Breathing has always done more than move oxygen in and carbon dioxide out. Every time you slow down, deepen, or hold your breath, you’re pulling levers connected to your heart, your blood vessels, your brain, and even your immune system. A new narrative review published July 8, 2026 in Frontiers in Psychiatry (Neuroimaging section), led by a team from Mayo Clinic including Pravesh Sharma and Paul H. Min, takes stock of how structured breathing practices act on the autonomic nervous system, cerebral blood flow, cerebrospinal fluid dynamics, and immune response. The review pulls together decades of physiology research to explain exactly how — and it’s a useful map for anyone who’s ever wondered what’s actually happening in the body during a breathing session.
Why six breaths per minute keeps showing up in breathwork research
Slow it down enough, and breathing starts to resonate with one of the body’s own regulatory rhythms. Around six breaths per minute — roughly 0.1 Hz — respiration synchronizes with the baroreflex, the feedback loop that keeps blood pressure stable by constantly adjusting heart rate and vessel tone in response to signals from baroreceptors in the carotid sinus and aortic arch. When breathing and baroreflex oscillations line up, the result is what researchers call cardiorespiratory coherence: heart rate variability rises, baroreflex sensitivity improves, and the whole system settles into a more flexible, better-regulated state.
This is measurable using heart rate variability (HRV), particularly the high-frequency component and root-mean-square measures tied to vagal activity. It’s worth being precise here, though: HRV is a useful window into autonomic balance, but it isn’t a direct readout of vagal tone, and the old low-frequency/high-frequency ratio — once treated as a clean marker of sympathetic versus parasympathetic activity — is now considered too blunt an instrument, since both branches of the autonomic nervous system contribute to both bands. Some effects don’t even show up in HRV at all: a well-known study on cyclic sighing (two nasal inhales followed by one long exhale) found reduced anxiety and improved mood with no detectable HRV change, suggesting breathing can influence the nervous system through channels HRV doesn’t capture, likely tied to CO2 sensitivity and interoception.
How carbon dioxide, not oxygen, does the heavy lifting in cerebral blood flow
It’s tempting to think of breathwork’s brain effects in terms of oxygen delivery. The more interesting story is carbon dioxide. When CO2 rises in the blood, it reacts with water to form carbonic acid, which dissociates and shifts blood pH — a cascade that triggers hyperpolarization in endothelial cells, opens potassium channels, closes calcium channels, and relaxes smooth muscle around cerebral blood vessels. The net effect is vasodilation. This is the same mechanism exploited in cerebrovascular reactivity testing, where clinicians have people inhale CO2-enriched air while scanning blood flow changes with BOLD fMRI.
Slow-paced breathing engages this pathway naturally. As respiratory rate drops, tidal volume rises to compensate, producing mild CO2 retention and, with it, cerebral vasodilation. This is a large part of why different breathing protocols aren’t interchangeable: eucapnic slow breathing, Buteyko-style hypoventilation aimed at CO2 tolerance, and hyperventilation-based cyclic breathing that drives CO2 down are pulling on three genuinely different physiological levers, not three flavors of the same one.
What real-time MRI reveals about breathing-driven cerebrospinal fluid flow
For most of the twentieth century, cardiac pulsation was assumed to be the main driver of cerebrospinal fluid (CSF) movement. Real-time MRI has overturned that assumption: respiration, not the heartbeat, now appears to be the dominant force behind CSF flow in humans. Yildiz and colleagues documented this directly, showing that yogic breathing techniques measurably shift pulsatile CSF dynamics, and more recent velocity-encoded MRI work in breathing-trained participants shows enhanced CSF displacement at the foramen magnum tracking closely with diaphragm movement — even during ordinary, untrained breathing.
The mechanism follows the Monro-Kellie doctrine: deep diaphragmatic inhalation drops intrathoracic pressure, drains venous blood from the brain and neck, and that shift in cranial volume gets compensated by CSF moving up from the lumbar spine into the skull. Exhalation partially reverses the flow, producing a rhythmic, bidirectional current synchronized with the breath.
Where the evidence gets more cautious is the next logical question: does more CSF movement mean better waste clearance? The glymphatic system — first mapped in rodents by Maiken Nedergaard’s group — handles clearance of metabolic byproducts, including amyloid-beta, through this same CSF-ISF exchange. Contrast-enhanced MRI has confirmed an analogous pathway exists in humans, but direct visualization of glymphatic clearance the way it’s done in rodents isn’t possible in living people. One study did find that slow-paced breathing reduced plasma amyloid-beta, possibly via noradrenergic modulation, but it’s a single finding awaiting replication. The flow mechanism is well established; the downstream clearance benefit in humans is still an open question, not a settled one.
Why the vagus nerve’s anti-inflammatory reflex matters beyond relaxation
Slow breathing’s effects aren’t confined to the cardiovascular system. The cholinergic anti-inflammatory reflex gives the vagus nerve a direct line to the immune system: efferent vagal signaling reaches the spleen and other immune organs, triggering acetylcholine release that activates α7 nicotinic receptors on macrophages and dials down production of pro-inflammatory cytokines like TNF-α, IL-1β, and IL-6.
This isn’t purely theoretical. A randomized controlled trial in patients hospitalized with moderate COVID-19 pneumonia found that six-breaths-per-minute paced breathing significantly lowered circulating IL-6 compared to controls. It’s a meaningful clinical data point, though the effect size was modest and the authors themselves flagged the need for larger trials. Broader mind-body intervention research — a meta-analysis spanning 89 randomized trials — shows similar reductions in inflammatory markers, but those programs typically combine breathing with meditation or movement, so the breathing-specific contribution can’t be cleanly separated out.
Where the Wim Hof Method fits among breathing-only and multicomponent protocols
One of the more useful distinctions in the current literature is methodological rather than physiological: separating breathing-only protocols (slow breathing, deep slow breathing, 4-7-8 breathing) from multicomponent protocols that pair breathing with something else. The Wim Hof Method sits firmly in the second category — cyclic hyperventilation combined with cold exposure — alongside practices like Sudarshan Kriya Yoga, which layers breathing into a broader meditative structure.
This matters for how research gets interpreted. When a study combines cyclic hyperventilation with cold immersion, any measured outcome — say, a shift in inflammatory markers or subjective stress reduction — reflects the combined protocol, not breathing in isolation. That’s not a limitation specific to WHM; it’s true of any stacked intervention, mindfulness programs included. It does mean that mechanistic findings from breathing-only studies — the baroreflex resonance, the CO2-driven vasodilation, the CSF flow patterns — describe the respiratory component of what’s happening in a WHM session, while the cold-exposure piece is doing separate, parallel work through its own pathways.
What the research still doesn’t know about dose, duration, and durability
The review is candid about where the evidence runs thin. Dose-response relationships — how much practice, how often, for how long — remain largely uncharacterized across almost every technique studied. Most trial protocols pick session lengths somewhat arbitrarily rather than testing a range systematically. And there’s an important distinction between three timescales of evidence that’s easy to blur: acute single-session effects (measurable immediately, well-documented), short-term training effects over weeks (reasonably well-supported, including sustained parasympathetic shifts after three months of regular practice), and long-term durability after a program ends, which is the least studied of the three by a wide margin.
The largest blinded trial to date comparing coherent breathing against a different paced-breathing rhythm found both approaches improved stress, anxiety, and depression equally — a result that points toward structured attention to breath itself, rather than any single named technique’s specific ratio, as a meaningful part of what’s driving the benefit. That’s a useful frame for thinking about breathwork broadly: the mechanisms are real and increasingly well-mapped, but claims of one protocol’s clear superiority over another still outrun what the comparative evidence actually supports.
