How to Read a Meditation Study: Questions to Ask

Headlines such as "meditation changes your brain" or "mindfulness works as well as medicine" can come from very different kinds of studies. This page gives you plain questions to ask before you trust a result. It is information only and cannot judge any one study for you. For the bigger picture of weaknesses in this field, see quality and bias in meditation research.

Start with how strong the source is

A single small study is a starting point. Systematic reviews and meta-analyses combine many studies, but they can only be as good as the studies inside them. NCCIH, the US government center for complementary health, says much meditation research has been preliminary or not scientifically rigorous, and that results were hard to analyze and may have been interpreted too optimistically. [1] So look for reviews, but also for how they rated their studies. NCCIH also points readers to PubMed, which contains publication information and brief summaries of articles from scientific and medical journals. [1]

Seven questions to ask

1. What exactly was done? "Meditation" covers many practices. Experts note that the specific practice or training in a study is often under-specified, making comparisons across studies difficult or impossible. [2] Look for the name of the practice or program, the number of sessions and the home practice.
2. Who was it compared with? Check whether the comparison was a waitlist, usual care, another credible program or a sham practice. Most studies have used waitlists, [3] and a review of trials with active controls found no evidence that meditation beat any active treatment. [4] More in placebo and active controls.
3. How were results measured? Many outcomes are self-reported. Experts warn that people with only brief training may rely substantially on demand characteristics when describing their experience, meaning they may report what seems expected. [2] Prefer studies that also use objective or independent measures, and note which is which.
4. How long did the benefit last? NCCIH reports that in an analysis of mindfulness-based treatments for anxiety, the few studies that followed people for more than 2 months found no long-term effects. [1] A result at the end of an 8-week course is not the same as a result a year later.
5. Who took part? Healthy volunteers, college students and patients with a diagnosis are different groups. NCCIH describes a review of studies on students that found most of the studies were of poor quality with a high risk of bias. [1]
6. Were harms looked for? The Van Dam review states that most current methods for assessing adverse events in meditation-related research are insufficient to produce an accurate estimate. [5] NCCIH says few studies have examined harm, so definite statements about safety cannot be made. [1] See adverse effects: what studies found.
7. Is the whole picture visible? Registered trials should publish. In one analysis of mindfulness trials, 13 of 21 trial registrations remained unpublished 30 months after completion. [6] Results can then look better than they are.

Spotting the common tricks

Match the claim to the type of evidence

Use these plain labels when you read: established (consistent results from well-controlled trials and guidelines), promising but limited (some good trials, but small, short or biased), mixed, insufficient evidence, and mechanism only (laboratory or imaging findings). A brain-imaging change or a blood marker is a lead for research, not proof of a health benefit.

When a study is about you

If a study seems relevant to a medical condition, talk with a health professional before changing treatment, and do not use meditation to replace conventional care or to postpone seeing a provider. [1] For well-supported options by condition, see meditation for stress, sleep and anxiety.