UCSF

Objective

Stress is a driver of depression, and people with depression often struggle to cope with stress and anxiety. This study directly compares the mental health effects of a Wim Hof Method intervention to an active control condition (slow breathing) in women with high stress and high depressive symptoms.

Methods

We randomized 84 healthy midlife women with high stress and high depressive symptoms to either: 1) the hormetic stress condition based on the Wim Hof Method (WHM) involving a breathing technique designed to induce intermittent hypoxia and cold showers (n = 41) or 2) an active comparison condition involving slow-paced breathing and warm showers (n = 43). We provided participants with daily audio instructions (15 min) for three weeks during the COVID-19 pandemic (2020-2021). Our primary outcomes were depressive symptoms, anxiety symptoms, and perceived stress collected at pre-intervention, post-intervention, and 3 months later. We also assessed daily stress rumination and affect with daily diary during the intervention, and participants completed a laboratory stressor, the Trier Social Stress Test, before and after the intervention, and provided samples for salivary cortisol reactivity.

Results

Participants in the active control condition perceived the intervention to be more credible and expected greater mental wellbeing benefits compared to those in the Wim Hof Method condition. Differential attrition was observed with six participants (7 %) dropping out -- all from WHM condition. Among the participants who completed the intervention, both groups improved on mental health outcomes immediately after the intervention with a 24 % reduction in depressive symptoms, a 27 % reduction in anxiety symptoms, and 20 % reduction in perceived stress. Improvements were maintained at the 3-month follow-up with 46 % of the sample reporting mild or no depressive symptoms. Participants in the WHM condition had significant reductions in rumination after daily stressful events compared to those in the active control group. Both conditions had reduced daily negative affect across the intervention and lower peak cortisol reactivity to the lab stressor post-intervention.

Conclusions

Counter to the preregistered predictions, and despite participants' differing expectations, the interventions led to equivalent reductions in depressive symptoms, anxiety symptoms, and perceived stress, which were sustained at three months. They also produced comparable reductions in cortisol reactivity and daily negative affect. However, the WHM condition was associated with greater reduction in reported rumination after daily stressful events than the active control, a finding that needs replication with larger and more diverse samples.