Tennis Elbow Treatment: What Works, What Backfires and What the Evidence Shows
Table of Contents
Key Takeaways
- In two randomised trials, 83–94% of patients managed with wait-and-see or physiotherapy reported success at 52 weeks [1][2].
- Physiotherapy combining elbow manipulation and exercise improved outcomes at six weeks compared with wait-and-see, but not at one year [2].
- In a placebo-controlled trial, corticosteroid injection led to lower one-year recovery (83% vs. 96%) and higher recurrence (54% vs. 12%) [3].
- Topical NSAIDs reduced short-term pain more than placebo, based on low-quality evidence [4].
- PRP and autologous blood injections did not differ from controls in the included trials [5], and a meta-analysis found no clinical benefit of shockwave therapy over sham [6].
- A small blinded trial failed to show additional benefit of tendon excision surgery over placebo surgery [7].
- In my clinical experience, active physiotherapy is often what tennis elbow patients are missing, and most of my patients find their symptoms ease with it, although the symptoms may recur.
Introduction: Tennis Elbow Treatment Starts With the Natural Course
Tennis elbow is a common cause of pain on the outer side of the elbow. In a representative Finnish population sample of 4,783 people aged 30–64, definite lateral epicondylitis was found in 1.3%, with no difference between men and women and the highest prevalence at ages 45–54 [8].
The treatment question is where things get interesting. The option that feels most decisive, a cortisone injection, gave the fastest early relief in randomised trials, yet it was followed by high recurrence and lower success at 52 weeks than wait-and-see or physiotherapy [1][2]. In meta-analyses, platelet-rich plasma and autologous blood injections did not differ from controls, and shockwave therapy showed no clinical benefit over sham [5][6].
As a physician and former competitive athlete, I look at tennis elbow from two angles: what helps patients get through a painful few months, and what avoids making the longer course worse. This article goes through the main tennis elbow treatment options in order of evidence: load management and patience, exercise, pain relief, braces, injections, shockwave and surgery.
In my practice, lateral epicondylitis comes up fairly often, and in my experience it is more common than its medial counterpart, golfer’s elbow. A fairly typical patient I see is a care worker who does a lot of manual work, with heavy lifting and many repetitions. Another is an installer doing physical, mechanical installation work with their hands.
What Tennis Elbow Is and Who Gets It
Tennis elbow (lateral epicondylitis, lateral elbow tendinopathy) is pain at the outer elbow where the wrist and finger extensor tendons attach to the lateral epicondyle. Pain typically flares with gripping and twisting movements.
Physical load is strongly associated with tennis elbow. In the Finnish population study, people whose work combined repetitive arm movements with forceful activities had 5.6 times the odds of possible or definite lateral epicondylitis compared with people exposed to neither, and current smoking was associated with an odds ratio of 3.4 for definite lateral epicondylitis [8]. The same study concluded that physical load factors, smoking and obesity are strong determinants of epicondylitis [8].
The most important fact for treatment decisions is the natural course. A Cochrane review summarises that the pain can last from 6 months to 2 years and may improve on its own [4]. That is why every treatment below has to be judged against a simple benchmark: does it beat time plus sensible advice?
In the patients I see, the pain usually starts gradually under load and localises to the outer side of the elbow. Handling tools, gripping, and lifting heavy objects or people are what tend to make it hurt. I have sometimes seen swelling around the epicondyle, and occasionally swelling in the forearm as well, which in my experience may then be linked to some form of tendon sheath inflammation. In many of my patients, the whole upper limb chain is symptomatic in one way or another: I fairly often see tennis elbow together with rotator cuff symptoms, and in some patients tendinopathy in several joints at once, typically load-related.
Wait-and-See Is a Real Treatment, Not Neglect
Two randomised trials of tennis elbow treatment tested a wait-and-see approach against physiotherapy and corticosteroid injections, and in both, most patients managed with wait-and-see or physiotherapy reported success at 52 weeks.
In a trial of 185 patients with symptoms for at least 6 weeks, success rates at 52 weeks were 83% with wait-and-see, 91% with physiotherapy and 69% with corticosteroid injections [1]. The difference between physiotherapy and wait-and-see was not statistically significant [1].
In a Brisbane trial of 198 participants, 56 of 62 people (90%) in the wait-and-see group and 59 of 63 (94%) in the physiotherapy group reported success at 52 weeks, compared with 44 of 65 (68%) after corticosteroid injection [2]. Importantly, wait-and-see in this trial was not passive: participants were reassured that the condition would settle, received an information booklet about the condition and guidance on adjusting daily activities to limit pain while staying active, and could use pain relief, heat, cold or braces as needed [2].
In my reading of these trials, explanation and activity adjustment are therefore part of tennis elbow treatment in their own right, not a way of doing nothing.
By the time patients reach my practice, their symptoms may have lasted several weeks or even months. In many of them, the condition has taken away their ability to work, and many no longer dare to exercise or go to the gym because they have learned that it makes the elbow sore; some have stopped strength training altogether because the elbow is tender. Some have already tried massage, and some have tried resting the arm.
Exercise as Tennis Elbow Treatment: Short-Term Benefit, Modest Long-Term Difference
Exercise is one of the active tennis elbow treatment options, and the evidence supports it with clear caveats.
Physiotherapy versus waiting. In the Brisbane trial, physiotherapy (eight 30-minute sessions over six weeks combining elbow manipulation and exercise, plus a home programme) was superior to wait-and-see at six weeks, with success in 65% versus 27% [2]. By 52 weeks, there was no difference between the two groups on any primary outcome [2]. Physiotherapy participants did, however, seek significantly less additional treatment: 13 of 63 compared with 34 of 62 in the wait-and-see group [2].
A 2022 meta-analysis from the Canadian Shoulder and Elbow Society pooled 5 randomised trials of strengthening-based physiotherapy against no active treatment and found no significant differences in pain or function [5]. In my view, the honest summary is that, in the Brisbane trial, physiotherapy combining elbow manipulation and exercise sped early improvement and reduced the need for other treatments, rather than changing where most patients ended up at one year.
Which type of exercise? A 2021 meta-analysis of 6 studies (429 participants) found that adding eccentric exercise to other therapy improved pain (SMD −0.63) and muscle strength (SMD 1.05) compared with the other therapy alone [9]. Compared with concentric or isotonic exercise, eccentric exercise improved pain slightly more (SMD −0.30) but did not differ for strength or function, and the authors note the small number of studies and varied exercise protocols as limitations [9].
Isometric exercise has become popular for tendon pain. In a randomised trial of 40 people with tennis elbow, one instruction session followed by 8 weeks of unsupervised daily progressive isometric exercise improved pain and disability scores compared with wait-and-see, but did not improve global rating of change (29% versus 26% success) or pain-free grip strength [10]. The authors suggested it may be doubtful as a sole treatment [10].
In my experience, active physiotherapy is often what these patients are missing, and what eases their symptoms. I usually refer patients to physiotherapy myself, early on, and tell them from the start that long-term rehabilitation will most likely require physical loading of the tendon. Many of my patients feel they get help from isometric and eccentric exercises. The exercises may, however, make the elbow sorer at first, and I tell patients not to be alarmed by this. What I have seen is that stretching alone, rest alone or anti-inflammatory medication alone usually does not resolve the symptoms; in my understanding, tendon cells need mechanical stimulus, known as mechanotransduction.
I’m also used to telling patients that they can continue going to the gym and do not need to avoid it. In my view, the hard gripping and the tension in the muscles during strength training may even benefit the healing process. Rather than telling patients to stay away from sport, I would say that in my experience, staying active in sport over the long term is more likely to ease the symptoms than to worsen them.
Pain Relief: Topical Versus Oral Anti-Inflammatories
For pain control during tennis elbow treatment, a Cochrane review of 15 trials (759 participants) found low-quality evidence from three trials (153 participants) that topical NSAIDs reduced pain more than placebo in the short term, by 1.64 points on a 0–10 scale [4]. Mild rash was reported in 2.5% of people using topical NSAIDs versus 1.3% with placebo [4].
Evidence for oral NSAIDs was conflicting: one placebo-controlled trial found greater pain improvement and the other found no difference, and neither found a difference in function [4]. In one placebo-controlled trial, oral NSAIDs were associated with an increased risk of gastrointestinal side effects [4]. The median follow-up across trials was only 2 weeks, so these data say nothing about the long-term course [4].
When the situation is very acute or difficult, I sometimes prescribe anti-inflammatory medication, and sometimes a rest splint for the first phase. Even then, I tell patients right away that in my experience these do not fix the problem on their own.
Braces and Counterforce Straps
Counterforce straps are worn just below the elbow and are a common form of self-directed tennis elbow treatment. In a randomised, placebo-controlled trial comparing a counterforce brace (n = 17) with a placebo brace (n = 14) in acute tennis elbow, both braces were followed by improvement, but the counterforce brace reduced the frequency and severity of pain more in the short term (2 to 12 weeks) and gave better overall elbow function at 26 weeks [11]. The trial was small, so in my view a brace is best seen as a way to make daily tasks more tolerable rather than a cure.
I see counterforce-type straps on patients from time to time, for example worn during work. In my experience, they are used mostly for tennis elbow and less often for golfer’s elbow. My guess is that this is because pictures almost always show the strap over the lateral epicondyle, so people associate it with tennis elbow, although in my view it might also work with medial epicondylitis.
Corticosteroid Injections: Fast Relief, Worse Long-Term Results
In the trials below, cortisone injections show a consistent pattern of short-term gain followed by long-term cost.
A Lancet systematic review of 41 randomised trials (2,672 participants) found that, for lateral epicondylalgia, corticosteroid injection had a large short-term effect on pain compared with no intervention (SMD 1.44), but no intervention was favoured at the intermediate term (SMD −0.40) and long term (SMD −0.31) [12]. Serious adverse events were rare: of 991 participants receiving corticosteroid injections in studies that reported adverse events, one (0.1%) had a tendon rupture [12].
The individual trials show the same pattern. In the trial of 185 patients, success at 6 weeks was 92% after injection, 47% after physiotherapy and 32% with wait-and-see, but the recurrence rate after injection was high [1]. In the Brisbane trial, 47 of 65 participants (72%) in the injection group deteriorated after three or six weeks, compared with 5 of 66 (8%) after physiotherapy and 6 of 67 (9%) with wait-and-see [2].
A placebo-controlled trial published in JAMA included 165 patients. At one year, complete recovery or much improvement was reported by 83% after corticosteroid injection versus 96% after placebo injection, and recurrence was 54% versus 12% [3]. The authors concluded that corticosteroid injection resulted in worse clinical outcomes after one year compared with placebo injection [3]. In the same trial, physiotherapy combined with placebo injection gave better short-term results at 4 weeks than placebo injection alone (39% versus 10%), though physiotherapy made no significant difference at one year [3].
Some patients ask me about a cortisone injection. It is not something I offer: in my view, there is no reason to inject cortisone into the tendon, and I have not done it myself. Nor have I seen anyone else give cortisone injections to treat tennis elbow.
PRP, Blood Injections and Shockwave
PRP and autologous blood. The Canadian meta-analysis of nonoperative tennis elbow treatment found no differences in pain or function between platelet-rich plasma and controls (2 studies) or between autologous blood injections and controls (3 studies), and concluded that the available evidence does not support these options [5].
Shockwave therapy. Here the evidence is genuinely conflicting. A 2021 meta-analysis of 27 studies (1,871 patients) concluded, with low to moderate certainty, that extracorporeal shockwave therapy has no clinical benefit compared with sham treatment or corticosteroid injection [6]. A 2025 umbrella review of 9 meta-analyses reported a statistically significant pain reduction versus placebo (mean difference −0.68 on the VAS) but no clinically important difference in grip strength, and still concluded that shockwave outperformed ultrasound therapy and corticosteroid injections [13].
Surgery: Little Evidence It Beats Placebo
A Cochrane review notes that surgery is sometimes recommended for persistent lateral elbow pain when less invasive treatments have failed. Its 5 trials (191 participants) were all highly susceptible to bias, and the authors concluded there was insufficient evidence to support or refute the effectiveness of surgery for lateral elbow pain [14].
One double-blinded trial randomised 26 patients who had tennis elbow for more than 6 months and had failed at least 2 nonsurgical treatments. Excision of the degenerated portion of the extensor carpi radialis brevis (ECRB) tendon was compared with skin incision and exposure of the tendon only [7]. Both groups improved, and the study failed to show additional benefit of surgical excision over placebo surgery [7]. In my view, although the trial was small, it supports trying conservative care first.
In my practice, tennis elbow patients are referred to an orthopaedist fairly rarely, as treatment is usually rehabilitation in any case. In my 10-year career, I have never seen a surgical plan or surgical treatment for tennis elbow, nor a patient whose tennis elbow had been operated on. I have seen orthopaedic consultations, but they usually end with a referral to physiotherapy and anti-inflammatory medication. Surgery is no doubt performed in some parts of the world, but I have not come across it once in Finland.
Practical Tennis Elbow Treatment Plan
Taken together, the trials above point to the following:
- Explanation and activity adjustment. With reassurance and advice to adjust daily activities, most wait-and-see participants reported success at 52 weeks [2].
- Exercise. Physiotherapy combining elbow manipulation and exercise improved six-week outcomes and reduced the use of other treatments [2], and adding eccentric exercise to other therapy improved pain and strength in a meta-analysis [9].
- Anti-inflammatories. Topical NSAIDs reduced short-term pain more than placebo, while evidence for oral NSAIDs was conflicting [4].
- Counterforce strap. In a small trial, a counterforce brace reduced short-term pain more than a placebo brace [11].
- Cortisone. Injection was followed by higher recurrence and a poorer one-year outcome than placebo injection [3].
- Add-on treatments. PRP and blood injections did not differ from controls in the included trials, and a meta-analysis found no clinical benefit of shockwave over sham [5][6].
- Surgery. A small blinded trial failed to show additional benefit of tendon excision surgery over placebo surgery [7].
In my occupational health work, the occupational physiotherapist may also look at ergonomic adjustments that can be made at the workplace to ease the symptoms.
Conclusion: Tennis Elbow Treatment in Perspective
In my reading of the evidence, the least dramatic tennis elbow treatment is often the most sensible one. In randomised trials, 83–94% of patients managed with wait-and-see or physiotherapy reported success at one year [1][2], while corticosteroid injection, despite the fastest early relief, led to 54% recurrence and worse one-year outcomes than placebo [3].
Physiotherapy, topical anti-inflammatories and a counterforce strap showed short-term benefits in the trials above. PRP and blood injections did not differ from controls, shockwave showed no clinically important benefit over sham despite some statistically significant pain reductions, and the one placebo-controlled surgical trial found no added benefit from tendon excision. In my view, understanding the natural course is itself part of the treatment.
This matches what I see in practice: despite how troublesome the symptoms can be, most of my tennis elbow patients find that their symptoms ease with good physiotherapy, which is encouraging. The symptoms may return, however, which is why I’m used to emphasising to patients that it is worth continuing the rehabilitation exercises even when the elbow is not bothering them, especially if the symptoms tend to recur.
References
- https://doi.org/10.1016/S0140-6736(02)07811-X
- https://doi.org/10.1136/bmj.38961.584653.AE
- https://doi.org/10.1001/jama.2013.129
- https://doi.org/10.1002/14651858.CD003686.pub2
- https://doi.org/10.1016/j.jseint.2021.11.010
- https://doi.org/10.1177/02692155211006860
- https://doi.org/10.1177/0363546517753385
- https://doi.org/10.1093/aje/kwj325
- https://doi.org/10.3390/jcm10173968
- https://doi.org/10.1249/MSS.0000000000002128
- https://doi.org/10.1016/j.jse.2018.10.002
- https://doi.org/10.1016/S0140-6736(10)61160-9
- https://doi.org/10.1186/s10195-025-00871-w
- https://doi.org/10.1002/14651858.CD003525.pub2
