Sabatella and colleagues published a network meta-analysis of randomised controlled trials comparing metabolic and bariatric surgery against GLP-1 receptor agonists in Obesity in April 2026. An indirect comparison, because the two are rarely randomised against each other.
Key facts
- Study
- Sabatella 2026 (PMID 41326176)
- Journal
- Obesity (Silver Spring), April 2026
- Method
- Network meta-analysis of RCTs
- Comparison
- Metabolic/bariatric surgery vs GLP-1 RAs
- Why indirect
- Direct randomisation is rare
- Key limitation
- Transitivity across dissimilar trials
Why the comparison is asked
Bariatric surgery has produced substantial and durable weight reduction for decades, with long-term outcome data no medicine can yet match. Incretin therapies now produce reductions approaching what surgery achieves. Whether an operation remains necessary is therefore a real clinical question rather than a rhetorical one.
Why randomising the two directly is rare
Blinding is impossible, recruitment is difficult because participants usually have a strong preference, and equipoise is hard to establish. A trial randomising someone to surgery or an injection asks them to accept an outcome they may find unacceptable, which is why so few such trials exist and why indirect methods are used instead.
Research material referenced
Retatrutide 10mg, third-party HPLC tested
What a network meta-analysis contributes here
It links surgical trials and drug trials through comparators they share, producing an indirect estimate with propagated uncertainty. That is more disciplined than reading the two literatures side by side, and it remains an indirect estimate resting on the assumption that the linked trials are comparable.
Why transitivity is strained in this particular comparison
Surgical trials and drug trials differ systematically. Surgical candidates typically have higher baseline BMI. Follow-up in surgical literature runs to years or decades where drug trials run to 68 or 72 weeks. Comparators differ. Those are the differences that make linking through a common arm questionable, and they should be examined rather than assumed away.
The dimension weight percentage does not capture
Durability and reversibility. Surgery is permanent and its effect persists without ongoing adherence; a medicine works while taken and weight returns when it stops. Two interventions producing the same percentage at one timepoint are not equivalent if one requires indefinite continuation and the other does not.
What this does not concern
Research material. This compares a surgical procedure with licensed medicines, both delivered under clinical supervision. Nothing supplied on this site is a GLP-1 receptor agonist, an alternative to surgery, or relevant to any treatment decision.
Frequently asked questions
Why not randomise surgery against a drug directly?
Blinding is impossible, participants usually have strong preferences, and equipoise is hard to establish, so few such trials exist.
What does the network method add?
It links surgical and drug trials through shared comparators with propagated uncertainty. More disciplined than reading the literatures side by side.
What does a weight percentage miss?
Durability and reversibility. Surgery persists without ongoing adherence; a medicine works while taken.
Continue researching
Related guides and reference material.
Primary sources & clinical trials
Published papers, trial registrations, regulator pages and company announcements. Links open the original.
- PubMedSabatella L et al., Comparative Efficacy of Metabolic/Bariatric Surgery Versus GLP-1 Receptor Agonists: A Network Meta-Analysis of Randomized Controlled Trials. Obesity 2026 (PMID 41326176)pubmed.ncbi.nlm.nih.gov
- PubMedFrederiksen KD et al., Bone structural changes after gastric bypass surgery evaluated by HR-pQCT. Calcif Tissue Int 2016 (PMID 26661530)pubmed.ncbi.nlm.nih.gov
- RefSURMOUNT-5: Tirzepatide as compared with semaglutide. NEJMnejm.org
Written and published by UK Peptides, the trading name of AURYXRESEARCH LTD (company 17388727, registered in England and Wales).
Questions or a correction? Email hello@ukpeptides.org or call +44 7426 987848.
More Research & Regulatory News articles
- Does Amylin Need a GLP-1 Partner?CagriSema pairs amylin with semaglutide and missed non-inferiority. A 2026 review surveys long-acting amylin peptides being developed on their own.
- The Trial Where Glucose-Dependence Shows Up in Numbers2,002 randomised, 52 weeks. Hypoglycaemia 6-9% with tirzepatide against 19% with glargine - and 1-3% versus 16% without a sulfonylurea.
- TRIUMPH-4: The First Retatrutide Phase 3 to Read OutTRIUMPH-4 randomised 445 adults with obesity and knee osteoarthritis. Co-primary endpoints, two estimands, and a pain result that did not track the dose.
- VESPER: What Is Actually Registered for BerobenatideSix VESPER trials, from a 239-person Phase 2b to a 3,578-person Phase 3. Registry records, enrolment, completion dates, and which acronyms are registered.
- SOLIS-1: Testing Two Compounds and Their Combination at OnceSOLIS-1 is a Phase 2b umbrella study of an estimated 872 participants, testing an amylin analogue, an ultra-long-acting GLP-1 agonist, and both in combination.