Steroid Injection vs PRP for Joint Pain: Which One, and When

The short answer: a corticosteroid injection is a fast, powerful anti-inflammatory. It usually reduces pain within a few days, and the relief typically lasts somewhere between a few weeks and a few months. PRP (platelet-rich plasma) is not an anti-inflammatory in the same sense. It is a concentrate of your own platelets injected into the joint or tendon with the aim of modulating the local healing environment. It works slowly, often taking four to twelve weeks to show its effect, and when it works the benefit tends to last longer than a steroid.

So the choice is rarely “which one is better”. It is: how quickly do you need relief, how many times has this joint already been injected, how advanced is the arthritis, is this a joint or a tendon, do you have diabetes, and can you afford a self-pay treatment that may or may not help you? Below is how I actually think through it in clinic, including the parts patients are right to worry about.

What a corticosteroid injection actually does

Corticosteroid is a potent suppressor of local inflammation. Injected into an arthritic knee, an inflamed subacromial space or a swollen small joint, it reduces the inflammatory chemicals that drive pain and swelling. Many patients feel better within 48 to 72 hours. In a genuinely inflamed joint the change can be dramatic.

What it does not do is change the structure of the joint. It does not thicken cartilage, repair a tendon or slow arthritis. It buys a window — a window in which physiotherapy becomes possible, sleep improves, and you can get back to loading the limb sensibly. If nothing is done inside that window, the pain usually returns when the drug wears off.

Duration is variable. In a mildly arthritic knee with a flare, three to four months of relief is common; in an advanced, bone-on-bone knee it may be six weeks. In some patients it does very little, which is itself useful information: a joint that does not respond to a well-placed steroid injection is often a joint whose pain is not primarily inflammatory.

The side effects patients are right to ask about

Post-injection flare

In a minority of patients the joint becomes more painful for 24 to 48 hours after the injection before it settles. This is a recognised reaction to the crystalline steroid preparation. Ice and simple analgesia usually manage it. The reason it matters is that a flare can be very difficult to distinguish from an early infection, so if the pain is worsening beyond 48 hours, or there is fever, spreading redness or you feel systemically unwell, that needs to be seen the same day and not waited out.

Skin depigmentation and fat atrophy

Where steroid leaks back along the needle track or is injected too superficially, it can cause a pale patch of skin and a dimple of lost subcutaneous fat at the site. In Indian skin tones the depigmented patch is visible and patients find it distressing. It is more likely with superficial injections — around the lateral elbow, the greater trochanter, small joints of the hand, the heel — than with deep intra-articular knee or hip injections. It often improves over many months but may not fully reverse. Careful placement, and using ultrasound guidance where the target is small or deep, reduces the risk.

Blood sugar rise in diabetics

This one matters a great deal in India, where a large share of my arthritis patients also have type 2 diabetes. A corticosteroid injection can raise blood glucose for roughly two to five days, sometimes substantially, even when the injection is into a joint rather than into a vein or muscle. If you are diabetic, tell the doctor before the injection, not after. Practical steps: check sugars more often for the next week, do not schedule the injection the day before an important fasting test, and if your HbA1c is poorly controlled, it is often better to get the diabetes in order first. Very poorly controlled diabetes also increases infection risk after any injection.

How often is too often

The conventional and sensible limit is not more than three to four injections into the same joint in a year, and spaced at least three months apart. Frequent repetition is a signal that the strategy is failing, not a reason to keep repeating it. There is reasonable concern from imaging studies that repeated intra-articular steroid over years may be associated with cartilage loss, and there is long-standing evidence that steroid weakens tendon and other collagen tissue locally. Whether repeated injection meaningfully accelerates arthritis in an average patient is still debated, but the direction of the concern is clear enough that I do not treat steroid as an indefinitely repeatable solution.

The absolute rule about tendons

Corticosteroid is never injected into the Achilles tendon, and I avoid injecting it into other high-load, load-bearing tendons such as the patellar tendon. Steroid can weaken tendon collagen, and rupture of a weight-bearing tendon is a far worse problem than the pain you were treating. Peritendinous injections around a tendon sheath are a different matter and are sometimes appropriate, but the substance of a major load-bearing tendon is off-limits. If someone offers you a steroid shot into a painful Achilles, decline it. Chronic tendon problems are better addressed with graded loading, and I have written separately about non-surgical care for tendinopathy.

What PRP actually does — and does not do

PRP is prepared by taking your own blood, spinning it in a centrifuge to concentrate the platelets, and injecting that concentrate into the target joint or tendon. Platelets carry growth factors, and the working theory is that delivering them in concentration shifts the local biology towards repair and away from persistent degenerative inflammation. It is your own tissue, so allergic reaction is not a concern, though a few days of ache and swelling after the injection is normal and expected.

The important honest statement: PRP does not regrow cartilage. If you have been told that an injection will rebuild a worn-out knee, that is not accurate. What the better evidence supports is symptom improvement — pain and function — in early to moderate knee osteoarthritis, and in certain tendon problems, particularly lateral epicondylitis (tennis elbow). In advanced, bone-on-bone arthritis with significant deformity, PRP is much less likely to give worthwhile relief, and I say so before taking anyone's money.

The second honest statement is about consistency. There is no single standardised PRP. Preparation systems differ in platelet concentration, in whether white cells are included, in spin protocol and in injected volume. Two clinics can both call it PRP and deliver quite different products. This variability is one reason the published trials disagree with one another. When you are choosing where to have it done, it is fair to ask what system is used, what concentration it achieves, and whether the injection will be performed under ultrasound guidance so it actually reaches the target.

The third is cost. In Bengaluru, PRP and similar regenerative injections are essentially self-pay. Insurance policies generally do not cover them, and a course is often more than one injection. That is real money spent on a treatment with a genuine chance of not helping you. I have set out the numbers and who is and is not a reasonable candidate in the pieces on PRP and GFC cost and eligibility in Bengaluru and PRP compared with GFC for knee pain. More detail on the treatment itself is on the regenerative treatment page.

Head to head

Feature	Corticosteroid	PRP
Onset of relief	2 to 5 days	4 to 12 weeks, gradual
Typical duration if it works	Weeks to a few months	Several months to about a year
Mechanism	Suppresses inflammation	Aims to modulate healing biology
Effect on diabetes	Can raise blood sugar for days	No effect on blood sugar
Repeatability	Limited; not more than 3 to 4 per joint per year	Repeatable, but cost limits it
Cost	Low; often covered in an OPD consult	Self-pay, significantly higher
Best evidence in	Inflammatory flares, bursitis, frozen shoulder, moderate to advanced OA flare	Early to moderate knee OA, lateral epicondylitis

How I decide

Steroid is usually the better choice when

• The joint is acutely inflamed, hot and swollen, and you need relief now rather than in two months.
• There is a clear bursitis, an impingement flare, or a frozen shoulder in the painful phase where a window of relief will let physiotherapy actually happen.
• Cost is a hard constraint and the alternative is doing nothing.
• You need to get through a specific short-term event — a wedding, travel, an exam period — and we both understand this is a bridge, not a cure.
• Arthritis is advanced and surgery is planned but some months away.

PRP is usually the better choice when

• Osteoarthritis is early or moderate on X-ray, joint space is preserved, and there is no significant deformity.
• You are diabetic and want to avoid a steroid-driven sugar spike.
• The problem is a chronic tendinopathy, particularly tennis elbow, that has not responded to a proper loading programme.
• You have already had steroid into that joint two or three times and it is working for shorter and shorter periods.
• You can afford it, and you accept in advance that it may not work.

Neither — do something else instead — when

• The knee is bone-on-bone with a fixed deformity and a real functional limitation. That is a conversation about arthritis management and, eventually, joint replacement, not an injection.
• There is a mechanical problem: a locked knee, a loose body, a genuinely displaced tear causing catching. No injection fixes a mechanical block.
• The joint might be infected. Fever, a hot swollen joint and feeling unwell is an emergency, not an injection candidate.
• The pain is referred from elsewhere — hip arthritis presenting as knee pain, or lumbar spine pain presenting as buttock and thigh pain. Injecting the wrong structure achieves nothing.
• Nothing has yet been tried in the way of weight management, strengthening and load modification.

The part that decides the outcome

Whichever injection you choose, it is an adjunct. It is not the treatment. The treatment for most degenerative joint pain is load management and strengthening: getting the quadriceps and hip abductors strong, reducing body weight where that applies, changing the activities that reliably flare the joint, and sustaining that for months rather than weeks. An injection makes that work possible when pain is blocking it. It does not replace it.

The patients who do best with either injection use the pain-free window to build strength. The ones who do worst feel better, resume everything that hurt them, and are back in three months asking for another shot. If you are considering an injection for knee pain, the wider set of non-surgical options is covered on the knee pain page.

When to see a doctor

Book an appointment before any injection if you have not had the joint properly examined and imaged — injecting a joint without a diagnosis is guesswork. Beyond that, see a doctor promptly if you have:

• A hot, swollen, very painful joint with fever or chills — this needs same-day assessment to exclude septic arthritis, particularly if you have had a recent injection.
• Pain after an injection that is increasing beyond 48 hours rather than settling.
• A joint that locks, gives way, or will not straighten fully.
• Sudden inability to push off or stand on tiptoe after a pop — this suggests tendon rupture and needs urgent review.
• Unexplained weight loss, night pain that wakes you consistently, or a history of cancer alongside new bone or joint pain.
• Rapidly worsening function despite reasonable conservative treatment over three months.

An injection is a decision worth making carefully and with a clear diagnosis behind it. Done for the right joint at the right stage, it is genuinely useful. Done repeatedly as a substitute for addressing the underlying load problem, it delays the treatment that would have worked.