An Online Mindfulness Intervention to Reduce Stress and Anxiety Among College Students

Abstract

This study evaluated the efficacy of two different web-based stress management programs among college students at a large Midwestern university. After completing the pretest, students (N = 401) were randomly assigned to a Mindfulness plus Present Control intervention, a Mindfulness only intervention, or a Stress-information only comparison group. Primary outcomes were stress, anxiety, depression, perceived stress and worry; hypothesized mediators of intervention efficacy were rumination, mindfulness and present control. Self-report measures were completed online at pre-intervention, post-intervention, first follow-up (2-3 weeks postintervention) and second follow-up (4-5 weeks postintervention). Ninety percent of the sample (n = 365) completed the pretest and comprised the intent-to-treat sample. Linear mixed modeling was used to assess significant change over time and hierarchical regression was used to test for mediation. Participants in all three groups reported significant decreases in all five primary outcomes across all time points (within group d's = -.15 to -.56). All time by intervention group interaction effects were non-significant suggesting that the three conditions were equally effective. With regard to the mediators, participants reported significant increases in present control and mindfulness and significant decreases in rumination from baseline to post-intervention and both follow-ups (within group d's = .01 to .71). There was one significant time by intervention group interaction effect in the analyses assessing change over time in the mediators specifically indicating a between-group difference in changes in rumination, F(8, 973) = 3.73, p = .0003. In this case, the Mindfulness plus Present Control group reduced rumination significantly more than the comparison group. Because there were few differences across conditions, mediation analyses were performed collapsing across conditions. In general, changes in present control were associated with changes in depression and changes in rumination were associated with changes in worry and perceived stress at the second follow-up controlling for baseline scores. Limitations and future direction are discussed.