Does Cracking Your Knuckles Cause Arthritis?

That inversion is the interesting part. The scary claim turns out to be well refuted; the boring mechanical question turns out to be open.

The sixty-year experiment

Donald Unger was told repeatedly as a child that cracking his knuckles would give him arthritis. He decided to test it, and ran a self-experiment for more than half a century.

He cracked the knuckles of his left hand at least twice a day and deliberately never cracked his right. Over the period he estimated the left hand had been cracked at least thirty-six thousand times.

He examined both hands and found no arthritis in either, and no difference between them. The result was published in Arthritis and Rheumatism in 1998 and earned him an Ig Nobel Prize.

A sample of one is not evidence in any ordinary sense, and Unger said so himself. What it did was force the question into the literature, where it has since been examined properly.

What the larger studies found

Subsequent work has looked at habitual crackers against non-crackers in much bigger samples, including a study published in the Journal of the American Board of Family Medicine examining knuckle cracking and hand osteoarthritis.

The consistent finding is that habitual knuckle cracking is not associated with increased osteoarthritis of the hand. Whatever the sound is doing, it is not degrading the joint in the way the folk warning claims.

There is a caveat worth keeping. Some older work reported associations with hand swelling and reduced grip strength in habitual crackers, and those findings have not been as thoroughly examined as the arthritis question. The arthritis claim is refuted; the broader claim that cracking is entirely without consequence is less well established than the internet suggests.

It is a good example of how a claim survives on repetition rather than evidence — the same mechanism that makes something you just learned about appear to be everywhere. Hearing it often feels like corroboration.

What actually makes the sound

This is where the article stops debunking and starts admitting uncertainty.

Your joints are wrapped in a capsule containing synovial fluid, which lubricates the surfaces and carries dissolved gases, mostly nitrogen. Pulling or bending the joint separates the surfaces and drops the pressure inside sharply — one study measured it falling to around minus three and a half atmospheres immediately before the crack.

At that pressure a gas cavity forms in the fluid. Everyone agrees on that much. The disagreement is about which moment produces the noise.

The 2015 study, and why it did not settle it

For over forty years the accepted answer, from Unsworth and colleagues in 1971, was that the sound comes from a cavity collapsing.

In 2015 Gregory Kawchuk and colleagues used real-time MRI to film ten knuckles being pulled until they cracked, capturing about three frames per second. They reported rapid cavity formation at the moment of the sound, with the cavity remaining visible afterwards — which suggested the noise comes from the cavity appearing, not collapsing. The paper described this as tribonucleation, and it was widely reported as having solved the question.

It had not. In 2018 a mathematical modelling study in Scientific Reports resolved the acoustics of a collapsing cavitation bubble and found the predicted waveform and dominant frequency consistent with recorded knuckle cracks — lending support to the older collapse account.

So the honest position is that the sound is associated with cavitation, that formation and collapse are both live candidates, and that the imaging and the acoustic modelling currently point different ways. Any article telling you this was settled in 2015 has read the press release rather than the follow-up. It sits alongside hiccups as a phenomenon everyone experiences and nobody has fully explained.

Other joints, same question

Knuckles are the studied case because they are easy to isolate, but the same cavitation occurs in the neck, back, jaw, hips and toes. Not every joint noise is cavitation, though, and the distinction is worth knowing.

A crack that can only be repeated after a delay is almost certainly cavitation. A sound that repeats immediately and every time — a knee on stairs, a shoulder through a particular arc — is usually a tendon or ligament sliding over a bony prominence, which is a different phenomenon entirely and involves no gas at all.

That refractory period is the cleanest diagnostic available to you. If you can do it again straight away, no cavity formed.

Why you cannot do it twice in a row

One detail is not disputed and is worth knowing, because it is the easiest thing to verify yourself.

After a joint cracks, it takes roughly twenty minutes before it will crack again. The gas that came out of solution has to redissolve into the synovial fluid before the conditions can be recreated.

The joint also physically separates slightly during the crack and stays that way briefly, which is part of why some people describe a feeling of release afterwards. That sensation is real; whether it constitutes any benefit is not established.

So should you stop?

This article is not going to tell you either way, and it is worth being clear why. The arthritis claim — the reason most people worry about it — has no support. That is a factual finding and it stands on its own.

Beyond that, the evidence thins quickly. If cracking is painful, or if a joint is swollen or restricted, that is a matter for a qualified clinician rather than for a trivia article about gas bubbles.

What the research does supply is a rare clean answer to a question almost everyone has been told the wrong answer to, from someone who spent six decades finding out.