GLP-1 drugs and knee replacement risk

A retrospective study of more than 42,000 propensity-matched patients found that people who used GLP-1 receptor agonists were significantly less likely to need a total knee replacement for osteoarthritis^[1]. Published on 2 June 2026 in Regional Anesthesia and Pain Medicine, the analysis drew on the TriNetX Global Research Network and reported a hazard ratio of 0.72 (95% CI 0.67 to 0.78) for patients on semaglutide or tirzepatide for at least three years. The effect appeared strongest with longer treatment and with newer agents in the class.

What the numbers show

Victoria Carter and colleagues used propensity-score matching to compare GLP-1 receptor agonist users with non-users in the TriNetX database, controlling for age, sex, BMI, diabetes status, and other confounders. They then divided the data by drug exposure window and compound. At one year of any GLP-1 receptor agonist use, the hazard ratio for total knee arthroplasty was 0.90 (95% CI 0.83 to 0.98), translating to an absolute risk difference of 2.80 percentage points over eight years of follow-up^[1]. When the analysis was restricted to three or more years of semaglutide or tirzepatide specifically, the association grew to a hazard ratio of 0.72 and an absolute risk difference of 4.71 percentage points. All primary comparisons reached statistical significance at p<0.001. Post-matching cohorts ranged from 13,351 to 42,062 patients depending on the exposure class and duration being tested.

These results land alongside a growing body of database studies pointing in the same direction. A 2025 Taiwanese cohort of 35,762 patients with type-2 diabetes reported that GLP-1 receptor agonist therapy reduced the risk of knee osteoarthritis itself (adjusted HR 0.85, 95% CI 0.78 to 0.93) and of total knee replacement (adjusted HR 0.91, 95% CI 0.89 to 0.98)^[2]. A US study of over 1.3 million obese patients published in the Orthopedic Journal of Sports Medicine found that among those with pre-existing osteoarthritis, GLP-1 receptor agonist users had a lower rate of total knee arthroplasty at one year (1.4% vs 2.1%, HR 0.8)^[3].

Why GLP-1 drugs might protect joints

The obvious explanation is weight loss. Excess body weight is the single largest modifiable risk factor for knee osteoarthritis, and the mechanical load on the knee joint rises roughly four times for every kilogram of body mass gained. GLP-1 receptor agonists like tirzepatide produce sustained weight reductions of 15 to 21% in clinical trials, which translates to a meaningful decrease in joint stress over time.

But the Carter study hints at something more than mechanical unloading. The effect was larger with newer dual-agonist agents and grew with longer treatment, which the authors say suggests "potential disease-modifying activity beyond weight loss alone"^[1]. Laboratory work has shown that GLP-1 receptors are expressed in chondrocytes (the cells that maintain cartilage), and preclinical models suggest that GLP-1 signalling may reduce inflammatory cytokines involved in cartilage degradation. This remains a hypothesis, not a confirmed mechanism.

What this study does not tell us

This is a retrospective database analysis, not a randomised controlled trial. The TriNetX network relies on electronic health records, which introduce the usual limitations: unmeasured confounders, inconsistent coding of osteoarthritis severity, and no standardised imaging to track cartilage changes over time. Propensity-score matching can balance known variables but cannot eliminate selection bias entirely. Patients who receive GLP-1 receptor agonists tend to have better access to healthcare and more regular clinical contact, which could itself reduce the likelihood of progressing to surgery through earlier non-surgical intervention.

The authors acknowledge this directly: prospective, randomised trials are needed to establish whether GLP-1 receptor agonists genuinely prevent or delay the need for knee replacement, or whether the association reflects other differences between treated and untreated populations^[1].

Where this fits for readers

If you are already on a GLP-1 receptor agonist for diabetes or weight management, this study adds to the evidence that joint health may be an additional benefit. It does not, on its own, justify starting a GLP-1 drug for osteoarthritis. No regulator has approved any drug in this class for joint protection, and no clinical practice guideline recommends it for that purpose. For country-specific prescribing rules, see the tirzepatide regulation pages. A separate 11-year analysis of total knee arthroplasty outcomes found that GLP-1 receptor agonist users had shorter hospital stays and higher rates of home discharge, though with more emergency department visits at 90 days^[4]. For questions about whether a GLP-1 receptor agonist is appropriate for your situation, talk to a clinician who knows your history.