Tennis Elbow and Golfer's Elbow: Tendinopathy, Not Inflammation

Written by Gisele J. Girault, MD

Updated September 2026 ·11 min read

Tennis elbow and golfer’s elbow are the same problem on opposite sides of the same joint, and the name both of them carry is misleading. The suffix -itis implies inflammation, and by the time most patients present there is very little inflammation to find.

As Dr. Gisele J. Girault puts it in her course on non-spinal injection, medial and lateral epicondylitis “are really not diseases of the epicondyle. Epicondylitis is a tendinopathy and it’s degenerative in nature.” That single reframing changes what the treatment is trying to achieve, and it explains why the condition so often outlasts the anti-inflammatory approaches aimed at it.

This reference covers the anatomy that separates the two, the examination that distinguishes lateral epicondylitis from the nerve entrapment that mimics it, the safety rule that governs every medial-side needle, and what the evidence says about steroid versus platelet-rich plasma. It sits under our clinical reference to joint and extremity injection.

Lateral epicondylitis and medial epicondylitis: which tendons are involved

The two conditions are anatomically symmetrical.

Lateral epicondylitis — tennis elbow. Girault describes it as involving the wrist extensor tendons, which originate at the elbow. Dr. Glenn Barnes adds the detail that matters at the needle: the lateral epicondyle of the humerus is the origination point of the forearm muscles responsible for wrist extension, wrist ulnar deviation, finger extension and forearm supination. Because those muscles share one tendon and one insertion, it is called the common extensor tendon, and it is the site of degeneration in lateral epicondylitis.

Medial epicondylitis — golfer’s elbow. The mirror image, involving the flexor tendons originating at the elbow. Barnes describes the common flexor tendon as the origin of the forearm muscles producing wrist flexion, wrist ulnar deviation and finger flexion.

Both are overuse injuries of muscles that originate on the humerus and cross the elbow joint. Girault is explicit that the presentation is almost never sudden: “it’s unusual for someone to have an acute epicondylitis unless they have a trauma associated with that.”

Why it is tendinopathy rather than tendonitis

Girault teaches the progression as a sequence, and it is worth following because each step explains a different clinical observation.

Repeated load produces micro tears at the tendon attachment. Those tears lead to tendonitis. Chronic tendonitis leads to tendinopathy — which, in her description, means a chronic process in which the tendon has thickened and in some cases shortened, so that it no longer has the same tensile strength or the same ability to stretch that it had before the injury.

A thickened, shortened, structurally altered tendon is not a problem an anti-inflammatory can reverse. This is the reason a patient can get real relief from an injection and still be back in six months, and the reason the regenerative literature has concentrated on this diagnosis in particular.

Examining the elbow, and the entrapment that imitates tennis elbow

The provocative tests are straightforward, but the second one exists because the diagnosis is not as obvious as it looks.

Cozen’s test evaluates lateral epicondyle pain. Dr. Robert Stall describes it as stabilising the elbow in 90 degrees of flexion, positioning the hand in pronation and radial deviation, then asking the patient to supinate the forearm against resistance. A positive test reproduces their usual symptoms.

Mosley’s test has also been proposed for lateral epicondylitis — but Stall flags the complication that makes it useful in a different way. Mosley’s may also be positive when the patient has radial tunnel syndrome, producing lateral proximal forearm pain from compression of the deep radial nerve. Some resources, he notes, suggest it may be more sensitive and specific for radial tunnel than for the tendinopathy itself, and therefore propose that if Cozen’s test is positive, Mosley’s should be performed, and if that is positive, radial nerve entrapment should be ruled out before a definitive diagnosis of lateral epicondylitis is made.

That sequence is the reason a proportion of “failed tennis elbow” referrals are not tennis elbow. Where the picture is ambiguous, electrodiagnostic testing is the tool that separates a tendon problem from a nerve problem.

For the medial side, Stall uses passive extension of both the elbow and the wrist with the forearm and hand supinated. The logic is mechanical: the flexor tendons cross the elbow joint to insert on the medial epicondyle, so extending both joints puts the relevant muscles and tendons under maximal tension.

The ulnar nerve rule that governs every medial-side injection

This is the most consequential safety point on the elbow, and it is specific to the medial epicondyle.

The ulnar nerve passes posterior to the medial epicondyle in the cubital tunnel, immediately behind the insertion of the common flexor tendon. Girault notes how superficial it is — “you can actually feel the ulnar nerve. It’s just under the skin.”

Barnes supplies the anatomical variation that turns proximity into risk. In approximately 10% to 15% of the population, the ulnar nerve dislocates out of the cubital tunnel and moves anterior to the medial epicondyle when the elbow is flexed — placing the nerve directly in the path of the needle. His conclusion is a rule with no exception attached to it: medial-side procedures should always be done with the elbow in full extension.

The same rule applies to injections of the ulnar collateral ligament, which sits deep to the common flexor tendon on the medial epicondyle and is the structure injured in valgus trauma and throwing.

Landmark injection technique

Girault performs these on palpation. The landmarks for the lateral side are the lateral epicondyle and the radial head, and most patients will identify the tender point themselves.

Her governing principle is that this is a superficial injection and that nothing hard should be struck: not the tendon, not the bone, and on the medial side not the nerve. She tents the skin and injects immediately beneath it, using a 30-gauge half-inch needle by preference — “you can’t really get into too much danger with that” — or a 25-gauge one-and-a-half-inch needle. The volume is roughly half to one cc of a 50/50 steroid and lidocaine mixture per site, and the injections can be repeated every two weeks for a series of three.

Stall teaches a variant worth knowing because it gives the operator a tactile check. The needle is inserted perpendicular to bone until contact is made, then withdrawn slightly — one to two millimetres — and injected slowly. If there is high resistance to depressing the plunger, withdraw slightly: that resistance means the needle is within tendon or ligament rather than beside it.

Which corticosteroid is used, and why the choice is not interchangeable, is covered in corticosteroid injections for joint and soft tissue pain.

What ultrasound adds at the elbow

Barnes makes a specific claim for this diagnosis rather than a general one. Comparing findings to the contralateral uninjured side is useful across musculoskeletal imaging, and he has found this “particularly true in lateral epicondylitis” — a degenerative tendon is recognised most reliably against the patient’s own normal one.

Elastography contributes to the same judgement by quantifying tissue elasticity, with lower elastic rates correlating with diseased connective tissue.

His technique for the lateral side begins with palpation rather than the probe: locate the lateral epicondyle, have the patient wiggle their fingers to reveal the orientation of the muscles as they originate, then place the linear probe along that orientation with its proximal edge on the epicondyle and rotate the distal end until the common extensor tendon is seen in long axis. The needle angle is shallow. Because the injection is intratendinous, low volumes are better tolerated, and he makes several passes to perform a small tenotomy and distribute the injectate.

The broader case for and against guidance is set out in our musculoskeletal ultrasound reference.

Steroid or PRP: what the evidence actually shows

Lateral epicondylitis is unusual among injection targets in having genuinely informative comparative data, and Stall is direct about its standing: it is one of the two most commonly studied diagnoses for platelet-rich plasma, and one of the two for which some of the best evidence supports clinical use.

The finding he highlights is that in one study of lateral epicondylitis, a single PRP injection not only produced significantly greater improvement than a corticosteroid group, but the result was maintained at two-year follow-up.

Two caveats belong with it. There is no consensus on which PRP formulation is optimal for treating tendinosis, though a significant number of studies used leukocyte-rich preparations. And the ideal injection regimen remains undefined, with contradictions in the published literature — the studies showing sustained benefit from repeated injections generally lacked a control group comparing them against a single treatment.

Read against the tendinopathy model, the direction of the evidence is coherent: a degenerative tendon responds better to something that provokes repair than to something that suppresses inflammation which has largely already resolved. Volumes are small, with Stall describing one to two millilitres of PRP for these targets. The wider evidence base is covered in our regenerative pain medicine reference.

For clinicians

Elbow injection is a landmark technique learned on live anatomy with a proctor present, which is why this reference carries no needle depths or trajectories. Empire teaches the elbow, wrist and hand injections alongside ultrasound-guided technique across the pain management academy.