Meditation is increasingly being used as a wellbeing tool, and a tool to support positive mental health. Of course, psychological therapies can help many people, but there is also the potential for inadvertent harm. Previous elf blogs have covered research that explores the adverse effects of psychotherapy (e.g. Crawford et al., 2016; Rozental et al., 2016), but there is no consensus on the definition of adverse effects, however it can refer to any undesirable effects associated with an intervention. Indeed, the language used in the primary paper, ‘adverse events’ is different to the author’s (EMG) preferred phrase ‘adverse effects’, however events will be used throughout the rest of this piece to refer to the same concept.
Trials of psychological interventions require reports of serious adverse events, such as self-harm, suicidality or death, however reporting of adverse events/effects in general is less consistent in the literature. Like yoga, meditation has its origins in spiritual practices but has a more discrete function in popular Western culture. Transcendental meditation and mindfulness are the most popular technique, with trends towards meditations around prosocial emotions, such as loving kindness and compassion-based meditations. There is a long history of recognising the potential of meditation to lead to unusual states of mind, and specifically, concerns were raised about potential for harm in the 1970’s (Farias et al, 2020). This review focuses on harmful and/or distressing effects of meditation and is reportedly the first to synthesise the literature on this topic (Farias et al, 2020).
This review was preregistered on PROSPERO and used a broad sweep approach to capture as many potentially relevant articles as possible. Any articles that referred to adverse events were included, whereas body-based practices (for example, yoga) were excluded.
All types of study (from case studies to RCTs) of meditation as an intervention for adults, were considered and those that reported at least one adverse event or deterioration in mental health were included. All were rated for quality according to their study type, with the two authors reaching a consensus on all ratings.
After a systematic review of over 6,742 search results, 83 unique studies were identified. The adverse events were then categorised, and prevalence estimates were calculated for 57 studies; case studies, and 11 other studies that did not include information about the number of participants experiencing events or only included those experiencing meditation adverse events were excluded. Most studies (65%) used an experimental design (2,673 participants in a meditation condition), with 17% of studies having an observational design (4,023 participants ) and 18% using case study designs (31 participants). Most of the observational studies included quantitative data, and just 2 of these studies accounted for 58% of the observational studies sample size. It was not possible to assess the quality of these two studies, as only summary information was available about the quality of each type of study, with most being of moderate to high quality. As expected, the majority of studies related to transcendental meditation or mindfulness.
Confusingly, the review reports that 65% of the studies described at least one type of adverse event, despite this being a criteria for inclusion in the analysis. Perhaps the inclusion criteria related to any mention of measuring adverse events, as opposed to the presence of.
The case studies all described serious adverse events, such as psychosis, depersonalisation and mania. The RCTs were more likely to report serious adverse events but was unclear in reports whether other less serious adverse events were measured and/or reported.
The pooled estimates were different across study design and were impacted by the wide variety of study design and lack of consistent measurement. RCTs had a prevalence of 3.7% and observational studies 33.2%. Adverse events were categorised into psychiatric, somatic and neurological/cognitive, with psychiatric being most commonly reported, occurring in almost half of all studies. Of the psychiatric symptoms, anxiety, depression, stress and visual/auditory hallucinations were most commonly cited, with suicidal behaviour included as an adverse event for 10% of studies. Participant wise, 1,102 of 6,464 (17%) were reported to have experienced adverse events.
There were mixed reports about the dosage effect, with some finding no relationship, and others finding that those who dropped out reported less adverse events. Other studies reported that more experienced meditators were found to be more likely to experience AEs.
This review suggests that adverse events are common during or after meditation and may occur for people who did not report prior mental health difficulties. Events such as anxiety, depression and stress were most commonly reported in studies that measured and reported adverse events, however the prevalence is unclear as measurement was inconsistent across studies, and the pooled estimates include observational and RCT studies.
Strengths and limitations
A wide range of adverse events were reported, however the wide range of study types obfuscated clarity on the prevalence of adverse events. For example, case studies all included only people experiencing adverse events, whereas RCTs did not. While the authors imply that the observational studies are more sensitive to identifying adverse events due to the uncontrolled nature of practice, or the lack of measurement in RCTs, it is also possible that the observational studies are ascribing difficulties to meditation that may not be related. In an RCT, randomisation provides an opportunity to account for everyday fluctuations in mood and wellbeing that are not specific to the intervention group only, and therefore, not likely to be caused by the intervention; essentially accounting for the nocebo effect as positive outcomes account for the placebo effect (Chavarria et al, 2017). For this reason, this author would have preferred that RCTs alone were included, particularly as there were many high-quality studies available to base more solid conclusions on.
Another critique is the lack of analysis based on clinical presentation; one would expect that meditation in the context of intervention for trauma symptoms or anxiety is likely to have a different impact than general meditation. Further, the experience of mediation as positive or negative may be influenced by mental health. Of most concern, is the finding that meditation can be potentially re-traumatising or lead to symptoms of psychosis, however as the samples are mixed, and methodologies so varied, it is difficult to draw clinical conclusions about who may be most impacted and by what form of meditation. Trauma informed mindfulness (Treleavan, 2018), adapted to reduce the likelihood of traumatisation, may provide an alternative. This author is not aware of, but would love to see, an RCT comparing traditional and trauma informed mindfulness that place the measurement of adverse events as central to the study.
Implications for practice
This study covers a wide range of potential adverse events associated with meditation, but much more specificity is required to inform clinical practice. For example, what are the identified adverse events for people with different mental health conditions? This study supports recognition of the potential of meditation to cause distress/harm, and so trauma informed mediation and indeed, clinically sensitive mediation is recommended. Indeed, sharing this information with clients/patients when suggesting meditation can support informed consent and validate the experiences of those who find meditation unpleasant, triggering or distressing.
As measurement and reporting of adverse events is not yet standardised or the norm, there will continue to be confusion about whether the lack of data about adverse events is due to them not being present, or not being measured. This review, by including observational and case studies along with RCTs highlights this issue; articles focused on adverse effects are likely to inflate their presence compared with RCTs whose primary focus is more often on the positive change anticipated from an intervention. Consistency in measurement and in reporting will improve this.
Statement of interests
Edel McGlanaghy leads a project on the adverse effects of psychotherapy, which is currently being prepared for publication.
Farias M, Maraldi E, Wallenkampf K C, Lucchetti G. (2020). Adverse events in meditation practices and meditation-based therapies: a systematic review. Acta Psychiatrica Scandanavica. Nov;142(5):374-393. doi: 10.1111/acps.13225. Epub 2020 Aug 21.
Chavarria V, Vian J, Pereira C, Data-Franco J, Fernandes BS, Berk M, Dodd S. (2017). The placebo and nocebo phenomena: their clinical management and impact on treatment outcomes. Clinical Therapeutics, 39(3). doi:10.1016/j.clinthera.2017.01.031
Laws, K & Huda, S. (2016). The harms of psychotherapy: are BME and LGBT communities more at risk? The Mental Elf, April 2016.
McGlanaghy, E. (2017). Negative effects questionnaire: measuring the potential harm of psychotherapy. The Mental Elf, March 2017.
Treleavan D. (2018) Trauma-sensitive mindfulness: practices for safe and transformative healing. W W Norton & Co.