Part 2 — HRV Biofeedback as a Foundation for Neurofeedback

1. Intro and brief history

In Part 1, I presented the notion that neurofeedback is not really brain training. It is the training of a brain embedded inside a constantly shifting physiological system, and when that system is not stable, the EEG learning loop becomes noisy. The three-layer framework — state regulation, interference reduction, external modulation — was an attempt to organize the adjunctive landscape by clinical function rather than by technology. HRV biofeedback sat at the base of that framework. Part 2 unfolds that base.

The lineage that gave us modern HRV-BF runs through Paul Lehrer at Rutgers and Evgeny and Bronya Vaschillo (Lehrer & Gevirtz, 2014). The discovery, simplified, was that each person has a breathing frequency at which the natural oscillations of the cardiovascular system — heart rate, blood pressure, baroreflex feedback — line up in phase. Breathe at that frequency, and the amplitude of heart rate variability swells dramatically. Breathe outside it, and the amplitude collapses back toward baseline.

That phenomenon had been observed in cardiology and pulmonology for decades. What Lehrer, Vaschillo, and Vaschillo did was treat it as a training target. If voluntary breathing at resonance frequency reliably maximizes the gain of the baroreflex loop, then sustained practice — over weeks, with feedback — should train the loop itself. The hypothesis turned out to be defensible. Sustained HRV-BF practice increases baroreflex sensitivity, raises cardiac vagal tone indices, and reshapes the autonomic response to challenge (Lehrer & Gevirtz, 2014; Goessl et al., 2017).

Around the same lineage, Stephen Porges' polyvagal theory offered an evolutionary account of how the vagus organizes safety, social engagement, and defense. That story has become enormously popular in clinical settings — and possibly easier to explain than the baroreflex one — but the mechanistic case for HRV-BF stands on its own without it. Baroreflex sensitivity, vagal tone, and the cardiorespiratory coupling resonance breathing engages are sufficient to explain why the training does what it does. Polyvagal framing is good as adjacent vocabulary; it is not the mechanistic spine of HRV-BF. I will name it once, here, and return to the cleaner story.

Forty years on, HRV-BF is the most evidenced of the biofeedback modalities outside surface EMG, the most clinically portable, and the modality NF practitioners most often add — or, as the rest of this post will argue, should add.

2. Alternate names

The literature uses several near-synonyms. Sorting them out matters because they index slightly different things and the terminology drift can mislead a careful clinician trying to read across studies.

Heart rate variability biofeedback (HRV-BF) and HRV biofeedback (HRVB) are the dominant terms. They refer to the training method as a whole — resonance-frequency breathing with real-time HRV feedback, over multiple sessions, with home practice. Resonance-frequency breathing names the breathing technique specifically and is sometimes used interchangeably with HRV-BF, though it really designates the active ingredient rather than the full protocol. Cardiac coherence training — popularized by the HeartMath organization — overlaps with HRV-BF in technique but adds proprietary metrics, equipment, and a meditation-style framing; the underlying physiology is the same baroreflex engagement, the packaging is different. Paced breathing is a more general term covering any slow, rhythmic breathing exercise; it captures the acute state shift but not the multi-session training arc. Slow-paced breathing and deep breathing tend to be used as catch-alls in the popular literature, often without specifying frequency, duration, or feedback.

Square breathing, box breathing, Wim Hof breathing and many others are tangential methods that have more-or-less similar mechanisms; each is complex and the field of HRV-biofeedback ironically deserves the same kind of clarification series that I’m running parallel to this one. The idea is to provide clarity and a better understand the confusing nomenclature surrounding the domain and the fact that it is a crowded modality with lots of spin-offs; neurofeedback and HRV biofeedback have this in common.

The distinction worth holding onto is between technique (resonance-frequency breathing, paced breathing) and training (HRV-BF as a multi-session learning protocol). A single ten-minute session of paced breathing produces a real, measurable state shift — calming sympathetic activation, briefly raising vagal indices, lowering subjective arousal. That is not nothing. But it is also not what the HRV-BF literature is talking about when it cites efficacy in anxiety or depression. The literature is talking about ten or twelve weeks of sustained practice, twice daily, with the goal of building a regulatory capacity that persists outside the practice itself. Conflating the two is one of the most common mistakes practitioners and clients make about this modality.

3. How HRV biofeedback works

The procedure, in clinical practice, is unusually well-defined for a biofeedback method. There are individual differences in how it is delivered, but the core sequence is stable across the lineage.

The first session is dedicated to resonance-frequency identification. The client is asked to breathe at each of several candidate paces — typically 4.5, 5.0, 5.5, 6.0, and 6.5 breaths per minute, sometimes 7.0 — for a minimum of two to three minutes per pace, with HRV monitored throughout. The pace that produces the highest LF-band amplitude, largest HR max-min spread, and cleanest phase alignment between respiration and heart rate is identified as the client's resonance frequency. For most adults it falls between 5.5 and 6.5 breaths per minute. The popular figure of 0.1 Hz, or six breaths per minute, is the population average, not a universal target. The individual variation counts — and a generic six-breaths-per-minute protocol applied without individualization is doing roughly-paced breathing, not HRV-BF.

Subsequent sessions train at the identified resonance frequency. A typical session structure looks like this: brief discussion of home-practice fidelity, a pre-training resting baseline of three to five minutes, four to six training runs of about five minutes each at resonance frequency with real-time HRV feedback on screen, brief discussion or coaching between runs, a self-guided transfer run of around five minutes without feedback, and a post-training resting period. Equipment is typically a clinical-grade BVP (Blood Volue Pulse; even if ECG is technically better) and respiration sensor pair (ProComp Infiniti with BioGraph, or equivalent) with software that displays heart rate, respiration waveform, and HRV indices — LF amplitude, LF percentage, and HR max-min are the indices most worth grading on.

Home practice runs in parallel and matters more than most clients realize. The standard recommendation is two daily practice sessions of about ten minutes each, five to seven days a week, across the in-clinic training period. Home practice is supported by a smaller-form-factor device — historically a fingertip pulse sensor with an app, more recently wearables like Optimal HRV, Polar, Garmin, the Eureka belt, or the Oura ring with an HRV-coherence app layered on top — that gives a simpler version of the same feedback. Without the home practice, the in-clinic sessions are too sparse to build a durable skill. With it, the protocol earns the training arc it claims.

The standard course runs ten to twelve weeks, ten to twenty sessions in clinic, with continuous home practice. That is the dose at which the autonomic restructuring the protocol promises tends to show up in measurement (Lehrer & Gevirtz, 2014). Shorter courses produce real state effects but more variable training-level gains. Longer courses, or maintenance practice continuing after the formal protocol ends, tend to consolidate the skill into something portable.

4. Mechanistic specifics

What is the training actually doing at the physiological level? Three intertwined mechanisms, each worth naming on its own terms.

The first is baroreflex amplification. The baroreflex is the closed-loop system that stabilizes blood pressure: pressure sensors in the carotid sinus and aortic arch detect changes in arterial pressure, signal the brainstem, and adjust heart rate and vascular tone accordingly. The loop has a natural resonance in the cardiovascular system — typically around 0.1 Hz in adults — at which feedback delay and amplitude alignment produce the largest oscillations of heart rate per unit of pressure change. Breathing at that frequency forces the loop to operate at resonance, which dramatically increases baroreflex gain. Over weeks of training, the baroreflex itself remodels. Sensitivity rises, response latency shortens, and the autonomic adjustments triggered by stress become both larger and faster.

The second is cardiac vagal tone enhancement. The vagus nerve provides the parasympathetic input to the heart that gives respiratory sinus arrhythmia (RSA) its amplitude — the rhythmic acceleration of heart rate on inhalation and deceleration on exhalation. Resonance-frequency breathing maximizes RSA amplitude, which over sustained practice translates into measurable increases in resting and challenge-state vagal tone indices. Cardiac vagal tone is one of the cleanest available indices of autonomic flexibility — the system's capacity to mount appropriate sympathetic-parasympathetic adjustments to whatever the moment requires.

The third is autonomic flexibility itself. Higher vagal tone, larger baroreflex gain, and trained voluntary capacity to engage the loop add up to a more flexible autonomic system — one that can move into sympathetic activation when challenge demands it, return to parasympathetic dominance when recovery is appropriate, and shift between them more cleanly. A common misreading of HRV-BF is that it makes the system more parasympathetic, full stop. It does not. It makes the system more responsive.

A distinction I want to lock down explicitly: HRV-BF is not primarily about lowering arousal. The acute state effect of paced breathing — momentary calming, blunted sympathetic activation — is real but is not the training target. The training target is the regulatory capacity itself, the deployable skill of engaging resonance-frequency physiology when it would help. Gao et al. (2026) made this point cleanly in their depression sample: the autonomic gains from ten weeks of HRV-BF showed up not in resting measurement but in the transfer stage, when participants were asked to engage their trained skill without real-time feedback. At rest, HRV-BF and active control looked similar. At transfer — when the skill was being deployed — HRV-BF participants produced substantially larger LF amplitude, SDNN, and HR max-min. The skill was real. It just needed to be asked for.

That is the cleanest summary of what HRV-BF actually trains. Not lower arousal. Not bigger HF at rest. A regulatory capacity that activates on demand.

5. Overview of the science base

The evidence base on HRV-BF is, by biofeedback standards, substantial. It is not uniform across indications, and the methodological quality of the underlying trials varies considerably. A fair summary requires distinguishing the domains where evidence is strong from those where it is thinner — and being honest about both.

Where the evidence is strongest: anxiety and depression. The Goessl et al. (2017) meta-analysis of HRV-BF for stress and anxiety, covering 24 studies and roughly 1,500 participants, found a large pre-to-post effect on self-reported stress and anxiety (Hedges' g ≈ 0.81), with significant effects retained in active-control comparisons. Subsequent randomized trials have continued to support efficacy in generalized anxiety disorder, depressive disorders, and mixed anxious-depressive presentations, though effect sizes against credible active controls are typically smaller than against waitlist or treatment-as-usual. The Gao et al. (2026) study I reviewed in detail in b202617 is a useful recent example: HRV-BF and an active progressive-muscle-relaxation comparator both reduced BDI-II and BAI scores comparably across five weeks, but only HRV-BF produced the deployable autonomic skill in the transfer stage.

The most methodologically careful recent contribution to the literature is the Minjoz et al. (2026) sham-controlled RCT in Biological Psychology (queued for upcoming NeuroBLOG coverage in BF-1). Sham-controlled trials are uncommon and difficult to execute in HRV-BF because credible blinding is hard — clients can usually tell whether they are being trained at their resonance frequency or not. The Minjoz design takes the methodological problem seriously and offers one of the cleanest comparisons of HRV-BF against a non-active feedback procedure. It is the kind of design the field has been quietly waiting for, and it deserves the careful read it will get when I work through it on its own.

Performance and stress management. HRV-BF has a robust applied literature in occupational and athletic settings — police, military, healthcare workers, executives, competitive athletes, musicians. The trials are smaller and the outcomes more variable, but the consistent finding is measurable improvement in subjective stress, perceived control, and physiological resilience indices, with effect sizes modest but durable.

PTSD. The PTSD literature on HRV-BF is promising but not yet mature. Multiple open trials and small RCTs show meaningful symptom reductions, particularly on hyperarousal and intrusion clusters, though the trials suffer from small samples, varying protocols, and inconsistent comparators. HRV-BF as part of a multi-component trauma protocol — alongside trauma-focused psychotherapy and sleep work — has stronger support than HRV-BF as a standalone intervention for PTSD.

Where the evidence is thinner: long-term durability beyond six months, comparative efficacy across populations (children, older adults, culturally diverse samples), dose-response relationships, and head-to-head comparisons with other autonomic-targeted interventions like mindfulness-based stress reduction.

Where the evidence is effectively silent: HRV-BF as a primer for neurofeedback. The pre-NF question — does HRV-BF, delivered before or alongside an EEG neurofeedback protocol, accelerate or stabilize EEG-side learning? — is the question I most want answered, and the published literature has barely addressed it. The clinical experience aligns with the mechanistic story (an unregulated autonomic system noises up the EEG learning loop; a trained one quiets it), but that alignment is suggestive, not meta-analytic. The next decade of biofeedback-and-neurofeedback research could productively spend a substantial fraction of its energy on this question. We will come back to it in Section 7.

6. Strengths and weaknesses

What does HRV-BF do well, and where does it fall short?

Strengths. The mechanistic story is unusually clean. Baroreflex amplification, vagal tone enhancement, and autonomic flexibility are all measurable, replicable across labs, and tied tightly to a defined intervention. The training is portable: once the resonance frequency is identified, the technique can be practiced anywhere, with or without a device, by anyone old enough to follow paced breathing. The intervention is low-risk. There are no serious adverse events in the literature beyond occasional dizziness in early sessions, easily managed by adjusting breath depth. The skill, once acquired, persists with light maintenance practice. It generalizes across stressors. It pairs well with other modalities — psychotherapy, neurofeedback, mindfulness, exercise — without obvious conflicts. It is one of the few autonomic-targeted interventions with a published meta-analysis showing meaningful average effects against credible comparators. And it is cheap. Once the practitioner has a clinical-grade system, the per-client cost is essentially the time itself. Home practice can be supported by a $40 wearable.

Weaknesses. The training arc is long and demanding. Ten to twelve weeks of in-clinic sessions plus twice-daily home practice is a meaningful ask of a depressed, exhausted, or chronically stressed client — the very populations the intervention most aims to help. Dropout rates and home-practice fidelity can become real clinical issues. The intervention is technique-sensitive in the early sessions: practitioners who do not identify resonance frequency individually, who do not check that the client is breathing diaphragmatically rather than thoracically, and who do not enforce the home-practice schedule are doing a substantially weaker version of the protocol than the literature reports. Beta-blockers, certain antiarrhythmics, and a few other medications attenuate the HRV gains the protocol can produce — Gao et al. (2026) had 83% of their HRV-BF participants on beta-blockers, and the autonomic effect sizes survived anyway, but in routine clinical work medication should be part of the conversation about expected response. The evidence base is meaningful in average but limited in stratification: predictor research — who responds, who does not, why — is only beginning to mature. And HRV-BF is not a substitute for the foundational treatments where they belong. It is not an antidepressant, not a trauma therapy, not a panic-disorder cure. It is a regulatory training that supports those treatments and that, in some clinical configurations, accelerates them.

A common practitioner failure: treating HRV-BF as if it were paced breathing with a fancier price tag. The acute state effects of paced breathing are real, can be delivered in a single session, and require no instruments. The training effects of HRV-BF — the durable, deployable autonomic skill — require the full multi-session protocol, the individually identified resonance frequency, and the home practice. Confusing the two undersells the modality and oversells the technique.

7. Brendan's perspective

The single sentence I most want to land in this post: HRV biofeedback is the most under-deployed lever in our field.

That is not a hit on neurofeedback practitioners. I am one of them. Most of us were trained in a tradition that put the EEG first and the autonomic substrate somewhere downstream — assumed, occasionally addressed, rarely targeted. Brainwave training was the differentiator. The autonomic story was a sentence in the intake interview.

What the past decade of clinical practice has gradually taught a lot of us — and what Part 1 of this series was trying to crystallize — is that the autonomic story is not downstream of the EEG story. It is upstream of it. An unregulated autonomic system noises up the EEG learning loop in several distinct ways: hyperarousal disrupts the sustained attention operant conditioning requires, dysfunctional breathing drags the EEG toward states the protocol is trying to move away from, and chronically dysregulated vagal tone leaves the system without the recovery capacity that makes between-session consolidation possible. The EEG protocol is being trained inside a physiology that may or may not be cooperating. If it is not cooperating, the learning curve flattens. If it is cooperating, the learning curve steepens.

Practitioners who add HRV-BF before or alongside neurofeedback often see faster, more durable EEG-side learning. I want to be specific about the epistemic status of that observation: it is clinical experience that aligns with the mechanistic story, not a meta-analytic claim. The controlled comparison — neurofeedback alone versus HRV-BF plus neurofeedback — has barely been studied. Until that work happens, the case for HRV-BF as a NF primer rests on three legs: a coherent mechanistic story, a strong general HRV-BF evidence base, and consistent practitioner experience. Those three legs are enough to make HRV-BF a defensible default for a wide range of clients. They are not enough to make it mandatory.

This is the brand-voice tripwire I want to be careful around. HRV-BF being the most under-deployed lever in our field is not the same as saying your protocol fails without it. Neurofeedback works for many clients without HRV-BF in the mix. Most under-deployed is a claim about the value of adding it. Mandatory would be a claim I cannot defend and would not want to.

The NeuroMASTER applications where HRV-BF earns its keep most clearly sit in Modules 2 and 3 — anxiety, panic, depression, PTSD, burnout, trauma — where autonomic regulation is centrally implicated and the symptom presentations themselves often interfere with EEG-side learning. It also has cross-module relevance in Module 6 performance work, sleep and recovery applications, and the general practice foundations that should sit underneath any of the more specialized work.

The line worth carrying forward into the rest of this series: the autonomic system is not the noise floor of neurofeedback. It is the substrate of neurofeedback. And substrate is the kind of thing you train deliberately, with a defined protocol, when you take it seriously.

8. Would I integrate HRV-BF into my NF practice? In what context?

Yes. I do with pretty much everyone. And the in what context question is where this section earns its keep as Part 2 of a series rather than a generic HRV-BF primer.

The honest clinical answer has three layers — who gets HRV-BF first, who gets it concurrent with neurofeedback, and what the actual workflow looks like in practice.

Who I would offer HRV-BF first, before any neurofeedback. The client whose presenting picture is dominated by autonomic dysregulation — chronic hyperarousal, panic-spectrum anxiety, somatic symptoms of stress (GI involvement, muscle tension, breath-pattern dysfunction), burnout with persistent fatigue-arousal mixing, PTSD with prominent autonomic features. The client whose intake interview makes it clear that they cannot sit still for forty-five minutes without their physiology pulling them out of whatever they are trying to attend to. The client who reports that their body is busy — heart racing in meetings, breath catching during sleep onset, GI churn around social anxiety. For these clients, asking the EEG to do the regulatory work before the autonomic substrate is stabilized is asking the wrong lever to lift the wrong load.

A typical pre-NF HRV-BF course for one of these clients runs eight to twelve sessions over six to ten weeks, with the standard protocol: resonance frequency identification in session one, training sessions twice weekly for the first three to four weeks, then weekly. Once the deployable skill is in place — measurable transfer-stage LF amplitude gains, subjective reports of using the technique outside the clinic, home-practice fidelity reasonably consistent — we move to the neurofeedback protocol with the autonomic substrate trained underneath it. In my experience the NF learning curve in these clients runs visibly cleaner after the HRV-BF foundation, and the maintenance is more durable.

Who I would offer HRV-BF concurrent with NF, rather than as a primer. The client whose presenting picture is more cognitive or behavioral than autonomic — ADHD with or without prominent emotional dysregulation, attention or memory difficulties without strong somatic anxiety, performance optimization in a relatively well-regulated baseline. For these clients the autonomic substrate is functional enough that the EEG can do its work; layering HRV-BF in parallel adds value at the margins (better between-session recovery, slightly more durable learning, useful self-regulation skill the client gets to carry out of the clinic) without earning the time cost of a full pre-NF course. A common configuration: a brief HRV-BF orientation in the first or second session — resonance frequency identification, basics of the technique, three to five minutes of practice — followed by short HRV-BF segments tucked into NF sessions and a lighter home-practice schedule. The skill is built more slowly but in parallel with the EEG work.

Who I would not push HRV-BF on. The client who is already over-protocoled. The depressed, exhausted, anhedonic client whose engagement reserves are the limiting factor — for whom adding ten weeks of twice-daily home practice is more likely to produce dropout than skill acquisition. Gao et al.'s (2026) responder data is the clearest published version of this clinical caution: the patients with severe anhedonia and energy depletion were the non-responders, and HRV-BF added on top of the wrong baseline does not earn its time cost. For these clients, foundational work first — psychiatric care, behavioral activation, sleep stabilization, low-dose engagement-building — and HRV-BF as a later move when the substrate of engagement is restored.

The actual workflow. Five handles worth being concrete about, because would I integrate this is a different question from how exactly:

In-clinic vs home practice ratio. The standard is a heavier in-clinic schedule for the first three to four weeks (twice weekly), then tapering to weekly, with home practice running continuously across the whole course. The in-clinic sessions front-load the training of correct technique, the resonance frequency targeting, and the early skill acquisition. The home practice is what consolidates the skill and gives it the dose for autonomic remodeling. Skip either piece and the protocol underperforms.

App-based vs instrument-based home practice. I will be honest — the consumer landscape has improved dramatically in the past five years. Wearable HRV trackers (Optimal HRV Monitor, Polar H10, Garmin chest straps, the Eureka belt, the Oura ring with HRV-coherence apps) now provide clinically usable feedback for home practice, with a $40 to $300 entry point. They are not as precise as a ProComp Infiniti, but they are precise enough for paced breathing at a known resonance frequency. The clinic still does the resonance frequency identification and skill calibration; the home practice runs on a consumer device. This is a meaningful improvement on the old configuration where home practice was either unmonitored or required an expensive home unit.

How many sessions before checking that resonance frequency has stabilized. My rule of thumb: re-check resonance frequency at session four and at session eight. Resonance frequency can drift in early training — typically toward slightly slower breathing as the diaphragmatic technique consolidates — and the protocol works substantially better if the breath pace tracks the actual resonance rather than the session-one estimate. If the client is on beta-blockers or has a cardiovascular condition that affects baseline HRV, more frequent rechecks are reasonable.

What I track and when I stop. LF amplitude and HR max-min during the transfer run, subjective reports of using the technique outside the clinic, and home-practice fidelity from the wearable data. Resting HF is the least informative of the routine metrics — Gao et al. (2026) made this point cleanly. I close out the course when the transfer-stage skill is in place, the client reports usable deployment outside the clinic, and the home-practice habit is established (the success case, week eight to ten); or, if four to six sessions in there is no measurable LF gain at transfer and no subjective uptake, I do a sequencing reset rather than a heroic continuation.

Where it sits in the NF formulation. HRV-BF is not the centerpiece of a neurofeedback practice. It is the substrate that makes the centerpiece work. Clients who understand HRV-BF as substrate rather than as primary intervention engage with it more reliably and integrate it more durably.

The series logic here is not subtle. Part 1 named the three-layer framework and identified state regulation as the foundation. Part 2 unfolds the foundation. Parts 3 through 7 will unfold the rest — skin conductance and arousal profiling next, then the interference layer (sEMG and peripheral temperature), then the external modulation entries (AVE, tVNS, rTMS and PBM). Each is a single lever. HRV-BF is the one with the strongest evidence base, the cleanest mechanistic story, the lowest barrier to integration, and the most consistent payoff. Start there.

Conclusion

HRV biofeedback is not a calmness intervention. It is a self-regulation skill — a deployable autonomic capacity, trained over weeks, that activates when the system needs it and rests when it does not. The mechanism is unusually clear (baroreflex amplification, cardiac vagal tone, autonomic flexibility). The evidence base is real (anxiety, depression, performance, PTSD — strong; durability and dose-response — thinner). The clinical fit for neurofeedback practice is the cleanest of all the adjunctive modalities mapped in Part 1, and the case for using it as a substrate for EEG work rests on a coherent mechanistic story, consistent practitioner experience, and a partial empirical literature that has not yet caught up with what clinicians have been observing.

For NF practitioners considering whether to integrate HRV-BF: yes, with care, with attention to who is ready for it and who needs foundational work first, with the standard protocol (individualized resonance frequency, twice-weekly in-clinic for the first month tapering to weekly, twice-daily home practice for ten to twelve weeks), with grading on the transfer run rather than resting measurements, and with the framing that HRV-BF is substrate rather than centerpiece.

The autonomic system is not the noise floor of neurofeedback. It is the substrate of neurofeedback. And the substrate, when trained deliberately, makes everything that sits on top of it work better.

That is what under-deployed looks like — a lever sitting in plain sight, doing real work for the practitioners who use it, and waiting for the rest of the field to notice.

References

• Gao, C., Mather, M., Liao, H.-Y., Lee, K.-L., Yeh, Y.-C., Ke, C. L. K., Yen, C.-F., & Lin, I.-M. (2026). Baseline depressive symptoms and heart rate variability indices predict HRV biofeedback outcomes in young adults with depression. Applied Psychophysiology and Biofeedback. https://doi.org/10.1007/s10484-026-09785-7

• Goessl, V. C., Curtiss, J. E., & Hofmann, S. G. (2017). The effect of heart rate variability biofeedback training on stress and anxiety: A meta-analysis. Psychological Medicine, 47(15), 2578–2586. https://doi.org/10.1017/S0033291717001003

• Lehrer, P. M., & Gevirtz, R. (2014). Heart rate variability biofeedback: How and why does it work? Frontiers in Psychology, 5, Article 756. https://doi.org/10.3389/fpsyg.2014.00756

• Minjoz, S., Jeanne, R., Pellissier, S., & Hot, P. (2026). Psychophysiological effects of heart rate variability biofeedback versus sham: A randomized controlled trial. Biological Psychology.