Explore this article's sources with AI
Follow PeptideMethods on Google
Tirzepatide vs Semaglutide in Arthroplasty
A records study found similar early outcomes after knee or hip arthroplasty, with two uncertain differences appearing at two years.
Why we wrote this. The two-year findings look favorable for tirzepatide, but the study design and multiple comparisons make a cautious read essential.
In this article (5 sections)
A retrospective US records study compared people prescribed tirzepatide with people prescribed semaglutide around primary knee or hip replacement. After matching 2,858 people in each group, the researchers found no statistically significant differences in the measured outcomes at 90 days or one year. At two years, the tirzepatide group had lower recorded odds of mechanical complications and periprosthetic joint infection. Reoperation was not significantly different[1]. These are associations from coded health records, not evidence that one medicine caused better surgical outcomes.
How the researchers built the comparison
The investigators used the TriNetX US Collaborative Network, a federated electronic health record database. The query ran on May 31, 2026, when the network held records from 66 participating US health care organizations. It identified adults who underwent primary total knee or hip arthroplasty and had at least two coded prescription fills for either drug within the six months up to surgery. Anyone with recorded use of both medicines was excluded. Of 669,698 people with an eligible procedure, 8,797 met that medication definition before matching[2]. A coded fill strengthens the exposure definition compared with a single prescription, but it does not prove that a person took the medicine or remained on it. The data were de-identified and aggregated, so the researchers did not review individual charts to validate each exposure or outcome.
Before matching, 2,864 records were in the tirzepatide cohort and 5,933 were in the semaglutide cohort. The groups differed in age, race, ethnicity and several recorded conditions. The researchers then matched patients one to one using age, sex, race, ethnicity and diagnoses that included obesity, type 2 diabetes, kidney disease and substance-related conditions. All reported standardized differences were below 0.05 after matching[2]. Matching can balance measured fields. It cannot remove differences that were not captured accurately in the database.
The early outcomes were similar
At 90 days, confidence intervals crossed 1.0 for surgical-site infection, sepsis, lower-limb cellulitis, wound disruption, emergency-department visits, readmission and early mechanical complications. At one year, the same was true for mechanical complications, periprosthetic fracture, instability or dislocation, reoperation and periprosthetic joint infection[2]. The study therefore did not identify a statistically significant advantage for either prescription group across its early comparisons.
That absence of a detected difference is narrower than proof that the drugs are equivalent. Event counts were limited for several implant outcomes, and wide confidence intervals left room for effects in either direction. The study's mechanical-complication category combined broken or dislocated prostheses, instability, loosening, osteolysis and bearing-surface wear. The paper also combined knee and hip arthroplasty in the main comparison, so its headline estimates do not provide a procedure-specific answer.
Two signals appeared at two years
At two years, tirzepatide exposure was associated with lower odds of any mechanical complication than semaglutide exposure (odds ratio 0.64, 95% confidence interval 0.42 to 0.96). The reported odds ratio for periprosthetic joint infection was also 0.64, with a 95% confidence interval of 0.41 to 0.999 and P = .048. Reoperation remained nonsignificant at an odds ratio of 0.68 (95% confidence interval 0.43 to 1.08). Periprosthetic fracture and instability or dislocation also did not differ significantly[2].
The infection result sits barely inside the paper's P < .05 threshold. The authors assessed numerous outcomes at three time points without adjusting for multiple comparisons, which raises the chance of finding at least one nominally significant result by chance. They also describe limited event counts. Taken together, those features make the two-year findings signals to test, not a basis for choosing a metabolic medicine around surgery.
Why the study cannot establish a drug effect
Treatment was not assigned randomly. People who received dual GIP and GLP-1 therapy may have differed from those receiving selective GLP-1 therapy in ways the matching fields did not capture. The database could not separate treatment for diabetes from treatment for obesity, stratify by dose or duration, or account fully for the amount and timing of weight change. Diagnoses and procedures were identified through codes, and the analysis could not tell whether a later event involved the operated joint or the opposite side[2].
The paper discusses possible explanations involving weight, inflammation and bone biology, but it did not measure a GIP-specific mechanism behind the surgical outcomes. A prospective comparison with dispensing or adherence data, weight and glycemic trajectories, procedure-specific results and a prespecified analysis plan would be needed to test whether the association reflects the medicine rather than patient selection or chance.
What remains unknown for patients and surgical teams
This analysis does not answer whether either drug should be started, stopped or switched before arthroplasty. It also does not evaluate anesthesia-related decisions, individual contraindications or current perioperative instructions. Those questions require a clinician who can review the reason for treatment, the planned procedure and the patient's current medication plan.
Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.
Frequently asked
Was tirzepatide randomized against semaglutide around joint replacement?
No. Researchers retrospectively compared electronic health records and matched the groups on recorded factors. The study can identify associations but cannot establish that either medicine caused an outcome.
Did the medicines differ at 90 days or one year?
No statistically significant differences were reported for the measured surgical, medical or implant-related outcomes at those time points. That does not prove exact equivalence, particularly for uncommon events.
What differed at two years?
The tirzepatide group had lower recorded odds of mechanical complications and periprosthetic joint infection. Reoperation, fracture and instability or dislocation were not significantly different.
Should this study change a perioperative medication plan?
The study does not support starting, stopping or switching either medicine for arthroplasty outcomes. Medication decisions around surgery require individualized guidance from the prescribing and surgical teams.
Sources
- [1]Davis SC et al. Dual GIP/GLP-1 receptor agonist versus selective GLP-1 receptor agonist in patients undergoing primary total knee or hip arthroplasty. European Journal of Orthopaedic Surgery & Traumatology. 2026. PMID 42905201.Tier 1 · primary↩
- [2]Davis SC et al. Full open-access article and supplementary information. Springer Nature. 2026.Tier 1 · primary↩
No revisions yet. First published .