The lean-mass problem on GLP-1s has been covered to death. The bone problem has not, and it is the one with five-year data behind it. A retrospective cohort of 73,483 matched patients presented at the 2026 Annual Meeting of the American Academy of Orthopaedic Surgeons reported osteoporosis in 4.1% of GLP-1 receptor agonist users versus 3.2% of matched controls at five years — a relative risk of 1.29. Osteomalacia showed the largest relative gap in the dataset: 2.0% versus 0.1%.
The reason this has not turned into a headline cycle is that the evidence points in two directions at once. A real-world cohort in type 2 diabetes reported a hazard ratio of 0.69 — a lower osteoporosis risk on GLP-1s. A separate 2-million-patient analysis found both results in the same dataset, split by whether the patient had diabetes. The signal is real, but it is conditional, and almost every summary circulating online picks one half of it.
Research-context information only. Compounds discussed below are research peptides and supplements; some are investigational drugs not approved by the FDA. Semaglutide and tirzepatide are active ingredients in FDA-approved prescription products; the research-peptide and compounded forms discussed and linked below are not FDA-approved and are sold for research purposes only. Protocols, doses, and reactions reported come from published research and self-reported community sources. Possession or use of investigational drugs outside an authorized clinical trial may be illegal in your jurisdiction. This article reports what has been documented, not what should be done. Consult a licensed physician for personal medical decisions.
What the five-year records show
The AAOS 2026 dataset is a matched retrospective cohort drawn from patients carrying both a type 2 diabetes and an obesity diagnosis, followed for five years, with GLP-1 users matched one-to-one against non-users. The musculoskeletal outcomes reported:
| Outcome at 5 years | GLP-1 users | Matched controls | Risk ratio |
|---|---|---|---|
| Osteoporosis | 4.1% | 3.2% | 1.29 |
| Gout | 7.4% | 6.6% | 1.12 |
| Osteomalacia | 2.0% | 0.1% | 2.55 |
All three differences were reported as statistically significant. The same AAOS session paired this with a separate retrospective analysis of ten common orthopaedic procedures, which reported better short-term surgical outcomes for GLP-1 users — lower odds of postoperative emergency department visits, reduced surgical site infection rates in total knee and hip arthroplasty, and lower knee revision rates. The authors' framing was that short-term perioperative benefit and long-term skeletal risk are not mutually exclusive, and that bone health surveillance is the gap.
The osteomalacia figure deserves a caveat that the coverage generally drops. A jump from 0.1% to 2.0% is a 25-fold absolute increase in a condition that is diagnosed largely on the basis of vitamin D status and bone biochemistry. In a population undergoing rapid weight loss with substantially reduced caloric intake — and often reduced fat-soluble vitamin absorption — a diagnostic-detection effect is difficult to rule out in a retrospective claims-style dataset.
The split almost nobody reports: diabetes status changes the sign
The most informative dataset on this question is an Epic Research analysis published in March 2026 that ran the same question across two separate populations: roughly 2 million patients with type 2 diabetes and 380,438 without. Stratifying by amount of weight lost, the direction of the effect reversed depending on diabetes status.
Among patients with type 2 diabetes who maintained stable weight, GLP-1 use was associated with an 8.7% lower osteoporosis risk than non-users, and the protective gap persisted as weight loss increased. Among patients without diabetes who maintained stable weight, GLP-1 use was associated with a 22% higher osteoporosis risk than other weight-loss medications.
That mirrors a real-world retrospective cohort out of Tangdu Hospital covering 2012 to 2023, which reported a fully adjusted hazard ratio of 0.69 (95% CI 0.45–0.84) for incident osteoporosis among GLP-1 receptor agonist users with type 2 diabetes (PMID 40469447). Type 2 diabetes is itself associated with impaired bone quality, so a treatment that improves glycaemic control has a plausible route to net skeletal benefit in that group that simply does not exist in a metabolically healthy person using the same drug purely to lose weight.
The practical read: the population most of the current research-peptide market resembles — people without diabetes, using a GLP-1 for body composition — sits on the side of the split where the reported risk is higher, not lower.
Why density falls: loading, lean mass, and turnover
The cleanest interventional evidence is a randomised, double-blinded, placebo-controlled phase 2 trial run across two Danish hospitals in 64 adults selected for increased fracture risk — a T-score below −1.0 at the total hip or lumbar spine, or a low-energy fracture within three years. Participants received once-weekly semaglutide 1.0 mg or placebo for 52 weeks (PMID 38737002).
At 52 weeks the semaglutide group had lower areal bone mineral density at both measured sites: an estimated treatment difference of −0.018 g/cm² at the lumbar spine (95% CI −0.031 to −0.005, p = 0.007) and −0.020 g/cm² at the total hip (95% CI −0.032 to −0.008, p = 0.001) — approximately 2.1% and 2.6% respectively. The turnover markers point at the mechanism: P1NP, a bone formation marker, showed no between-group difference, while CTX, a bone resorption marker, was higher on semaglutide. Formation held steady; breakdown accelerated.
Three mechanistic threads run through the literature, and none of them requires the drug to be directly toxic to bone:
- Mechanical unloading. Bone remodels to the load placed on it. Removing 15–25% of body mass removes a proportional share of the daily compressive load, and the skeleton adapts downward.
- Lean mass goes with it. A network meta-analysis of 22 randomised trials and 2,258 participants reported GLP-1 receptor agonists reducing total body weight by 3.55 kg, fat mass by 2.95 kg and lean mass by 0.86 kg — with lean mass accounting for roughly 25% of the total loss (PMID 39719170). A 2026 meta-analysis in the International Journal of Obesity covering 7 studies and 821 patients reported absolute lean mass down 1.74 kg even as lean mass as a proportion of body weight rose 1.81%. Muscle pulling on bone is a primary osteogenic stimulus; less muscle means less of it.
- Intake falls with appetite. Substantially reduced food intake reduces calcium, vitamin D and protein intake at exactly the point the skeleton is remodelling fastest. This is the thread that also makes the osteomalacia number plausible rather than anomalous.
Every one of those mechanisms is shared with any rapid weight loss — bariatric surgery cohorts show the same directional pattern. What is specific to the current moment is the number of people losing that much mass that fast, without a surgical team monitoring their bone biochemistry on the way down.
What it changes for anyone running a GLP-1 protocol
Nothing here changes availability, legality or pricing for any compound. What it changes is what belongs on the monitoring list next to the usual metabolic panel.
The interventions with the most documentation behind them in this literature are unglamorous and non-pharmacological: progressive resistance training, adequate protein, and repletion of vitamin D and calcium. A clinical trial currently recruiting (NCT07554417) is testing precisely that — whether a 12-week structured exercise and individualised nutrition programme reduces muscle and bone loss in adults with obesity on a GLP-1. That trial has not reported.
For the compound side, our ranked breakdown of what is actually documented for protecting lean mass on a GLP-1 is in Preserve Muscle on Semaglutide & Tirzepatide, and the growth-axis evidence specific to bone turnover is below. Current per-vendor $/mg and COA coverage for the compounds named here are on Best Semaglutide Vendors, Best Tirzepatide Vendors and Best MK-677 Vendors, with active codes on the deals page.
