Breathwork is one of the few biohacking interventions that costs nothing, works within minutes, and has a growing base of randomized trial evidence. It is also a category where the strongest RCT and the loudest marketing rarely point to the same protocol. Wim Hof method has the biggest brand and the most oversold claims. Box breathing has military and clinical evidence for a narrow but real acute effect. Physiological sighing has a recent, credible Stanford trial that is quietly the most useful protocol in the group. Coherent breathing at 5 to 6 breaths per minute has strong heart rate variability (HRV) evidence but is boring enough that few brands have promoted it hard.
This article walks through what each of the four protocols is, what the actual trials show, which claims to trust and which to discount, and a practical 5-minute daily protocol you can run today. Muse 2 is the one tool that makes acute-stress breathwork measurable in real time if the biofeedback matters to you.
Box breathing: the military-derived acute regulator
Box breathing — inhale for 4 seconds, hold for 4, exhale for 4, hold for 4 — was popularized by Mark Divine and others with SEAL backgrounds, and has been formally taught in the U.S. military as a stress management tool. The mechanism is straightforward: the deliberate lengthening of the exhale and post-exhale hold increases parasympathetic tone via vagal afferent signaling, which reliably reduces heart rate and subjective arousal within 60 to 90 seconds.
The trial evidence for box breathing specifically is smaller than for its close cousins (paced breathing, resonance breathing), but the mechanism generalizes. Zaccaro and colleagues (2018, Frontiers in Human Neuroscience) reviewed the paced breathing literature and found consistent short-term effects on HRV and self-reported anxiety across protocols with extended exhales. Röttger and colleagues (2020) showed measurable reduction in cortisol and self-reported stress in an acute stressor paradigm using box-style paced breathing.
What box breathing does reliably: reduces acute physiological arousal in one to five minutes, useful for pre-performance anxiety, high-pressure meetings, or the ten minutes before a hard conversation. What it does not do: change your baseline stress response, treat clinical anxiety on its own, or produce lasting physiological adaptations without regular practice.
The equal-count structure (4-4-4-4) is easy to remember. Some practitioners lengthen the exhale (4-4-6-2, or 4-2-8-2) to increase the parasympathetic pull. The extended-exhale variants are consistent with the underlying vagal biology.
Wim Hof method: a real paper, oversold marketing
Wim Hof method (WHM) combines three components — controlled hyperventilation cycles, breath retention (typically after passive exhale), and cold exposure. The breathwork alone is a rapid cyclic hyperventilation followed by an extended breath hold, repeated in rounds. It produces respiratory alkalosis, a hypocapnic (low CO2) state, and a period of relative hypoxemia during the breath hold — physiological perturbations that are subjectively intense and psychologically satisfying.
The Kox and colleagues (2014, PNAS) study is the paper the marketing rests on. In a well-designed trial, participants trained in WHM were injected with endotoxin (a bacterial component that triggers a controlled immune response) and compared to untrained controls. The trained group showed reduced pro-inflammatory cytokine release and reduced sickness symptoms compared to controls. That is a real finding published in a top journal, and Kox's group has done credible follow-up work.
What the Kox paper does not show is that WHM prevents illness, treats autoimmune disease, or produces the broader immune modulation the marketing implies. The endotoxin model is a specific, acute, controlled challenge. Generalizing from that to real-world infection resistance or chronic inflammation is a leap the data does not support. The published human evidence for WHM is thin outside the endotoxin work and a few small pilot trials on inflammation markers.
There are also real safety concerns. The hyperventilation phase produces respiratory alkalosis and cerebral vasoconstriction; breath holds during or after can produce syncope. Several drowning deaths have been documented from people doing WHM breathwork in or near water. Do not do WHM breathwork in a pool, bathtub, or cold plunge. Do not do it while driving. Do not do it if you have unmanaged cardiovascular disease, epilepsy, or a history of syncope.
The honest read: the Kox paper is real; the marketing overshoots; the safety envelope is narrower than a fitness protocol implies. WHM is a legitimate practice for people who want the specific altered-state experience it produces, done in a safe setting. It is not a general-purpose breathwork recommendation for most people.
The physiological sigh: Stanford's short and useful trial
Andrew Huberman's lab at Stanford published one of the cleanest recent breathwork trials — Balban and colleagues (2023, Cell Reports Medicine). The trial compared three brief daily breathwork practices — cyclic sighing, box breathing, and cyclic hyperventilation — against mindfulness meditation, over 30 days, in 108 participants. Each intervention was five minutes per day.
Cyclic sighing came out ahead. Participants doing 5 minutes of daily physiological sighing showed the largest improvement in positive affect, the largest reduction in respiratory rate, and modestly larger reductions in anxiety compared to meditation, box breathing, and cyclic hyperventilation. The effect sizes were small to moderate but consistent, and the study was well-designed with a real active control.
The physiological sigh itself is simple. Inhale deeply through the nose. Before exhaling, take a second short inhale on top of the first to fully expand the lungs. Then exhale slowly and completely through the mouth. Repeat. The double-inhale mechanism has physiological rationale — it recruits alveolar surface for gas exchange and the extended exhale produces the vagal pull that drives the parasympathetic shift.
For acute stress in the moment — a spike of anxiety, a difficult moment before a call, a sudden overwhelm — one to three physiological sighs will reliably lower arousal in 30 to 60 seconds. For a daily practice, five minutes of cyclic sighing has the strongest recent RCT evidence of any of the protocols in this article. It is boring, cheap, and effective. That combination makes it the protocol most worth adopting.
Coherent breathing: HRV at 5 to 6 breaths per minute
Coherent breathing, also called resonance breathing, is paced breathing at approximately 5 to 6 breaths per minute — an inhale-exhale cycle of about 10 to 12 seconds. The name comes from the observation that this pace produces the largest and most coherent oscillation in heart rate variability (HRV), because it matches the intrinsic frequency of the baroreflex, the blood-pressure control loop.
The RCT evidence for coherent breathing on HRV is strong. Lehrer and colleagues (multiple studies from 2003 onward) established the physiological rationale and showed reliable increases in HRV amplitude at resonance frequency across dozens of trials. Steffen and colleagues (2017) showed that 5 minutes of resonance breathing measurably increases HRV during and immediately after the practice.
HRV is a real biomarker — higher HRV is associated with better autonomic flexibility and stress tolerance. Whether increasing HRV during a breathing session translates to durable improvements in baseline HRV, and whether that produces the health outcomes people care about, is where the evidence gets thinner. Regular resonance breathing practice over 8 to 12 weeks has been shown to modestly increase baseline HRV in some studies, but the clinical significance is uncertain.
Coherent breathing is essentially the same as slow box breathing at the low end. If you extend box breathing to a 6-6-6-6 pattern (24 seconds per cycle, 2.5 breaths per minute), you are inside the resonance range. Most people find a 5-5 or 6-4 pattern (10 to 12 seconds per cycle, 5 to 6 breaths per minute) comfortable and effective. The equal or extended-exhale versions produce the largest HRV oscillation.
Which have real RCTs and which are anecdotal
Sorting the four honestly:
- Physiological sigh — best current RCT. Balban 2023 is the cleanest recent comparative trial, and cyclic sighing came out ahead of box breathing, cyclic hyperventilation, and meditation on affect and anxiety. Practical, safe, and free.
- Coherent (resonance) breathing — strongest HRV evidence. Two decades of Lehrer's work and follow-ups show reliable HRV effects. Clinical significance is modest but the physiological signal is unambiguous.
- Box breathing — good acute-effect evidence, generalized from paced breathing. Not a lot of trials on the 4-4-4-4 pattern specifically, but the mechanism and the paced-breathing literature support the acute anxiety and arousal reduction claim.
- Wim Hof method — real endotoxin trial, overhyped extrapolations. Kox 2014 is real. The broader immune, illness-resistance, and disease-treatment claims are not currently supported by the human data. Safety envelope is narrower than a fitness protocol implies.
A 5-minute daily protocol worth running
If you are going to adopt one breathwork practice, run the Balban protocol as tested.
- Sit or lie in a quiet spot. Set a 5-minute timer.
- Inhale slowly through the nose. Take a comfortable deep breath.
- Add a second short inhale. On top of the first, without exhaling, sip in a bit more air to fully expand the lungs.
- Exhale slowly through the mouth. Aim for the exhale to be roughly twice the length of the total inhale.
- Repeat. Continue for the full five minutes. There is no counting required and no perfect pace.
For acute stress interruption — mid-day spike, before a call, in the moment — one to three physiological sighs will produce a measurable arousal drop in under a minute. Add a slow four-count coherent breathing session if you want a longer recovery, or a box breathing session before a presentation.
If you want the biofeedback of watching your physiological state change in real time, an EEG-based device that measures brain activity during breathing sessions can help. Muse 2 does that.
Safety and honest limits
Breathwork is generally low-risk when done seated or lying in a safe environment. The exceptions matter.
- Do not do hyperventilation-based breathwork in or near water. This includes WHM breathwork, bhastrika, and any protocol that involves cyclic hyperventilation followed by breath holds. Loss of consciousness in water is fatal.
- Do not do intense breathwork while driving or operating machinery. Lightheadedness and syncope are documented risks.
- Consult a physician if you have epilepsy, unmanaged cardiovascular disease, or a history of dissociation or panic disorder. Some breathwork protocols can trigger seizures or panic in susceptible individuals.
- Breathwork is not a substitute for treatment of clinical anxiety, depression, or PTSD. It can be a useful adjunct. It is not a primary intervention.
The honest framing on the whole category: breathwork is a real, cheap, evidence-supported family of tools for acute state regulation. The Stanford cyclic-sighing trial is the current strongest evidence for a daily practice. The Wim Hof paper is real but has been dramatically overinterpreted in the marketing. Coherent breathing and box breathing both have supportable acute-effect claims. None of these will fix a life that is producing chronic stress; they will help you meet the stress with a nervous system that is not maximally activated.
Five minutes a day of cyclic sighing is the highest-yield addition most readers can make to a wellness routine. It costs nothing. It has the best current RCT. It is worth trying for two weeks before deciding whether it earns a permanent place.
