Radical Health · Radical Attune · The evidence
The evidence · graded · including the nulls

Every claim, with its receipt.

All eleven breathing patterns and all five sound methods in Radical Attune, each with an evidence grade and its primary sources — chosen to include the studies that found nothing. 25 peer-reviewed references. Sources verified June 2026.

This page exists because the app's own ABOUT screens carry the same material, and we'd rather you could read it before you download anything. Two honest framings before you start.

First, a grade describes the research behind a practice, not a promise about you. "Well studied" means the method has held up across trials — not that it will work for you, and not that Radical Attune has been studied. No trial has tested this app.

Second, the sound half of this app is the humble half. Rhythmic sound can measurably nudge the body and brain, but much of the entrainment field is early, mixed, or thin, and we say so below in each case rather than in a disclaimer you'd skip.

Radical Attune's in-app breathing evidence board: a preamble describing the eleven guided patterns, the founder's favourites marked ungraded, and researched patterns badged Well studied or Some evidence with tappable citations.
The same board, in the app. Every grade and citation on this page is also under ABOUT → Breathing.
How to read a grade

Five words,
used strictly.

These are the app's own badge words, shown verbatim on the session screen and the in-app evidence boards. We use exactly these five here — no clinical-sounding synonyms like "strong" or "proven".

Well studied

Supported by meta-analyses or multiple randomised trials pointing the same way.

Some evidence

Supported by randomised trials or a systematic review, with real limitations.

Early evidence

Small, acute, or mixed. Plausible, not established.

Very limited

Mechanism or correlation only, with little or no outcome evidence in humans.

Ungraded

The founder's own hand-tuned patterns. No study claims attached, by design.

Section 1 — the founder's own patterns

Three we like.
No study attached.

These three are hand-tuned by the app's founder because they feel good, and they're labelled ungraded for exactly that reason. They borrow their shapes from the researched patterns below — slow even breathing, a long exhale, alternate nostrils — but no specific trial tests these timings, and we don't dress them up as though one does.

PatternTimingDefault lengthGrade
Calm5.5s in · 5.5s out — about 5.5 breaths a minute5 minUngraded
4·4·7·24s in · 4s hold · 7s out · 2s hold5 minUngraded
Alternating nostril4·4·7·2, right in / left out, swapping each round10 minUngraded

Calm is the founder's everyday pace. 4·4·7·2 is a longer exhale with gentle holds, to wind all the way down. Alternating nostril combines the two. The two hold-containing patterns show a safety note in the app: they include breath-holds, so sit or lie down, and you can pause at any point.

Section 2 — the researched patterns

Eight patterns.
Eight grades.

Slow-paced breathing

Well studied
5s in · 5s out · 6 breaths a minute · 5 min default · rate is tunable

The most-studied pace in the whole field, and the app's everyday default. Slow breathing at roughly six breaths a minute reliably increases heart-rate variability and shifts the balance toward the parasympathetic "rest and digest" state. This is the one pattern the app will calibrate to you personally, because the resonant rate is individual — usually between 4.5 and 6.5 breaths a minute.

What it doesn't say: raising HRV in a session is not the same as treating anxiety, and the effect does not clearly outlast the practice.

Sources

Pursed-lip breathing

Well studied
2s in through the nose · 4s out through pursed lips · 3 min default · short rounds

A long exhale through pursed lips reduces the sense of breathlessness and helps recovery after exertion. The strongest evidence base here comes from pulmonary rehabilitation.

What it doesn't say: this is an ease-and-recovery exercise, not a stress-relief tool, and the app keeps it to short rounds for that reason. Radical Attune does not treat, manage or monitor COPD or any respiratory condition — if you have one, your clinician's guidance comes first.

Sources

Cyclic sighing

Some evidence
3s inhale · 1s top-off · 6s exhale · 5 min default

A deep inhale, a short top-off, then a long letting-go exhale. The fastest reset in the catalog. In a 2023 randomised trial comparing breathwork arms against mindfulness meditation, cyclic sighing came out ahead on mood and respiratory-rate reduction.

What it doesn't say: one trial, modest size, short follow-up.

Sources

Belly (diaphragmatic) breathing

Some evidence
4s in · 6s out · breathe low · 5 min default

Breathing low into the belly rather than high in the chest — the mechanical foundation under every other pattern here. A randomised trial found improvements in attention and cortisol against a control condition.

What it doesn't say: single-trial, small sample, and cortisol is a marker rather than an outcome you'd feel.

Sources

Alternate-nostril breathing

Some evidence
4s each · close one side, then the other · 10 min default

Breathe one nostril at a time, closing the other by hand. A deliberate, seated, meditative practice. A 2024 meta-analysis of 14 randomised trials found reductions in blood pressure, and a much older systematic review of 44 trials pointed the same way.

What it doesn't say: trial quality in this literature is uneven, blinding is close to impossible, and a blood-pressure change in a study session is not a treatment for hypertension.

Safety note shown in the app: needs a free hand to close each nostril, so it's best seated. Skip it if your nose is congested.

Sources

4-7-8

Early evidence
4s in · 7s hold · 8s out · 4 min default · capped at 4 cycles per round

A structured wind-down: a long exhale with a hold in the middle. Popular, and the acute measurements are real — but small.

What it doesn't say: the evidence is short-term and acute, with no good long-term trials. Best before sleep rather than all day.

Safety note shown in the app: the 7-second hold can feel light-headed. Sit or lie down, and you can pause at any point. The app caps it at four cycles per round for the same reason.

Sources

Box breathing

Early evidence
4 · 4 · 4 · 4 · 5 min default

Equal counts in a square: in, hold, out, hold. Steady and easy to remember, which is most of its appeal — good for composure before something demanding.

What it doesn't say: when it's been compared head-to-head against 4-7-8 and six-breaths-a-minute paced breathing, box breathing has not come out ahead. Choose it because you can remember it under pressure, not because it's stronger.

Safety note shown in the app: the empty-lung hold can feel air-hungry. Keep the counts comfortable.

Sources

Power breaths (cyclic hyperventilation)

Early evidence · gated
30 fast breaths · 45s empty hold · inhale · 15s full hold · repeat · 10 min default

The one pattern here that isn't calming. It's high-arousal and stimulating — the opposite of everything above — and it's gated behind a safety screen you have to read, and unavailable in Duet.

What it does: the physiology is well-documented. Cyclic hyperventilation followed by breath-holds measurably moves blood oxygen, CO₂ and pH, and raises adrenaline.

What it doesn't say: "measurably changes your blood chemistry" is not "is good for you." The outcome evidence is thin, and this pattern is genuinely unsuitable for some people.

Safety: never practise it in or near water, while driving, or while standing — you can faint. Sit or lie down. Avoid it during pregnancy, or if you have epilepsy, a heart condition, or a history of fainting. If you feel unwell, stop.

Sources
Section 3 — the sound methods

The humble
half.

This is where we'd rather be straight with you than sell you something. Rhythmic sound can nudge the body and brain measurably — and much of this field is early, mixed, or thin. Note that no sound method reaches the top grade; the best any of them manages is early evidence. Each entry below pairs the encouraging findings with the sobering ones, because that's the actual state of the research.

Binaural beats

Early evidence

What it is: play two pure tones a few hertz apart — 200 Hz in one ear and 207 Hz in the other — and the brain, hearing both, conjures a third pulsing beat at the difference, here 7 Hz. That phantom beat exists only in your head, which is why headphones are required: each ear must get its own tone. The idea is to set that beat to a target brain rhythm.

What the science honestly shows: this is the most-studied delivery method, and the picture is genuinely mixed. Encouragingly, a meta-analysis of 22 studies found small-to-moderate benefits for anxiety, memory and pain, and a surgical randomised trial roughly halved pre-operative anxiety versus the same audio without beats. Soberingly, careful EEG work often finds the brain doesn't actually lock onto the beat at all — so any calm you feel may come from slow, pleasant sound and relaxed expectation rather than true entrainment. Worth using. Worth not over-believing.

Sources

Isochronic tones

Very limited

What it is: a single tone switched cleanly on and off at a steady rate — a rhythmic pulse rather than a phantom beat. Because the rhythm is in the sound itself, it doesn't need headphones and works fine on a speaker, which also makes it easy to layer under the breath pacer.

What the science honestly shows: these sharp pulses give the auditory system a strong, clear rhythm to follow, and a 2024 EEG study found they shift brain-wave power more than binaural beats do — the cortex really does track them. The honest gap: that's a measurement of brain activity, not of feeling better. There are essentially no clinical trials testing isochronic tones on mood, sleep or focus, so treat the stronger brain response as promising mechanism, not proven outcome.

Sources

Brain-wave bands (delta · theta · alpha · beta · gamma)

Early evidence

What it is: rhythmic sound pulsed at a target rate, aiming to nudge brain rhythms toward sleep, calm, focus, or alertness. In the app these are the named frequencies — Deep sleep at 2.5 Hz, Deep drift at 4 Hz, Meditation at 6 Hz, Relaxed focus at 10 Hz, Alert calm at 12 Hz, Flow at 16 Hz, Clarity at 40 Hz.

What the science honestly shows: the brain measurably tracks a steady rhythm — a real, replicable EEG effect. But "the EEG synced" is not "the person improved." A 40 Hz study produced textbook entrainment with zero cognitive gain; effects rarely outlast the sound; and the trials are small and mixed. One alpha-entrainment randomised trial did find lowered anxiety markers, which is why this sits at early evidence rather than very limited.

Sources

Solfeggio frequencies (528 Hz and others)

Early evidence

What it is: a set of warm, consonant tones with evocative names — foundation, renewal, liberation, change, transformation, connection, expression, intuition, awakening. Despite the lore, they were devised in the late 20th century. Not ancient, and with no established physiological basis of their own.

What the science honestly shows: the encouraging, well-supported part is that these are pleasant, consonant tones — and pleasant, slow music is one of the best-evidenced ways to unwind. A meta-analysis of 104 trials across 9,617 people found music reliably lowers stress hormones, heart rate and anxiety, and Cochrane reviews echo it. So as a calming backdrop for breathing and rest, they genuinely help. What isn't established is that the specific numbers do anything beyond what any pleasant music does: the most-cited 528 Hz study had nine people and no control group. Enjoy them for how they sound, not as a magic frequency.

Sources

Schumann resonance (7.83 Hz)

Very limited

What it is: a real, well-established physics phenomenon — the natural electromagnetic "heartbeat" of the planet. Global lightning, tens of strikes every second, rings the cavity between the Earth's surface and the ionosphere like a bell, creating standing waves. The fundamental sits near 7.83 Hz, with harmonics around 14, 20 and 27 Hz — the same series the app offers. Because 7.83 Hz is below the roughly 20 Hz floor of human hearing, it's delivered indirectly, as a binaural beat or a pulse.

What the science honestly shows: it's an appealing target because 7.83 Hz falls in the brain's theta / high-delta range, the rhythm of deep calm and meditation, and observational research has linked higher Schumann and geomagnetic field power to greater heart-rate variability — a marker of the parasympathetic state. But that work tracks an ambient electromagnetic field over weeks, and the association runs both ways across studies. It is not a sound through earbuds. No human trial shows a 7.83 Hz audio tone reproduces any of it, and a 2023 systematic review casts doubt on audio brainwave entrainment generally. Honest read: a beautiful organising idea and a sensible calm-band target — not a proven remedy.

Sources
Section 4 — what the app measures, and what those numbers aren't

Estimates,
labelled as such.

01

Breath–heart coherence ("lock")

A non-clinical estimate of how closely your heart-rate rhythm is following your breathing rhythm — respiratory sinus arrhythmia, in the literature. It's computed on device from live heart rate and the breath pacer. It is not a validated index, not a clinical measurement, and not comparable to anyone else's number.

02

Calm

A derived, non-clinical figure summarising a session, not a physiological state anyone has validated.

03

Camera heart rate

An experimental estimate from the rear camera and torch, used only when no Apple Watch reading is available. Frames are processed on device and never stored or transmitted. It is not validated and not intended for diagnosis, treatment, or medical monitoring.

04

Calibration

With an Apple Watch, the app measures how much your heart rate swings at five breathing rates and recommends the largest swing, which is a reasonable proxy for your resonant rate. It's a within-session comparison of your own readings, not a diagnostic test — and the app tells you to trust your own sense too.

05

Insights

Comparisons across your own past sessions. The app refuses to turn a comparison into a recommendation until it has at least three readings, and labels small samples as hints rather than proof. Your own history is a sample size of you, which is both the point and the limitation.

Section 5 — what we'll never claim

The list we
hold ourselves to.

What we'll never claim
  • ✕ That these tones treat, prevent or diagnose anything. It's a relaxation tool.
  • ✕ That brainwave entrainment is settled science. The careful studies disagree with each other.
  • ✕ That a measured EEG shift means you feel better. Those are two different findings.
  • ✕ That 528 Hz or any specific number has a proven physiological effect of its own.
  • ✕ That an audio tone reproduces what the Schumann resonance research measured. No trial shows it.
  • ✕ That camera heart rate is medical-grade. It's an experimental estimate, and unvalidated.
  • ✕ That a breath–heart coherence number is a clinical measurement. It's a non-clinical estimate.
  • ✕ That a "Well studied" badge means it will work for you. It means the research held up, not that you will respond.
  • ✕ That Radical Attune has been studied. It hasn't. The evidence backs some of the methods inside it — slow breathing most of all — and nothing more than that.
  • ✕ That any of this replaces a professional. For anything medical, see one.
References

All 25 sources.

Every source cited above, once each, with a working link. Twelve for the breathing patterns, thirteen for the sound methods.

  1. Laborde et al. 2022 — slow breathing & HRV meta-analysis Neurosci Biobehav Rev
  2. You, Laborde et al. 2024 — slow-paced breathing meta-analysis (31 studies, n=1133) Mindfulness
  3. Roberts et al. 2009 — pursed-lip breathing systematic review in COPD (11 studies)
  4. European Respiratory Society review 2024 — breathing techniques Eur Respir Rev
  5. Balban, Spiegel, Huberman et al. 2023 — breathwork vs mindfulness meditation RCT Cell Reports Medicine (cyclic sighing, box breathing, cyclic hyperventilation)
  6. Ma et al. 2017 — diaphragmatic breathing RCT Frontiers in Psychology
  7. Nam et al. 2024 — alternate-nostril breathing & blood pressure meta-analysis (14 RCTs)
  8. Ghiya 2017 — alternate-nostril breathing systematic review of 44 RCTs Int J Res Med Sci
  9. Vierra, Boonla & Prasertsri 2022 — 4-7-8 acute HRV & BP effects Physiological Reports
  10. Aktaş & Karabulut 2024 — 4-7-8 scoping review (15 studies)
  11. Marchant et al. 2025 — square breathing vs 4-7-8 vs 6 breaths/min head-to-head
  12. Kox et al. — Wim Hof breathing physiology (SpO₂/CO₂/pH pilot)
  13. Garcia-Argibay et al. 2019 — binaural beats meta-analysis (22 studies): cognition, anxiety, pain Psychological Research
  14. Padmanabhan et al. 2005 — binaural beats & pre-operative anxiety RCT Anaesthesia
  15. López-Caballero & Escera 2017 — controlled study found no EEG or arousal change from binaural beats (null result)
  16. Orozco Perez et al. 2020 — only weak cortical entrainment, no mood change eNeuro (null result)
  17. dos Anjos et al. 2024 — isochronic tones shift EEG power more than binaural beats Neuroscience
  18. Huang & Charyton 2008 — review of brainwave-entrainment effects (incl. isochronic/photic)
  19. Wu et al. 2025 — 40 Hz entrains EEG but produces no cognitive benefit (null result)
  20. Shehani et al. 2024 — alpha entrainment lowers anxiety markers (RCT)
  21. de Witte et al. 2020 — music & stress: meta-analyses of 104 RCTs (9,617 people) Health Psychology Review
  22. Bradt, Dileo & Shim 2013 — Cochrane review: music for pre-operative anxiety (26 trials)
  23. Akimoto et al. 2018 — 528 Hz, cortisol & oxytocin (n=9, no control) Health
  24. Alabdulgader et al. 2018 — Schumann/geomagnetic power & HRV over 5 months (correlational) Scientific Reports
  25. Ingendoh et al. 2023 — systematic review casts doubt on audio brainwave entrainment (null result)

Links open the source as published. Where a study is cited by more than one pattern or method above — Balban 2023 and Huang & Charyton 2008 both are — it is listed once here. Nothing was dropped for being unflattering.

Closing note

25 sources.
Nulls included.

Every reference on this page is linked, and the list was assembled to include the studies that found no effect — López-Caballero and Escera finding no EEG or arousal change from binaural beats, Orozco Perez finding only weak cortical entrainment and no mood change, Wu and colleagues getting textbook 40 Hz entrainment with zero cognitive benefit, Ingendoh's review casting doubt on the whole audio entrainment premise. They're here because leaving them out is how this genre usually works, and we'd rather you could check.

Sources verified June 2026. If you find one we've read wrong, tell us: help@radicalhealth.app.

← Back to Radical Attune