CIMH Study Shows High Ventilation Breathwork Cuts Brain Blood Flow by 45.5% in 30 Adults
Researchers examined 30 practitioners using fMRI to measure acute cerebral perfusion and connectivity changes during voluntary hyperventilation.

Researchers at the Central Institute of Mental Health, the Medical Faculty Mannheim of Heidelberg University, and the MIND Foundation examined the neurobiological effects of high ventilation breathwork in 30 experienced practitioners with a mean age of 44.97 ± 11.80 years. The study, published in Progress in Neuro-Psychopharmacology and Biological Psychiatry, tracked acute cerebrovascular shifts and resting-state functional connectivity using magnetic resonance imaging doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. During the sessions, participants engaged in continuous circular breathing inside the scanner for 36 to 96 minutes while listening to music.
Perfusion and Subjective Ratings
Arterial spin labelling revealed that high ventilation breathwork induced a 45.5% reduction in global grey matter cerebral blood flow doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. Perfusion fell from a baseline of 29.81 ± 7.58 to 16.24 ± 4.35 milliliter/100 g/min at peak intensity, alongside increased heart and respiratory rates
doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. Despite the physiological shift, statistical analyses showed that cerebral blood flow reductions were not associated with altered states of consciousness ratings assessed by the 5-Dimensional Altered States of Consciousness scale or real-time Likert ratings
doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗.
Brain Network Dynamics
Resting-state scans evaluated functional connectivity within the default mode network before and after the breathing exercises doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. At the group level, default mode network functional connectivity did not change significantly
doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. In exploratory analyses using global signal regression, greater decreases in default mode network functional connectivity predicted higher ratings on the altered states scale (β = -4.88, p < 0.05), though this association was not significant without global signal regression (p = 0.75)
doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗. Cerebral blood flow decreases did not correlate with functional connectivity shifts
doi.org
Progress in Neuro-Psychopharmacology and Biological Psychiatrydoi.orgOpen source ↗.



