The Fuller Picture: What 2026's GLP-1 Tendon and Muscle Data Actually Shows Your Post-Op Patients

Last time we talked GLP-1s on this blog, the news wasn't great — activity down, some lean mass and bone density going with it. A batch of new 2026 research on tendon healing and muscle quality specifically — as opposed to just "the scale" and "the DEXA" — is more encouraging in some genuinely important places, especially once surgery is already happening. But it's not a clean "good news" story either: there's a real, honest wrinkle on the front end that's worth knowing before you reassure a patient too quickly. Worth fifteen minutes before your next post-op eval.

The rotator cuff picture: reassuring after surgery, more complicated before it

Obesity and type 2 diabetes have a well-documented track record of making rotator cuff tendons heal worse — more re-tears, worse outcomes after repair. A scoping review in BMC Musculoskeletal Disorders this June (Elmaraghy et al.) set out to map what's actually been studied on GLP-1RAs and rotator cuff disease specifically. Read directly from the full paper rather than the abstract, the real answer is two-sided — not simply reassuring, and worth presenting to patients that way.

On the post-op side, the data is genuinely good. Across four human retrospective cohort studies covering 1.48 million patients (391,716 of them GLP-1RA users), preoperative GLP-1RA use was not associated with increased postoperative complications following rotator cuff repair. One study (Rasmussen & Ilyas) found no significant difference in reoperation risk at 3, 6, or 12 months, and no differences in septic arthritis, adhesive capsulitis, blood clots, or readmission. Another (Seddio et al.) found semaglutide users with type 2 diabetes undergoing arthroscopic repair actually did better: any 90-day adverse event occurred in 11.0% of semaglutide users vs. 27.4% of non-users, ED visits were lower (24.2% vs. 41.4%), and the 2-year re-tear rate was lower too (12.5% vs. 18.3%, P < .001).

On the incidence side — whether GLP-1 patients develop a rotator cuff tear in the first place — the signal runs the other way. A separate matched-cohort study (Davis, Bank, Lauck, et al.) found a higher 5-year incidence of atraumatic rotator cuff tears among GLP-1RA users across all three metabolic subgroups studied: nonobese diabetic (3.7% vs. 2.3%), obese diabetic (4.5% vs. 3.3%), and obese nondiabetic (3.2% vs. 2.3%). A second study using SGLT2 inhibitors as an active comparator found the same directionality — SGLT2i users had a lower risk of both tear and subsequent repair than GLP-1RA users.

Why the split? The review's own authors offer two competing explanations rather than picking one. The biologically plausible one: rapid weight loss and improved metabolic health from GLP-1RAs may let patients become more physically active and load their shoulders more than before — combined with some lean muscle loss, that could genuinely raise tear risk. The more mundane one: GLP-1RA patients see providers more often for weight and medication monitoring, which means more opportunities for a shoulder problem to get imaged and coded — so part of this signal could be detection bias rather than a true increase in tendon disease. The review is explicit that current evidence can't distinguish between these two explanations yet, and flags surveillance bias and confounding by indication as real limitations across every included study.

The preclinical piece is more straightforwardly encouraging, with real caveats. Two rat studies of liraglutide in acute rotator cuff repair found improved biomechanical strength and stiffness at the tendon-bone interface, better collagen organization, reduced inflammatory infiltration, and dramatically reduced fatty infiltration — more than 10-fold lower at 12 weeks compared to controls. That's a genuine mechanistic signal for why liraglutide might support healing once a repair is already underway. The catch: both studies used young, healthy rats with acute surgical injuries, which doesn't necessarily generalize to older humans with chronic degenerative tendon disease — the review's authors say so themselves.

Shoulder surgery: neutral for replacement, genuinely better for rotator cuff work

If you've got patients asking whether their GLP-1 prescription is going to complicate upcoming shoulder surgery, the full picture is more interesting than a flat "no difference" — and it splits by procedure type.

A systematic review and meta-analysis in the Journal of Shoulder and Elbow Surgery (Moews et al., June 2026 issue, PMID 41619919) pooled six studies covering 43,415 patients: four looked at total shoulder arthroplasty (TSA), one at arthroscopic rotator cuff repair, and one at manipulation under anesthesia/capsular release for adhesive capsulitis.

For TSA specifically, the numbers were genuinely neutral either direction: pooled 90-day complications were 18.1% in GLP-1 users vs. 15.9% in non-users (OR 0.86, 95% CI 0.36–2.07, P = .74), and 2-year complications were 3.8% vs. 3.7% (OR 1.24, 95% CI 0.73–2.00, P = .42). No meaningful signal either way.

The arthroscopic cohorts told a different, more favorable story — and the numbers line up closely with the Seddio et al. post-op data cited in the rotator cuff section above. GLP-1 users had significantly lower 90-day complications after arthroscopic rotator cuff repair (11.0% vs. 27.4%) and after capsular release for adhesive capsulitis (2.5% vs. 4.8%), plus a lower 2-year re-tear rate (12.5% vs. 18.3%). The authors' conclusion: GLP-1 use should not be treated as a contraindication to proceeding with TSA — and for arthroscopic rotator cuff work specifically, the signal isn't just neutral, it's favorable, consistent with what the post-op side of the rotator cuff picture already showed above.

Now the more complicated part: muscle mass vs. muscle quality vs. muscle strength

Here's where it gets genuinely nuanced, and where you want to be careful not to oversimplify for a patient in either direction.

A systematic review and meta-analysis in the International Journal of Obesity (Laverde et al., published June 19, 2026) pooled seven RCTs (821 patients) using GLP-1RAs specifically at obesity-treatment doses — liraglutide 3 mg, semaglutide 2.4 mg, tirzepatide 5/10/15 mg — against placebo. Two things were true at once, both rated high-certainty evidence: lean mass as a proportion of total body weight improved significantly (+1.81 percentage points, 95% CI 1.1–2.52) because fat loss outpaced lean loss, while absolute lean mass still dropped by an average of 1.74 kg (95% CI −3.04 to −0.45) and 3.06% from baseline.

Semaglutide showed the steepest absolute loss in the pooled analysis (−5.44 kg). Tirzepatide's SURMOUNT-1 data suggested an even larger effect (−10.9% lean mass vs. −2.0% on placebo) but couldn't be pooled into the formal meta-analysis because of how that trial reported its results. The authors' own read: lean mass loss "should not be considered a limitation" for using these drugs in patients with obesity, since the net effect on body composition is favorable — but they're explicit that resistance exercise and adequate protein intake need to go alongside the medication to actually preserve muscle, not just improve the ratio on a scan.

One limitation worth flagging directly to a patient or colleague: this review measured lean mass, not muscle function — none of the seven trials included grip strength, physical performance, or any other functional outcome. The authors say so themselves, and note that functional decline is known to happen earlier and faster than mass changes show up on a DXA scan. That's exactly the gap the next paper was written to address.

That functional gap is what a January 2026 commentary in the British Journal of Pharmacology (Prokopidis) takes on directly — and it's worth being precise about what kind of paper this is: a narrative commentary, not a pooled meta-analysis, so there are no effect sizes or confidence intervals attached to it. Its actual thesis is broader than a simple mass-vs-strength contrast: lean soft tissue loss is not a reliable surrogate for muscle strength change, and strength outcomes need to be built directly into clinical monitoring and trial design, especially for older adults. That framing is the more useful takeaway — it's also what explains why the short-term and long-term evidence below aren't actually contradicting each other.

The evidence the commentary draws on: short-to-mid-term trials of semaglutide and liraglutide in adults with obesity showed statistically preserved handgrip strength despite measurable lean tissue loss, and tirzepatide combined with resistance and aerobic training in young men produced no additional strength benefit beyond exercise alone. But longitudinal and retrospective data in older adults with type 2 diabetes on prolonged semaglutide told a different story — reduced handgrip strength and accelerated sarcopenia markers. Once you separate population (young vs. older) and duration (short trial vs. prolonged real-world use), the two findings stop looking like a contradiction and start looking like two different points on the same timeline.

A separate systematic review in Liver International, looking specifically at patients with MASLD (metabolic dysfunction-associated steatotic liver disease), adds a third data point: GLP-1RA therapy was not associated with a clinically meaningful decline in muscle mass or strength in that population, with early evidence pointing toward improved muscle quality via reduced myosteatosis (less fatty infiltration inside the muscle tissue). And a separate systematic review in Aging and Disease (December 2025) lays out the biological case for why GLP-1RAs might be protective in some contexts — anti-inflammatory signaling and metabolic effects that plausibly support muscle quality even while overall mass drops.

Translation for your plan of care: mass, quality, and strength are three different things that don't always move together. Duration of use and baseline patient age look like the variables that most reliably predict which way an individual patient trends — a 45-year-old on semaglutide for eight months post-op is a different risk profile than a 72-year-old two years into treatment, even though both are "a GLP-1 patient" on your intake form. And since none of the mass-focused trials measured strength directly, grip strength or a simple functional test is worth adding to your own monitoring rather than assuming DXA or bioimpedance trends tell the whole story.

What this actually changes about your plan of care

Don't hold off recommending resistance training or delay surgical timelines out of reflexive GLP-1 caution. Once a rotator cuff repair is already happening, GLP-1 use tracks with better re-tear and complication numbers, not worse, across the post-op literature.

Do mention the incident-tear signal to patients ramping up activity fast on a GLP-1. The same scoping review found a real (if not fully explained) increase in new atraumatic rotator cuff tears among GLP-1RA users — possibly from rapidly increased activity levels outpacing tendon/muscle capacity, possibly from more frequent imaging catching things that would've gone undetected otherwise. Either way, a patient who's suddenly much more active after months of appetite suppression and weight loss is worth a conversation about progressive loading, not just enthusiasm.

Still load the tissue deliberately, especially post-op. Even with encouraging healing data, none of this replaces progressive loading as the actual driver of tendon and muscle adaptation. The mechanism story (less inflammation, better metabolic environment) supports healing capacity — it doesn't substitute for the rehab dosage itself.

Ask how long they've been on the medication, not just whether. The "preserved strength short-term, possible decline long-term" split from the BJP commentary means duration of use is now a genuinely useful intake question, especially for patients over 65.

Track strength, not just activity tolerance. The IJO review's own authors flag that none of the trials behind the reassuring body-composition numbers measured grip strength or physical function — those outcomes lag behind mass changes. A simple functional baseline is worth having on a GLP-1 patient regardless of what their DXA or bioimpedance scan says.

Reassure without overpromising. "The newer data on tendon healing is actually looking pretty good" is honest and true. "This drug protects your tendons" is not — this is early evidence in a fast-moving literature, and today's reassuring numbers came from small cohorts that need replication.

Bottom line

The GLP-1 tendon and surgical-outcome data this year is genuinely mixed, not uniformly reassuring — and that split is the useful part. Once a rotator cuff repair or shoulder replacement is already happening, GLP-1 use tracks with fewer complications and fewer re-tears, not more, across multiple studies. But the same body of evidence shows a real signal for more new atraumatic rotator cuff tears among GLP-1 users in the first place, for reasons that aren't yet fully sorted out. The muscle quality-vs-quantity-vs-strength picture adds a third layer: body composition improves on net, absolute lean mass still drops, and nobody's actually measured whether strength follows mass down. All three things are worth knowing before your next eval — and worth explaining to a patient with the same nuance, not a flattened "good news" or "bad news" headline.