Mindfulness Eases Distress by Quietly Undoing Experiential Avoidance, Behavioral Trial Shows

Emotional distress is one of the most common burdens in modern life, and mindfulness-based interventions have become one of the most widely prescribed answers to it. Yet a stubborn question has shadowed the field for decades: what actually changes in the mind during the first weeks of mindfulness training, before the full eight-week program has run its course? A new randomized controlled trial, published in BMC Psychology by researchers at Peking University together with Stefan G. Hofmann of Philipps University Marburg, offers an unusually concrete answer. The team tracked experiential avoidance, the reflexive tendency to push away uncomfortable thoughts and feelings, using not just questionnaires but a behavioral reaction-time paradigm, and found that reductions in this avoidance pattern emerged early and statistically mediated the intervention’s effects on distress and anxiety.

Experiential avoidance is a core transdiagnostic construct in clinical psychology. Rather than being willing to remain in contact with unpleasant inner experiences, avoidant individuals engage in efforts to escape or suppress them, from rumination and suppression to substance use and procrastination. Decades of research have linked high avoidance to anxiety, depression, and general emotional dysfunction, and acceptance-oriented therapies such as acceptance and commitment therapy explicitly target it. Mindfulness-based interventions, including mindfulness-based stress reduction, are thought to work partly by loosening this avoidant stance, training participants to observe sensations, emotions, and thoughts with curiosity instead of fleeing them. But most evidence for this mechanism has relied on self-report questionnaires, which are vulnerable to demand effects, social desirability, and the simple fact that people often cannot introspect accurately on their own avoidance habits.

The new study, led by Zitong Xin, Mengyao He, Jialu Jin, and Xinghua Liu, was designed to break that dependence on introspection. The researchers recruited 81 participants experiencing high emotional distress and first validated a behavioral task intended to measure avoidance objectively. In the image rating task, participants viewed pictures drawn from the International Affective Picture System, a standardized library of emotionally evocative images, and rated them while the experimenters recorded reaction times. The logic is elegant: people who habitually avoid negative experiences should respond differently to negative versus neutral images, and the difference in reaction times between the two categories should track their avoidance tendencies.

That is precisely what the baseline data showed. Participants responded significantly faster to negative images than to neutral ones, a p < .001 difference that the authors interpret as an avoidance-related response pattern, consistent with the idea that emotionally threatening stimuli capture and mobilize attention in avoidant individuals. More importantly for the study’s aims, the size of the reaction-time difference between negative and neutral images correlated significantly with scores on established self-report measures of experiential avoidance, including the Acceptance and Action Questionnaire-II and the Brief Experiential Avoidance Questionnaire, with correlations reaching p < .01. In other words, the behavioral paradigm showed convergent validity: it measured the same underlying construct that the questionnaires measure, but through performance rather than self-description.

With the behavioral measure validated, the team moved to the intervention phase. Participants were randomly assigned either to an eight-week online Mindfulness Intervention for Emotional Distress, abbreviated MIED, with 41 participants, or to a waitlist control group of 40. The trial was prospectively registered at ClinicalTrials.gov as NCT05891015, and the researchers analyzed the data on an intent-to-treat basis. Crucially, the team measured everything twice: once at baseline and again at week 3, only three weeks into the eight-week program. This early time point is the study’s methodological signature, because it allows the researchers to ask whether the hypothesized mechanism, reduced avoidance, changes before the clinical outcomes do, which is what a mediational account requires.

The results at week 3 were striking across the board. Compared with the waitlist controls, participants in the MIED group showed significant reductions in general emotional distress as measured by the Kessler Psychological Distress Scale, in anxiety as measured by the Beck Anxiety Inventory, and in depression as measured by the Beck Depression Inventory-II, all at p < .01. Just as importantly, the mindfulness group also showed significant reductions in both self-reported and behaviorally measured experiential avoidance at the same early time point. Three weeks of online mindfulness training had already shifted the reaction-time signature of avoidance, not merely the way participants described themselves on questionnaires.

The mediation analyses then delivered the study’s central finding. Reductions in the behavioral measures of experiential avoidance partially mediated the effects of the mindfulness intervention on reductions in general emotional distress and anxiety at week 3. In plain terms, part of the reason the intervention reduced distress and anxiety so quickly was that it had already reduced the participants’ avoidant response pattern as captured by the reaction-time task. Notably, the behavioral avoidance measure did not mediate the reduction in depression, a result the authors report transparently and which suggests that depression may respond to mindfulness through additional or alternative pathways that unfold on longer timescales.

The technical significance of this design deserves emphasis. Mechanism studies in psychotherapy research are notoriously difficult because self-report measures of a mechanism can change simply because participants expect the therapy to change them, a confound that behavioral measures largely sidestep. By validating the image rating task against the AAQ-II and BEAQ at baseline and then deploying it inside a randomized controlled trial, the researchers created a chain of evidence: the task measures avoidance, the intervention changes avoidance, and the change in avoidance statistically accounts for part of the change in symptoms. The partial mediation pattern, emerging at week 3 of an 8-week program, positions experiential avoidance as an early mechanism rather than a late byproduct of symptom improvement.

The practical implications are equally compelling. If avoidance is an early and measurable target, then clinicians and digital health developers can monitor it in real time, using brief behavioral tasks rather than lengthy questionnaires, and identify participants who are not engaging with the mechanism during the first weeks of an online program. That opens the door to adaptive interventions, where non-responders receive booster sessions or alternative exercises before dropping out, addressing the adherence problems that plague digital mental health. The findings also suggest that the acceptance component of mindfulness, the willingness to remain in contact with discomfort, is not decorative but mechanistically load-bearing.

The authors are careful to frame the evidence as preliminary. The sample of 81 participants, the waitlist control design, and the single behavioral paradigm all leave room for refinement, and the mediation is partial rather than complete, meaning other mechanisms certainly contribute. Still, the study stands out for bringing a randomized controlled trial, a behavioral paradigm, and an online intervention together in a single test of how mindfulness works. As online mental health interventions proliferate globally, knowing not just whether they work but what changes first inside the mind may prove to be the difference between another wellness app and a genuinely mechanistic therapy.

Subject of Research: Psychology

Article Title: Experiential avoidance as an early mechanism of an online mindfulness-based intervention for emotional distress: evidence from a behavioral paradigm in a randomized controlled trial

Article References: Xin, Z., He, M., Jin, J., Hofmann, S. G., & Liu, X. (2026). Experiential avoidance as an early mechanism of an online mindfulness-based intervention for emotional distress: evidence from a behavioral paradigm in a randomized controlled trial. BMC Psychology. https://doi.org/10.1186/s40359-026-05121-0

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05121-0

Keywords: mindfulness, experiential avoidance, emotional distress, randomized controlled trial, behavioral paradigm, anxiety, depression, online intervention, mechanism research, mental health, Experiential, avoidance