Physical Therapy, Injections or Both for Osteoarthritis: How the Options Fit Together

The physical therapist’s gym and the injection room sit two doors apart in many outpatient buildings, and people with an aching knee often stand in the corridor between them wondering which door they were supposed to walk through. One promises months of homework. The other promises a needle and, maybe, a quieter joint by the weekend. It is a fair thing to feel torn about.

The physical therapy vs injections osteoarthritis question comes up in almost every joint consultation, and the honest answer is less dramatic than a duel. Guidelines from the NHS and others treat exercise-based therapy as the base layer for osteoarthritis and treat injections as a tool with a specific, time-limited job.

This explainer walks through what each option actually does inside the joint, what the best head-to-head evidence shows, who tends to be offered which, and how clinicians combine them. The decision, at every step, belongs to you and your treating team.

Physical therapy vs injections for osteoarthritis: why this isn't an either/or question

Osteoarthritis is the wearing and remodeling of a joint over years: cartilage thins, the bone underneath thickens, the joint lining can become inflamed, and the muscles around the joint weaken because moving hurts. Each of those changes responds to a different lever, which is the first reason the two treatments are rarely rivals.

An injection works on one lever, the inflamed lining, and works fast. Physical therapy works on the others: the muscles that unload the joint, the stiffness that shortens your stride, the confidence that keeps you climbing stairs. Those gains arrive slowly and tend to stay. The NHS lists exercise as a core treatment for everyone with osteoarthritis, with injections reserved for pain that is severe or is blocking other treatment, and that ordering reflects how the two behave over time rather than a ranking of importance.

Think of it the way a gardener thinks about a dry summer. Watering today keeps the plants alive; improving the soil keeps them alive next year. Neither replaces the other, and the most useful conversations with a clinician are about sequence, not selection.

Three practical questions tend to settle the sequence for a given person:

• Is pain currently so high that you cannot exercise at all, even gently?
• Have you already had a structured, supervised exercise program, or only been told to keep active?
• Are there health reasons, such as diabetes or a recent joint infection, that make a steroid injection a more careful decision?

The rest of this article unpacks each of those, starting with what happens inside a physical therapy program, because it is the part most people picture least accurately.

What actually happens in physical therapy for osteoarthritis

A physical therapist is a licensed clinician who assesses movement and prescribes exercise and hands-on techniques; for osteoarthritis, the first visit is mostly measurement. How far does the knee bend? Can you rise from a chair without using your arms? How does the hip on the painful side behave when you stand on one leg? Those baselines matter because progress in osteoarthritis is measured in function, not on X-rays.

From there, a typical program has four strands. Strengthening targets the muscles that act as shock absorbers, especially the quadriceps at the front of the thigh and the gluteal muscles at the side of the hip; a stronger quadriceps takes load off the knee joint surface with every step. Range-of-motion work keeps the joint moving through its full arc, because a stiff joint concentrates force on a smaller patch of cartilage. Manual therapy, meaning the therapist mobilizing the joint and soft tissue by hand, can ease pain enough to make the exercise tolerable. Education covers pacing, footwear, walking aids and, where relevant, weight, since the NHS notes that losing excess weight reduces joint load and pain.

The sessions themselves are unglamorous. Sit-to-stands with a slow lowering phase. Step-ups onto a low block. Side-lying leg raises that look easy and are not. The therapist adjusts the difficulty each week so the work stays hard enough to build tissue and easy enough that flare-ups stay small.

Home practice carries most of the benefit. In the largest recent head-to-head trial comparing therapy with injection for knee osteoarthritis, the therapy arm received a handful of clinic visits over roughly four to six weeks with booster sessions later, and the daily work happened at home in between. What you do on the days you do not see the therapist is the treatment.

What an osteoarthritis joint injection involves, and what each type does

An intra-articular injection means medicine placed directly into the joint space, usually with a local anesthetic to numb the skin first. The procedure itself is brief. The clinician cleans the skin, may use ultrasound to guide the needle in a deep joint like the hip, and delivers the medicine in a matter of seconds. You are typically asked to rest the joint for a day or two afterward.

The most common medicine is a corticosteroid, a synthetic version of the body’s own anti-inflammatory hormone cortisol. Inside the joint it dampens the activity of immune cells in the synovium, the thin lining that produces joint fluid and becomes inflamed in osteoarthritis flares. Less inflammation means less swelling, less heat and, for many people, less pain. Mayo Clinic describes the relief as beginning within a few days and typically lasting up to several months, with clinicians generally spacing repeat injections and limiting how many are given in a year because of concern about cartilage and surrounding tissue.

Hyaluronic acid injections, sometimes called viscosupplementation, add a gel-like substance similar to a component of natural joint fluid. The theory is lubrication and shock absorption; the evidence is mixed, and NICE-based guidance followed by the NHS does not recommend them for osteoarthritis.

Platelet-rich plasma, made by concentrating a person’s own blood platelets, and so-called stem cell injections are also offered in some settings. Mayo Clinic notes that evidence for these is limited, and mainstream guidelines do not currently endorse them as proven treatments for osteoarthritis.

Every injection carries small risks: a temporary flare of pain for a day or two, a rise in blood sugar in people with diabetes, thinning or lightening of skin at the site, and, rarely, joint infection. Your clinician weighs those against the likely benefit for your joint.

Cortisone shot vs physical therapy for the knee: what the head-to-head trial found

For years the comparison rested on indirect evidence, until a randomized trial published in the New England Journal of Medicine put the two directly against each other in 156 adults with knee osteoarthritis. Half received a corticosteroid injection, with the option of repeat injections during the year. Half received physical therapy built around supervised exercise and manual treatment.

At one year, the physical therapy group reported less pain and less functional disability on a standard osteoarthritis questionnaire than the injection group. The difference was not enormous, but it was consistent, and it pointed in the direction guidelines had already been leaning: the slower treatment produced the more durable result.

Several details in that trial deserve a plain reading rather than a headline. The injection group did improve; steroid injections are not ineffective, they simply faded. The therapy group’s advantage appeared over months, not weeks, which matches how muscle strength develops. Participants in the therapy arm also had more contact time with a clinician, and the trial could not fully separate the exercise from the attention. Finally, the result applies to knees; hips and hands were not studied, and hip injections are technically different because the joint is deep and usually needs imaging guidance.

What the trial did not show is that injections are a mistake. It showed that when someone can exercise, exercise pays off more at the one-year mark. The practical question that flows from that is whether a person can exercise right now, or whether pain has to come down first, and that is where the two options begin to fit together rather than compete.

How long do knee injections last, realistically?

The answer people hope for is months; the honest range is wide. Mayo Clinic’s patient guidance on cortisone injections describes relief that can last up to several months, and clinical experience includes people who feel little change and others who get a long stretch of comfort from a single injection. There is no reliable way to predict in advance which group a given knee will fall into.

A few patterns help set expectations. Relief usually starts within days rather than instantly; the local anesthetic mixed into the injection can produce a brief numbness on the day, followed by a lull, and then the steroid effect builds. A short flare of soreness in the first day or two is common and settles on its own. People whose knee is visibly swollen and warm, signs of active synovitis, often notice a bigger difference than people whose pain is mainly mechanical, meaning it comes with weight-bearing and eases at rest.

The effect wears off because the medicine is gradually cleared from the joint while the underlying cartilage change remains. That is why repeat injections are common and why clinicians limit them. Mayo Clinic notes concern that repeated steroid exposure may weaken cartilage and nearby tendons over time, so most clinicians space injections out and reserve them for flares rather than using them on a fixed schedule.

Timing an injection deliberately can stretch its usefulness. Some people ask for one ahead of a period when function matters most, such as before starting a supervised exercise program or before a trip that will involve a lot of walking. Those are reasonable conversations to have with the clinician who knows your joint, and the decision about timing and frequency rests with them.

Does physical therapy help osteoarthritis if the cartilage is already worn?

People often arrive at their first therapy appointment holding an X-ray report full of words like severe joint space narrowing and bone spurs, quietly convinced that exercise cannot fix a joint that looks like that. They are right about one thing: therapy does not regrow cartilage. It does not need to.

Pain in osteoarthritis correlates surprisingly poorly with what an X-ray shows. Some people with dramatic changes on imaging walk comfortably; others with mild changes are limited daily. Much of that gap is explained by tissues the X-ray does not capture: the inflamed lining, the weak muscles, the stiff capsule, the sensitized nerves that have been firing for months. Those are precisely the things exercise changes.

Stronger thigh and hip muscles reduce the load that reaches the joint surface and control the small side-to-side wobbles that irritate it. Regular movement keeps joint fluid circulating, which is how cartilage receives nutrients, since it has no blood supply of its own. Steady, tolerable activity also appears to turn down pain sensitivity over time, the reverse of what happens when a joint is rested for weeks.

The NHS positions exercise as the main treatment for osteoarthritis regardless of stage, and the National Institutes of Health includes exercise and physical therapy among its first-line recommendations. The trial described earlier enrolled people with a range of X-ray severity, and the therapy benefit was not restricted to mild disease.

There is a limit. When a joint is end-stage, meaning bone is grinding on bone, pain is constant including at night, and function has collapsed despite a proper trial of exercise and other measures, surgery becomes part of the conversation. Even then, therapy before and after surgery matters, and the people who go in stronger generally find rehabilitation easier. Your surgeon and therapist together will judge where your joint sits on that spectrum.

Who is usually offered each option, and who is usually asked to wait

Exercise-based therapy is offered to nearly everyone with osteoarthritis, which makes the more interesting question who gets an injection and who is asked to hold off.

An injection tends to be suggested when pain is severe enough that a person cannot begin or continue exercising, when a joint is in an obvious inflammatory flare with swelling and warmth, or when other measures have been tried and have not been enough. The NHS frames steroid injections in exactly those terms: for severe pain, or when other treatments have not worked. A clinician may also suggest one to bridge a specific gap, such as the weeks before a planned surgery or the start of a rehabilitation program.

Several situations usually lead to waiting, adjusting or choosing another route:

• A suspected or recent joint infection, or an infection anywhere in the body, because steroids suppress local immunity and a needle can introduce bacteria.
• Diabetes that is poorly controlled, since steroids raise blood sugar for several days; the clinician may want glucose monitored more closely around the injection.
• A recent steroid injection into the same joint, because of the spacing limits described earlier.
• Blood-thinning medicines, which do not always rule out an injection but may change how it is done.
• A planned joint replacement within the coming weeks or months, since some surgeons prefer a steroid-free interval beforehand to reduce infection risk.

Therapy has fewer hard stops, but a therapist will modify a program around a recent fall, a heart or lung condition that limits exertion, or a flare severe enough that even gentle range-of-motion work is not tolerated. In that last case, the two options meet: the injection settles the flare, and therapy resumes.

None of these are rules a reader should apply alone. They are the factors a clinician weighs, and that clinician has the whole picture.

Physical therapy vs injections osteoarthritis: side-by-side

Laying the two options next to each other makes the complementary shape obvious. The table summarizes what mainstream guidance and the trial evidence describe; individual experience varies, and none of these rows is a promise.

Feature	Physical therapy (exercise-based)	Corticosteroid injection
What it targets	Muscle strength, joint mobility, movement patterns, pain sensitivity	Inflammation of the joint lining
Speed of effect	Gradual; gains build over weeks	Usually within days
Durability	Persists while activity continues; one-year advantage in head-to-head trial	Weeks to several months, then fades
Guideline position	Core treatment for all with osteoarthritis	For severe pain or when other measures have not helped
Main risks	Temporary soreness, small flare-ups if progressed too fast	Short flare, raised blood sugar, skin changes, rare infection, concern about cartilage with repeated use
Effort required	High and ongoing	Low; a brief procedure
Typical role	Foundation	Bridge or flare control

Two rows explain most real-world decisions. Speed of effect is why injections are chosen when a person is in too much pain to move. Durability is why therapy is chosen when a person wants to still be walking comfortably next year. Most clinicians, and most guidelines, end up recommending the second while sometimes using the first to get there.

Hyaluronic acid, platelet-rich plasma and stem cell injections are deliberately left out of the table because the evidence does not support placing them alongside the two established options; they are discussed separately below.

How the two fit together: using an injection as a bridge to exercise

The most common way clinicians combine the two treatments is sequential, and the logic is simple. A joint in a hot flare will not tolerate the sit-to-stands and step-ups that build strength. An injection can lower the inflammation enough that, a week or two later, those exercises become possible. The window of relief is then spent building the muscle that will carry the joint after the medicine fades.

Timing matters in that plan. Starting therapy the day after an injection is rarely useful, because the joint is often sore and the steroid effect has not fully built. Waiting until the relief is wearing off wastes the window. Most clinicians aim for a start once the initial post-injection soreness has settled, typically within the first week or two, and then progress steadily through the months when the joint is quieter.

The reverse order also happens. Someone who has been exercising well for months and then hits a flare, perhaps after a long day on their feet, may be offered an injection to get through the flare without losing the strength they have built. Here the injection protects the therapy rather than enabling it.

A third pattern is less about sequence and more about honesty: a person who has had several injections with steadily shorter benefit is often at the point where more injections add little, and a structured exercise program, if not yet properly tried, becomes the more promising path. Conversely, a person who has done everything right in therapy and still cannot manage stairs may be someone for whom the conversation turns toward surgical options rather than more injections.

Each of those is a judgment about one joint in one life. The clinician who has examined the joint, seen the imaging and heard how the pain behaves through a day is the person to make it with you.

What the following days and weeks usually look like

After an injection, the first day is often uneventful and occasionally worse before it is better. The local anesthetic wears off within hours, and some people feel a distinct increase in pain for a day or two, which Mayo Clinic describes as a common short-term flare. Resting the joint for a couple of days is usually advised; a hot bath or vigorous activity the same day is not. People with diabetes are often asked to check blood sugar more frequently for several days, since readings can rise. By the end of the first week, most people who are going to respond have noticed a difference.

The middle weeks are the payoff period, and the period most easily wasted. This is when the joint is quietest and exercise is most tolerable. If therapy is part of the plan, this is when it should be under way.

Toward the end of the effect, pain creeps back rather than returning overnight. Noticing that creep is useful information for your clinician, because a very short response sometimes changes the plan.

Therapy follows a different arc. The first two weeks can feel like nothing is changing, and mild muscle soreness the day after sessions is normal and different from joint pain; therapists teach people to tell the two apart. Strength gains become noticeable to most people somewhere in the first month or two, first as easier stairs or a longer comfortable walk rather than as a lower pain score. The trial described earlier tracked improvement out to one year, and the advantage for the therapy group was clearest at that late point, which is worth remembering during the early weeks when the injection group in that study was feeling better.

Either way, keeping a simple note of pain, swelling and what you managed to do each day gives your team far better information than memory at the next appointment.

Hyaluronic acid injections for knee osteoarthritis, PRP and stem cells: what the evidence actually shows

Three other injections come up in almost every conversation, usually because someone has read about them or been offered them. They deserve a clear-eyed summary rather than either enthusiasm or dismissal.

Hyaluronic acid is a natural component of joint fluid, and injecting a gel version into the knee is meant to restore cushioning and lubrication. Trials have produced inconsistent results, with some showing small benefits and others none beyond placebo. Mayo Clinic describes the evidence as mixed. The NHS does not offer these injections for osteoarthritis, following national guidance that concluded the benefit was not sufficient. Some people report relief; the difficulty is that any injection into a painful joint tends to produce some short-term improvement, and separating the substance from the procedure has proven hard.

Platelet-rich plasma involves drawing a person’s blood, spinning it to concentrate platelets, and injecting that concentrate into the joint. Platelets release growth factors, and the hope is tissue repair. Studies so far are small, vary in how the plasma is prepared, and have not shown consistent benefit; mainstream guidance treats PRP as investigational for osteoarthritis rather than proven.

Stem cell injections, usually meaning cells taken from a person’s own fat or bone marrow, are marketed in some places as regenerative. There is no good evidence that these cells regrow cartilage in a human osteoarthritic joint, and the procedures are largely unregulated in many settings. Mayo Clinic advises that evidence is insufficient and that people should discuss such offers with their own clinician before proceeding.

The honest summary is that none of the three has earned a place beside exercise and corticosteroid injection in current guidelines. That may change as evidence accumulates. Until it does, a person weighing physical therapy vs injections for osteoarthritis is weighing the two options this article has spent most of its time on.

What people often get wrong about therapy and injections

Some misunderstandings are so common that they shape decisions before anyone reaches a clinic, and correcting them changes what people choose.

Rest protects the joint. It feels intuitive and is largely backwards. Prolonged rest weakens the muscles that unload the joint, stiffens the capsule and, over weeks, increases pain sensitivity. The NHS and NIH both list regular activity as central to managing osteoarthritis. Rest is for the first day or two after an injection or during a sharp flare, not as a strategy.

An injection means the joint is getting worse. Injections are chosen for pain and inflammation, not as a marker of stage. Plenty of people with modest X-ray change have injections during flares, and plenty with advanced change never need one.

Cortisone shots wear out the cartilage, so they should be avoided at all costs. There is genuine concern, reflected in Mayo Clinic guidance, about repeated steroid exposure and cartilage, which is why frequency is limited. An occasional injection to break a flare or enable exercise is a different proposition from a standing schedule, and clinicians weigh the trade-off case by case.

Physical therapy is just being told to exercise. Generic advice to keep active produces far less than a supervised, progressive program with a therapist adjusting load week by week. Many people who say therapy did not work for them never had the second kind.

If the injection worked, that proves inflammation was the whole problem. Steroids also have a general pain-dampening effect and are usually mixed with anesthetic; a good response is encouraging, not diagnostic.

Newer means better. Hyaluronic acid, PRP and stem cell injections are newer than either steroid injection or exercise. They are not, on current evidence, better.

Each of these corrections points in the same direction: the least glamorous option is the one with the strongest evidence, and the quick option has a real but narrow job.

Questions to ask your care team

A consultation about an aching joint moves quickly, and the questions that matter most are easy to forget once the conversation turns to imaging and appointment dates. Bringing a short list changes the meeting. The questions below are meant to open a discussion, not to be answered from a webpage.

• Based on my examination and imaging, how much of my pain do you think is coming from inflammation and how much from weakness and stiffness?
• Have I had a proper supervised exercise program yet, or only general advice to stay active? If not, is that where we should start?
• If you are suggesting an injection, what is its specific job right now: breaking a flare, making exercise possible, or getting me through a particular period?
• How will we judge whether the injection worked, and what would a poor or very short response tell you about next steps?
• How many steroid injections into this joint would you be comfortable with over time, and what would make you want to stop?
• Do I have any health conditions, such as diabetes or a bleeding tendency, that change how or whether an injection should be done?
• When after an injection should I begin or resume exercise, and what should I avoid in the first days?
• What signs after an injection would you want to hear about the same day?
• If therapy and injections together are not enough, what does the next stage of the conversation look like, and how would we know we are there?
• Is there anything I am doing day to day, footwear, walking aids, weight, activity pattern, that is loading this joint more than it needs to be?

Write down the answers, or ask whether a family member can sit in. The answer to the second question in particular, whether you have genuinely had structured therapy, decides more treatment plans than any imaging finding, and it is a question only you and your clinician can answer together.

When to call your doctor

Osteoarthritis is a slow condition, and most of what it does can wait for a scheduled appointment. A few things cannot, and they are worth knowing before you need them.

After a joint injection, contact the clinician who performed it, or seek urgent care, if the joint becomes increasingly hot, red and swollen after the first two days rather than settling; if you develop a fever or chills; if pain becomes severe and unrelenting rather than easing; or if there is spreading redness, pus or a persistently leaking puncture site. These can be signs of joint infection, which is rare but needs same-day assessment. People with diabetes should also call if blood sugar readings climb well above their usual range and do not come back down within a few days.

During a physical therapy program, a small increase in muscle soreness after sessions is expected. Call your therapist or doctor if the joint swells markedly and stays swollen, if pain after exercise lasts beyond a day or two and is getting worse week on week rather than better, if the joint locks or gives way so that you fall or nearly fall, or if you notice new numbness, tingling or weakness in the leg or arm below the joint.

At any time, regardless of treatment, seek prompt medical advice for a joint that suddenly becomes hot, swollen and too painful to move without an injury to explain it; for pain that wakes you every night and is no longer related to activity; for unexplained weight loss or feeling generally unwell alongside joint symptoms; or for a sudden inability to bear weight after a fall. These are not typical of osteoarthritis and need evaluation for other causes.

If you are unsure whether something counts, calling is the right choice. Your treating team would rather hear about a false alarm than miss an early sign, and every decision about what happens next stays with them.

Frequently asked questions

Cortisone shot vs physical therapy for the knee: which one should I try first?

Guidelines place exercise-based physical therapy first for nearly everyone with knee osteoarthritis, with a cortisone injection reserved for severe pain or a flare that prevents exercise. The one large head-to-head trial found therapy gave more benefit at one year, while the injection worked faster. Your clinician will decide the order based on how much pain you have now and whether you have already had a structured program.

How long do knee injections last for osteoarthritis?

Corticosteroid injection relief typically begins within a few days and can last up to several months, according to Mayo Clinic, though the range is wide and some people notice little change. The effect wears off because the medicine is gradually cleared from the joint. Clinicians space repeat injections and limit how many are given because of concern about cartilage and nearby tissue with repeated use.

Does physical therapy help osteoarthritis when the X-ray already shows severe damage?

Yes, in most cases. Pain in osteoarthritis correlates poorly with X-ray appearance because much of it comes from inflamed lining, weak muscles and stiffness that imaging does not show, and those respond to exercise. Therapy does not regrow cartilage; it reduces the load reaching the joint and improves function. When a joint is truly end-stage despite proper therapy, surgery enters the conversation, but strength gained beforehand still helps recovery.

Are hyaluronic acid injections for knee osteoarthritis worth trying?

The evidence is mixed and current guidance does not recommend them. Trials show inconsistent results, some with small benefits and others none beyond placebo, and the NHS does not offer hyaluronic acid injections for osteoarthritis following a national evidence review. Some people do report relief, but any injection into a painful joint tends to produce short-term improvement, which makes the substance hard to judge. Discuss it with your clinician rather than deciding alone.

Can I have physical therapy and an injection at the same time?

Yes, and combining them is common. A typical approach is to have the injection first, allow a few days for post-injection soreness to settle, and then begin or resume therapy while the joint is quieter, usually within the first week or two. The relief window is spent building strength that outlasts the medicine. Your therapist and the clinician who gave the injection will coordinate the timing.

What are the risks of a steroid injection into an arthritic joint?

Common short-term effects include a flare of pain for a day or two and, in people with diabetes, raised blood sugar for several days. Less common effects include thinning or lightening of the skin at the injection site and weakening of nearby tendons. Joint infection is rare but serious. Repeated injections raise concern about cartilage, which is why clinicians limit how often they are given.

Why did my injection stop working sooner the second time?

Shorter relief with repeat injections happens for some people and has several possible explanations, including progression of the underlying joint change, less inflammation being present to treat, or simply natural variation between injections. A very brief response is useful information for your clinician, because it may suggest more injections will add little and that a structured exercise program or a different conversation about options is the better next step.

Is it safe to exercise with osteoarthritis if it hurts?

Mild discomfort during and after exercise is expected and generally safe, and the NHS and NIH both list regular activity as central to managing osteoarthritis. The signals that matter are joint swelling that persists, pain that is markedly worse the next day and worsening week on week, or the joint locking or giving way. A physical therapist teaches people to tell muscle soreness from joint irritation and adjusts the program accordingly.

Do PRP or stem cell injections repair cartilage in osteoarthritis?

No good evidence currently shows that either regrows cartilage in a human osteoarthritic joint. Platelet-rich plasma studies are small and inconsistent, and stem cell procedures are largely unregulated with insufficient evidence, according to Mayo Clinic. Mainstream guidelines treat both as investigational rather than established treatments. If you are offered one, discuss it with your own clinician before deciding.

How do I know when injections and therapy are no longer enough?

The usual markers are pain that is constant including at night, a joint that has lost most of its function, and no meaningful improvement despite a genuinely supervised exercise program and reasonable use of injections. At that point the conversation with your treating team typically widens to include surgical options. It is a judgment about one joint in one life, and the team that has examined you is best placed to make it with you.

References

• NHS: Osteoarthritis: Treatment
• NIH NIAMS: Osteoarthritis
• MedlinePlus: Osteoarthritis

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.