# Tendinopathy: why the injection that helps this month leaves you worse next year

slug: tendinopathy · https://miscsubjects.com/a/tendinopathy · updated 2026-08-05T09:08:40.383Z

The name lies to you. For a century this was called tendinitis, the "-itis" meaning inflammation, and that single syllable sent millions of people toward anti-inflammatories and steroid injections. When pathologists finally looked at the tissue they took out of chronically painful tendons, the inflammatory cells largely were not there. What they found was disorganised collagen, disordered blood vessels growing in where they do not belong, and a repair process that had started and never finished.

Hence tendinopathy — a deliberately vague word meaning "something is wrong with this tendon" — and hence the fact that the most intuitive treatment for it is the one with the worst long-term record in the literature.

Start with that finding, because everything else follows from it.

## The injection makes this month better and next year worse

This is the most important sentence on the page, and it comes from a meta-analysis in the Lancet that pooled 41 randomised trials covering 2,672 people across every common tendon site.

**Steroid injections beat the alternatives in the short term. At six months and at a year, that advantage had not just faded — it had reversed.**

[[embed:source:s1]]

Reversed means the injected group ended up worse than the people who did not get injected. Not equal. Worse.

A separate randomised trial in patellar tendon pain watched the same thing happen and looked inside the tendon while it did. Steroid injection produced good early results and poor late ones. Heavy slow resistance training produced good early results *and* good late ones — and, unlike the injection, the tendon tissue itself improved and collagen turnover went up.

[[embed:source:s2]]

So the choice is not "fast relief versus slow relief". It is: an intervention that borrows relief from your future, or an intervention that is slower at first and repays you.

A 2025 review of the same question in the rotator cuff lands in the same place, in gentler words: steroids are useful for rapid symptom control, the benefit wanes, and repeated use carries risk.

[[embed:source:s3]]

There is a legitimate use for that trade. If you cannot sleep, or you cannot begin loading the tendon at all because of pain, buying a window in which rehabilitation becomes possible is a reasonable purchase. Buying relief and then doing nothing with the window is how people end up in the worse-at-a-year group.

## Loading the tendon is the treatment, and there is no way around it

Everything with good long-term evidence is a version of the same thing: apply controlled, progressive load to the tendon, repeatedly, for months.

A living systematic review of 68 randomised trials in the common lower-limb tendinopathies reached a blunt recommendation.

[[embed:source:s4]]

Three months of exercise on its own, before considering anything added to it. That is the recommendation from the largest current synthesis, and it is roughly the opposite of the usual sequence, which is an injection first and exercise if that fails.

The same review examined the most popular add-on and found it does not earn its place: adding shockwave therapy to eccentric exercise in patellar tendon pain produced no short-term benefit on pain or function, on moderate-strength evidence.

[[embed:source:s5]]

A UK military rehabilitation consensus — a group with a strong operational interest in getting people back to load-bearing work quickly — came to the same conclusion.

[[embed:source:s6]]

When the organisation that most wants a shortcut says exercise is the primary treatment and everything else is case by case, that is worth more than an enthusiastic clinic website.

## What the loading actually looks like

The best-studied protocol is Alfredson's: 3 sets of 15 slow lowering repetitions, twice a day, for twelve weeks. For an Achilles tendon that is standing on a step and lowering the heel below the step, using the good leg to get back up.

Two details decide whether it works.

**It is meant to hurt, within limits.** Discomfort during and shortly after is expected. The usual rule is pain up to about 5 out of 10 during the exercise, settling by the next morning. Pain that is worse the following day means the load was too high, not that the exercise is wrong.

**It takes twelve weeks and people quit at four.** The tendon does not turn over quickly. Every week you skip is a week the collagen was not asked to reorganise.

A 2023 review confirms the protocol holds up and — usefully — that intensity and pace can be adjusted to the person rather than followed rigidly.

[[embed:source:s7]]

That matters because the original protocol is brutal, and a modified version done for twelve weeks beats the strict version abandoned in week three.

## Things that sound like they should help and do not

**Topical nitroglycerin patches.** A well-designed 2024 randomised trial added them to eccentric exercise for mid-portion Achilles tendon pain. Both groups improved substantially at six, twelve and twenty-four weeks. The difference between them was not significant at any timepoint.

[[embed:source:s8]]

Read that carefully, because it contains both a negative and a positive: the patch added nothing, and the exercise worked in both arms.

**Repeated steroid injections.** See the first section. The evidence does not say one injection is a catastrophe; it says the long-term direction is wrong and repetition compounds it.

**Rest.** Complete rest reduces pain while you rest, and the tendon deconditions further. The load is the medicine.

## Ten people with a painful tendon

Ten people present with a chronically painful tendon. On current evidence, if all ten receive an injection and no rehabilitation, most feel better within weeks and more of them are worse at a year than if they had been left alone. If all ten do twelve weeks of progressive loading, the improvement is slower to arrive and still there a year later, and the tendon tissue itself looks better. Somewhere between three and five of the ten will stop the exercise before week eight, because it is dull and it hurts and the first month feels like nothing is happening.

The failure mode of this condition is not choosing wrong. It is choosing right and stopping early.

## What a repair peptide would have to prove here

Repair peptides come up constantly for tendon problems. The state of the evidence is simple and is not improved by discussing it at length: none of them has randomised human trial evidence in tendinopathy. Laboratory and animal work exists for some, and where it does the argument belongs on a page whose title names both the compound and the tendon problem, with its evidence tier attached. The operator of this site has a commercial interest in compounds of that kind, which is exactly why the statement on this page is the absence of the trials.

## Twelve weeks, and how not to quit in week four

1. **Stop thinking of it as inflammation.** The tissue findings do not support it, and the word drives the wrong treatment.
2. **Start loading, today, and put twelve weeks in the calendar.** Slow repetitions, progressive, with pain up to about 5 out of 10 that settles overnight.
3. **If you are offered an injection, ask what it is for.** Buying a window in which you can start rehabilitation is a real reason. Buying relief instead of rehabilitation is the path with the worse one-year outcome in the largest pooled analysis available.
4. **Decline the add-ons for the first three months.** Shockwave on top of exercise showed no short-term benefit; the nitroglycerin patch added nothing over exercise alone.
5. **Adjust the protocol rather than abandoning it.** A version you will actually complete beats the textbook version you quit.
6. **Judge progress at eight weeks, not at two.** Nothing in this condition moves quickly, and expecting it to is the single commonest reason people stop.


## Sources

1. Efficacy and safety of corticosteroid injections and other injections for management of tendinopathy: a systematic review of randomised controlled trials — https://pubmed.ncbi.nlm.nih.gov/20970844/
2. Corticosteroid injections, eccentric decline squat training and heavy slow resistance training in patellar tendinopathy — https://pubmed.ncbi.nlm.nih.gov/19793213/
3. Short-Term Relief or Long-Term Repair: A Narrative Review of Corticosteroid and Platelet-Rich Plasma Injections in Rotator Cuff Tendinopathy — https://pubmed.ncbi.nlm.nih.gov/41268031/
4. Effectiveness of Exercise Treatments with or without Adjuncts for Common Lower Limb Tendinopathies: A Living Systematic Review and Network Meta-analysis — https://pubmed.ncbi.nlm.nih.gov/37553459/
5. Effectiveness of Exercise Treatments with or without Adjuncts for Common Lower Limb Tendinopathies: A Living Systematic Review and Network Meta-analysis — https://pubmed.ncbi.nlm.nih.gov/37553459/
6. UK Defence Rehabilitation consensus agreement for the conservative management of Achilles and patellar tendinopathy: a modified Delphi approach — https://pubmed.ncbi.nlm.nih.gov/39824541/
7. Comparability of the Effectiveness of Different Types of Exercise in the Treatment of Achilles Tendinopathy: A Systematic Review — https://pubmed.ncbi.nlm.nih.gov/37628466/
8. Topical glyceryl trinitrate (GTN) and eccentric exercises in the treatment of mid-portion achilles tendinopathy (the NEAT trial): a randomised double-blind placebo-controlled trial — https://pubmed.ncbi.nlm.nih.gov/39013615/