This meta-analysis of randomized controlled trials (RCTs) on the Transcendental Meditation® (TM) technique updates previous meta-analyses and assesses the effects of initial anxiety level, age, duration of practice, regularity of practice, research quality, author affiliation, and type of control group on effect sizes. This systematic review of the literature used the Comprehensive Meta-Analysis (CMA) program for core analyses of effect sizes, bias analysis, meta-regression, and moderator variable analysis. Comprehensive literature searches included databases devoted to meditation research. More than 600 TM research papers were identified; 14 of these addressed trait anxiety and reported on 16 studies among 1295 participants with diverse demographic characteristics. No adverse effects were reported. The standardized difference in mean, d, for the TM technique compared with controls receiving an active alternative treatment (10 studies) was d=-0.50 (95% CI, -.70 to -0.30; p=0.0000005). Compared with controls receiving treatment as usual (wait list or attention controls, 16 studies), d=-0.62 (95% CI, -0.82 to -0.43; p=1.37E-10). Meta-regression found that initial anxiety level, but not other variables, predicted the magnitude of reduction in anxiety (p=0.00001). Populations with elevated initial anxiety levels in the 80th to 100th percentile range (e.g., patients with chronic anxiety, veterans with post-traumatic stress disorder, prison inmates) showed larger effects sizes (-0.74 to -1.2), with anxiety levels reduced to the 53rd to 62nd percentile range. Studies using repeated measures showed substantial reductions in the first 2 weeks and sustained effects at 3 years. Overall, TM practice is more effective than treatment as usual and most alternative treatments, with greatest effects observed in individuals with high anxiety. More research is needed in this area, especially with high-anxiety patients, conducted under medically supervised conditions.
