GLP-1s and shoulder replacement safety

A 2026 retrospective database study found higher 90-day readmission and overall complication rates among non-diabetic adults with obesity who had a recent glucagon-like peptide-1, or GLP-1, agonist prescription before shoulder arthroplasty. It did not find significant differences in the measured two-year outcomes, including revision. This is an association in claims data, not proof that a GLP-1 medicine caused the early outcomes. It should not prompt a change to prescribed treatment without the surgical and prescribing teams.^[1]

The study is useful because it focused on a group that earlier shoulder-replacement research often did not isolate: people with obesity but without diabetes. It also leaves practical questions open, including why readmissions were more frequent and whether the finding would appear in prospectively collected data. If you take a medicine such as semaglutide, the relevant decision is individual perioperative planning with clinicians who know your operation and prescription history.^[1]

What the researchers compared

The authors used the PearlDiver Mariner administrative-claims database to identify people having anatomic or reverse total shoulder arthroplasty in the United States between 2010 and 2023. They included adults with a body mass index of at least 30 kg/m² within six months of surgery and excluded people with diabetes. They also excluded operations performed for cancer, infection, or pathologic fracture.^[1]

A GLP-1 user was defined by a prescription within three months of surgery. The investigators matched each user with up to ten non-users using age, sex, smoking status, shoulder-replacement type, comorbidity measures, and listed conditions. The final matched sample included 462 GLP-1 users and 4,377 non-users. Matching can make groups more alike on recorded factors, but it cannot account for every difference that may influence care or readmission.^[1]

The 90-day finding: readmissions drove the difference

At 90 days, 107 of 462 GLP-1 users were readmitted, compared with 571 of 4,377 non-users. The adjusted odds ratio for readmission was 2.034, with a 95% confidence interval from 1.594 to 2.578. The reported rate of any complication was also higher in the GLP-1 group: 43.94% versus 34.59%, with an adjusted odds ratio of 1.474.^[1]

That broad complication measure needs context. The paper reports no statistically significant between-group difference for the other individual 90-day outcomes it evaluated, including surgical-site infection, prosthetic joint infection, venous thromboembolism, acute kidney injury, pneumonia, emergency-department visits, reoperation, or transfusion. The authors describe the higher overall rate as largely driven by readmissions, not as evidence of a demonstrated increase in each complication on that list.^[1]

What the two-year results did and did not show

The study found no statistically significant differences at two years for the outcomes measured: prosthetic joint infection, fracture, loosening, dislocation, osteolysis, or all-cause revision. For revision, the adjusted odds ratio was 0.710, but its confidence interval was wide and crossed 1.0. That result is compatible with no detected difference in this dataset; it does not establish that GLP-1 use protects a shoulder replacement from revision.^[1]

The operation matters here. Shoulder arthroplasty includes anatomic and reverse designs, and the study adjusted for arthroplasty type rather than presenting a separate causal answer for every surgical situation. A reader planning shoulder replacement should not transfer these population-level estimates to their own risk without a discussion with the care team. For broader background, see our semaglutide overview and the semaglutide regulation information.

Why this study cannot settle perioperative medication decisions

This was not a randomised trial. It used prescription and billing records, so it cannot show whether patients took the medicine as recorded, identify every reason for treatment selection, or establish the reason for a readmission. The authors could not determine why the observed readmissions occurred. They discuss gastrointestinal effects and medication reinitiation as possibilities, but those explanations were not tested by the database analysis.^[1]

The result also applies to a defined population: non-diabetic adults with obesity who had a recent GLP-1 prescription and underwent elective shoulder arthroplasty in the study period. It does not directly answer outcomes for people with diabetes, people using a specific individual GLP-1 medicine, or patients whose treatment timing differs from the study definition. It also does not provide a medication-holding schedule.^[1]

What we do not yet know

We do not yet know what explains the readmission signal, whether it can be reproduced in other health systems, or whether a prospective study would find the same pattern. We also do not know from this paper how individual medicine, dose, duration, symptoms, nutritional status, anaesthetic plan, or surgical factors contributed to outcomes. Those gaps matter more than a single odds ratio when care is being planned around an operation.^[1]

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Always consult a qualified healthcare professional before making decisions about a prescribed medicine or planned surgery. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.