2026 ECNP Congress

PS04-4028

Towards personalised biofeedback: assessing individual differences in optimal respiratory frequency within a virtual reality paradigm

Benjamin Cook — G.Y. Leung, A.M. Wiley, A. Brendler, N. Subramaniam, T. Woolley, C. Powell, S. Sitnikov, D. Matthews, P.C. Fletcher · University of Cambridge, Department of Psychiatry, Cambridge, United Kingdom

2026 ECNP Congress

Placing respiration under volitional control is an effective means of modulating autonomic nervous system activity, with upstream effects on psychological functioning [1]. Slow-paced breathing (SPB) thus forms an integral part of biofeedback. Most protocols implement predefined frequency instructions, with the aim to maintain respiration near ‘resonance’ (~0.1Hz) for augmenting vagal tone [2], improved cognition, and emotional regulation [3]. Whilst successful, this standardised approach may fail to achieve optimal psychophysiological effects, as resonance (i.e. where respiratory, cardiac, and vascular rhythms become temporally aligned) is determined by the unique biological properties of one’s vascular system [4], and subsequently, optimal respiratory frequency for maximal autonomic efficiency varies across individuals [5]. Research has recently endeavored to explore the potential value in individually tailored SPB protocols which incorporate pre-training optimal respiratory frequency assessments, obtaining mixed conclusions on efficacy. To quantify individual differences in optimal respiratory frequency within an engaging and immersive context, we used a custom-designed virtual reality (VR) biofeedback game developed in collaboration with Ninja Theory Ltd, with the aim to inform personalised SPB protocols for therapeutic applications.

86 participants experienced a VR-based biofeedback game, in which they were serially presented with instructions for discrete, counterbalanced breathing frequencies spanning 4-6.7 breath cycles/min.  During each condition, participants were guided through inhale-hold-exhale cycles within a relaxing boat environment with written and verbal prompts, and the game dynamics adjusting to the participant’s real-time physiological input. Respiration, photoplethysmography-derived heart rate variability, and electrodermal activity were recorded continuously, with post-condition visual analogue scales assessing subjective experience. Natural log transformed (ln) SDNN was baseline corrected relative to a pre-task phase, and then participant’s optimal breathing condition was identified as the condition producing their largest lnSDNN response. To examine between-participant variability, the distribution of optimal frequencies was compared against a uniform distribution using a chi-squared goodness-of-fit test. To confirm breathing condition influenced HRV response, a repeated-measures ANOVA compared baseline-corrected lnSDNN across permutations. To quantify within-participant optimisation, optimal-condition lnSDNN was compared the mean response across the six non-optimal conditions using a Wilcoxon signed-rank test.

Following the exclusion of extreme outlying participants (+/- 3SD), the distribution of lnSDNN-defined optimal respiratory frequencies from across all seven breathing conditions did not significantly depart from uniformity, χ²(6) = 7.75, p = .257, suggesting no evidence for a single consistently optimal respiratory frequency. A repeated-measures ANOVA showed a significant effect of breathing condition on baseline-corrected lnSDNN, Greenhouse–Geisser corrected F(4.13, 330.19) = 4.05, p = .003, ηp² = .048. When comparing participants’ lnSDNN-defined optimal respiratory frequency against their mean response across the six non-optimal conditions, baseline-corrected lnSDNN was significantly greater during optimal than non-optimal frequencies, Wilcoxon signed-rank test (V = 3321, p < .001).

Preliminary analyses suggest that optimal respiratory frequency varies between individuals, and that no single breathing frequency is consistently optimal for maximising HRV. Moreover, within-participant analyses showed that baseline-corrected lnSDNN was significantly greater during each participant’s optimal frequency. These findings support the value of identifying individual optimal respiratory frequency prior to biofeedback training, as this may maximise HRV response and enhance associated regulatory effects.

References

  1. Thayer, J.F., Lane, R.D., 2000. A model of neurovisceral integration in emotion regulation and dysregulation. Journal of Affective Disorders 61, 201–216. https://doi.org/10.1016/S0165-0327(00)00338-4
  2. Lehrer, P.M., Vaschillo, E., Vaschillo, B., 2000. Resonant frequency biofeedback training to increase cardiac variability: rationale and manual for training. Applied Psychophysiology and Biofeedback 25, 177–191. https://doi.org/10.1023/A:1009554825745
  3. Mather, M., Thayer, J.F., 2018. How heart rate variability affects emotion regulation brain networks. Current Opinion in Behavioral Sciences 19, 98–104. https://doi.org/10.1016/j.cobeha.2017.12.017
  4. Shaffer, F., Meehan, Z.M., 2020. A practical guide to resonance frequency assessment for heart rate variability biofeedback. Frontiers in Neuroscience 14, 570400. https://doi.org/10.3389/fnins.2020.570400
  5. Sévoz-Couche, C., Laborde, S., 2022. Heart rate variability and slow-paced breathing: when coherence meets resonance. Neuroscience and Biobehavioral Reviews 135, 104576. https://doi.org/10.1016/j.neubiorev.2022.104576