Abstract:  Mindfulness meditation (MM) has enjoyed a growing popularity in healthcare in recent years when bio-psycho-social approaches are becoming more and more emphasized in modern medicine. There has been mounting empirical evidence showing MM’s significant effectiveness in alleviating anxiety for both nonclinical and clinical populations. However, the effect size of the available empirical investigation results has remained inconsistent and possible moderators have yet to be explored comprehensively. In order to determine the immediate and long-term efficacy of MM in overcoming anxiety, we conducted a meta-analysis based on a systematic and comprehensive review of the published studies on mindfulness-based interventions for anxiety. We also examined whether some characteristics of research participants (e.g. age, geographic areas) and interventions (e.g. format, duration, at-home practice), and specifics of the study (i.e. types of control, quality of the study) and data analysis (e.g. attrition rate) moderate the magnitude of the effectiveness of MM interventions (11variables). The review was performed following the rigorous PRISMA Protocol. Published studies using randomized controlled trial were selected from major databases worldwide to include investigations conducted in both Western and Eastern countries. Databases used include VIP Journal Integration Platform, Wan fang, China National Knowledge Infrastructure, Scopus, Web of Science, Embase, the Cochrane Library, and PubMed. Keywords used are mindfulness, meditation, MBSR, MBCT, anxi*, mood, intervention, therapy, program. Using the random effect model, we pooled the effect size (Hedge’s g), and conducted a publication bias evaluation, a moderating effect analysis and an interaction analysis in CMA 3.3. Results of our analyses revealed fifty-five RCTs from both Eastern and Western countries (k = 68 samples, N = 4595 participants). Mindfulness meditation for treating anxiety is shown to be efficacious with a medium to large immediate effect (g = 0.60), but the effect is not reliably shown at follow-up assessments. The post-heterogeneity test result suggests that using the random effect model is reasonable. Univariate meta-regression analysis yielded that study quality, geographic areas, participants’ age, intervention format (Group vs. Individual), amount of at-home practice, and attrition rate shown in data analysis remarkably influenced the effect size of MM’s immediate effect, while types of control, health condition, mindfulness practice experience, intervention duration, or statistical analysis methods used (ITT vs. PP) did not appear to moderate MM’s immediate effectiveness for reducing anxiety. Additionally, geographic areas affect the effect size the most. Multiple meta-regression models suggested that type of control and geographic areas, as well as statistical analysis methods significantly moderate the effect size of intervention effectiveness. Overall, the study results demonstrated high immediate effect of mindfulness-based practices for alleviating anxiety, but the effect did not seem to last. In addition, geographic area turns out to be the strongest moderator, and practitioners in the East seem to benefit more than those who are in the West. Study quality, participants’ age, mindfulness practice experience, intervention format, at-home practice quantity and attrition rate also affect the effect size to a certain extent. Future research is warranted to improve methodological quality of outcome studies, to provide more clear and replicable evidence of MM efficacy, and to explore more underlying moderators for the intervention effect size, such as participant satisfaction and so forth.

Key words: mindfulness meditation, anxiety, intervention, moderating effect, cultural differences, meta-analysis