Tirzepatide pulls ahead of semaglutide in real-world weight loss data

A retrospective study circulated on 2 July 2026 has reopened a question that has quietly shaped the obesity-drug market since the launch of Mounjaro: when patients stay on therapy in the real world, does tirzepatide deliver a sturdier weight loss curve than semaglutide? The early read, drawn from a chart posted by X user @cremieuxrecueil, is yes — and by enough margin to be commercially uncomfortable for Novo Nordisk.

The study is a retrospective look at GLP-1 users comparing the two dominant molecules in the category. Tirzepatide, sold as Mounjaro for type 2 diabetes and Zepbound for chronic weight management, is the dual GIP/GLP-1 receptor agonist developed by Eli Lilly. Semaglutide, the GLP-1-only molecule sold by Novo Nordisk as Ozempic and Wegovy, has until now owned the popular imagination around the class. The new data suggest that the gap between the two is wider than the early head-to-head trials implied — and that the gap is widening where it matters most: at the bottom of the responder curve.

What the chart actually shows

The chart tracks weight change over the period users remained on each drug. Tirzepatide users, on average, lose more weight than semaglutide users over comparable durations, the data indicate. The second finding is the one with the longer tail: tirzepatide users are less likely to be classified as non-responders and less likely to be late responders — that is, patients who initially appear to plateau and then break through, or who fail to break through at all.

That second point matters because the obesity-drug market is increasingly a retention story. A patient who loses 5 percent of body weight and stops is, from a health-economics standpoint, a write-off; a patient who loses 15 percent and stays on therapy for eighteen months is the unit of value the manufacturers are selling to insurers, employers, and pharmacy-benefit managers. If tirzepatide is producing fewer non-responders at the front end of treatment, the lifetime value of an average prescription rises even before you consider list price.

The counter-read

Retrospective comparisons are not head-to-head trials. They rely on the data each patient brought with them when they started therapy — baseline weight, comorbidities, prior GLP-1 exposure, insurance status, and the decision their clinician made about dosing. Patients who received tirzepatide rather than semaglutide in 2024 and 2025 were not assigned at random; clinicians steered sicker, more refractory, or better-insured patients toward the newer molecule, particularly during the period of supply constraint that followed Mounjaro's U.S. launch. That selection effect tends to suppress apparent efficacy in the new drug and inflate it in the incumbent.

There is also the question of adherence. Semaglutide was, for most of 2023 and 2024, the drug patients could actually fill. Tirzepatide spent long stretches on allocation, with patients forced to skip doses or ration pens. A retrospective study that compares users "on therapy" will, by construction, under-weight the population that was bounced off semaglutide for supply reasons before they could register as a late responder or a non-responder. The chart's clean line is doing a lot of uncredited work.

The believers in the framing counter that the size of the gap is hard to explain by selection alone — that a 2-to-4-percentage-point difference in average weight loss would not, on its own, be news; that the headline is the compression of the lower tail, and that the lower tail is precisely where selection bias should work against tirzepatide, not for it.

Why the structural read still favours tirzepatide

What the retrospective data sit inside is a longer arc. Eli Lilly's phase 3 programme for tirzepatide in obesity — the SURMOUNT trials — produced average weight losses in the 16-to-22 percent range, against Wegovy's roughly 15 percent in STEP 1. Novo Nordisk's response has been a generation-2 molecule, CagriSema, and an oral semaglutide formulation at higher doses. Both are bets that the GLP-1 backbone can close the gap. The new retrospective is consistent with the trial data; it does not replace them, but it also has not been contradicted by anything published since.

The structural question is whether the obesity-drug market becomes a duopoly with a durable leader, or a duopoly in which the molecules trade places every few years. A more granular way to ask it: does the next formulary cycle at a large U.S. insurer or PBM treat tirzepatide and semaglutide as interchangeable, or does it steer preferentially toward tirzepatide for new starts on the strength of responder-distribution data like this? The chart is the kind of artefact that payer actuaries read in the same week it appears.

The stakes for the rest of 2026

Novo Nordisk's commercial position rests on Wegovy's first-mover advantage, its pen-based delivery device, and the supply chain it built before the GLP-1 wave hit. Eli Lilly's rests on a molecule that, on every available cut of the data, produces more weight loss per patient. If retrospective studies continue to point in the same direction, two things tend to follow: price competition tightens, and the prescriber habit shifts.

For patients, the near-term read is that the question of which drug to start is being answered more decisively in tirzepatide's favour. The longer-term read is the one the manufacturers are most anxious about: that a third mechanism — oral GLP-1s, oral triple agonists, the amycretin class — eventually resets the leaderboard again, and that the current head-to-head becomes a footnote in the next formulary war.

The sources do not specify the sample size, the duration of follow-up, or whether the comparison adjusted for baseline BMI and prior GLP-1 exposure. Until those numbers are published, the chart is a direction-of-travel indicator, not a verdict. The direction, for now, is clearly north.

This article framed the comparison as a real-world efficacy question rather than a market-share question; the wire coverage that followed treated it as the latter, and the duopoly framing above is Monexus's own.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://x.com/cremieuxrecueil/status/
• https://en.wikipedia.org/wiki/Tirzepatide
• https://en.wikipedia.org/wiki/Semaglutide