You were told tendonitis. The tendon is not inflamed.

PRP FOR TENDONITIS

The word is wrong, and the word is why the treatment keeps failing. Anti-inflammatories, rest and a steroid injection are all aimed at inflammation that stopped being there months ago.

What the tissue actually shows

When a chronically painful tendon is biopsied, the thing everybody expects to find is not there. There are no meaningful numbers of inflammatory cells. What there is instead is collagen disarray — fibers that should run in parallel lying in disorganized bundles — along with an increase in ground substance, and neovascularization, new small vessels growing into the tendon from outside, each accompanied by nerve fibers.

That last detail is the one that explains the pain. Nerves arrive with the vessels. A tendon can therefore hurt a great deal while looking only moderately abnormal on a scan, and the severity of the picture and the severity of the pain routinely fail to match.

The correct term for this is tendinopathy, or tendinosis where the emphasis is on the degeneration itself. “Tendonitis” describes an inflammatory condition that a chronic tendon largely is not. The suffix is not pedantry: -itis points every treatment decision that follows toward suppressing inflammation, and that is the wrong target.

Why the wrong word produces the wrong treatment

Follow the logic that the word sets up. If the problem is inflammation, then a non-steroidal anti-inflammatory should fix it, rest should let it settle, and a corticosteroid injection should fix it decisively. Each of those is a reasonable inference from a false premise.

Anti-inflammatories reduce pain for a while and do nothing to the disorganized collagen. Rest is worse than neutral: a tendon that is not loaded gets weaker, and the thing that reorganizes collagen along the line of force is force. A corticosteroid injection is the most consequential of the three, because it works — briefly and convincingly — and the twelve-month data in lateral elbow tendinopathy has it performing worse than doing very little at all. It is a loan against the tendon, and the repayment is real.

So people arrive here having done three sensible things, none of which addressed the problem, and having concluded that nothing works. What actually happened is that nothing tried was aimed at the right thing.

Where this stalls, and why it does not simply heal

Two brakes hold a degenerate tendon in that state. The first is mechanical and local: tendons at the classic sites — the Achilles midportion, the common extensor origin at the elbow, the gluteal insertion — have marginal blood supply exactly where they fail. There is little margin for repair to begin with.

The second is systemic and gets missed almost universally. Insulin resistance and raised blood sugar glycate collagen, cross-linking it so it becomes stiffer and fails at lower strain. A tendon under that metabolic load lays down worse tissue than the tissue it is replacing. Chronic metabolic inflammation — the low-grade kind that has nothing to do with the acute inflammation the name implies — holds the tissue in a degradative rather than a rebuilding state. This is set out at length on metabolic health and tissue repair, and it is the most common reason a technically correct treatment underperforms.

The systemic number behind every tendon on this site

Tendinopathy is treated almost everywhere as a loading problem, and load is genuinely half of it. The other half shows up when you compare populations rather than activities.

A systematic review with meta-analysis pooled 31 studies examining tendinopathy against diabetes. Tendinopathy was substantially more common in people with diabetes — odds ratio 3.67 (95% CI 2.71 to 4.97) across 17 of those studies. The relationship also ran in the other direction, with diabetes more common among people with tendinopathy. People who had both had carried diabetes about 5.26 years longer than people with diabetes and no tendinopathy, and their tendons were measurably thicker than controls on imaging.

Duration of exposure mattered. That is what you would expect if the mechanism is glycation progressively cross-linking collagen — stiffening the tendon, lowering the strain it tolerates before it fails, and slowing the repair once it has. It is not what you would expect if the explanation were simply that some people do more with their arms.

This is the general finding. Individual tendons then have their own evidence on top of it, and it differs by site rather than repeating: the Achilles has the fluoroquinolone association and the loading data, the rotator cuff has a glycemic-control finding that changed retear rates after repair, and the gluteal tendons have a hormonal trial whose effect was gated by body mass. We put the numbers on the page they belong to instead of repeating one number everywhere. What we measure is on the metabolic health page.

What PRP is aimed at here

The rationale is specific rather than general. Concentrated platelets release growth factors into a zone that has poor delivery and has stopped remodelling. The intention is to restart a repair sequence that stalled, not to suppress anything. That is the opposite intention to a steroid, which is why the two are not interchangeable and why choosing between them is a real decision rather than a preference. That comparison is laid out on PRP compared with cortisone.

The evidence is genuinely mixed and it is worth saying so in the same breath. Systematic reviews of PRP across tendon and ligament healing report benefit that varies substantially by site, by preparation and by whether a loading program accompanied it. Ultrasound-guided delivery performs better than blind injection, which is why nothing here is injected by landmark. And the comparison against corticosteroid tends to favor PRP at twelve months while favoring steroid at six weeks — which is exactly what you would expect from a treatment aimed at rebuilding rather than at silencing.

The loading program is not the aftercare

This is the part most people get wrong, and it is worth being blunt about. The injection supplies biology. Load supplies direction. Collagen laid down in a tendon that is not being loaded organizes badly and produces tissue no stronger than what it replaced.

Heavy slow resistance work, progressed over roughly twelve weeks, is the treatment. The injection is what makes that work possible in a tendon that had stopped responding to it. A program done nominally — a sheet of exercises, done sometimes, at a weight that never increases — is the single most common reason a correctly placed injection produces nothing.

Which tendon, and what that changes

“Tendonitis” is not one condition, and the site changes the answer more than anything else does.

• Lateral elbow — the site where the twelve-month evidence most clearly favors PRP over a steroid injection.
• Medial elbow — where the ulnar nerve complicates the picture and has to be ruled in or out before anything is injected.
• Achilles — where the best-known trial is negative, where midportion and insertional behave differently, and where we say so rather than bury it.
• Patellar — where the calendar matters as much as the tendon, because the rehabilitation cannot be served mid-season.
• Gluteal — called bursitis for decades, mostly is not, and driven by compression that the usual stretching advice makes worse.
• Rotator cuff — where what is torn decides everything and symptoms will not tell you which.

When it is not the tendon at all

A meaningful share of people sent here for a tendon injection do not have a tendon problem. Referred pain from the neck can present as a shoulder or elbow that hurts on use; that is covered on shoulder or neck. Lateral hip pain is frequently attributed to the tendon when the lumbar spine is driving it. A nerve entrapment reproduces tendon pain closely enough to fool a scan reading. And an inflammatory arthropathy — several tendons in a younger person, with morning stiffness — needs a rheumatological opinion and not a needle.

The assessment exists to answer that before anything is drawn. If the honest answer is that your problem is elsewhere, that is what you will be told.

What people ask when the word turns out to be wrong

If it is not tendonitis, why does everyone call it that?

Because the name predates the histology. The tissue studies that established the degenerative picture came after the term was already in every chart, every discharge note and every search box. Clinicians largely moved to tendinopathy; the world did not follow, and there is no harm in the word as long as it is not steering the treatment.

This tendon usually fails for reasons that are not in the tendon. explains what that looks like.

Should I stop taking anti-inflammatories?

Not necessarily, and not on our say-so if something else is being managed with them. They are reasonable for short-term pain control. What they are not is a treatment for the underlying tissue, and taking them for months on the assumption that they are working on the tendon is the mistake worth avoiding.

There is more on this in The one place where the twelve-month data goes against the steroid..

I have already had a steroid injection. Does that rule me out?

It does the opposite. A steroid injection that helped and then wore off is useful information — it tells us the pain is coming from where we thought. What we will not do is repeat it indefinitely, because the twelve-month picture for repeated steroid in tendon is not good.

The one place where the twelve-month data goes against the steroid. goes through it in detail.

How long before I know whether it worked?

Longer than people expect. Week two commonly feels like nothing, and that is normal rather than a failure — the recovery timeline sets out why. Meaningful change is judged at around three months, alongside the loading program, not before it.

Related reading

• Injured at work
• PRP compared with cortisone
• Recovery timeline
• Metabolic health and tissue repair
• What we treat
• Am I a candidate?
• Risks and side effects
• The ankle sprain that never came right

In a tendon the question of duration partly dissolves.