Three years of semaglutide cut surgeries by nearly 5% over eight years, researchers found
Adults with knee osteoarthritis who stayed on a newer GLP-1 drug for three years recorded 4.71 fewer total knee replacements per 100 patients over the following eight years.
A study published June 2 in Regional Anesthesia & Pain Medicine reported the finding, which Harvard Health Publishing described as nearly 5 percent fewer knee replacements across that window.
The finding matters for anyone tracking the downstream costs of a drug class that has strained benefit plans.
The researchers, led by Victoria Carter, set out to test whether GLP-1 medications do more than ease arthritis pain, examining whether patients who took them ultimately avoided or delayed surgery.
Drawing on the TriNetX Global Research Network, the study identified adults diagnosed with knee osteoarthritis between 2010 and 2024, then matched those who took GLP-1 drugs against similar patients who did not.
After propensity score matching, the cohorts ranged from 13,351 to 42,062 patients, the authors reported.
Some participants used older agents such as liraglutide, dulaglutide, exenatide or lixisenatide, while others used the newer semaglutide (sold as Ozempic, Rybelsus and Wegovy) or tirzepatide (Mounjaro and Zepbound), Harvard Health Publishing noted.
Any GLP-1 use, whether over one or three years, was linked to a lower rate of total knee replacement across every follow-up interval, the study found (all p<0.001).
One year of exposure to any GLP-1 drug produced an absolute risk difference of 2.80 percentage points at eight years, as per the paper, with a hazard ratio of 0.90.
The effect grew with longer treatment and newer agents: three years of semaglutide or tirzepatide yielded a 4.71 percentage point reduction at eight years and a hazard ratio of 0.72.
The researchers put the potential scale in context.
A 1.44 percentage point reduction seen after three years of a newer GLP-1 drug would translate to roughly 14,400 fewer knee replacements a year in the US, the authors estimated, alongside lower healthcare spending and less surgical morbidity.
The authors cautioned against reading too much into the numbers.
The study was observational, and the team could not rule out residual confounding from factors such as frailty, physical activity or disease severity.
The TriNetX database lacked radiographic data, validated pain scores, inflammatory markers and longitudinal weight measures, the authors wrote, and exposure was defined by prescription orders rather than confirmed dispensing.
The findings should be read as associations consistent with possible disease-modifying effects rather than proof of causation.
Earlier research had already tied GLP-1 use to reduced knee pain, likely because weight loss relieves stress on the joints, Harvard Health Publishing reported.
This study went further by testing surgical outcomes, and the authors argued the pattern, stronger with longer use and newer drugs, points to benefits beyond weight loss alone.
Confirming that will require prospective trials that measure dosing, duration and disease progression directly.
