Spirituality
Mindfulness: What the Trials Show
By Beyond Growth · 14 November 2026
Does mindfulness work? Modestly, and it depends on what you compare it with. Against doing nothing, courses reduce distress by a small to moderate amount (effect size 0.32 in one large analysis). Against active treatments, a major review found no advantage. In a trial of 84 schools, the programme did no better than normal lessons. And some people report bad effects, usually temporary, occasionally lasting. The honest summary is "a reasonable option with modest benefits", which is not what the marketing says.
Mindfulness has gone from a fringe research topic to a school lesson, a corporate wellbeing tool and, at times, a replacement for therapy (Van Dam et al., 2018). A spirituality site is a fair place to ask what the trials show, because the claims have grown faster than the evidence. This essay is fair to both sides: the research is real and positive in places, and the sales pitch is bigger than the research.
A 2012 essay at this address, by a different writer, took a sceptical look at mindfulness (2012 original, via the Internet Archive). We take a different route: what do the controlled trials say?
What do the big reviews find?
A widely cited review is a 2014 meta-analysis in JAMA Internal Medicine. It pooled 47 randomised trials with 3,515 participants, all with active control groups, which matters because it filters out the effect of simply doing something. It found "moderate evidence" of improved anxiety (effect size 0.38 at 8 weeks), depression (0.30 at 8 weeks) and pain (0.33), and low evidence for stress and quality of life. It also reported "no evidence that meditation programs were better than any active treatment (ie, drugs, exercise, and other behavioral therapies)" (Goyal et al., 2014).
Against passive controls, who received no programme, the results are better. A pre-registered analysis of individual data from 13 trials and 2,371 adults found that mindfulness programmes reduced distress by 0.32 standard deviations (95% interval 0.24 to 0.41) one to six months after the course (Galante et al., Nature Mental Health). That is a "small to moderate" effect for people who volunteered for a course. It is a real result, and it is not a revolution.
There is one place where the case is stronger. A 2016 analysis of nine trials with 1,258 patients found that mindfulness-based cognitive therapy lowered the risk of depressive relapse (hazard ratio 0.69 against all comparison groups combined, and 0.79 against active treatments including antidepressants, over 60 weeks) (Kuyken et al., JAMA Psychiatry). That is a structured, manualised therapy for people who have already been depressed. It is not the same thing as downloading an app.
Does it work for children and at work?
The best test so far for schools is the MYRIAD trial: 85 UK schools were randomised, and 84 (8,376 students aged about 12) were analysed. Ten mindfulness lessons were compared with normal teaching. At one year, the authors found "no evidence that SBMT was superior to TAU", with differences between groups of 0.005 for depression risk and 0.02 for both behavioural functioning and wellbeing. No intervention-related adverse events were observed (Kuyken et al., Evidence-Based Mental Health). That is a null result from a large, well-run trial, and it is worth saying plainly.
Workplaces and apps look better, with caveats. A 2023 meta-regression of 91 workplace trials found stress reductions of 0.72, but the authors flagged heterogeneity and risk of bias, and effects were stable over 1 to 12 weeks but not over 13 to 52 (Mindfulness, 2023). A review of mindfulness apps for college students found stress reduced by 0.435 standard deviations, with moderate certainty, but noted that "sustained usage may be a challenge" (Journal of Affective Disorders, 2024). Apps help some people some of the time, and they depend on people actually opening them.
What are the reported bad effects?
Meditation is often presented as harmless. That is not a safe assumption. A systematic review of 83 studies with 6,703 participants found at least one adverse event reported in 65% of the studies, and a prevalence of 8.3% overall: 3.7% in experimental studies and 33.2% in observational ones. The most common were anxiety, depression and cognitive anomalies, and the authors noted these "may occur in individuals with no previous history of mental health problems" (Farias et al., Acta Psychiatrica Scandinavica).
The widest range comes from asking people directly. In an interview study of 96 people on an 8-week mindfulness-based cognitive therapy course, 83% reported at least one meditation-related effect, 58% an effect with a negative valence or impact, and 37% a negative impact on functioning. Lasting bad effects occurred in 6 to 14%. The authors' conclusion is the fair one: practice "is associated with transient distress and negative impacts at similar rates to other psychological treatments" (Britton et al., Clinical Psychological Science). So mindfulness is not uniquely dangerous. It is also not a risk-free relaxation exercise.
Is the app the same as the ancient practice?
In an early Buddhist text, the Magga-vibhanga Sutta, "right mindfulness" is the seventh of eight path factors, after right view, resolve, speech, action, livelihood and effort. It is described as being "ardent, aware, & mindful", and it is practised while "putting away greed & distress with reference to the world" (SN 45.8, trans. Thanissaro Bhikkhu). It is one part of a path with ethics and a goal.
The clinical version began somewhere else. Jon Kabat-Zinn's 1982 paper described a 10-week outpatient programme for 51 chronic pain patients who "had not improved with traditional medical care". At 10 weeks, 65% showed at least a one-third reduction in a pain rating index. The paper itself discusses "the limitations of the present uncontrolled study" (Kabat-Zinn, 1982). It was a careful translation of a technique into medicine, and later trials mostly treat it that way.
The gap opens when the stand-alone technique is sold as the whole thing: a path to calm in ten minutes a day, to better leadership, to resilient children. A critical review in Perspectives on Psychological Science warns that "misinformation and poor methodology associated with past studies of mindfulness may lead public consumers to be harmed, misled, and disappointed" (Van Dam et al.). That is a peer-reviewed warning about hype and weak studies, not a blogger's.
Myth versus fact
| The claim you hear | What the evidence supports |
|---|---|
| "Mindfulness is proven to reduce stress." | Small to moderate reductions against no treatment; no clear edge over other active treatments. |
| "It is good for kids." | One large school trial found no benefit over normal lessons at one year. |
| "It cannot hurt." | Bad effects are reported by a minority, mostly temporary, occasionally lasting. |
| "It is ancient wisdom, so it is deep." | The apps teach a technique; the original sat inside an ethical path. |
What should you take from it?
If you like it, and it helps, there is no reason to stop. The evidence says it is a reasonable option with modest benefits, in a field with real trials and honest researchers. If you have a history of trauma or severe anxiety, choose a trained teacher and mention it before you start. Bad effects can also appear in people with no history of mental health problems, so if practice leaves you more anxious or detached, stop and get advice. If an app or course promises transformation, treat it as you would any other promise: ask for the comparison group.
More in the spirituality section. For a related critique of push-button inner peace, see the hollow sink of push-button enlightenment; for the same habit of testing a promise against its evidence, read is manifestation real?
