Two things happen to joints on a GLP-1, and they pull in opposite directions.
First, carrying less weight means less load through every step, and the evidence that this eases pain is good. Second, feeling better makes people start moving again, often enthusiastically and on tissue that hasn't been asked to do much in years. Meanwhile the same rapid weight loss is taking muscle that was helping to stabilize the joint.
Which of those you notice first depends on where you started and how you go about it.
The good news is well-evidenced
The strongest evidence comes from a trial designed to answer exactly this question. STEP 9 randomized 407 people with obesity and moderate knee osteoarthritis to weekly semaglutide 2.4 mg or placebo for 68 weeks, on top of diet and activity counseling. Weight fell 13.7% with semaglutide versus 3.2% with placebo, and knee pain scores improved substantially more: a 41.7-point drop on the WOMAC pain scale versus 27.5 with placebo. Physical function improved more too (Bliddal et al., New England Journal of Medicine, 2024, DOI (external link)).
Keep two things in mind about that result. It's a large, well-run trial, and it was funded by the medication's manufacturer; that's worth knowing, though it doesn't invalidate the finding. And look at the placebo arm. That group lost only about 3% of its weight but still improved 27.5 points, with counseling, time, and the usual placebo response. Semaglutide added roughly 14 points on top. The extra weight loss is the likeliest explanation for that gap, but the trial wasn't designed to separate the drug's effect from the weight loss.
The mechanism is straightforward, and older work measured it directly. The IDEA trial put 454 overweight and obese adults with knee osteoarthritis through 18 months of intensive diet, exercise, or both. Diet-based weight loss measurably reduced the compressive force going through the knee compared with exercise alone, and the group that did both diet and exercise ended up with less pain and better function than either alone (Messier et al., JAMA, 2013, DOI (external link)).
That combination result is the practical headline. Weight loss and movement together beat either by itself.
What the evidence does not show
The same IDEA participants had X-rays and MRIs at 18 months. There was no statistically significant difference between any of the three groups in joint space narrowing or cartilage thickness loss (Hunter et al., Osteoarthritis and Cartilage, 2015, DOI (external link)).
So substantial weight loss made knees hurt less and work better, without measurably changing the joint's structure over a year and a half. Feeling better is a real and worthwhile outcome, and it's the one patients care about. But "weight loss reverses arthritis" or "regrows cartilage" isn't what was found, and you should be skeptical of anything sold on that promise.
Why your joints might hurt more right now
If your knees, hips, or shoulders feel worse a few months in, you're not doing it wrong, and it usually isn't the medication. Three ordinary explanations cover most cases.
You started moving again. Losing 20 or 30 pounds makes activity feel possible, so people take up walking, hiking, the gym, and tennis all at once. Cartilage adapts to load reasonably well, but tendon adapts slowly, and only in response to gradually increasing load. In a meta-analysis of exercise studies lasting at least 8 weeks, tendons did get stiffer and stronger with training, with somewhat larger changes in programs of 12 weeks or more (DOI (external link)). Doing in week two what you'd like to be doing in month six is how you find your Achilles tendon.
You lost muscle along with the fat. GLP-1 weight loss takes lean tissue with it: roughly 25-40% of the weight lost in recent trials (DOI (external link)). Muscle around a joint is part of what protects it. In a cohort of over 3,000 knees followed for 30 months, people in the highest third for knee extensor strength had roughly half the odds of developing symptomatic knee osteoarthritis compared with the weakest third, though thigh strength did not predict structural, X-ray-visible OA (Segal et al., Arthritis & Rheumatism, 2009, DOI (external link)).
That finding mirrors the IDEA results: strength protects how your knee feels, not what it looks like on film. It also means muscle loss and joint pain are closely linked, which is why strength training shows up as the answer to both.
Your body changed shape faster than your mechanics adjusted. Gait, balance, and load distribution all shift as body mass drops. Some aches are the joint learning a new pattern.
What actually helps
Strength training, started early and progressed slowly. This is the most valuable thing on the list. It protects the muscle that rapid weight loss is taking, and stronger thighs go with fewer symptomatic knees. If sore joints are why you've been avoiding it, that's backwards: for most activity-related aches, sensible loading helps. Our beginner resistance-training plan covers sets, reps, and how progressive overload works.
Respect the tendon timeline. Muscle gets stronger in weeks; tendon takes months. When something aches, the usual fix is to reduce the load and keep moving, not to stop entirely, because complete rest deconditions the tissue further. A rule of thumb many physiotherapists use: pain that settles within about 24 hours after activity is generally acceptable, while pain that's worse the next morning means you did too much.
Hit your protein. You can't build or defend muscle without it, and appetite suppression makes this the hardest target on a GLP-1. See high-protein meals and snacks for low appetite.
Get help with technique if you're new to this. A knee that hurts under a bad squat pattern often stops hurting under a good one. How to find a personal trainer covers what credentials mean and what to ask.
About the supplements
Joint supplements are a large, noisy market. The fairest available comparison is a network meta-analysis that ranked seven of them across 39 trials and 4,599 patients with knee osteoarthritis. Boswellia came out highest for pain and stiffness; curcumin, collagen, ginger, and krill oil showed benefits on some outcomes; and none increased side effects versus placebo (Zhang et al., Nutrients, 2025, DOI (external link)).
Collagen has real trial support. In 250 people with knee osteoarthritis, 10 g daily of hydrolyzed collagen improved joint comfort over six months (Benito-Ruiz et al., International Journal of Food Sciences and Nutrition, 2009, DOI (external link)), and undenatured type II collagen at just 40 mg daily outperformed glucosamine plus chondroitin over 180 days (Lugo et al., Nutrition Journal, 2016, DOI (external link)).
One trial may fit this audience better than the osteoarthritis studies do. 139 athletic people with knee pain but no arthritis took 5 g of collagen peptides daily for 12 weeks and reported significantly less activity-related pain than placebo (Zdzieblik et al., Applied Physiology, Nutrition, and Metabolism, 2017, DOI (external link)). Knees that complain during new activity, without diagnosed arthritis, describe a lot of people on a GLP-1.
All of these are modest effects on top of the things that matter more. Our collagen guide explains why the Type I/II/III numbers mean less than the label implies, and the supplements overview puts the rest in order.
When it's not just an ache
Talk to a clinician instead of training through it if you have joint swelling, redness, or warmth; a joint that locks, catches, or gives way; pain that wakes you at night; pain following a specific injury; or a single joint that's suddenly and severely painful. Rapid weight loss can also shift uric acid levels. After bariatric surgery, for example, uric acid tends to rise in the first month and gout flares are more common early on, before both fall below pre-surgery levels (DOI (external link)). GLP-1 drugs and tirzepatide tend to lower uric acid over time (DOI (external link)), but a sudden hot, exquisitely painful joint (classically the big toe) is worth having looked at instead of assuming it's wear and tear.
The short version
For most people who bring joint pain into GLP-1 treatment, it gets better, and the trial evidence for that is solid. It shows up as less pain and better function, not a repaired joint, and in STEP 9 much of the improvement also happened in the placebo group.
If yours got worse first, the usual culprits are doing too much too soon and losing the muscle that was helping hold the joint together. Both have the same answer: start strength training earlier than feels necessary, progress more slowly than feels satisfying, and eat enough protein to keep what you're defending. Supplements come last.
Research findings above are attributed to peer-reviewed articles indexed in PubMed. This is general education, not medical advice; new, severe, or persistent joint pain deserves an assessment from your prescriber or a musculoskeletal clinician.