Hypnosis for Chronic Pain: What the Evidence Supports

Chronic pain is the condition with the most hypnosis research, and it is where the field’s strengths and weaknesses show most clearly. There is enough evidence to take hypnosis seriously as a pain intervention, but not enough to know reliably who it will help, how much and for how long.

What the laboratory evidence shows

The cleanest data come from experimental pain studies in healthy volunteers, where conditions can be tightly controlled. A 2019 meta-analysis of 85 such studies, involving 3,632 participants, found that hypnosis with direct pain-relief suggestions produced clinically meaningful reductions in pain: about 42% in highly hypnotizable people, 29% in medium scorers and minimal change in low scorers [1].

These are useful numbers, but they describe brief, controlled pain in healthy people. Chronic pain involves sensitization, disability, mood and meaning, none of which a cold-pressor test captures.

What the clinical evidence shows

Fewer trials have tested hypnosis in people living with chronic pain, and they are generally smaller and more varied.

• Against standard care. A meta-analysis of 12 clinical trials found moderate effects of hypnosis for chronic pain compared with standard care or no treatment [2]. The authors noted that most trials were small and methodologically limited.
• Against other active treatments. A three-arm trial randomized 328 veterans with chronic pain to eight group sessions of hypnosis, mindfulness meditation or pain education. At the end of treatment there was no significant difference between groups in average pain intensity. At three and six months, hypnosis showed greater reductions in pain intensity and pain interference than the education group, but did not differ from mindfulness at any time point [4].
• Prediction intervals. When a surgical hypnosis meta-analysis reported pooled effects, the 95% prediction intervals included no effect for every outcome except procedure time [6]. That means the average was positive but a given program might not help. The same caution likely applies to chronic pain.

What happens in the brain

Brain imaging studies show that hypnotic suggestions targeting pain can alter activity in regions involved in pain processing. When participants are given suggestions to reduce the unpleasantness of pain, activity in the anterior cingulate cortex changes; suggestions targeting pain intensity affect somatosensory cortex activity [5]. These findings support the idea that hypnotic pain relief is not simply distraction or relaxation, though the studies are small and the relevance of lab imaging to clinical practice is indirect.

A narrative review concluded that hypnosis likely works through multiple pathways and that no single brain mechanism accounts for it [3].

The suggestion matters

Not all hypnotic interventions are the same. Research distinguishes between direct suggestions for pain reduction, suggestions for changes in the experience of pain (such as reframing or dissociation) and general relaxation. The laboratory meta-analysis found that direct analgesic suggestions produced the largest effects [1]. Whether this holds in clinical practice, where sessions are longer and more varied, is less clear.

How much hypnotizability matters

In the laboratory, hypnotizability strongly predicts pain relief. In clinical settings, the picture is different. A meta-analysis of 10 clinical studies found that hypnotizability accounted for about 6% of the variance in outcomes, a small contribution [2]. For more on this, see What Is Hypnotizability?.

The implication is that a low score on a suggestibility scale does not rule out benefit, but the odds of a strong response are lower.

How long effects last

Most trials measure outcomes at the end of treatment and sometimes at a follow-up a few months later. The veterans trial found that differences between hypnosis and education were larger at six months than at the end of treatment [4], which is encouraging. But long-term data beyond six to twelve months are scarce, and many trials do not report whether participants practiced self-hypnosis between sessions.

Limitations worth naming

Limitation	Why it matters
Most trials are small	Small trials tend to report larger effects
Blinding is usually impossible	Self-reported pain is vulnerable to expectation bias
Comparison groups vary	Beating a wait-list is less informative than beating a sham
Follow-up is often short	Chronic pain is, by definition, long-term
Types of pain are pooled	Low back pain, fibromyalgia and headache may respond differently

What this means in practice

Hypnosis for chronic pain has a plausible evidence base, particularly as an adjunct to established treatment. It is not a first-line therapy, and a single positive trial is a reason to look closer, not a proof of efficacy. The best evidence is for people with moderate to high hypnotizability receiving direct analgesic suggestions, but that describes a subset of those who might try it.

For how to assess any individual hypnosis trial, see What Is a Randomized Controlled Trial?. For why the comparison group shapes the conclusion, see Why Comparison Groups Matter.

When to see a clinician first. Chronic pain should be assessed by a doctor before adding any complementary approach. See a clinician promptly about new pain, pain with unexplained weight loss, numbness, weakness or bowel or bladder changes. Use hypnosis alongside medical care, not instead of it.