The retatrutide news cycle has been noisy all year, and it is easy to lose the plot when the headline number gets repeated so many times it stops meaning anything. Yes, the 12 mg group lost 25 percent of their body weight. We have said that before, most recently when the topline numbers dropped and again when Lilly filed with the FDA. But now the full paper is in print, and reading it properly is a different experience from skimming a press release. There is a lot in this dataset that the one-line coverage skipped.
The paper landed in the New England Journal of Medicine on 29 September, from Ania Jastreboff and the TRIUMPH-1 investigators, and it is the single longest and largest look at the triple agonist yet. If you care about how these compounds behave over real timeframes rather than 24-week snapshots, this is the dataset worth sitting with.
Why 80 weeks is the number that matters
Most obesity trial coverage you see is based on short studies. Eli Lilly’s SURPASS and SURMOUNT programmes, and the big semaglutide trials before them, mostly used 40 to 72 week windows. TRIUMPH-1 followed 2,339 adults with obesity, none of them with type 2 diabetes, for 80 weeks - about a year and a half - with doses of 4 mg, 9 mg, or 12 mg once weekly, or placebo.
That matters because these drugs keep working for longer than people expect. The weight-loss curves in this trial had not flattened out by the end. The mean weight change at 80 weeks was minus 17.6 percent at 4 mg, minus 23.7 percent at 9 mg, and minus 25.0 percent at 12 mg, against minus 3.9 percent on placebo. In other words, the 12 mg arm did not just beat the placebo group, it beat every single obesity medicine that has ever completed a phase 3 trial.
But the interesting structural point is the shape of the curve. A plateau at week 80 means the body has settled at a new set point, and maintenance becomes the whole game. No plateau means the drug keeps finding more to lose, which raises a different question: where does it end, and what does the extension study show?
The co-benefits were built into the trial design, not bolted on
Here is the part that made the TRIUMPH-1 design genuinely clever, and that most coverage glossed over. This was what is called a basket trial. Nestled inside the main obesity study were two nested sub-studies: 574 participants with knee osteoarthritis, and 243 with obstructive sleep apnoea. The primary endpoints of the paper were not just weight. They were also the change in knee pain on the WOMAC scale in the arthritis subgroup, and the change in apnoea-hypopnea index in the sleep apnoea group.
Both came back strongly positive, and both are worth unpacking.
In the knee osteoarthritis subgroup, pain scores on the WOMAC index dropped by about 3.2 to 3.6 points in the retatrutide groups, versus roughly 1.9 points on placebo. To be clear about what that means in the real world: these were people whose knees hurt badly enough to qualify, most of them also carrying a lot of extra load on those joints. Losing 20 to 25 percent of body weight mechanically unloads the knee. But the researchers asked whether there was also a direct anti-inflammatory or pain-modulation effect beyond what weight loss alone would explain, and the data suggest there is more going on than pure mechanical unloading. That is a big deal, because knee OA affects millions of people and the current toolbox, injections and painkillers and eventually joint replacement, is not great.
In the sleep apnoea group, the apnoea-hypopnea index - the count of breathing disruptions per hour - fell by about 31.7 to 34.3 events per hour in the 9 mg and 12 mg arms, versus roughly 9.9 events per hour on placebo. For context, most people in these studies started with an AHI of around 55 to 60 events per hour, which is severe sleep apnoea territory. A drop of 30-odd events is enough to take many participants out of severe range entirely. There was a separate companion paper in the same issue of the NEJM covering the sleep apnoea data in full, and it found similar results in a second trial that included people with type 2 diabetes.
The odd detail that nobody in mainstream coverage picked up: at 12 mg, the sleep apnoea benefit was actually slightly smaller than at 9 mg. The 9 mg arm hit the best AHI reduction. Whether that is a statistical quirk of a relatively small subgroup or a real signal about dose selection, we do not know yet, but it is the kind of nuance that gets lost when everyone just reads the headline.
What the Research Says
- The TRIUMPH-1 phase 3 trial randomised 2,339 adults with obesity but no diabetes to retatrutide (4 mg, 9 mg, or 12 mg weekly) or placebo for 80 weeks. Mean weight change was minus 17.6 percent, minus 23.7 percent, and minus 25.0 percent respectively, versus minus 3.9 percent with placebo. Source: Jastreboff AM et al., “Retatrutide, a Triple Hormone Receptor Agonist, for Treatment of Obesity,” New England Journal of Medicine, 29 September 2026, DOI: 10.1056/NEJMoa2604169.
- Nested sub-studies showed significant knee osteoarthritis pain reduction on the WOMAC scale (differences versus placebo of 1.4 to 1.8 points, p less than 0.001 under the intention-to-treat estimand) and significant apnoea-hypopnea index reduction in obstructive sleep apnoea (differences of 22 to 24 events per hour versus placebo, p less than 0.001). Same source, same paper.
- A companion phase 3 analysis in the same issue, “Retatrutide Once Weekly for Treatment of Obstructive Sleep Apnea,” DOI: 10.1056/NEJMoa2611477, covered 691 participants across two trials and reported the largest AHI reductions in the 9 mg arm.
- An independent analysis of self-reported experience among Reddit users taking unapproved retatrutide found a symptom profile that diverges noticeably from the trial data - more fatigue, insomnia, energy and muscle-related reports, fewer dominant gastrointestinal complaints - which the authors attribute in part to unregulated product quality and dosing. Source: medRxiv, “Self-Reported Side Effects Among Reddit Users Taking Unapproved Retatrutide,” 2026, https://www.medrxiv.org/content/10.64898/2026.05.28.26352819v1.full-text.
The gap between the trial and the grey market
That last source deserves its own paragraph, because it is the reality check this community needs. The TRIUMPH-1 results come from a trial where product identity, purity, and dose are controlled by the manufacturer. What people buy online as “retatrutide” is often none of those things, as the TGA’s lab testing of seized vials has already shown in Australia. The medRxiv analysis combed through dozens of subreddits and found that unapproved users report a different pattern of effects - and the authors are careful to say the differences could come from anything: product quality, stacking with other peptides, reconstitution errors, or just a different population.
We have covered this territory before, including the Victoria liver toxicity cases and the oesophagus tear incident. But it is worth repeating the core lesson: trial data for a compound does not transfer to whatever is actually in the vial you bought online. Different molecule, different dose, different risk profile.
If you want to follow how the retatrutide story develops - the FDA review, the 104-week extension data that will eventually answer the plateau question, the knee and sleep apnoea follow-ups - the Grey Highway Telegram group is where we track it as it lands. Our retatrutide compound page has the background on the GLP-1, GIP, and glucagon receptor triple agonist mechanism if you want the pharmacology primer.
What to watch next
Three things on the horizon:
- The extension phase of TRIUMPH-1 takes people out to 104 weeks, comparing those who went straight from placebo onto retatrutide with those who had 80 weeks on drug first. That is the closest thing we have yet to an answer on whether these losses hold or rebound.
- Regulatory review in the US is under way following Lilly’s mid-2026 filing, and a decision would reshape the whole competitive landscape against tirzepatide and the CagriSema programme.
- Whether the knee osteoarthritis and sleep apnoea effects become labelled indications in their own right, which would open a second pathway for prescribing beyond weight.
Long-term data is always the least exciting news and the most useful. Eighty weeks is not a lifetime, but it is a lot closer to it than most of what gets hyped in this space.
Disclaimer: This article is for educational and informational purposes only. It does not constitute medical advice, therapeutic recommendations, or endorsements of any compound. Grey Highway is a research-education community. We do not sell, supply, or promote the use of research compounds. Always consult a qualified healthcare professional regarding health decisions. For Australian regulatory information, visit the TGA website.