Perceived stress reduction accounts for mindfulness intervention effects on anxiety and depression

Study 2 enrolled 607 participants (mean age 30, 82% female) with similar inclusion criteria, randomizing 304 to MBI and 303 to waitlist, with measurements at weeks 3, 5, and postintervention. Both studies measured perceived stress, anxiety, and depression using validated scales. Linear mixed-effects models showed that the MBI group experienced significantly greater reductions across all three measures compared to waitlist. In Study 1, mediation analysis found that changes in perceived stress explained part of the intervention's effect on both anxiety and depression. Study 2, using more sophisticated latent growth curve modeling, found that perceived stress changes accounted for substantial proportions of the MBI effect—meaning much of how mindfulness helped came through reducing how stressed people felt. Cross-lagged analyses further revealed that perceived stress acted as a temporally prior process for depression, suggesting that early reductions in perceived stress preceded and predicted later improvements in depressive symptoms, whereas the relationship with anxiety was more bidirectional. The integrated evidence suggests perceived stress is a key mechanism through which MBI alleviates emotional distress.

The two-trial design strengthens confidence in the findings. Study 2's frequent measurement (weeks 3, 5, postintervention) clarified the temporal sequence: perceived stress reductions appeared early and preceded depression improvement, whereas anxiety showed reciprocal change with stress across the intervention timeline. This suggests different pathways—depression improvement may flow primarily from reduced stress perception, while anxiety and stress reciprocally reinforce each other during mindfulness practice. The effect sizes were clinically meaningful: in both studies, the between-group differences in perceived stress, anxiety, and depression all favored the MBI group substantially compared to waitlist. A potential limitation is that both samples were predominantly female (82–83%) and relatively young (mean age ~30), so findings may not fully generalize to older or more balanced-gender populations. Additionally, while the mediation analyses were rigorous, they cannot prove causation—only that stress reduction and emotional improvement co-occur and follow a plausible temporal order.