Abstract
High ventilation breathwork with retention (HVBR) might be beneficial for mental/physical health and has been growing in popularity over the past decade. However, little research has investigated the potential therapeutic effects of brief, remotely delivered HVBR and the tolerability profile of this technique. This study (NCT06064474) was the largest blinded randomised-controlled trial on HVBR to date in which 200 young, healthy adults balanced for gender were randomly allocated in blocks of 2 by remote software to 3 weeks of 20 min daily HVBR (fast breathing with long breath holds) or a placebo HVBR comparator (15 breaths/min with short breath holds). The trial was concealed as a ‘fast breathwork’ study wherein both intervention and comparator were masked, and only ~ 40% guessed their group assignment with no difference in accuracy between groups. Both groups reported analogous credibility and expectancy of benefit, subjective adherence, positive sentiment, along with short- and long-term tolerability. At post-intervention (primary timepoint) for stress level (primary outcome), we found no significant group x time interaction, F(1,180) = 1.98, p = 0.16, ηp2 = 0.01, d = 0.21), nor main effect of group, (F = 0.35, p = 0.55, ηp2 < 0.01) but we did find a significant main effect of time, (F = 13.0, p < 0.01, ηp2 = 0.07). There was a significant improvement in stress pre-post-intervention in both groups, however there was no significant difference in such improvement across groups. In addition to stress at follow-up, we found no significant group x time interactions for secondary trait outcomes of anxiety, depression, mental wellbeing, and sleep-related impairment. This was also the case for state positive and negative affect after the first session of breathwork and at post-intervention. Brief remote HVBR therefore may not be more efficacious at improving mental health than a well-designed active comparator in otherwise healthy, young adults. Funders: The Ryoichi Sasakawa Young Leaders Fellowship Fund, and PsyDAO.