The tear is only half the diagnosis. The other half is measurable.

PRP FOR THE SHOULDER AND ROTATOR CUFF

Shoulder pain that wakes you when you roll onto it, and makes reaching into a back seat briefly impossible, is usually the rotator cuff. What is far less certain is whether the structure your scan describes is the structure generating the pain — and whether the body it sits in can repair it.

Start with what the scan cannot tell you

Rotator cuff tears are extremely common in people with no symptoms at all. In the Chingford cohort — 1,003 women aged 64 to 87, imaged as a general population sample rather than as a clinic sample — full-thickness tears were present in 22.2%, and 48.4% of those full-thickness tears caused no symptoms. Comparative work on asymptomatic and symptomatic shoulders points the same way: the anatomy on the report and the reason a shoulder hurts are two different questions.

A report describing a tear therefore tells you what your shoulder looks like, not why it hurts. That matters for the decision, because a patient shown the word “tear” reasonably concludes something torn must be repaired. Often the tear had been there for years and something else changed — load, sleep, the neck, or the tendon's metabolic environment.

What the tissue is doing

The supraspinatus fails at a hypovascular zone near its insertion. Chronic disease shows the familiar picture: collagen disorganization, increased ground substance, neovascular and neural ingrowth, and a repair process that started and did not finish. Subacromial bursal inflammation frequently accompanies it and may be generating more of the pain than the tendon itself.

The measurable half: three numbers about the body, not the shoulder

A systematic review and meta-analysis looked specifically at whether metabolic comorbidity associates with rotator cuff disease. Across twelve studies of diabetes, five of hypertension and eight of hyperlipidemia:

• Diabetes — odds of rotator cuff disease increased, OR 1.49 (95% CI 1.43 to 1.55), and specifically associated with cuff tears at OR 1.28 (1.07 to 1.52).
• Hyperlipidemia or dyslipidemia — OR 1.48 (1.42 to 1.55).
• Hypertension — OR 1.40 (1.19 to 1.65).

The authors' own caveat belongs here rather than in a footnote: they judged the epidemiological evidence plausible for diabetes and hyperlipidemia but not for hypertension, where the risk of bias is highest. We report it that way because a number quoted without its limitation is advertising.

Widen the lens from the cuff to tendon generally and the signal gets larger. Pooling 31 studies, tendinopathy was substantially more common in people with diabetes — odds ratio 3.67 (2.71 to 4.97) — with the relationship running in both directions, and people who had both had carried diabetes about 5.26 years longer than people with diabetes alone. Duration of exposure matters, which is what you would expect if the mechanism is glycation of collagen rather than bad luck.

The finding that makes this actionable rather than merely interesting

Association is not the same as leverage. So here is the study that closes the gap: a meta-analysis of six studies and 4,395 patients asked whether glycemic control, measured by hemoglobin A1c, changed what happened after arthroscopic rotator cuff repair in diabetic patients.

Better control was associated with a markedly lower retear rate — odds ratio 0.242 (95% CI 0.128 to 0.454). Revision rates did not differ. Read that plainly: among people with diabetes, the ones whose glucose was controlled tore again at roughly a quarter the odds of the ones whose glucose was not. The included studies were mostly retrospective cohorts rather than randomized trials, and that is a real limitation. It is still the clearest available demonstration that a modifiable systemic variable changes a structural outcome in this exact tendon.

That is the whole argument of this practice, expressed in one odds ratio. The tendon does the repair. What we can change is the environment it repairs in and the precision of what we deliver. What gets measured, and why it happens before rather than after, is on the metabolic health page.

Nicotine, with the number attached

A meta-analysis of fourteen studies covering 73,817 people, of whom 8,553 were smokers, found smoking associated with roughly double the risk of retear after cuff repair (risk ratio 2.06, 95% CI 1.30 to 3.28) and a higher reoperation rate (1.29, 1.20 to 1.40). Interestingly, most patient-reported scores did not differ. The structure failed more often; the questionnaires did not notice.

If you are still using nicotine in any form, that is the single highest-yield change available to you before anything is injected, and it is worth more than the choice of preparation.

Sleep, which runs in both directions

A painful shoulder destroys sleep architecture, and fragmented sleep raises inflammatory tone and lowers pain thresholds, which makes the shoulder hurt more. People arrive exhausted, and the exhaustion is not incidental — it is part of the loop holding the problem in place. It is also one of the few variables that can be moved in weeks rather than months.

What a biologic is aimed at here, and what it is not

For tendinopathy without a significant structural tear, the target is a stalled repair in poorly perfused tissue, restarted and then organized with load. Small partial tears with pain out of proportion to the structural damage belong in the same category.

For a large full-thickness retracted tear it is not. A tendon that has pulled off its footprint does not reattach because growth factors were placed near it, and anyone telling you otherwise is selling something. That is a mechanical problem with a mechanical answer, discussed on the surgery page.

Where published results in this field disagree with each other, the reasons are usually procedural rather than biological: trials rarely characterize what was actually injected, they inject different targets — tendon, bursa or joint — under the same label, and almost none of them record the metabolic state of the person receiving it. Pooling studies like that compares treatments that were never the same treatment given to patients who were never the same patients. The longer version is on why the trials disagree.

Before reaching for an operation

It is worth knowing that when arthroscopic subacromial decompression was tested against placebo surgery in a multicenter randomized trial, the operation did not deliver the advantage its popularity implied. That finding is not an argument against surgery in general — it is an argument for being precise about which shoulder problem is being operated on, and for exhausting the reversible variables first.

What else makes a shoulder hurt

• Adhesive capsulitis — loss of passive external rotation is the giveaway, and injecting a cuff tendon will not touch it.
• Acromioclavicular joint disease — pain localized to the top of the shoulder, worse on cross-body reach.
• Cervical radiculopathy — a C5 or C6 problem presenting as shoulder pain with a normal shoulder examination, covered on the shoulder-or-neck page.
• Glenohumeral osteoarthritis — stiffness and crepitus rather than impingement.
• Calcific tendinopathy — a distinct entity with its own treatment, sometimes dramatically painful, often responsive to needling.

How it is done and what follows

Ultrasound guidance, with the target chosen deliberately: intratendinous fenestration for focal degeneration, or a peritendinous and bursal approach where the bursa is the likelier pain source. This is not a preference. A systematic review and meta-analysis found ultrasound-guided shoulder girdle injections both more accurate and more effective than landmark-guided ones, and the shoulder girdle is several possible targets sitting within a few centimeters of each other.

• Weeks 0 to 2. Pendulum and passive range. Expect a flare, sometimes for several days, and expect it to disturb sleep before it improves it.
• Weeks 2 to 6. Isometrics into cuff activation, then light resisted external rotation. Scapular control work starts here.
• Weeks 6 to 12. Progressive resistance, with overhead loading reintroduced gradually.
• Months 3 to 6. Where the answer arrives. Cuff remodeling is slow and this joint tolerates rushing badly.

What we establish before agreeing to treat a shoulder

More shoulders are injected than should be, and the reason is usually that the assessment stopped at the scan.

• Does passive range match active range? If passive external rotation is lost, this is a capsule problem and no cuff injection will help it.
• Does the neck reproduce the pain? A cervical screen takes two minutes and reclassifies a meaningful number of shoulders.
• Where does the tenderness actually sit? The acromioclavicular joint is a few centimeters from the cuff and a completely different treatment.
• What does the scapula do when you lift? A cuff working against a scapula that stops rotating will keep failing whatever is injected into it.
• What is the metabolic picture? It changes the odds we quote you and the sequence we recommend, and on the numbers above it changes them more than most people expect.

If that assessment points somewhere other than a needle, we say so. It is a better outcome than a well-performed injection into the wrong structure.

When a shoulder does not respond

The first question is almost always whether the pain source was correctly identified, because the differential above is long and shoulders are frequently mislabeled. The second is whether the scapular and thoracic contribution was addressed. The third is what has changed in the metabolic picture since the first visit — a shoulder treated at an A1C of 9 and reassessed at 9 has not had the plan completed. Only then does a surgical opinion or a second injection belong in the conversation.

What people ask about a torn cuff

My scan shows a tear. Do I need surgery?

Not necessarily, and the population data is the reason — nearly half of full-thickness tears in a general population sample caused no symptoms at all, so the tear on your report may not be what hurts. Where an operation genuinely is the right step is set out on the surgery page.

Does my diabetes really affect my shoulder?

Measurably, and in the direction you would not want: diabetes raises the odds of cuff disease, and among diabetic patients having a repair, better glycemic control was associated with about a quarter the odds of retearing. That is the most actionable finding on this page, and what we measure is on the metabolic health page.

I smoke. Should I deal with that first?

Yes, and it outranks the choice of preparation — smoking roughly doubled the retear risk after cuff repair in a meta-analysis of nearly 74,000 people. How the rest of the sequence is ordered is on the candidacy page.

Is an injection better than a cortisone shot here?

Cortisone often works faster and repeated exposure weakens tendon, so the trade is the same one made everywhere else on this site and it depends on what you need the next three months to look like. The comparison is on the cortisone page.

Why does it hurt more at night?

Side-lying compresses the cuff and inflammatory tone peaks overnight, so it is typical rather than a bad sign — though the sleep it costs you feeds back into the pain. What the first weeks feel like is on the recovery timeline.

Related reading

• Is it my shoulder or my neck?
• The shoulder after a collision or a fall
• Why we ask for your bloodwork
• Why the word tendonitis produces the wrong treatment
• How PRP works
• Am I a candidate?