Is a placebo-anchored comparison worthless?

No. It is how most indirect comparison works and it can be accurate - in this very network the placebo-anchored tirzepatide-versus-semaglutide estimate was later matched to within a quarter of a percentage point by a head-to-head trial. The point is not that it fails but that its accuracy depends on an assumption you should know is being made.

What would break transitivity here?

Any factor that differs systematically between the trials and also modifies treatment effect. Different eras of background lifestyle support, different baseline severity, different rates of discontinuation, or different definitions of the analysis population would all do it. Comparing placebo arms across the trials is the usual informal check.

Why does the number of trials in a network matter?

Because a sparse network has few redundant paths, so there is little opportunity to check whether direct and indirect evidence agree. With six trials and only one active-comparator edge, this network has almost no capacity for that internal consistency check.

Does this apply to other network meta-analyses?

The structural questions do. Sparse networks anchored on placebo, dominated by one large old trial, with a single small bridge between two of the drugs, are common across therapeutic areas. Reading the network diagram before the forest plot is a general habit worth having.

GLP-1 & Incretin Science

Tirzepatide Has No Active Comparator in the Network

JMWritten & reviewed by Jack Muncaster · Founder, UK PeptidesLast reviewed 2026-08-237 cited sources

A network meta-analysis is only as good as the links between its trials. In the 2026 obesity network, tirzepatide connects to semaglutide and liraglutide entirely through placebo arms, and more than half the participants come from a single trial published in 2015.

Key facts

Network: Ciudin et al., Adv Ther 2026 (PMID 41820778)
Trials in network: 6
Total tabulated participants: 9,355
Trials with an active comparator: 2 (O'Neil 2018, STEP 8)
Active-comparator link: Semaglutide vs liraglutide only
Tirzepatide trials in network: 1 (SURMOUNT-1, placebo-controlled)
Largest contributor: SCALE Obesity and Prediabetes, 4,974 tabulated
Publication years spanned: 2015 to 2022

Draw the network before reading the result

Every network meta-analysis has a shape, and the shape carries more information than the headline estimate. In this one, six trials connect three drugs. SURMOUNT-1 randomised tirzepatide against placebo. STEP 1 and STEP 5 randomised semaglutide against placebo. SCALE Obesity and Prediabetes randomised liraglutide against placebo. Two trials - the O'Neil 2018 dose-ranging study and STEP 8 - randomised semaglutide directly against liraglutide. Add those up and one fact stands out: there is no trial anywhere in this network in which tirzepatide is randomised against an active drug.

What that means for the tirzepatide comparisons

Every estimate of tirzepatide against semaglutide, and of tirzepatide against liraglutide, is produced by comparing SURMOUNT-1's placebo arm with the placebo arms of the semaglutide and liraglutide trials, and assuming those placebo arms are exchangeable. That assumption is called transitivity, and it is doing the entire load-bearing work. It is not obviously wrong here - all six trials enrolled adults without type 2 diabetes on a background of reduced-calorie diet and increased physical activity, with similar baseline weight and BMI - but it is an assumption about unobserved comparability, and no amount of statistical machinery can test it directly.

The semaglutide-liraglutide link is the only direct evidence, and it is small

The two active-comparator trials contribute 239 and 338 participants in the analysis's own tabulation - 577 people out of 9,355. Everything the network says about how semaglutide and liraglutide compare is anchored, in part, on those 577. STEP 8 is also the one trial the authors flag as different in design: it was double-blind between each treatment and its matched placebo but not between the active treatments, because their dosing schedules differ, and it ran in a single country. Small, partially blinded, single-country trials are not disqualifying. They are the kind of thing worth knowing sits at the join.

More than half the network is one trial from 2015

SCALE Obesity and Prediabetes contributes 4,974 of the 9,355 tabulated participants - 53% of the network - and it is the sole anchor for liraglutide. It was published in 2015, seven years before SURMOUNT-1, STEP 5 and STEP 8. Trial vintage matters more than it looks. Background care changes, participant expectations change, and analytical conventions change: the SCALE report used last-observation-carried-forward imputation, a method largely superseded by the estimand framework that the later trials use. The authors note that SCALE Obesity and Prediabetes is the one study in the network that did not report an efficacy estimand at all, which is a substantive gap given that the efficacy estimand is where the headline comparisons against liraglutide come from.

A sample size that does not match the registry

The analysis tabulates 4,974 participants for SCALE Obesity and Prediabetes. The trial's registry record, NCT01272219, records 3,731 as actual enrolment, and the New England Journal report describes 3,731 patients randomised in a 2:1 ratio to liraglutide 3.0 mg or placebo over 56 weeks. The two figures differ by more than a thousand people and the text available does not explain the difference; a plausible reading is that the analysis counts a broader population than the 56-week randomised phase, but that is inference, not something the paper states. Anyone relying on this network should check it against the supplementary material. The O'Neil figure moves the other way - 239 tabulated against 957 randomised in NCT02453711 - which is straightforwardly explained by a dose-ranging trial contributing only the arms relevant to the network.

How to use any of this

None of it makes the analysis wrong. It makes the analysis legible, which is different and more useful. When a network estimate is quoted at you, three questions get you most of the way: which trials are actually in it, where the direct evidence sits, and whether the drug you care about is connected to anything except placebo. In this network the answers are six, in two small trials on the semaglutide-liraglutide edge, and no.

Extended research context

The GLP-1 & Incretin Science deep dive

Deep dive: the two routes to a bigger effect

Every compound trying to beat GLP-1 alone has taken one of two routes. The first adds more receptors from the same hormone family — GIP in tirzepatide, GIP and glucagon in retatrutide. The second adds a non-incretin satiety hormone, which in practice means amylin: CagriSema combines cagrilintide with semaglutide, and amycretin engages both receptors from one molecule. Both routes work, because they recruit signalling pathways that do not fully overlap. Neither has escaped the constraint that binds all of them, which is that gastrointestinal tolerability worsens as effect size grows.

Deep dive: why a percentage is not a result

The most-quoted numbers in this field are the least comparable. REDEFINE 1 reported 22.7% and 20.4% for the same compound in the same trial — the first among participants who adhered to treatment, the second across everyone randomised. TRIUMPH-1 reported 28.3% in an uncomplicated obesity population while TRIUMPH-3 reported up to 22.6% in adults with established cardiovascular disease, using the same compound. Before any two figures can be compared they have to match on estimand, population, duration, comparator and whether the number is placebo-adjusted. Most published comparisons match on none of them.

Deep dive: what happens after the trial stops

Every headline figure describes weight while treatment continues. The STEP-1 extension found that a year after semaglutide was stopped, participants had given back roughly two-thirds of what they lost, moving from 17.3% mean reduction to a net 5.6% — though average weight remained below baseline and nearly half stayed at least 5% down. Meta-analysis puts regain at around 0.8 kg per month. This is why maintenance studies such as TRIUMPH-6 matter more to the field's future than another two points of peak reduction.

Research applications

▸Comparing incretin and amylin compounds on a like-for-like basis
▸Interpreting estimands, thresholds and placebo-adjusted figures in trial reports
▸Tracking the obesity pipeline across sponsors and jurisdictions
▸Understanding receptor pharmacology behind GLP-1, GIP, glucagon and amylin
▸Distinguishing licensed medicines from investigational compounds

Handling checklist

✓Identify which estimand a quoted percentage comes from before citing it
✓Check the trial population and baseline BMI against the comparison you are making
✓Confirm the duration and whether the reduction curve had plateaued
✓Read discontinuation rates alongside efficacy figures
✓Verify every NCT identifier against ClinicalTrials.gov rather than secondary reporting

Common research-handling mistakes

Learnt from thousands of researcher orders across our UK labs.

✗ Comparing headline percentages across different trials

Fix: Population, duration, estimand and comparator all differ; the numbers are not interchangeable.

✗ Quoting the larger of two figures from the same trial

Fix: Name the estimand. Efficacy and treatment-policy answer different questions.

✗ Treating peak reduction as a durable outcome

Fix: Substantial regain follows cessation across the class; peak figures describe a maintained state.

✗ Assuming an oral route means a weaker mechanism

Fix: Route and receptor count are independent. Orforglipron is weaker because it hits one receptor, not because it is a tablet.

✗ Reading investigational compounds as available treatments

Fix: Most of this pipeline holds no authorisation anywhere; mazdutide is approved only in China.

Continue researching

Peer-reviewed guides, comparators and matched reference materials.

Frequently asked questions

Is a placebo-anchored comparison worthless?: No. It is how most indirect comparison works and it can be accurate - in this very network the placebo-anchored tirzepatide-versus-semaglutide estimate was later matched to within a quarter of a percentage point by a head-to-head trial. The point is not that it fails but that its accuracy depends on an assumption you should know is being made.
What would break transitivity here?: Any factor that differs systematically between the trials and also modifies treatment effect. Different eras of background lifestyle support, different baseline severity, different rates of discontinuation, or different definitions of the analysis population would all do it. Comparing placebo arms across the trials is the usual informal check.
Why does the number of trials in a network matter?: Because a sparse network has few redundant paths, so there is little opportunity to check whether direct and indirect evidence agree. With six trials and only one active-comparator edge, this network has almost no capacity for that internal consistency check.
Does this apply to other network meta-analyses?: The structural questions do. Sparse networks anchored on placebo, dominated by one large old trial, with a single small bridge between two of the drugs, are common across therapeutic areas. Reading the network diagram before the forest plot is a general habit worth having.

Primary sources & clinical trials

Peer-reviewed research and registered trials from PubMed, ClinicalTrials.gov, PubChem, FDA and NIH. All links open in a new tab and point to the primary source, so every claim can be verified at origin.

Written and reviewed by

Jack Muncaster · Founder, UK Peptides

Jack founded UK Peptides in Manchester after repeatedly receiving research compounds with missing or recycled paperwork. He is responsible for supplier selection, batch release decisions and the content published in this research library. Every article here is sourced to primary literature and every product page to a signed third-party certificate.

How we write and check this content Report a factual error in this article

More GLP-1 & Incretin Science articles

Popular across the research hub

One flagship guide from every other research category — keep exploring.

← Back to Forum