Abstract
Understanding why interventions work is essential to optimizing them. Although mechanistic theories of meditation-based interventions (MBIs) exist, empirical evidence is limited. We randomly assigned 662 adults (79.9% reported clinical levels of anxiety or depressive symptoms) to a four-week smartphone-based MBI or wait-list control condition early in the COVID-19 pandemic. Psychological distress and four theory driven preregistered psychological mediators of well-being (mindful action, loneliness, cognitive defusion and purpose) were assessed five times during the intervention period and at three-month follow-up. In preregistered analyses, assignment to the intervention predicted significant gains on all mediators which in turn significantly mediated follow-up distress (21.9%-62.5% of intervention effect on distress). No significant mediation pathway was observed in an exploratory multiple mediator analysis, but reduced loneliness accounted for 61.7% of the combined indirect effect. Multiple psychological pathways may mediate reduced distress in a digital MBI.