When Mindfulness Makes Things Worse
Mindfulness is now routinely recommended for stress, anxiety, depression, chronic pain and general wellbeing. It appears in psychotherapy, hospitals, schools, workplaces, smartphone applications and wellness programs. For many people it is useful. Research supports small-to-moderate improvements in several psychological outcomes, particularly anxiety, depression and pain (Goyal et al., 2014). However, the popular idea that mindfulness is intrinsically benign is not supported by the evidence.
Meditation can sometimes produce anxiety, panic, emotional destabilisation, traumatic re-experiencing, depersonalisation, derealisation, perceptual abnormalities, insomnia and cognitive disruption. More rarely, the literature contains cases involving mania, hallucinations, delusional beliefs and psychotic episodes. Some people report functioning worse after meditation than before it.
A systematic review by Farias et al. (2020), covering 83 studies and 6,703 participants, estimated an overall prevalence of meditation-related adverse events of approximately 8.3%, although estimates varied dramatically according to study design and how adverse effects were assessed. Anxiety, depression and cognitive abnormalities were among the most frequently reported categories. Importantly, adverse events were also reported in individuals without a known previous history of mental illness.
This does not mean mindfulness should be abandoned. It means mindfulness should be treated like any psychological intervention: dose, technique, context, vulnerability and monitoring matter.
Where Did Mindfulness Come From?
Modern clinical mindfulness did not suddenly appear in Western psychology.
Its historical roots lie primarily in Buddhist contemplative traditions, although the relationship between traditional Buddhist mindfulness and modern therapeutic mindfulness is more complicated than is sometimes presented.
The English word mindfulness became associated with the Buddhist term sati. Historian and Buddhist scholar Rupert Gethin notes that T. W. Rhys Davids translated sati as “mindfulness” in 1881. Traditional Buddhist accounts involve considerably more than simply “living in the present moment”; mindfulness existed within broader systems involving attention, ethical behaviour, memory, insight, concentration and specific understandings of suffering and the self (Gethin, 2011).
During the twentieth century, elements of Buddhist contemplative practice increasingly entered secular Western culture. One major clinical development was Jon Kabat-Zinn's Mindfulness-Based Stress Reduction (MBSR), originally developed within medical settings. Mindfulness practices subsequently became incorporated into Mindfulness-Based Cognitive Therapy (MBCT), Acceptance and Commitment Therapy, Dialectical Behaviour Therapy and numerous other interventions.
Kabat-Zinn (2003) explicitly acknowledged the Buddhist origins of mindfulness while adapting the practice for medicine and psychology. Modern clinical mindfulness therefore represents an adaptation rather than an exact reproduction of historical Buddhist practice.
That history matters when discussing adverse effects. Traditional contemplative systems did not necessarily assume that meditation would simply make practitioners relaxed and happy. Historical Buddhist literature describes unusual perceptual experiences, profound changes in self-experience and periods of psychological difficulty during intensive practice. Modern wellness culture has sometimes removed those warnings while retaining the practices themselves (Lindahl et al., 2017).
What Exactly Is Mindfulness?
There is no single universally accepted psychological definition. One influential operational model proposed by Bishop et al. (2004) describes mindfulness as involving two broad components:
self-regulation of attention toward immediate experience; and
an orientation toward that experience characterised by curiosity, openness and acceptance.
In practical terms, mindfulness involves deliberately observing sensations, thoughts, emotions or perceptions while attempting not to become automatically absorbed in, avoid or judge them.
Mindfulness is therefore not simply relaxation. A person practising mindfulness may become more aware of unpleasant sensations, frightening thoughts, traumatic memories or unfamiliar changes in consciousness. Relaxation may occur, but it is not guaranteed. Researchers often distinguish at least two major meditation styles. Focused-attention meditation repeatedly directs attention toward a particular object, such as breathing. Open-monitoring meditation involves observing whatever thoughts, sensations, emotions and perceptions arise without maintaining one specific object of attention (Lutz et al., 2008).
Common clinical exercises include:
attending to breathing;
systematically observing bodily sensations during a body scan;
noticing thoughts as mental events rather than facts;
observing emotions without immediately changing them;
mindful walking or movement;
directing attention toward sounds or other sensory experiences; and
open awareness of moment-to-moment consciousness.
The intensity of these practices varies enormously. Ten minutes of guided breathing at home is not psychologically equivalent to meditating for eight or ten hours per day during a silent residential retreat. This distinction becomes critical when discussing serious adverse effects.
What Do We Mean by Psychosis and “Psychotic-Like” Experiences?
Psychosis refers broadly to a disturbance in reality testing. Core psychotic symptoms include hallucinations, delusions and disorganised thinking or behaviour. A psychotic episode can occur in schizophrenia-spectrum disorders, bipolar disorder, severe mood disorders, substance-related conditions, neurological or medical disorders, and sometimes transiently in other circumstances (American Psychiatric Association, 2022).
A hallucination is a perception-like experience occurring without a corresponding external stimulus. A delusion involves a strongly held belief that is inconsistent with available evidence and cannot adequately be explained by the person's cultural or religious context.
Psychosis is therefore not synonymous with having an unusual experience.
This distinction is especially relevant to meditation.
A meditator might temporarily experience:
flashes of light;
an unusual feeling that the body has expanded;
loss of the ordinary distinction between self and environment;
intense emotional experiences;
time distortion;
depersonalisation;
derealisation; or
vivid internally generated imagery.
These can resemble aspects of psychopathology without constituting a psychotic disorder.
Researchers sometimes describe these as psychotic-like experiences: perceptual, cognitive or belief-related phenomena resembling psychotic symptoms but occurring with less severity, persistence, conviction or functional impairment.
A 2026 systematic review by Mysko, Gear and Ellett provides an important counterweight to claims that mindfulness itself generates psychosis. Across nonclinical populations, greater trait mindfulness was actually associated with fewer psychotic-like experiences, although mindfulness interventions themselves did not significantly reduce such experiences in the smaller intervention literature (Mysko et al., 2026). The issue therefore appears to involve particular practices in particular people under particular conditions, rather than mindfulness being intrinsically psychotogenic.
What Adverse Effects Have Actually Been Reported?
The most informative literature extends far beyond psychosis. Schlosser et al. (2019) surveyed 1,232 regular meditators. Approximately 25.6% reported having experienced a particularly unpleasant meditation-related experience that they believed may have been caused by meditation. Participants who had attended meditation retreats and those using exclusively more “deconstructive” meditation practices showed higher odds of reporting unpleasant experiences. This study cannot establish causality and the sample was self-selected. Nevertheless, it demonstrates that unpleasant meditation experiences are not exceptionally rare. Lindahl et al. (2017) conducted an extensive mixed-methods investigation of challenging contemplative experiences. They identified 59 categories spanning cognitive, perceptual, emotional, bodily, motivational, self-related and social domains.
Reported phenomena included:
fear and panic;
perceptual hypersensitivity;
visual phenomena;
hallucination-like experiences;
derealisation;
altered body perception;
disrupted self-world boundaries;
loss of agency;
traumatic memories;
agitation;
racing thoughts;
insomnia;
paranoia;
depression; and
unusual or delusion-like beliefs.
Some experiences were brief and interpreted positively. Others were distressing, persistent or functionally impairing. A small number of situations developed into states described by practitioners or experts as resembling mania or psychosis.
Britton et al. (2021) examined adverse effects more systematically following mindfulness-based programs. Among 96 participants, 83% reported at least one meditation-related side effect when systematically questioned. Negative-valence effects occurred in 58%, negative effects on functioning in 37%, and depending on the duration threshold used, approximately 6–14% reported more persistent negative effects.
Those numbers should not be interpreted as meaning mindfulness harms 83% of people. The study deliberately used sensitive interviewing to detect any meditation-related change, including transient experiences that participants might never spontaneously report. Persistent clinically relevant problems were much less common.
Nevertheless, it demonstrates an important methodological point: researchers find substantially more adverse effects when they actually ask about them (Britton et al., 2021).
Can Meditation Actually Trigger Psychosis?
There are published cases suggesting that it can sometimes act as a precipitating factor.
Sharma et al. (2022) reviewed published reports describing psychosis occurring in association with meditation. They identified 28 individual cases across 19 publications. Diagnoses included acute psychosis, schizophrenia, mania with psychotic features and schizoaffective disorder. However, half of the reported individuals also had identifiable potential precipitating factors including insomnia, reduced food intake, previous psychiatric illness, stress or psychoactive substance use.
The authors therefore reached a cautious conclusion: an association exists, but the existing literature cannot establish that meditation itself caused the psychosis. That qualification is essential. Many reported episodes occurred around intensive meditation, rather than ordinary brief mindfulness exercises. Residential retreats may combine prolonged meditation with silence, social isolation, major changes in routine, restricted stimulation, altered sleep, fasting or reduced food intake. Each of these factors may independently affect mental state. For example, sleep disruption itself is associated with psychotic-like, dissociative and hypomanic experiences. Experimental evidence suggests that sleep loss can contribute causally to at least some of these phenomena (Barton et al., 2018).
How Could Mindfulness Make Someone Worse?
No single mechanism has been established. Several overlapping pathways are plausible.
1. Attention can amplify experience
Mindfulness deliberately trains attention. This is normally useful because people learn to notice thoughts, emotions and bodily sensations before automatically reacting to them. Neurocognitive models emphasise attention regulation, emotion regulation and increased awareness of internal experience as important mechanisms of meditation (Tang et al., 2015).
But greater awareness does not guarantee that what becomes more noticeable will be pleasant. Someone prone to panic may become intensely aware of heartbeat, breathing or dizziness. Someone prone to obsessive monitoring may begin examining every internal fluctuation. Someone experiencing subtle perceptual abnormalities may become progressively more attentive to them. The practice can therefore accidentally increase the salience of experiences that were previously background noise.
2. Interoception can become overwhelming
Body scans and breath-focused practices increase attention toward internal bodily signals.For many people this produces grounding. For others, particularly people with panic symptoms or trauma-related hyperarousal, sustained inward attention can amplify sensations that are already interpreted as threatening.
Instead of:
“I notice my breathing.”
the sequence may become:
“My breathing feels strange. Why does it feel strange? Something is wrong. I am becoming detached. I cannot breathe normally.”
Mindfulness has then shifted from non-reactive observation into escalating self-monitoring.
3. Decentring can resemble depersonalisation
One intended objective of mindfulness is decentring: learning that a thought is a mental event rather than an absolute representation of reality. That distinction can be therapeutic. But intensive observation of consciousness can also destabilise ordinary experiences of selfhood. Lindahl et al. (2017) documented changes in embodiment, agency and self-world boundaries. Some participants experienced these as liberating; others experienced fear, unreality or loss of control. Phenomenologically, the border between:
“I am observing my thoughts without identifying with them”
and
“These thoughts do not feel like mine”
can become clinically important.
Likewise:
“The self is not fixed”
is philosophically very different from:
“I no longer feel real.”
The latter can represent depersonalisation or derealisation.
4. Previously avoided material can emerge
Mindfulness reduces avoidance and increases contact with ongoing experience. That may include material a person has spent years trying not to experience. Lindahl et al. (2017) found reports of traumatic memories and emotionally charged material emerging during meditation. In appropriately structured therapy this may become workable psychological material. Without preparation or support, it may instead produce flooding, panic or dissociation. Mindfulness is therefore not always emotionally neutral observation. It can function partly like exposure. And exposure delivered at the wrong intensity can worsen symptoms rather than reduce them.
5. Arousal can move in the wrong direction
Meditation is commonly marketed as calming. It can also produce increased arousal. Participants have reported agitation, intense energy, reduced perceived need for sleep, emotional lability, racing thoughts and heightened sensory perception. At extremes these experiences may overlap phenomenologically with hypomanic or manic states. Britton et al. (2021) found that persistent adverse effects were particularly associated with dysregulated arousal and dissociative phenomena.
6. Sleep disruption can create a feedback loop
An intensive meditator may begin sleeping less. The person may interpret reduced sleep as evidence that meditation is increasing awareness or energy. More meditation follows. Sleep decreases further. Sleep loss then increases emotional instability, perceptual anomalies and cognitive dysregulation. The unusual experiences may subsequently be interpreted as evidence of spiritual progress, leading to still more practice. This creates a potential positive-feedback loop:
intensive meditation → reduced sleep → altered perception/arousal → increased meaning attributed to the experience → more meditation → further sleep disruption.
The mechanism is plausible and consistent with case literature, although it has not been established experimentally as a specific meditation-to-psychosis pathway.
7. Interpretation matters
Meditation does not occur in a psychological vacuum. If a person sees a flash of light during meditation and thinks:
“Interesting visual phenomenon.”
the experience may end there. If the same experience becomes:
“I have activated a supernatural ability and the universe is communicating directly with me,”
its psychological consequences may be different. Clinicians therefore need to assess not only unusual experiences but also conviction, meaning, reality testing, behavioural consequences and functional impairment.
Case Study 1: When Mindfulness Produces More Anxiety and Derealisation
(The following examples are fictional composites designed to illustrate clinical principles. They are not reports of individual clients.)
Consider a 34-year-old person experiencing chronic anxiety after several traumatic experiences. They download a mindfulness application and begin practising a 40-minute body scan every night. Initially they feel calmer. After several weeks they become increasingly aware of subtle changes in breathing, heart rate and muscle tension. During meditation they occasionally feel detached from their body. They begin thinking about the sensations during the day and checking whether they “feel real.” The meditation session increases from 40 minutes to an hour because they believe they are failing to practise correctly. Eventually they report:
increased anxiety;
derealisation;
fear of losing control;
intrusive memories;
disrupted sleep; and
reduced concentration at work.
The simplistic recommendation would be:
“Keep meditating. The discomfort is part of the process.”
That may be precisely the wrong intervention.
A more appropriate clinical response could involve reducing or temporarily stopping prolonged internally focused meditation, restoring sleep, assessing trauma and dissociation, introducing externally oriented grounding, and using shorter practices with eyes open.
Walking, identifying objects in the environment, orienting toward sounds or engaging in paced behavioural activities may be more appropriate than repeatedly scanning internal sensations.
Mindfulness has not necessarily “failed.” The type and dose of mindfulness were mismatched to the person's current nervous-system state.
Case Study 2: When Intensive Meditation Begins to Resemble Mania or Psychosis
Consider a 28-year-old person with no diagnosed psychotic disorder who attends an intensive silent meditation retreat. They meditate for many hours each day. Their sleep decreases progressively. They eat substantially less than usual. After several days they experience unusually intense colours, feelings of unity with the environment and surges of energy. Initially these experiences are pleasant. They then begin sleeping approximately three hours per night while reporting that sleep is unnecessary. Thoughts accelerate.
They become convinced that coincidences occurring around them contain personal messages. They believe meditation has allowed them to perceive a hidden structure controlling reality. Eventually they report hearing a voice commenting on their spiritual role and become increasingly suspicious of retreat staff who suggest they should rest. At this point the clinical issue is no longer whether the experience could have a spiritual interpretation.
The relevant features are:
major reduction in sleep;
increasing arousal;
perceptual abnormalities;
ideas of reference;
possible auditory hallucinations;
increasing conviction;
deteriorating reality testing; and
behavioural change.
Meditation should not simply be intensified. The person requires prompt mental-health assessment, restoration of ordinary sleep, nutrition and routine, reduction of stimulation from intensive contemplative practice, and evaluation for mania, psychosis, substance effects and other possible causes. This example resembles the multifactorial pattern repeatedly described in meditation-related case literature: meditation intensity interacting with sleep disruption, isolation, physiological stress and individual vulnerability rather than a simple one-variable causal model (Sharma et al., 2022).
What Should Clinicians and Clients Actually Do?
The evidence suggests a middle position between two unhelpful extremes.
The first extreme is:
“Mindfulness is harmless, so everyone should meditate.”
The second is:
“Meditation causes psychosis and should be avoided.”
Neither position fits the data.
Mindfulness-based interventions have actually been studied specifically in people with psychotic disorders. Meta-analytic evidence suggests that appropriately modified mindfulness and acceptance-based interventions can be feasible and beneficial for some people with schizophrenia-spectrum disorders (Jansen et al., 2020).
More importantly, a 2024 systematic review examining safety across mindfulness-based interventions for psychosis found no significant increase in adverse events, symptom deterioration, deaths, treatment noncompletion or dropout compared with control conditions. Hospitalisation and crisis-service use were actually higher in control groups in the pooled analyses.
However, the authors also identified a major problem: adverse effects were inconsistently assessed and reported. None of the included studies systematically reported intervention side effects in the same way that medication research typically monitors adverse reactions (O'Brien-Venus et al., 2024).
For clinical practice, several principles follow:
Start with formulation rather than prescription. Mindfulness should be selected because it fits the psychological mechanism being targeted, not because it is assumed to be universally beneficial.
Consider dose. Five minutes of externally anchored mindful attention is different from an hour of open-monitoring meditation.
Monitor function, not just meditation experiences. Reduced sleep, inability to work, escalating anxiety, social withdrawal and impaired reality testing matter more clinically than whether an experience feels unusual.
Take dissociation seriously. Increased unreality, detachment or loss of embodiment is not necessarily evidence that someone needs to “go deeper.”
Protect sleep. A sustained reduction in sleep—particularly when accompanied by elevated energy, accelerated thinking or unusual beliefs—is a warning sign.
Distinguish observation from rumination. Repeatedly examining internal sensations can become compulsive self-monitoring rather than mindfulness.
Be careful with intensive retreats. Prolonged practice combined with fasting, isolation and disrupted sleep represents a substantially different psychological exposure from ordinary outpatient mindfulness.
Adapt rather than automatically abandon mindfulness. Shorter sessions, eyes-open practice, walking, movement, sounds and external sensory anchoring may be preferable to intensive internal monitoring for some people.
Most importantly, emerging hallucinations, fixed unusual beliefs, marked loss of reality testing, severe insomnia, manic symptoms, suicidal thinking or major functional deterioration require clinical assessment rather than further unsupervised meditation.
Limitations of the Evidence
The adverse-effects literature has substantial methodological limitations.
First, meditation is not one intervention. Studies frequently combine mindfulness, Vipassana, Zen, transcendental meditation, concentration practices and other contemplative methods. Findings from one practice cannot automatically be generalised to another.
Second, adverse-event definitions vary dramatically. A transient unpleasant emotion during meditation is fundamentally different from persistent functional deterioration. This partly explains why studies produce prevalence estimates ranging from single digits to much higher percentages.
Third, retrospective surveys are vulnerable to self-selection and attribution bias. Someone experiencing a dramatic meditation-related difficulty may be more likely to participate in research about meditation difficulties.
Fourth, case reports establish possibility but not incidence or causality. Twenty-eight published psychosis cases do not tell us the probability that a meditator will develop psychosis.
Fifth, serious events may also be underreported. Historically, mindfulness trials have focused much more carefully on potential benefits than harms.
Sixth, people undertaking intensive meditation may simultaneously experience altered sleep, diet, social contact and sensory stimulation. Separating the effect of meditation from these variables is extremely difficult.
Finally, the strongest evidence available does not indicate that ordinary mindfulness-based therapy systematically increases psychotic symptoms. Evidence from psychosis populations is generally reassuring when interventions are adapted and clinically supervised (O'Brien-Venus et al., 2024; Mysko et al., 2026).
Where Research Needs to Go Next
Future mindfulness research requires the same safety standards expected of other psychological and medical interventions. Trials should prospectively measure adverse events rather than waiting for participants to spontaneously report them. Researchers should distinguish:
transient discomfort;
negative emotional experiences;
functional deterioration;
dissociation;
mania;
psychotic-like experiences; and
diagnosable psychotic episodes.
Studies also need to measure meditation dose, technique, previous experience, trauma history, psychiatric history, family vulnerability, substance use, sleep, fasting and retreat conditions. Most importantly, researchers need longitudinal designs capable of determining sequence.
Conclusion
Mindfulness is neither a miracle cure nor a uniquely dangerous practice.
It is a psychological intervention capable of changing attention, emotional processing, bodily awareness and experiences of the self. Those changes can be therapeutic. They can also occasionally be destabilising. Systematic reviews confirm that meditation-related adverse experiences exist, including anxiety, depression, cognitive disruption, dissociation and functional deterioration. Case literature also documents rare episodes of mania and psychosis occurring in temporal association with intensive meditation.
What remains unknown is the precise incidence of meditation-induced psychosis and the degree to which meditation itself, rather than associated factors such as sleep deprivation, fasting, isolation or underlying vulnerability, is causal. The most scientifically defensible conclusion is therefore not that mindfulness is dangerous. It that mindfulness is psychologically active. And anything psychologically active can have indications, contraindications, dosage effects, individual variability and adverse outcomes. Treating mindfulness with that level of clinical seriousness does not undermine the practice.
It makes its use more evidence-based.
References:
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.).
Barton, J., Kyle, S. D., Varese, F., Jones, S. H., & Haddock, G. (2018). Are sleep disturbances causally linked to the presence and severity of psychotic-like, dissociative and hypomanic experiences in non-clinical populations? A systematic review. Neuroscience & Biobehavioral Reviews, 89, 119–131. doi:10.1016/j.neubiorev.2018.02.008
Bishop, S. R., Lau, M., Shapiro, S., Carlson, L., Anderson, N. D., Carmody, J., Segal, Z. V., Abbey, S., Speca, M., Velting, D., & Devins, G. (2004). Mindfulness: A proposed operational definition. Clinical Psychology: Science and Practice, 11(3), 230–241. doi:10.1093/clipsy.bph077
Britton, W. B., Lindahl, J. R., Cooper, D. J., Canby, N. K., & Palitsky, R. (2021). Defining and measuring meditation-related adverse effects in mindfulness-based programs. Clinical Psychological Science, 9(6), 1185–1204. doi:10.1177/2167702621996340
Farias, M., Maraldi, E., Wallenkampf, K. C., & Lucchetti, G. (2020). Adverse events in meditation practices and meditation-based therapies: A systematic review. Acta Psychiatrica Scandinavica, 142(5), 374–393. doi:10.1111/acps.13225
Gethin, R. M. L. (2011). On some definitions of mindfulness. Contemporary Buddhism, 12(1), 263–279. doi:10.1080/14639947.2011.564843
Goyal, M., Singh, S., Sibinga, E. M. S., et al. (2014). Meditation programs for psychological stress and well-being: A systematic review and meta-analysis. JAMA Internal Medicine, 174(3), 357–368. doi:10.1001/jamainternmed.2013.13018
Jansen, J. E., Gleeson, J., Bendall, S., Rice, S., & Alvarez-Jimenez, M. (2020). Acceptance- and mindfulness-based interventions for persons with psychosis: A systematic review and meta-analysis. Schizophrenia Research, 215, 25–37. doi:10.1016/j.schres.2019.11.016
Kabat-Zinn, J. (2003). Mindfulness-based interventions in context: Past, present, and future. Clinical Psychology: Science and Practice, 10(2), 144–156. doi:10.1093/clipsy.bpg016
Lindahl, J. R., Fisher, N. E., Cooper, D. J., Rosen, R. K., & Britton, W. B. (2017). The varieties of contemplative experience: A mixed-methods study of meditation-related challenges in Western Buddhists. PLOS ONE, 12(5), e0176239. doi:10.1371/journal.pone.0176239
Lutz, A., Slagter, H. A., Dunne, J. D., & Davidson, R. J. (2008). Attention regulation and monitoring in meditation. Trends in Cognitive Sciences, 12(4), 163–169. doi:10.1016/j.tics.2008.01.005
Mysko, K., Gear, E. Q., & Ellett, L. (2026). Mindfulness and psychotic-like experiences in nonclinical populations: A systematic review and two meta-analyses. Psychological Medicine, 56, e39. doi:10.1017/S0033291725103061
O'Brien-Venus, B., Ellett, L., Burgess-Barr, S., & Chadwick, P. (2024). Systematic review of the safety of mindfulness-based interventions for psychosis. Clinical Psychology Review, 112, 102445. doi:10.1016/j.cpr.2024.102445.
Schlosser, M., Sparby, T., Vörös, S., Jones, R., & Marchant, N. L. (2019). Unpleasant meditation-related experiences in regular meditators: Prevalence, predictors, and conceptual considerations. PLOS ONE, 14(5), e0216643. doi:10.1371/journal.pone.0216643
Sharma, P., Mahapatra, A., & Gupta, R. (2022). Meditation-induced psychosis: A narrative review and individual patient data analysis. Irish Journal of Psychological Medicine, 39(4), 391–397. doi:10.1017/ipm.2019.47
Tang, Y.-Y., Hölzel, B. K., & Posner, M. I. (2015). The neuroscience of mindfulness meditation. Nature Reviews Neuroscience, 16, 213–225. doi:10.1038/nrn3916.
Comments