Cracking Joints: Is Knuckle Cracking and Knee Popping Harmful?

'Joint cracking' hides at least two different phenomena. The first is a single loud 'crack' that can be produced on demand, typically in the finger joints after a pull or a bend, and that cannot be repeated for roughly a quarter of an hour or more. The second is a repeatable creaking, grinding or scraping during movement, most often in the knee, shoulder or neck, which medicine calls crepitus.

For decades it was assumed that a finger crack arises when a gas bubble bursts in the joint fluid. That interpretation was challenged by real-time imaging. The authors of a study published in PLoS One pulled a participant's finger inside an MRI scanner and recorded the joints at the moment of the crack.

The practical message is that the mechanism is physical and involves a small gas cavity forming in the joint fluid, not 'bone hitting bone'. It is worth remembering the scale of the evidence, though: this is an experiment on ten joints with no long-term follow-up, so it describes a mechanism, not the safety of a decades-long habit. That is why we move on to studies that asked directly about arthritis.

Knee creaking when squatting or climbing stairs is usually a different mechanism: friction or snapping of soft tissues (tendons, retinacula, synovial folds) over bony prominences, or minor irregularities of the joint surfaces. The source of the sound is often hard to pinpoint, which partly explains why research on crepitus is so heterogeneous.

This study has clear strengths: it compares radiographically confirmed osteoarthritis with controls and asks about total exposure, not only whether someone cracks their knuckles at all. It also has the typical limits of a retrospective design: it relies on participants' recollections of a habit over many years and includes a fairly small group. A similar conclusion comes from a short clinical review in the Journal of Family Practice, according to which habitual knuckle popping over decades was not associated with clinical or radiographic signs of osteoarthritis.

Not every result fits one reassuring story, and that is worth knowing. An older short report in the Annals of the Rheumatic Diseases evaluated 300 patients aged at least 45, of whom 74 habitually cracked their knuckles. Hand arthritis was not more common, but knuckle crackers more often had hand swelling and lower grip strength, and the habit correlated with manual labour, nail biting, smoking and alcohol. The authors concluded that the habit is linked to functional impairment of the hand, but it was soon pointed out that this is a cluster of accompanying factors (manual work, other habits) rather than proof of causation.

A small Turkish study of 35 young people (aged 19–27) who cracked their knuckles at least 5 times a day and 35 matched controls found no difference in grip strength, and ultrasound-measured cartilage of the metacarpal head was, if anything, slightly thicker in the crackers. It is a cross-sectional study, so it cannot say what is cause and what is effect, and it says nothing about many years of the habit. Taken together the picture is fairly consistent: there is no convincing evidence that knuckle cracking causes arthritis, but also no studies that would allow us to say the habit is perfectly harmless for everyone.

Knee creaking worries people in the clinic more than knuckle cracking does, because the knee is a weight-bearing joint and people associate the sound with 'cartilage wearing away'. Yet a recent meta-analysis in the British Journal of Sports Medicine shows that crepitus is an almost universal phenomenon in the population.

Two things are worth remembering from this meta-analysis at the same time. First, knee crepitus is very common even in completely healthy people without pain, so the sound alone is not a good reason to panic. Second, it is clearly more frequent in osteoarthritis, and the association with a radiographic diagnosis is more than threefold (OR 3.79). The two facts do not contradict each other: the sound is 'non-specific', meaning it does not separate a healthy knee from a diseased one well, but it carries a small amount of statistical information.

In addition, the large variation between results (heterogeneity above 90%) and the very low certainty of evidence mean that the quoted percentages are more of an order of magnitude than an exact figure. Differences in how researchers defined and measured crepitus (self-report, auscultation, devices) make comparing studies difficult.

Interpreting these results needs care in two directions. On the one hand, in people around 60, frequent crepitus really was a signal of higher risk of developing symptomatic osteoarthritis. On the other, the effect was concentrated in the group that already had radiographic changes, so the sound acted more like a marker of existing, still 'silent' disease than a factor that causes it. This is an observational study, so it does not prove that crepitus 'causes' osteoarthritis, and the authors also note that the patellofemoral joint was not systematically assessed in the cohort.

For younger people with painless, perpetually 'popping' knees the conclusion is reassuring but does not remove the need for attention: the results concern middle-aged and older adults, so they cannot be mechanically transferred to twenty-year-olds. There are also no studies showing that deliberately 'treating' the sound alone, with exercises, supplements or procedures, changes the course of disease.

The data above concern mainly finger and knee joints, because those are best studied. In other places, such as the shoulder, hip, elbow, jaw or neck, the noises have similar mechanisms (cavitation, tendon snapping, crepitus), but we lack equally good data on their long-term significance. That is a gap in the literature that deserves a candid mention.

A typical example is the 'snapping hip', where the iliotibial band or the iliopsoas tendon snaps over a bony prominence. It is usually painless and needs no treatment, but when pain is present it can be part of a bigger picture of pain on the outer side of the hip; we write about it in a separate article on bursitis and lateral hip pain. Likewise, clicking in the temporomandibular joint should be assessed together with jaw pain, limited mouth opening and bruxism, which also have their own article.

As for the neck and spine, self-induced 'cracking' is for some people a soothing habit. The studies above do not cover it, and sudden, forced rotational movements of the neck are not neutral: rare but serious vascular complications after neck manipulation have been described in the literature, although we cannot reliably estimate them for self-stretching. It is sensible to avoid abrupt, forceful twists.

A secondary myth has grown around this one: that you must 'lubricate' joints with supplements because they creak. None of the studies cited here tested such an approach. If someone reaches for collagen or hyaluronic acid because of knee pain, it is worth checking what the research on pain actually shows rather than treating creaking as an indication.

A sensible approach to cracking joints comes down to distinguishing between a noise and a symptom. A sound that has been with you for years, does not hurt and does not limit movement is, in the light of current studies, usually a normal, very common phenomenon that needs no intervention. A symptom, such as pain, swelling, locking or instability, deserves a separate assessment, whether or not the joint makes any noise at all.