TL;DR: In the phase 3 TRIUMPH-4 trial, 445 adults with obesity and knee osteoarthritis took retatrutide 9 mg, 12 mg, or placebo for 68 weeks. Weight loss reached up to 28.7%, and WOMAC pain scores improved roughly 74 to 76% on both active doses, with about 1 in 7 patients on the lower dose becoming completely pain-free. The trial cannot tell us how much of that relief is unloaded joints versus a direct anti-inflammatory effect of the drug itself, and that question is genuinely open.
The trial: TRIUMPH-4 in one paragraph
Eli Lilly’s TRIUMPH program is a set of phase 3 trials testing retatrutide, a triple agonist that activates GLP-1, GIP, and glucagon receptors, across different weight-related conditions. TRIUMPH-4 is the arm built specifically around knee osteoarthritis. It enrolled 445 adults with obesity or overweight who also had moderate-to-severe knee osteoarthritis; people with type 2 diabetes were excluded, so the results reflect a metabolic-but-not-diabetic population. Participants were randomized 1:1:1 to subcutaneous retatrutide 9 mg, retatrutide 12 mg, or placebo, once weekly, for 68 weeks. Lilly announced topline results in December 2025, and it was the first of the TRIUMPH trials to report.
This is worth sitting with for a second: osteoarthritis trials are usually run by rheumatology and orthopedics groups testing joint-specific drugs. TRIUMPH-4 is a metabolic company testing a metabolic drug and measuring a joint outcome as a primary endpoint. That framing alone tells you something about where the field thinks the disease actually starts.
What WOMAC actually measures
If you haven’t seen it before, WOMAC (the Western Ontario and McMaster Universities Osteoarthritis Index) is the standard yardstick for knee and hip osteoarthritis in clinical trials. It’s a patient-reported questionnaire covering three areas: pain (walking, using stairs, resting, in bed), stiffness, and physical function (getting out of a chair, putting on socks, getting in and out of a bathtub). Higher scores mean worse symptoms.
The WOMAC pain subscale specifically asks patients to rate pain across five everyday movements on a numeric scale, then sums to a total. A “74 to 76% improvement” in this context means the average patient’s reported pain during those activities dropped by roughly three-quarters from where they started. It’s not a proxy or a lab value; it’s people saying stairs hurt less, walking hurts less, and standing up from a chair hurts less than it did a year and a half earlier.
The results, dose by dose
Across the two active arms, the weight loss was substantial even by GLP-1 standards. Retatrutide 12 mg produced average weight loss up to 28.7% of body weight at 68 weeks; the 9 mg arm reached roughly 26.4%. Placebo, as usual in these trials, moved only slightly.
The pain outcome tracked right alongside it. WOMAC pain scores improved by roughly 74 to 76% on both active doses, a reduction of a little over 4 points on the WOMAC pain subscale, compared with a smaller improvement on placebo. Physical function scores moved in the same direction. And then there’s the number that tends to stick with people: at 68 weeks, about 14.1% of patients on 9 mg and 12.0% on 12 mg reported being completely free of knee pain, versus about 4.2% on placebo.
Share of patients reporting zero knee pain on the WOMAC pain subscale at week 68.
Notice that the 9 mg dose edged out the 12 mg dose on the pain-free rate, even though 12 mg produced more weight loss. That’s a small detail in a topline release, not a statistically confirmed dose-response reversal, but it’s a genuinely interesting seam in the data, and it’s one of the reasons the mechanism question below doesn’t have a tidy answer yet.
| Outcome at 68 weeks | Retatrutide 9 mg | Retatrutide 12 mg | Placebo |
|---|---|---|---|
| Mean body weight change | About 26.4% | Up to 28.7% | About 2.1% |
| WOMAC pain score improvement | About 75.8% | About 74.6% | Smaller improvement |
| Completely pain-free | 14.1% | 12.0% | 4.2% |
Load reduction or direct drug effect? The open question
Here’s the part that makes this trial more than just another weight-loss headline. There are two plausible, non-exclusive explanations for why a metabolic drug relieved joint pain, and TRIUMPH-4 was not designed to tell them apart.
The mechanical story is the simple one. Knee joints carry roughly three to six times body weight with each step. Lose 28% of your body weight and the load through an arthritic knee drops substantially with every stair and every stride. Less compressive force on already-damaged cartilage plausibly means less pain, independent of anything happening at the receptor level. This is the default explanation, and it doesn’t require any new biology.
The direct-effect story is more speculative but not far-fetched. GLP-1 receptors have been identified on chondrocytes (the cells that make up cartilage) and in synovial tissue, and lab studies with GLP-1 receptor agonists like liraglutide have shown suppressed inflammatory signaling in cartilage cells, protection against inflammation-induced cell stress, and slowed cartilage degeneration in animal models of osteoarthritis. Separately, GLP-1 drugs are known to lower systemic inflammatory markers such as IL-6 and CRP in people with type 2 diabetes, and both cytokines are implicated in osteoarthritis progression. So there’s a real, published biological basis for a joint-specific, weight-independent effect. It just hasn’t been isolated in humans at trial scale.
The 9 mg versus 12 mg pain-free discrepancy noted above is a small hint that mechanics alone may not explain everything, since more weight loss at 12 mg didn’t translate into a higher pain-free rate. But it’s one data point from a topline release, not a mechanistic proof, and it could just as easily be noise. Future analyses, including any correlation between change in inflammatory markers and change in WOMAC score independent of weight change, would be the kind of evidence that could actually move this question forward.
Where this fits with retatrutide overall
TRIUMPH-4 doesn’t stand alone. It’s one readout in a broader retatrutide development program that includes trials in general obesity and type 2 diabetes, and the weight-loss magnitude here is consistent with what retatrutide has shown elsewhere. If you’re unfamiliar with the drug itself, our what is retatrutide explainer covers the triple-agonist mechanism, and the retatrutide vs tirzepatide comparison lays out how it stacks up against the closest approved alternative.
The 9 mg and 12 mg doses used in TRIUMPH-4 sit toward the higher end of retatrutide’s studied dose range, which means titration matters. Ramping too fast into these doses is a common way people front-load side effects they didn’t need to have; our retatrutide dosing and titration guide walks through the standard schedule, and the GLP-1 side effect management guide covers the nausea, constipation, and appetite changes that showed up as the most common adverse events in this trial too.
There’s also a muscle question worth flagging for anyone losing 25%+ of body weight over 68 weeks on a triple agonist: a meaningful share of rapid GLP-1 weight loss can come from lean tissue, and losing muscle around an arthritic knee is not obviously a win even if the scale looks great. Our preventing muscle loss on GLP-1s piece covers the protein and resistance-training levers that matter most during a cut this size.
Where the program goes next
Lilly has said it plans to include knee osteoarthritis among the initial indications in its regulatory submission for retatrutide, expected with the FDA in late 2026. That’s a plan, not an approval, and it sits well outside the scope of what this article is trying to cover. The more interesting near-term development, from a research standpoint, is whether follow-up analyses of TRIUMPH-4 or a purpose-built mechanistic study can actually separate the load-reduction effect from a direct joint effect. Until then, what we have is a large, well-controlled trial showing that retatrutide reduced knee pain by a lot, alongside a mechanistic case for why that might be more than gravity doing the work.
Frequently asked questions
What is the TRIUMPH-4 trial?
TRIUMPH-4 is a phase 3, 68-week, placebo-controlled trial testing retatrutide in 445 adults with obesity or overweight and knee osteoarthritis, without type 2 diabetes. Participants were randomized 1:1:1 to retatrutide 9 mg, 12 mg, or placebo. Lilly announced topline results in December 2025, reporting weight loss up to 28.7% and WOMAC pain score improvements of roughly 74 to 76% on the active doses.
How much did WOMAC pain scores improve on retatrutide?
WOMAC pain scores, a standard patient-reported measure of knee pain during activities like walking and stairs, improved by about 75.8% on the 9 mg dose and about 74.6% on the 12 mg dose at 68 weeks, compared with a smaller improvement on placebo. In practical terms, that’s roughly a three-quarter reduction in reported pain during everyday movements.
Were any TRIUMPH-4 patients completely free of knee pain?
Yes. At 68 weeks, about 14.1% of patients on retatrutide 9 mg and 12.0% on 12 mg reported zero pain on the WOMAC pain subscale, versus about 4.2% on placebo. That means roughly 1 in 7 patients on the lower dose reached a pain-free state, a result that doesn’t happen by coincidence at that frequency in an osteoarthritis population.
Does retatrutide relieve knee pain by reducing joint load or through a direct effect?
Both are plausible and TRIUMPH-4 wasn’t designed to separate them. Weight loss of up to 28.7% meaningfully cuts the mechanical load an arthritic knee carries with every step, which alone could explain much of the improvement. But GLP-1 receptors are present in cartilage and synovial tissue, and lab studies show GLP-1 receptor agonists can suppress joint inflammation and lower systemic markers like IL-6 and CRP independent of weight change, so a direct anti-inflammatory contribution is also biologically plausible.
Is retatrutide approved for knee osteoarthritis?
No. TRIUMPH-4 topline results were announced in December 2025, and Lilly has said it intends to include knee osteoarthritis among the initial indications in its planned FDA submission for retatrutide in late 2026. That’s a submission timeline, not a current approval or clinical use.
How does the TRIUMPH-4 weight loss compare to other retatrutide trials?
The up to 28.7% weight loss seen in TRIUMPH-4’s knee osteoarthritis population is in line with the magnitude retatrutide has produced in its broader obesity trial program, reinforcing that this is a consistent, dose-dependent effect of the drug rather than something specific to an osteoarthritis subgroup.