Cortisone - Band-aid or lasting damage?
A question I get often from patients with tennis elbow: does a cortisone shot actually hurt the tendon in the long run? A newly published MRI study gives us some of the clearest evidence yet, and the answer leans toward yes.
What the study looked at
In a retrospective, multicenter cross-sectional study published in BMC Musculoskeletal Disorders, researchers reviewed elbow MRIs from 147 patients with clinically diagnosed lateral epicondylitis (LE), all with symptoms lasting six months or longer, imaged between 2015 and 2023. Of those patients, 63 (42.9%) had received at least one prior corticosteroid (glucocorticoid) injection at the common extensor origin before their MRI.
Tendon damage was graded using the Walz classification, an MRI-based staging system radiologists and orthopedists use to score the severity of common extensor tendon pathology in lateral epicondylitis, ranging from Grade 0 (normal) to Grade 3 (severe or complete tear). The researchers also measured lesion size directly on MRI, and adjusted their analysis for age, sex, occupational workload, treatment center, and symptom duration.
What the researchers found
- Patients with a prior corticosteroid injection had significantly higher odds of a more severe Walz grade than patients who had never been injected (adjusted OR 2.78, 95% CI 1.30–5.97; p = .008).
- Longer symptom duration was independently associated with higher Walz grades as well (adjusted OR 1.02 per month; p = .031).
- Prior injection was also associated with larger lesions at the tendon origin: +2.07 mm in the coronal plane, +1.49 mm in the axial plane, and a 94% larger sagittal-plane diameter.
- Interestingly, the number of injections a patient had received was not significantly associated with Walz grade — meaning even a single injection was linked to worse tendon findings, not just repeated dosing.
The authors were careful to note that because this was a retrospective, cross-sectional design, it can only show an association between corticosteroid injection and tendon damage, not prove that the injections caused it — patients with more severe disease may simply be more likely to have been offered an injection in the first place. Still, this adds meaningful imaging-based evidence to a growing body of literature raising concern about the effects of local corticosteroid use on tendon structure.
My take: dose matters, and so do the alternatives
I've argued for years that most patients don't need a high dose of steroid to get the clinical result they're after. That's not just clinical opinion — a 2023 randomized, controlled, double-blind study found that 10 mg of triamcinolone was non-inferior to 40 mg for treating knee osteoarthritis, meaning the lower dose worked just as well with presumably less tissue exposure to steroid (Utamawatin K, et al. BMC Musculoskeletal Disorders. 2023;24:92).
If a corticosteroid injection is going to be used at all, using the lowest effective dose is a reasonable way to help modulate some of its potential downsides.
But the bigger picture here is that we now have a widening set of treatment options for lateral epicondylitis that are actually designed to promote tendon healing rather than simply suppress inflammation, including physical therapy, PRP (platelet-rich plasma), Tenex/TenJet percutaneous tenotomy, and energy-based modalities like EMTT, ESWT, and laser therapy. For most patients with chronic tennis elbow, I believe these regenerative and healing-focused options deserve serious consideration before reaching for a cortisone shot.
Bottom line
This study doesn't prove that corticosteroid injections cause tendon damage in tennis elbow, but it does add meaningful imaging-based weight to the argument that CSI may come with a real structural cost. Patients and clinicians should weigh that risk against the short-term symptom relief CSI provides, and consider whether a healing-oriented alternative might be the better first move.
Source: Henze AS, Eckl L, Kadantsev P, Ambacher T, Kuehn P, Schoch C, Siebenlist S, Geyer S. MRI severity of common extensor origin damage associated with prior glucocorticoid injection in chronic lateral epicondylitis: a retrospective multicenter analysis. BMC Musculoskeletal Disorders. 2026. https://doi.org/10.1186/s12891-026-10450-7