In the first piece of this publication, I laid out the architecture of nervous system resiliency training, or NSRT: a substrate (sleep) and two training axes (bottom-up and top-down) that together build the capacity referred to as autonomic flexibility. That piece was the map. A broad layout of the topographical map. This one is the first deep dive.

We are starting with the bottom-up axis for a specific reason. The architecture said bottom-up enters first. The capacity to attend usefully to your internal state, which is what the top-down axis trains, depends on having an internal state regulated enough to be worth attending to. Bottom-up builds the stability and top-down trains what to do with it. So the bottom-up axis is where the work begins, and the primary stimulus on that axis is the subject of this piece.

That stimulus is resonance frequency breathing.

A note on emphasis before we go deeper. Resonance frequency breathing is the primary stimulus on the bottom-up axis. It is not the only one. The bottom-up axis includes additional physiological modalities that share a common mechanistic family with resonance breathing, working on related autonomic loops through different inputs. These secondary stimuli have their own roles, their own dosing logic, and their own use cases, and future pieces will introduce each of them on its own terms. But they sit on the axis as secondaries for a reason: resonance frequency breathing is the most mechanistically central, the most extensively researched, and the most accessible of the bottom-up modalities, which is why it is the foundation and why this piece focuses on it. The rest of the axis builds out from here.

What Resonance Frequency Breathing Actually Is

Breath is the primary conscious input to the autonomic nervous system. You can think your way into a calmer state with mixed success. You can will your heart to slow down with no success at all. But you can change your breath at will, and your breath, in turn, changes everything downstream of it: heart rate, blood pressure, baroreflex activity, vagal output. The autonomic nervous system was designed to operate without conscious access. Breath is the door. And many are just simply not leveraging this innate tool.

Not all breathing uses that door the same way. The vast majority of breathing practices, whether they are taught as breathwork, pranayama, or paced breathing apps, produce acute changes in state. You breathe a certain way for a few minutes, you feel different, the effect fades. That is real, and sometimes useful, but it is not training.

Resonance frequency breathing is different. It is a specific paced-breathing protocol, practiced at the individual’s unique resonance rate, that drives the cardiovascular system into maximum oscillation. When you breathe at this rate, heart rate and blood pressure begin to oscillate in large, coherent waves that are synchronized with the breath itself. The baroreflex, a feedback loop connecting heart rate and blood pressure regulation, enters a state of peak activation. Over weeks and months, repeated exposure to this state produces measurable adaptations in baroreflex gain, vagal tone, and autonomic regulation broadly (Lehrer et al., 2003; Vaschillo et al., 2006).

The keyword is resonance. Every physical system has a resonance frequency, the frequency at which it oscillates most efficiently. Push a swing at its resonance frequency and the amplitude grows with minimal effort. Push it at the wrong frequency and the swing fights you.

Think about this example…

You can attempt to push a kid as hard as you want on a swing, but if your timing is off, your efficiency is off and the strength or force of your push is negated by poor timing.

Your cardiovascular system has a resonance frequency too. Breathe at that rate, and the system enters maximum oscillation. Breathe at a nearby rate, and you get something that looks similar on the surface but produces meaningfully less of the underlying physiological response.

Why Generic Paced Breathing Falls Short

This is where the wellness conversation about breathing gets imprecise, and where most readers have probably been operating without knowing it.

The most common recommendation in popular paced-breathing content is six breaths per minute, often presented as a five-second inhale and a five-second exhale. That number is not arbitrary. It is roughly the population average for resonance frequency, and at six breaths per minute, most people will get some of the benefit of resonance breathing. This is not nothing. It is a reasonable starting approximation and a defensible default for someone who has no other information.

But it is not the trained practice.

Do not get me wrong, and I want to be clear, slow-paced breathing (around 6 breaths per minute) that is not at your resonance frequency rate is still effective and helpful. However, this is enough evidence to suggest that finding that resonance rate is the best way to maximize the output of the nervous system.

Individual resonance frequency varies. The literature consistently finds that resonance rates fall between roughly 4.5 and 6.5 breaths per minute, with meaningful variation across individuals based on cardiovascular geometry, body size, and other factors (Shaffer & Meehan, 2020; Vaschillo et al., 2006). Six breaths per minute is the average. Your resonance frequency is probably not six. It might be 5.5. It might be 4.5. Until you find it, you are doing generic paced breathing, which is useful, but you are not training at your resonance frequency, which is the practice that produces the documented adaptations.

This matters more than it might sound. The difference between breathing at your resonance frequency and breathing close to it is the difference between large coherent oscillations in the cardiovascular system and smaller, less coherent ones. The adaptation accrues from the coherent state. Training near your resonance frequency is useful. Training at it is what the research has actually validated.

A recent randomized trial by Sumińska et al. (2026) compared four weeks of HRV biofeedback at individually determined resonance frequency against HRV biofeedback at a fixed rate of 0.1 Hz (6 breaths per minute), with a waitlist control group. Both training groups showed significant reductions in self-reported stress, anxiety, and depressive symptoms relative to control, with no statistically significant differences between the individualized and fixed-rate conditions.

The study has begun circulating in wellness-adjacent commentary as evidence that fixed-rate slow breathing is “just as good” as individually calibrated resonance frequency training, and that the entire effort of finding your resonance frequency is therefore unnecessary. That reading is wrong, and importantly, it is not the reading the authors themselves offer. The authors are careful. They explicitly note that the Bayesian analysis provided only anecdotal evidence favoring the null hypothesis, that the sample was underpowered to detect small differences, and that future studies should include larger samples. They do not claim the two approaches are equivalent. The popular interpretation does. The popular interpretation is what needs dismantling, and two design features make the case.

First, the sample is small. Eighty participants completed the protocol across three groups, leaving roughly 26 per training arm. That is adequate to detect large between-group differences and inadequate to detect the small-to-moderate differences that would actually be expected between two slow-paced breathing protocols targeting overlapping mechanisms.

Second, and more critically, the intervention lasted only four weeks. This is the more serious problem. The HRV biofeedback literature consistently demonstrates that the autonomic adaptations produced by resonance frequency training, changes in baroreflex gain, vagal tone, and resting HRV, accrue over training periods longer than four weeks. Four weeks is sufficient to produce acute state effects and self-reported symptom improvements driven substantially by nonspecific factors. It is not sufficient to produce the trait-level autonomic remodeling that is the actual mechanism by which individual resonance frequency would be expected to outperform fixed-rate breathing. Consistent with this, the intervention did not produce significant changes in resting HRV in either group, which is exactly what you would expect from a training dose this short.

The study did what it set out to do. It tested whether four weeks of slow-paced breathing produces self-reported symptom improvement, and confirmed that it does. The question of whether individualized resonance frequency training produces superior autonomic adaptation over a meaningful training period remains open. This study was not designed to answer it. The popular reading pretends otherwise.

What the Practice Looks Like

Setting aside the assessment question for a moment, here is what a starting practice in resonance frequency breathing looks like in structural terms.

Frequency: four to five sessions per week. The training effect is dose-dependent and cumulative. Fewer than three sessions per week produces noticeably less adaptation. More than five sessions per week produces diminishing returns relative to time invested.

Duration: eight to twelve minutes per session for trait adaptation (cut the difference at 10 minutes). This is the consensus figure across the clinical biofeedback literature (Lehrer & Gevirtz, 2014). Shorter sessions, three to six minutes, produce acute state changes but less trait adaptation. Longer sessions are tolerable but not necessary.

Rate: at your individual resonance frequency, if you know it. If you do not, slow paced breathing at roughly five to six breaths per minute is the starting approximation. This produces some of the benefit of resonance practice and is a reasonable default for the first weeks of practice.

(Note…you can go through the whole protocol developed by Dr. Paul Lehrer and colleagues; however, this is cumbersome and will require either a trained professional or will make you go through a specified protocol on an app. Instead, head on over to https://ohm.health and get the Ohm Resonance Lamp. This is hands-down the best way to ensure you are training at your resonance frequency and the lamp finds it for you each time you use it! Make sure to use code: DRJAY for 10% off).

Mechanics: smooth, continuous breathing through the nose when possible. No breath holds. No forceful breathing. The goal is not effortful, it is precise. Resonance breathing should feel almost passive when you find the rate.

Progression: consistency first, duration second, precision of resonance detection third. Most people do not need to extend duration. They need to find their actual resonance frequency and train at it.

You will feel acute effects in the first session. Heart rate slowing. A sense of steadiness. A softening of whatever mental static was present when you started. Those are the state effects, and they are real. The trait effects, the ones that make your nervous system fundamentally more flexible, emerge over weeks of consistent practice.

The Assessment Problem

Everything in the previous section is calibrated to a generic protocol. The trained practice requires more.

Finding your individual resonance frequency is not difficult in principle. The standard clinical protocol, developed by Lehrer, Vaschillo, and colleagues, involves breathing at six different rates for several minutes each (typically 6.5, 6.0, 5.5, 5.0, 4.5, and 4.0 breaths per minute), measuring the resulting heart rate oscillations, and identifying the rate that produces the largest, most coherent oscillation. The rate that produces the maximum response is your resonance frequency (Shaffer & Meehan, 2020).

In principle, simple. In practice, almost no one does this.

The assessment requires three things that have, until recently, been hard to combine outside of a clinical setting. First, a sensor capable of capturing heart rate variability with enough fidelity to detect the oscillation differences across rates. Second, real-time biofeedback so the person being assessed can see the oscillation response as it happens, which is what makes the rate identifiable. Third, a structured protocol that guides the person through the assessment without requiring a clinician in the room.

This is the gap I joined Ohm (code DRJAY for 10% off) to close. The Ohm Resonance Lamp is the first device I have seen that handles all three requirements in a screen-free, daily-use form factor: a high-fidelity PPG sensor captures the cardiovascular response, an ambient light interface provides real-time biofeedback as you breathe, and the assessment protocol runs in the device itself. The device identifies your individual resonance frequency, then guides you to train at that rate in subsequent sessions, with the light signal reinforcing the coherent state when you find it.

Full disclosure: I am the Chief Health and Performance Officer at Ohm Health. I joined the company because the device solves a problem I had been trying to solve in clinical practice for more than a decade: making the trained practice of resonance frequency breathing accessible outside of a lab. The framework is independent of the tool, and the practice can be done with other equipment, but Ohm is the tool I currently recommend for daily use because it is the only one that brings clinical-grade assessment and biofeedback into a form factor people will actually use every day.

Why This Is the Foundation

Resonance frequency breathing is the bottom-up primary for reasons that go beyond convenience. Three features set it apart from every other candidate.

First, it is mechanistically central. The baroreflex is one of the core regulatory loops in autonomic function, and resonance frequency breathing is the most direct, most validated way to train it. Other stimuli on the bottom-up axis act on related mechanisms, but the baroreflex is the load-bearing target, and resonance breathing is the direct route to it.

Second, the literature is unusually mature. HRV biofeedback at resonance frequency has been studied for more than two decades across populations including anxiety disorders, depression, asthma, hypertension, athletic performance, and stress-related conditions broadly (Lehrer & Gevirtz, 2014). The adaptations are reproducible. The protocol is stable. The dosing is understood. Most interventions in the autonomic regulation space have a fraction of this evidence base.

Third, the input is accessible. Breath is the one autonomic-relevant variable that humans can change at will, with no equipment, with full conscious control. That makes it uniquely viable as a training stimulus for a system that is otherwise inaccessible. Every other modality on the bottom-up axis requires equipment or external conditions. Breath does not, which is why it is the foundation.

What This Builds Toward

Resonance frequency breathing is the primary stimulus on the bottom-up axis, but it is not the only stimulus. Future pieces will introduce the secondaries on this axis, the modalities that extend or support the primary work, with their own mechanisms and use cases. The top-down axis, with mental attunement training as its primary, will get its own series of deep dives. The substrate, sleep, deserves its own treatment as well.

Before any of that, the bottom-up primary needs to be on the page. This is the piece that does that.

If you are starting somewhere, start here. Slow paced breathing at five to six breaths per minute, four to five sessions per week, ten minutes per session, for the next four weeks. That is the generic protocol, and it will produce real effects. If you want the trained version, the next step is finding your individual resonance frequency, which is a different conversation about assessment, equipment, and what it actually takes to train the bottom-up axis at full precision.

Reading the gauge is not training the engine. Generic paced breathing is closer to training, but it is not yet the trained practice. The trained practice is what produces the adaptation. That is what we are building toward here.

—

References

Lehrer, P. M., & Gevirtz, R. (2014). Heart rate variability biofeedback: How and why does it work? Frontiers in Psychology, 5, 756.

Lehrer, P. M., Vaschillo, E., Vaschillo, B., Lu, S. E., Eckberg, D. L., Edelberg, R., Shih, W. J., Lin, Y., Kuusela, T. A., Tahvanainen, K. U. O., & Hamer, R. M. (2003). Heart rate variability biofeedback increases baroreflex gain and peak expiratory flow. Psychosomatic Medicine, 65(5), 796–805.

Shaffer, F., & Meehan, Z. M. (2020). A practical guide to resonance frequency assessment for heart rate variability biofeedback. Frontiers in Neuroscience, 14, 570400.

Sumińska, S., Rynkiewicz, A., & Szulczewski, M. (2026). Resonance frequency versus fixed 0.1 Hz breathing in HRV biofeedback: A four-week randomized comparison. Scientific Reports. https://doi.org/10.1038/s41598-026-53333-6

Vaschillo, E. G., Vaschillo, B., & Lehrer, P. M. (2006). Characteristics of resonance in heart rate variability stimulated by biofeedback. Applied Psychophysiology and Biofeedback, 31(2), 129–142.

—

Dr. Jay Wiles is a clinical health and performance psychologist, BCIA-certified in HRV biofeedback, and the originator of nervous system resiliency training. He serves as Chief Health and Performance Officer at Ohm Health and Director of Absolute Rest. He works with athletes across the NHL, NFL, MLB, and F1, alongside executives and operators in high-performance environments.

On this site:Nervous System Resiliency Training, the full framework.

First published in The Resilient Engine, my newsletter on training the nervous system. Read the original, or subscribe there to get new pieces as they land.