Meditation and the Autonomic Nervous System: What the Evidence Shows

What is meditation, for the autonomic nervous system?

"Meditation" covers practices that ask very different things of the body. Researchers still classify them in varied and largely subjective ways [S1]. The main families studied are:

• Focused attention — holding attention on one object, such as the breath or a sound, and returning to it when the mind wanders.
• Open monitoring — noticing whatever arises, without choosing an object.
• Loving-kindness and compassion — deliberately generating warm feelings towards oneself and others.
• Mantra and Transcendental Meditation (TM) — silent or spoken repetition of a word or sound.
• MBSR and MBCT — eight-week group courses that combine body scans, sitting meditation, gentle yoga and teaching.

These are different interventions, and they should not be pooled into one conclusion. In experienced practitioners, some styles were accompanied by a relaxation response and others by clear arousal [S2]. A course like MBSR adds movement, group support and homework; a mantra recited aloud changes the breath. A verdict on "meditation" as a whole hides these differences.

What the autonomic nervous system is, and why heart rate variability is only an indirect window onto it, is covered on its own page. This page asks what meditation does to the signals that studies measure.

How does it work?

The honest answer is that the routes are not settled. Three candidates appear in the research:

• Attention and appraisal. Calmer thinking and less rumination may lower arousal. This is the route most people have in mind.
• Breathing. Many practices slow the breath, and slow breathing changes heart rhythm on its own (see below).
• Posture, stillness and setting. Sitting quietly for a while lowers heart rate compared with daily activity, whatever the mind is doing.

Brain imaging finds patterns that mostly differ between traditions [S1]; the brain side is covered in how fast meditation changes the brain and gamma waves in meditators. This page stays with the autonomic system.

How is it measured?

Studies of meditation and the autonomic system rely on a few signals:

• Heart rate variability (HRV), during a session or at rest before and after a course.
• Resting heart rate.
• Blood pressure, at the clinic or over a day and night with an ambulatory monitor.
• Cortisol and other stress markers in blood or saliva, and sometimes inflammatory markers.
• Breathing rate — measured in some studies, ignored in many.

Two time frames matter. A session effect is a change while someone is meditating. A trait effect is a lasting change after weeks of practice, or a stable difference between practitioners and non-practitioners. They answer different questions and are easy to confuse.

What affects it?

• The style of practice. Calming and arousing styles move the same signals in opposite directions [S2].
• Breathing rate during practice. Slower breathing raises HRV regardless of what the mind is doing [S10].
• The comparison condition. Wait-list controls flatter the effect; active controls such as relaxation, health education or listening to a story shrink it [S3] [S6].
• Who is studied. Effects on cortisol appeared mainly in people who were already ill or under strain [S16].
• How the signal is measured. Ambulatory and office blood pressure gave different answers in the same meta-analysis [S13].

What does the evidence show?

Heart rate variability: during a session and after a course

During a session: context-dependent. HRV often changes while people meditate, but not always in the same direction. A body scan raised the breath-linked heart rhythm more than other relaxing activities in two small studies, without changing heart rate [S6]. Mantra and prayer recited slowly enough to bring breathing near about 5.5–6 breaths per minute (around 0.1 Hz) produced large swings in heart rhythm [S8]. Some Chi and Kundalini Yoga techniques produced very large oscillations tied to slow breathing [S9].

"Meditation always raises HRV." The evidence does not support it. In one small study, overall HRV fell during deep heartfulness meditation compared with a breathing-matched control [S7], and arousing styles from the Vajrayana tradition were accompanied by arousal rather than relaxation [S2]. Meditation is not only an autonomically quiet state [S9].

After a course: unknown. A meta-analysis of standardized mindfulness-based interventions found mixed and inconclusive effects on HRV, because there were few studies and no large rigorous trials [S4]. The authors called for larger trials with active controls and longer follow-up [S4]. A single uncontrolled study before and after an intensive Vipassana retreat found changes during meditation that its own authors read as a shift in breathing and attention rather than a simple calming effect [S5].

Resting heart rate: context-dependent

A meta-analysis of 45 randomized trials with active controls found that open monitoring practices reduced heart rate, and that heart rate fell when all forms were pooled [S3]. These are small changes in mixed populations. Resting heart rate is a broad signal, and a modest fall says little about which part of the practice caused it.

Blood pressure: context-dependent

This is the outcome with the most trials, and the answer depends on the practice.

• Guideline view. The American Heart Association's statement on lowering blood pressure (Brook et al. 2013) rated Transcendental Meditation Class IIB and other meditation techniques Class III, describing modest, mixed or inconsistent evidence [S11]. It considered it reasonable for people above normal blood pressure to try such approaches as additions when clinically appropriate [S11]. Its later statement on meditation and cardiovascular risk (Levine et al. 2017) found a possible benefit from studies of modest quality, and said meditation may be considered as an adjunct to guideline-directed care, with the benefits still to be better established [S12].
• Against active controls, all meditation subtypes reduced systolic blood pressure [S3].
• TM and non-TM. A meta-analysis of randomized trials found lower office blood pressure with both. For TM, the effect on ambulatory systolic pressure was not significant, and the authors call for more ambulatory-measured TM trials [S13].
• TM overviews. An overview of reviews estimated a small reduction in both systolic and diastolic pressure with TM [S14]. Two of the reviews came from a TM-related institution, most were of fair quality, and the authors call for long-term trials by independent researchers [S14].
• A single small trial of mindfulness training in people with high-normal blood pressure or mild hypertension found lower blood pressure at the end of the course compared with health education, but the difference was no longer significant at follow-up [S15].

The practical side is in MBSR and mindfulness: the clinical evidence.

Cortisol and other stress markers: context-dependent

Against active controls, meditation reduced cortisol and C-reactive protein when all forms were pooled, and focused attention practices reduced cortisol on their own [S3]. A separate meta-analysis found a medium effect on cortisol in blood samples that appeared only in people with a somatic illness; in saliva samples the effect was small and not significant [S16]. Cortisol changes with the time of day, sleep and sampling method, so these results are harder to compare than blood pressure.

Sleep with autonomic measures: emerging

Few meditation trials measured autonomic signals and sleep together. The evidence here is one study: in one randomized trial against a wait-list, an app-based mindfulness programme improved ring-measured sleep efficiency, and heart rate fell and HRV rose during the sessions [S17]. HRV was measured during practice, not across the night, and personal burnout rose right after the programme before returning toward baseline [S17]. One study — emerging.

The main trap: breathing

Many meditation practices slow the breath, and slow breathing by itself raises vagally mediated HRV — during the breathing, right after a session and after weeks of practice [S10]. The mechanism is described in how breathing changes HRV, and the outcome evidence in slow breathing. So a rise in HRV during meditation does not show that attention, acceptance or compassion did anything beyond changing the breath.

What is known.

• Reciting a mantra or a rosary prayer at a slow pace brings breathing near the resonance rate, and the swings in heart rhythm grow large [S8].
• In a Vipassana study, the only change during meditation before the retreat was the kind expected from a change in breathing [S5].
• Large heart-rhythm oscillations in some Chi and Kundalini Yoga techniques were tied to slow breathing [S9].

What is debated. Few studies hold breathing constant. One design that does uses a paced-breathing control matched to each person's breathing rhythm during meditation; in that study, HRV fell rather than rose during deep meditation [S7]. A body scan raised the breath-linked heart rhythm more than other relaxing activities [S6], which hints at an effect beyond rest, but the abstract does not report whether breathing differed. Whether meditation changes heart rhythm through attention at all, once breathing is matched, remains open. How meditation and breathwork compare in practice is covered in meditation vs breathwork; training with real-time heart-rhythm feedback is a separate method, covered in HRV biofeedback.

Session vs long-term effects

What is shown. Changes during a session are well documented, and their direction depends on the style and on breathing [S2] [S6] [S8] [S9]. Courses with active controls show modest falls in blood pressure, heart rate and cortisol [S3].

What is not shown. Lasting changes in resting HRV after a course have not been shown consistently [S4]. Studies of experienced practitioners compare people who already meditate with people who do not [S2] [S7] [S9]. These are cross-sectional comparisons. People who keep meditating for years may differ from the start — in health, temperament, lifestyle or breathing habits — so a difference between groups does not show that meditation caused it. Pre–post studies without a control group, such as the Vipassana retreat study [S5], cannot separate practice from time, setting or expectation.

For tracking your own practice over time, see measuring meditation progress and the hub meditation with measurable progress.

What it does not tell you

• Expectations. Participants know they are meditating, so a true placebo is impossible, and hopeful expectations can lower blood pressure and heart rate on their own.
• The control group decides the answer. Many studies in this field lack an active control [S3]. A wait-list shows the effect of doing something versus nothing; an active control such as relaxation or health education shows what meditation adds [S6] [S15].
• Small samples. Most single studies here are small [S5] [S6] [S7] [S8] [S9] [S15] [S17], and the HRV meta-analysis had too few rigorous trials to conclude [S4].
• Different practices pooled. Pooled results mix styles that push the autonomic system in different directions [S1] [S2] [S3].
• Publication bias and independence. Positive small studies are more likely to be published, and part of the TM evidence comes from a TM-related institution [S14].
• A physiological change is not a health outcome. A lower reading during practice does not show fewer heart attacks or strokes; none of the sources here measured those.
• It is not a treatment on its own. Meditation does not replace medication or medical care for high blood pressure or any other condition. Nothing here means "stop your treatment".

Safety

Meditation is generally low-risk, but not risk-free: some people report anxiety, low mood or unsettling experiences, especially with intensive practice or a history of trauma or psychosis. One trial in this page saw personal burnout rise briefly after a mindfulness programme [S17]. The full picture is in meditation adverse effects and safety. Stop and seek help if practice leaves you more distressed, and talk to a clinician before intensive retreats if you have a mental-health condition.

In ONDA

ONDA shows your pulse live during a practice, from the iPhone camera or an Apple Watch. That lets you see how your pulse moves during a session; it does not measure your autonomic nervous system, and it cannot tell whether a change came from attention or from slower breathing. ONDA has no study of its own effectiveness, and the findings on this page do not show what ONDA does or does not achieve. ONDA does not diagnose or treat any condition. See what ONDA measures.