Evidence Trace, complete trial record
In plain terms
In plain terms: HRV Biofeedback has some human trial evidence supporting it, but the evidence is not yet strong enough for EBL to call it established. It is worth knowing about and worth discussing with your clinician.
HRV Biofeedback
B
Supported: a good trial, backed by more
The endpoint this grade is based on: self_report
The evidence as the record states it
Lehrer et al. (Applied Psychophysiology and Biofeedback, 2020; PMID 32385728): meta-analysis restricted to randomly controlled studies, 58 included; HRV biofeedback beat both inactive and active controls, with the largest effects on anxiety, depression and anger. Minjoz et al. (Biological Psychology, 2026; PMID 41905438): 47 adults, HRV biofeedback against sham biofeedback; psychological distress improved. Goessl et al. (Psychological Medicine, 2017; PMID 28478782) pooled 24 studies, 484 participants, not stated to be randomized, and found a large reduction in self-reported stress and anxiety. Vann-Adibe et al. (Applied Psychophysiology and Biofeedback, 2026; PMID 41310318): remote HRV biofeedback, 18 studies; no significant effect on stress. Grade B on self-reported anxiety, which caps at B.
PMID 32385728 PMID 41905438 PMID 41310318 PMID 28478782
Evidence by specific use
Self-reported anxiety, HRV biofeedback (resonance breathing) training
Grade B
Lehrer et al. (Appl Psychophysiol Biofeedback, 2020; PMID 32385728) and Minjoz et al. (Biol Psychol, 2026; PMID 41905438). Self-report caps the letter at B.
PMID 32385728 PMID 41905438
Vann-Adibe et al. (Appl Psychophysiol Biofeedback, 2026; PMID 41310318) found no significant effect on stress; Goessl et al. (Psychol Med, 2017; PMID 28478782) found a large effect from studies not stated to be randomized. Mixed: C.
PMID 41310318 PMID 28478782
Why Grade B and not Grade A
Grade B reflects moderate human evidence -- typically a single strong trial, multiple smaller trials, or surrogate endpoint data without hard clinical outcome replication. To reach Grade A, HRV Biofeedback would require at least one additional independent randomized trial demonstrating a hard clinical endpoint or objectively measured functional outcome in a comparable population.
What this grade covers and what it does not
Grade B on self-reported anxiety. A meta-analysis of randomized controlled studies (Lehrer et al., 2020; PMID 32385728) and a sham-controlled trial (Minjoz et al., 2026; PMID 41905438) agree. WHY NOT A: self-reported anxiety caps the grade. WHAT THIS GRADE DOES NOT COVER: perceived stress, where the pooled results disagree and the use is graded C; wearable accuracy; and clinical cardiovascular outcomes.
What would move this grade
An independent randomized trial demonstrating a hard clinical endpoint -- mortality, major adverse event, or objectively confirmed functional improvement -- in an adequately powered sample, with results independently replicated, would support Grade A.
Citations withdrawn from this record
PMID 29548298, withdrawn July 2026A CLINICAL TRIAL PROTOCOL for spider-fearful individuals. A protocol has no results and cannot support a grade in either direction.
PMID 32713257, withdrawn October 2026Miller et al. 2020 validated a wrist strap against polysomnography for sleep measurement in 12 adults. It says nothing about biofeedback and supports no letter.
How to track whether this is working
Resting HRV (RMSSD) using a validated device (Polar H10, Whoop, Oura) at baseline and monthly. The PSS-10 or STAI at baseline and 8 weeks for stress and anxiety endpoints. Resonance frequency breathing is occurring during sessions: this varies between 4.5 and 7 breaths/minute and must be identified individually.