Should We Prescribe Meditation to Our Patients?
Many people claim that meditation helps them reduce stress, anxiety, or depression, but little quality evidence exists to support those anecdotes. Add to that the difficulty in designing a controlled trial of meditation—eg, how can people be blinded to their treatment group?—and it’s hard to know how to counsel patients on the effectiveness of this strategy.
A recent meta-analReeysis examined 47 trials with a total of 3,515 participants to see if meditation was truly efficacious for a variety of conditions. Researchers included studies of various types of mindfulness and transcendental, or “mantra,” meditation for their influence on positive and negative affect, health-related behaviors, and outcomes like pain level or body weight.
All studies included a control group, defined as a program “matched in time and attention” but not involving meditative practice. Most trials were for short-term interventions (between four and 15 weeks), while a few were conducted up to five years. Twenty of these trials specifically enrolled subjects with psychiatric or substance abuse disorders, while the rest enrolled patients with a range of other medical problems.
So what were the results? When meditation was compared with a specific active control intervention (such as an exercise program or progressive muscle relaxation), no type of meditation showed any advantage over the control condition. When meditation was compared with a nonspecific control like an educational program, subjects engaged in mindfulness meditation (but not “mantra”-based meditation) showed improvement in anxiety (effect size 0.38 at eight weeks), depression (0.30), and pain level (0.33), but not in other measures.
These effect sizes are comparable to those found in antidepressant trials, particularly for mild to moderate depression.
The majority of subjects in the meditation trials were drawn from a primary care, not psychiatric, population. Risks were virtually nil, as compared with the well-known adverse effects of medications (Goyal M et al, JAMA Intern Med 2014; online ahead of print).
TCPR’s Take: This meta-analysis supports the use of mindfulness meditation as slightly more efficacious than a nonspecific control, especially for mild or moderate depression or anxiety, but shows that it is no more effective than other therapeutic interventions. Furthermore, the authors emphasize that meditation is a skill to be “learned and practiced over time,” and whose benefit cannot be readily assessed in a brief clinical trial.
CBT for Schizophrenia: Is Talk Cheap?
Schizophrenia, a condition that is thought to affect 1% of the world’s population, remains one of the most challenging psychiatric disorders to treat.
Lately there’s been a resurgence of interest in cognitive behavioral therapy (CBT) for this disorder. Treatment guidelines in the United Kingdom, for instance, recommend the use of CBT for all patients with schizophrenia, as it can reduce the risk of hospitalization compared to treatment as usual.
A newly published study may put a damper on that enthusiasm for CBT. In a comprehensive meta-analysis of fifty international clinical trials of CBT for schizophrenia published between 1993 and 2013, researchers found that the effect of CBT on overall symptoms—and on positive symptoms in particular—may not be as high as once thought.
Specifically, when the researchers limited their analysis to studies which were not biased in favor of CBT, they found that the benefits of CBT seemed to vanish.
The most significant bias was found in the presence or absence of “blinding.” When subjects’ symptoms were rated by clinicians who didn’t know what treatment the subjects were receiving (ie, the clinicians were “blinded”), CBT was found to be less effective than when the evaluators knew which group the subjects were in. Average effect sizes relative to placebo decreased from 0.62 (in unblinded studies) to 0.15 (in blinded studies) for overall symptoms, and from 0.57 to 0.08 for positive symptoms.
The authors did not identify any differences in outcome caused by other potential biases, such as a high dropout rate, the lack of a control group, or publication bias—in which positive results are more likely to be published than negative results (Jauhar S et al, Br J Psychiatry 2014;204:20–29).
TCPR’s Take: This meta-analysis suggests that the efficacy of CBT for schizophrenia may be overstated and that evaluators in clinical trials may be biased in favor of outcome. In other words, when they know which patients are receiving CBT, the evaluators may be more likely to rate them as “improved.” In practice, however, the outcomes we seek often have more to do with our patients’ overall function than with symptom burden, and the integration of CBT principles with other supports—including medication—may pay off according to these outcome measures.