Breathwork

Breath regulation in premodern yoga texts

Regulation of the breath appears in several premodern South Asian yoga traditions. A review of Mallinson and Singleton's 2017 sourcebook Roots of Yoga notes that the book gathers passages from more than a hundred premodern works and traces how practices such as breath control recur across different systems. A neurophysiological review reports that the Yoga Sutras of Patanjali treat breath practice as inhalation, exhalation, retention and prana, and that the Hatha Yoga Pradipika, attributed to Svatmarama, describes eight kumbhaka (breath-retention) techniques.

A scholar of yoga places the composition of the Haṭhapradīpikā in the fifteenth century, around its middle in one account and early in the century in another; its oldest surviving manuscript is dated 1496. The earliest known textual counterpart of hatha yoga is an eleventh-century Buddhist work, the Amṛtasiddhi. Medieval yoga texts disagreed about breath control: the thirteenth-century Dattātreyayogaśāstra teaches retention at length, while the twelfth-century Amanaska warns that many breath retentions can cause disease. The sources reviewed do not date the Yoga Sutras.

Other traditions and general statements

One review of high-ventilation practices states that intense breathing was probably used for spiritual and healing purposes in many cultures since early times. It also describes Tibetan Buddhist g-tummo ('inner fire'), a practice whose teaching is restricted to advanced practitioners. These are general statements cited second-hand, and they are not evidence of an unbroken line from those traditions to modern breathwork.

Twentieth-century therapeutic and commercial styles

The same review lists several named styles: Holotropic Breathwork, developed by Stanislav and Cristina Grof; Rebirthing, developed by Leonard Orr and popularised by Sondra Ray; Sudarshan Kriya Yoga, associated with Sri Sri Ravi Shankar; and the Wim Hof Method, developed from the 1990s or 2000s.

Sources differ on dating Holotropic Breathwork. An obituary of Christina Grof places it in the late 1970s, when she was resident at Esalen, while Esalen describes it as born there in the 1980s and the Grofs' training organisation dates facilitator training to the 1980s and a modular programme to 1989. Orr's own account dates his first rebirthing experience to 1962 and the connected-breathing method to 1974 and 1975.

The review adds that it has been suggested that the psychoanalyst Wilhelm Reich made the first prominent Western attempt to use hyperventilation therapeutically. Holotropic Breathwork now trades under the name Grof Breathwork, and Sudarshan Kriya and the Wim Hof Method are trademarked, which the review notes may limit access to them.

An umbrella term

A 2023 review quotes the Oxford English Dictionary as defining breathwork broadly as exercises, techniques and therapies that involve manipulating how one breathes. The label therefore covers many different practices, and naming the practice alone says little about what a person actually does. A 2023 meta-analysis counted a trial as breathwork only if breathwork made up at least half of the intervention.

Slow-paced and high-ventilation styles

One systematic review defined slow breathing as fewer than ten breaths per minute, with most protocols tested at about four to six. Fincham and colleagues use 'high ventilation breathwork' for practices that raise the rate, depth or both above the normal range. This is one research group's label, not a universal standard. In their meta-analysis, ten of twelve stress trials used slow-paced breathing and two used fast-paced breathing.

The high-ventilation review groups Rebirthing, Holotropic Breathwork and Conscious Connected Breathing as continuous fast and deep styles. It places Wim Hof-type breathing and the pranayama techniques kapalabhati ('skull shining') and bhastrika ('bellows') among styles that include pauses or breath holds.

Pranayama and programme names

Pranayama is usually translated as regulation or control of the breath or vital energy (prana plus ayama), although the etymology here is given second-hand and should be confirmed with an Indological source. Sudarshan Kriya Yoga is a programme of five techniques followed by meditation and rest, so it is not interchangeable with generic breathwork.

Slow-paced breathing exercises

The US National Center for Complementary and Integrative Health (NCCIH) describes breathing exercises as focusing on slow, deep breaths, also called diaphragmatic breathing, and lists them among relaxation techniques. In a US trial of military veterans, Sudarshan Kriya Yoga was delivered to groups as a 15-hour workshop followed by two one-hour sessions a week for five weeks.

Intense and group sessions

Modern breathwork is often, but not always, led by a trained facilitator and accompanied by evocative or ambient music. A Holotropic or Grof Breathwork group session is described as about three hours of very fast, deep breathing with no set pace, with optional bodywork, a sitter, mandala drawing and discussion afterwards; one-to-one sessions can last about an hour. A Wim Hof-style round is described as three or four sets of 30 to 40 deep breaths with breath holds, taking ten to fifteen minutes or more. These descriptions come from a table in a review, not from teaching manuals, and formats vary between schools and facilitators.

What participants may experience

Intense styles can produce marked effects during and after a session, including light-headedness, tingling, palpitations, sweating, cramping of the hands and feet (tetany), crying, laughing and other emotional release, and changes in perception. The review says facilitators should explain risks and contraindications before a person consents. A person can ask what the session involves and may choose to slow down or stop.

Uses that have been studied or proposed

Research on breathwork has concentrated on self-reported stress, anxiety and depressive symptoms. Pranayama has also been studied for bronchial asthma, symptoms of chronic obstructive pulmonary disease, cardiorespiratory function and quality of life in people with cancer; one review located eighteen small controlled trials.

Proponents of Sudarshan Kriya Yoga have proposed it for stress, anxiety, post-traumatic stress, depression and substance misuse. In 2005 they described it as a beneficial, low-risk adjunct, which was the authors' own assessment of early evidence and not a trial result. A 2007 hypothesis paper proposed Holotropic Breathwork as a possible adjunct to psychotherapy for anxiety and depressive disorders and called for better-designed research.

Popular claims without support in the sources reviewed

Popular claims that breathwork releases stored trauma, resets the nervous system or detoxifies the body were not supported by any source reviewed for this draft. They should not be repeated unless a source supports a specific claim in a specific population. The sources reviewed also do not document why people seek breathwork in everyday life, so this section describes research targets and proponents' claims only.

Meta-analysis of randomised trials

A 2023 meta-analysis of twelve randomised trials (785 adults), searched to February 2022, found lower self-reported stress with breathwork than with non-breathwork controls. The pooled effect was small to medium (Hedges g -0.35, 95% confidence interval -0.55 to -0.14). Pooled effects for anxiety (20 trials, g -0.32) and depressive symptoms (18 trials, g -0.40) were similar in size.

Three of the twelve stress trials were at high risk of bias and nine had some concerns. Outcomes were self-reported, too few trials reported follow-up to judge whether effects last, and only two tested fast-paced breathing. The authors called for low-risk-of-bias studies.

Two later blinded randomised trials by the same group in non-clinical adults (400 people using coherent breathing; 200 using fast breathing with long breath holds) found no greater improvement in stress, anxiety or depression than a placebo breathing pattern, so the pooled result is not consistently reproduced.

Slow breathing and pranayama

A systematic review of slow breathing in healthy adults included fifteen small studies, none randomised, and reported higher heart rate variability and more EEG alpha power, but results could not be pooled. A systematic review of pranayama found eighteen controlled trials with 16 to 160 participants; quality was rated poor in six, fair in seven and good in five, and heterogeneity prevented meta-analysis. NCCIH reports preliminary evidence that diaphragmatic breathing reduces stress.

A Cochrane review of 22 trials in 2,880 adults with mild to moderate asthma found that breathing exercises may improve quality of life at three months (mean difference 0.42, 95% confidence interval 0.17 to 0.68, moderate certainty), with inconclusive asthma control and lung function results; it did not assess adverse effects. A 2025 meta-analysis of six trials in 517 people with mental disorders (mainly post-traumatic stress and depression) found a short-term reduction in symptom severity with pranayama against passive controls (standardised mean difference -0.27, 95% CI -0.52 to -0.03), with high risk of bias in four trials.

Specific programmes and intense styles

In one US trial of 85 military veterans with post-traumatic stress symptoms, Sudarshan Kriya Yoga was non-inferior to cognitive processing therapy on PTSD checklist scores at six weeks, one month and one year. On depression and affect measures it was non-inferior at six weeks and one month but inferior at one year. The trial had an active comparator only and no untreated group.

For Holotropic Breathwork, one Danish uncontrolled study of 20 participants reported small improvements on some temperament and interpersonal scores alongside an increase in paranoid ideation. A search of PubMed titles and abstracts found no randomised trial of Holotropic Breathwork or Rebirthing; trial registries were not searched. A 2014 randomised trial of 24 healthy men combined cyclic hyperventilation with breath retention, meditation and cold exposure and measured immune markers, not disease outcomes.

Reported effects and adverse-event reporting

The 2023 meta-analysis found no adverse events or lasting harm attributed to breathwork, but only four of the twelve stress trials actively reported on adverse events. Not reporting harms is not evidence that none occur. In one fast-paced trial in depression, a small subgroup reported hot flushes, shortness of breath or sweating. Experimental hyperventilation can cause nausea, tetany, palpitations, light-headedness and anxiety, and it can provoke panic in vulnerable people. In the 2025 pranayama meta-analysis, adverse events were more frequent with fast than slow techniques and were insufficiently reported.

Precautions for intense styles

The high-ventilation review names epilepsy, panic disorder, psychotic disorders and pregnancy as conditions needing specific contraindication. It advises caution for people with cardiovascular, cerebrovascular or respiratory conditions and recommends physician screening where warranted. These cautions are expert inference from physiology, and the sources reviewed give no incidence data for serious events in breathwork sessions. NCCIH reports rare cases in which relaxation techniques caused or worsened symptoms in people with epilepsy, certain psychiatric conditions or a history of trauma.

A review of yoga case reports includes pneumothorax linked to kapalabhati and pneumomediastinum linked to forceful straining. Its authors state that case reports cannot estimate how often such events occur, and they advise beginners to avoid forceful breathing.

Breath holding in or near water

Hyperventilation followed by breath holding carries a specific hazard in or near water. The Royal Life Saving Society UK defines shallow water blackout as loss of consciousness from lack of oxygen to the brain after hyperventilation and breath holding, notes that it can be fatal without rapid rescue, and states that hyperventilation should not be permitted before breath holding. Small laboratory studies of 9 and 18 volunteers found that hyperventilating before breath holds lengthened them and, in serial holds, was followed by lower blood oxygen. This guidance does not mention breathwork classes, so applying it to them is an inference.

Training, regulation and when to seek care

The UK Breathwork Association (UKBA) distinguishes breathwork from legally regulated professions such as medicine and states that it intends to become a voluntary register accredited by the Professional Standards Authority, whose list of accredited registers has no breathwork category. In Ireland, the Department of Health says complementary therapy providers are not subject to statutory professional regulation, and the regulator CORU's list of regulated professions does not include breathwork facilitators. Neither Irish page names breathwork, so this is an inference. Nothing verifiable was found for the European Union as a whole or for US states.

Chest pain, fainting, a seizure or severe psychological distress during or after a session are reasons to seek medical help. Breathing practice is not a replacement for assessment or for prescribed treatment.