Spine Cracking — What Actually Happens

04 / the technique Anchored counter-rotation

This is a self-mobilization, not traction. Instead of pulling a joint straight apart (panels 1–3), you load the thoracic spine with rotational torque — and the facet joints separate along their angle of glide.

Anchor. One arm grips something fixed. This pins the scapula and shoulder girdle on that side — your stable axis.
Counter-swing. The free arm sweeps across the body in the opposite direction, driving rotation through the rib cage.
Separation. Because one side is fixed, the rotation forces a glide at the costovertebral and facet joints near the medial scapular border — the same negative-pressure cavitation fires here.
The stretch you feel between shoulder blade and spine is the rhomboids and mid-trapezius lengthening as the scapula wings outward — that's the muscle relief, separate from the joint pop.

05 / the surprise The pop isn't the point

listen: the pop fires on the top track — the H-reflex drops in silence below

Here's what catches people: the sound and the benefit are two separate events on two separate tracks.

The thrust sends a volley up the Ia afferents into the cord. That barrage briefly depresses the α-motoneuron pool — measurable as a drop in H-reflex amplitude (the electrical analogue of the stretch reflex). Excitability falls, the segment "quiets down," and that maps onto the relief.

The audible pop is just gas leaving solution in the joint. It rides the top track and contributes nothing to the bottom one.

Dishman & Bulbulian (2000): spinal manipulation transiently attenuates the H-reflex — a real but short-lived neuromotor effect.
Bialosky et al.: manual therapy works largely through neurophysiological, not structural, mechanisms.

Pre-manipulation

Post-thrust · H attenuated

And the part that really lands — the pop is optional

● pop heard

→

H-reflex drops · outcome improves

○ no pop

→

same drop · same outcome

Flynn et al. (2003): whether or not a cavitation was audible did not predict change in pain, disability, or range of motion. The thrust still loaded the joint and still fired the afferents — the sound was just along for the ride. The chasing of "the satisfying pop" is chasing a side-effect.

Sequence of events — ~0.1 seconds

T = 0 ms

Joint surfaces are pulled apart. Negative pressure drops inside the synovial capsule — dissolved gases (mostly CO₂, some N₂ and O₂) can no longer stay in solution.

T = 0–3 ms

Tribonucleation: a gas cavity nucleates between the joint surfaces. The bubble grows rapidly as dissolved gas rushes out of solution into the low-pressure void.

T = 3–7 ms

Bubble reaches peak size (~3–5 mm in facet joints). Pressure differential is maximum. The joint capsule is stretched to its limit.

T ~ 7 ms

Pop. The bubble collapses — or the capsule snaps back — generating a fast pressure wave. This is the audible crack. MRI has confirmed the gas cavity persists post-crack.

T = 7–30 min

Gas slowly re-dissolves. You cannot crack the same joint until this refractory period ends. Mechanoreceptors in the capsule fire, briefly flooding the CNS — the neurological basis for perceived relief.

long term

No evidence cracking causes arthritis. Habitual cracking may mildly stretch ligaments over years. The relief is real but temporary — the underlying tension that made you want to crack returns.

06 / the bottom line So… should you?

Mostly, it's harmless — just don't mistake the relief for a repair.

✓ what's true

For most people, the occasional gentle crack does no harm. The relief is genuine — but it's the neurological quieting from Panel 05, not a joint being "put back."

Nothing slips out of place and clicks back in. There's no realignment. The bubble forms, the mechanoreceptors fire, the muscle guarding eases for a while — then it returns.

✗ the oldest myth

That it causes arthritis. The cleanest disproof is one man's 60-year experiment: Dr. Donald Unger cracked only his left hand, twice a day, for six decades — keeping the right as a control.

cracks, one hand, ~60 years. Arthritis in neither hand. Published 1998; Ig Nobel Prize, 2009. Larger studies (DeWeber, 2011) since agree — no link to osteoarthritis.

When it's not just a harmless habit

If you keep needing to crack the same spot, the crack is treating a symptom — the underlying tension, posture, or movement load is what actually needs attention, and that responds to loading and mobility work, not the pop. And see a clinician, rather than self-managing, if pain radiates down a limb, comes with numbness, tingling, or weakness, follows an injury, wakes you at night, or arrives with fever.

One real caution — the neck. Forceful high-velocity twisting of the cervical spine is the single place where the risk stops being trivial: rare but documented arterial dissection. Be gentle there, and leave aggressive neck manipulation to someone trained — the relief isn't worth the tail risk.

The honest takeaway: it's a satisfying, mostly safe habit that buys real but temporary relief. If you find yourself chasing the pop, that's the signal to look at what's creating the tension — not to crack harder.

07 / sources The evidence

Everything above rests on published work, not folklore — including the parts that overturn what most people (and a few practitioners) still believe. Here's what's settled, what's still open, and where it all comes from.

● settled

The crack is gas-cavity formation, not bubble collapse — watched live on MRI.
Knuckle cracking shows no link to osteoarthritis, across decades and large samples.
The relief from manipulation is largely neurophysiological, not a structural correction.
The audible pop is incidental to the clinical effect — present or absent, the outcome holds.

○ still debated

How long the motoneuron-excitability drop truly lasts — seconds, or minutes?
Whether cavitation occurs at the targeted segment or simply wherever the joint gaps first.
The long-term effect of habitual, forceful self-manipulation of the spine.
Why relief is durable for some and fleeting for others.

references & further reading

[1]

Unsworth, Dowson & Wright (1971). "Cracking joints": a bioengineering study of cavitation in the metacarpophalangeal joint. Annals of the Rheumatic Diseases.mechanism The original cavitation model — which had the physics backwards, attributing the sound to bubble collapse.

[2]

Kawchuk, Fryer, Jaremko et al. (2015). Real-time visualization of joint cavitation. PLOS ONE.panels 01–02 The MRI study that settled the 60-year debate: the crack is the bubble forming, and the cavity persists afterward.

[3]

Ross, Bereznick & McGill (2004). Determining cavitation location during lumbar and thoracic spinal manipulation. Spine.specificity Found the pop frequently occurs at a different segment than the one targeted — often several at once.

[4]

Dishman & Bulbulian (2000). Spinal reflex attenuation associated with spinal manipulation. Spine.panel 05 Measured the transient H-reflex drop — the neuromotor fingerprint of a manipulation.

[5]

Bialosky, Bishop, Price et al. (2009). The mechanisms of manual therapy in the treatment of musculoskeletal pain: a comprehensive model. Manual Therapy.panel 05 The framework arguing manual therapy works through neurophysiological, partly central, mechanisms — not bones moving.

[6]

Flynn, Fritz, Wainner & Whitman (2003). The audible pop is not necessary for successful spinal high-velocity thrust manipulation in individuals with low back pain. Archives of Physical Medicine and Rehabilitation.panel 05 The decoupling result: the cavitation did not predict change in pain, disability, or range of motion.

[7]

Unger (1998). Does knuckle cracking lead to arthritis of the fingers? Arthritis & Rheumatism.panel 06 The 60-year, one-hand self-experiment. No arthritis in either hand. Awarded the Ig Nobel Prize in 2009.

[8]

DeWeber, Olszewski & Ortolano (2011). Knuckle cracking and hand osteoarthritis. Journal of the American Board of Family Medicine.panel 06 A larger sample reaching the same verdict: no association between cracking and hand OA.