Do Braces and Supports Actually Help?

A brace is the one thing in pain care you can buy on the way home from the appointment where nobody offered you anything. No prescription, no waiting list, thirty dollars at the pharmacy, and a photograph on the box of somebody your age walking a trail without a limp. It is also, by a wide margin, the pain device most likely to end up in a drawer within a month.

Both of those things are true for the same reason. A brace does something real and quite narrow, and almost nothing sold in a pharmacy explains what that narrow thing is. Fitted to the right joint for the right reason it is one of the better-evidenced items in this whole field. Bought off a shelf on a guess, it is a strap that makes your skin itch.

What a brace is actually doing

Every brace on the market is some combination of four mechanisms, and knowing which one you are buying is most of the decision.

It moves load off a painful part of a joint. This is the unloader or valgus knee brace, and it is the only mechanism that is genuinely mechanical. If arthritis has worn the inner compartment of your knee, a rigid hinged frame can lever the joint a few degrees outward, shifting force onto the side that still has cartilage on it. It is the same idea as a high tibial osteotomy, done with straps instead of a saw.

It stops a movement that hurts. A wrist splint, a thumb spica, a rigid ankle brace. The joint is held somewhere it cannot get into the position that provokes the symptom — most usefully while you are asleep and have no say in the matter.

It applies compression and warmth. The elastic sleeve. There is no structural support in a tube of neoprene, whatever the packaging says. What there is: a bit of warmth, a bit of swelling control, and a constant, even sensation across the skin over the joint. That last one is not nothing — it is the same gate-control mechanism a TENS machine uses, delivered by fabric instead of a battery.

It changes what you do. A brace is a thing on your body that you can feel all day. People move differently in one — sometimes more, because they trust the joint; sometimes less, because they have been told to be careful. Nobody sells a brace on this mechanism, and for several joints it is probably the largest of the four.

Where the evidence is strong, and where it isn’t

The honest picture is joint by joint. There is no general finding about bracing, and anyone offering you one is selling something.

The base of the thumb

This is where guidelines are most enthusiastic. The 2019 American College of Rheumatology and Arthritis Foundation guideline strongly recommends an orthosis for osteoarthritis at the first carpometacarpal joint — the joint at the base of the thumb where the pinch grip lives, and one of the most reliably miserable places in the body to have arthritis. Strong recommendations are rare in that document. Most of it is conditional.

Then the best trial in the field went the other way. OTTER II randomised 349 people across 17 hospital departments into three groups: a therapist-supported self-management programme on its own, that programme plus a real thumb splint, or that programme plus a placebo splint. At eight weeks, all three groups had improved and none had improved more than another. The real splint beat self-management alone by half a point on a twenty-point hand pain scale, a difference that was neither statistically significant nor large enough to feel, and it did not beat the placebo splint either.

That result is narrower than it sounds, and the comparator is why. Everyone in the trial got 90 minutes of hand therapy from a therapist over eight weeks — joint protection, grip strategies, a proper exercise programme. Against that, a splint added nothing. Against nothing at all, which is what most people in Canada are actually choosing between, it is a different comparison, and one that trial did not run. The practical reading is not “thumb splints don’t work”; it is that if you can get in front of a hand therapist, do that first, because the splint is the part of the package with the least evidence behind it.

The wrist, at night

Carpal tunnel syndrome is the classic splinting indication, and the 2023 Cochrane review of splinting for it pooled 29 trials and 1,937 adults. Short-term symptom relief against no treatment came out at 0.37 points on a scale running 1 to 5, against the full point the reviewers had set in advance as the smallest difference a person would actually notice — and even that came at low certainty. Longer-term results were too inconsistent to conclude anything at all.

But sitting inside that unimpressive average is a more useful number: people given a splint to wear at night were nearly four times as likely to report themselves overall improved. The certainty on that is also low, and it comes from a single study. Still, it points at the same thing hand clinics have said for years — the splint is for the small hours, when your wrist curls up under the pillow and you wake with a dead hand.

The angle matters more than anything else about the splint, and it is where most drugstore purchases go wrong. Pressure inside the carpal tunnel is at its lowest with the wrist held straight — neutral, or within a few degrees of it. Bend it either way and the pressure on the median nerve climbs. Plenty of cheap wrist supports are built around a metal stay that holds the wrist in noticeable extension, which is the position a bench presser wants and the exact opposite of what a compressed nerve needs. Put it on and look at your forearm from the side: your hand should carry straight on from it, not tilt back.

Splinting is cheap and its recorded harms are trivial — some discomfort, some swelling, nothing serious across the whole review. When a treatment is that safe and that inexpensive, a small and uncertain benefit is still worth four weeks of your time before anyone talks about surgery.

The knee

The knee is where the disagreement is sharpest. That same ACR guideline strongly recommends a tibiofemoral brace for people whose knee arthritis is affecting their walking or joint stability enough to justify the bother. The Cochrane review of braces and orthoses for knee osteoarthritis, covering 13 studies and 1,356 people, found that at twelve months the braced group’s pain was, to two decimal places, identical to the unbraced group’s — and rated the evidence low quality.

Individual trials look better. The ROTOR trial gave 67 people with medial knee arthritis an unloading brace or usual care for six weeks, and the brace group’s 24-hour pain fell 26mm further on a 100mm scale, which is a large difference by any standard. It also had no sham brace, so everyone knew which group they were in — the same problem that dogs every trial of a device you can see and feel.

The number that best predicts how this goes for you is not from an efficacy trial at all. In a follow-up study of people fitted with unloader braces, roughly a quarter were still wearing theirs regularly two years later. The reasons the rest gave are worth reading before you spend anything: it didn’t help enough, it was uncomfortable, it didn’t fit, it rubbed, it was too bulky to wear under trousers. Only the first of those is about whether bracing works.

One clear negative from the same Cochrane review, since these get sold hard: laterally wedged insoles, the shoe inserts marketed for knee arthritis, showed no benefit over plain neutral insoles, and that finding was rated moderate quality — which is as close to settled as this literature gets. Save the money.

The lower back

Elastic back belts and lumbar corsets are the most-sold and least-supported item in the category. The evidence does not show them preventing back pain, and it does not show them treating it. Britain’s NICE guideline on low back pain and sciatica tells clinicians not to offer belts or corsets, and its chronic pain guideline takes the same line. If a lumbar support helps you get through a shift, it is not doing you harm — but do not expect it to change anything, and do not let it stand in for the loading and walking that does.

A rigid brace prescribed after spinal surgery or for a vertebral fracture is a completely different object with a completely different job. Nothing here applies to one of those. Wear it exactly as the surgeon said.

The neck

Soft collars after whiplash have been studied repeatedly and the answer has not moved: acting as usual — keeping the neck moving, within what you can tolerate — does better than immobilising it. A collar for a few days after a significant injury is one thing. A collar as a management strategy for ongoing neck pain is a way of getting stiffer.

How a guideline can say “strongly recommend” about weak evidence

The knee contradiction above looks like one of them must be wrong. Neither is.

The Cochrane reviewers were asked one question: how confident are we that the effect is real and this large? Small trials, no blinding, high dropout — the answer is not very, and they said so.

The guideline panel was asked something else: given what we know and don’t know, what should a clinician do? A knee brace costs a few hundred dollars, harms almost nobody, and can be abandoned on a Tuesday if it isn’t helping. When the downside of being wrong is that small, a panel can reasonably recommend trying something the evidence has not nailed down. Nobody would do that for a drug with a side-effect profile.

Underneath both sits the problem that keeps this literature stuck: you cannot blind someone to wearing a brace. There is no convincing placebo strap. Every trial in the field is contaminated by participants knowing what they got, which is why the certainty ratings stay low no matter how many studies get added — and it is why OTTER II, which actually built a placebo splint, was such an unusual piece of work.

Will it weaken the muscle underneath?

This is the first question most people ask, and the honest answer is that the evidence is thinner than the confidence on either side of it.

The worry traces back to functional bracing after knee ligament reconstruction in athletes, where some studies found reduced muscle output. Generalising that to a 70-year-old wearing a support for two hours at the shops is a stretch. When it has been looked at directly in knee arthritis, more brace-wearing hours did not translate into weaker legs — it was associated with more physical activity, which is the opposite of the feared direction.

Notice also who publishes the reassurance. Search this question and you will get page after page from brace manufacturers, all of them confidently debunking it. That is not evidence either. The fair summary is that nobody has shown ordinary brace use wasting a muscle, and nobody has run the study that would settle it.

The real risk is not atrophy. It is substitution — the brace working well enough that the strengthening never happens, and a year later the joint is no stronger and the brace is no longer optional. Braces belong alongside loading the joint, not instead of it. If a brace is what lets you get through a walk you would otherwise skip, it has earned its place. If it is what lets you skip the exercises, it hasn’t.

Fit is most of the result

Nearly every reason people give for abandoning a brace is a fitting problem, not a bracing problem. A few things are worth getting right:

• Measure, don’t guess the size. Small/medium/large on a knee brace refers to a thigh circumference measured a set distance above the kneecap, and the number is printed on the packet. Take a tape measure to the pharmacy. A brace one size too big migrates down your leg all day and one size too small cuts off the back of your knee.
• Line the hinge up with the joint, not the middle of the brace. On a hinged knee brace the pivot goes level with the top of the kneecap, roughly. Get that wrong and the brace fights your knee through every step, which is exactly what “it rubbed” means.
• Straps go on in the order printed on them. They are numbered for a reason — the sequence is what positions the frame before it gets tightened. Doing them up top to bottom because that is the order they fall to hand is the single commonest fitting error.
• Snug, not tight. Two fingers should slide under a strap. Deep grooves in the skin, pins and needles, a cold or dusky foot or hand: too tight, every time.
• Look at the skin every time it comes off. Redness that fades in twenty minutes is fine. Redness that is still there an hour later, or any broken skin, means it stops until it is refitted. Wash the sleeve — most of the itching people blame on neoprene allergy is sweat and detergent residue.
• Wear it for something, not for the day. The joints with the best evidence are braced for a purpose: a night’s sleep, a walk, an afternoon of gardening. All-day wearing is where the fit problems and the substitution problem both come from.

Buying one in Canada

There are two very different products under the same word, and the price gap tells you which is which. An off-the-shelf sleeve, splint or soft support runs roughly $30 to $150 and you can walk out with it. A custom-fabricated brace — cast or scanned to your leg, built to your alignment — runs from several hundred dollars into the low thousands, is made by a certified orthotist, and takes a few appointments.

Most people should try the cheap one first. The exception is the unloader knee brace, where the whole mechanism depends on the frame sitting in a specific relationship to your joint. An off-the-shelf unloader on a leg it does not match is not a weaker version of the treatment; it is a hinge in the wrong place.

On paying for it: no provincial health plan covers braces the way it covers a doctor’s visit, and the funding routes are provincial, so what follows is Ontario’s and the details differ elsewhere.

Ontario’s Assistive Devices Program pays up to 75% of an approved price for orthotic devices, and 100% for people receiving Ontario Works, ODSP or Assistance to Children with Severe Disabilities. You need a valid Ontario health card, a physical disability expected to last six months or more, and a need that is not solely for sport, work or school. That six-month rule is the important one: the program is built around long-term devices, so a support you want for a bad fortnight is not what it is for. The paperwork goes through an authoriser and a registered vendor, so ask the clinic fitting you to start it — do not buy the brace first and apply afterwards.

If ADP will not fund the item, ODSP recipients have a second route: braces, casts and splints sit in the Mandatory Special Necessities benefit, which covers things needed to prevent surgery, to recover from surgery or injury, or to restore function lost to a neuro-muscular-skeletal condition — but only where no other source, public or private, will pay.

Private extended health plans very often cover bracing, usually with a prescription and often with an annual cap that is easy to exceed if you also claim orthotics or physiotherapy against it. Check the cap and the wording before the fitting appointment rather than after. Veterans Affairs, provincial workers’ compensation boards and motor vehicle insurers all fund bracing under their own rules if your injury falls under one of them.

Whoever is paying, get fitted by someone whose job it is — a certified orthotist, a physiotherapist or an occupational therapist. Arthritis Society Canada can point you at services in your province, and if the cost is the obstacle, say so out loud at the appointment: clinics know the funding routes and will not offer them to someone who doesn’t ask.

Giving one a fair test

Four weeks is enough to know, and the test is more informative if you decide the terms in advance.

Pick the activity it is for — the walk, the night’s sleep, the two hours at the stove — and wear it for that, not from breakfast to bed. Rate the pain during that activity before you start bracing and again each week; a memory of how bad last month was is not a measurement. Give a night splint the full four weeks before judging it, because the wrist evidence that exists is about accumulated nights rather than immediate relief. And take it off between times, so that at the end you can still tell what the joint does without it.

If it isn’t helping by week four, the fit is worth checking once before you conclude the brace is useless — but if a refitted brace still isn’t helping by week six, believe it. Persisting with a brace that is doing nothing is how a joint ends up braced for a decade by nobody’s actual decision.

A brace is a tool for doing something, not a treatment you wear. The ones that work are the ones with a job and a time to come off.

So, do they help?

For the base of the thumb and the wrist at night, yes — modestly, cheaply, and with almost nothing to lose by trying. For knee arthritis, an unloader brace properly fitted to the right knee helps some people a great deal and most people not enough to keep wearing it, and there is no way to find out which you are except to try. For the lower back and the neck, the evidence says no, and a belt is most useful as a reminder that you were going to do the walking.

What braces do not do, anywhere, is treat the thing causing the pain. Nothing about a strap regrows cartilage, decompresses a nerve or reverses arthritis. They make a specific hour more bearable, which is worth having — the same way heat and TENS are worth having, and for the same limited reasons.

The best result usually comes from the least interesting combination: the right brace for the specific activity, and the loading and movement that make the joint able to do more of it. If you are still assembling that side of it, the morning mobility routine and the walking build-up are free and printable, managing arthritis pain covers the wider ground for a worn joint, and Pain BC runs free programs across the country for people building the rest of the plan.