If meditation clearly increases your anxiety, stop the session. A difficulty is not automatically harm, but neither is it proof that the method is “working at a deeper level.” What matters is what happens afterward: can you return to your activities, sleep, and stay connected to your surroundings?
Distressing experiences have been documented during or after certain practices. Their frequency depends heavily on definitions and how information is collected. The NCCIH, a US public agency, emphasizes that safety studies remain too limited to draw definitive conclusions. This guide is for adults and offers pointers for caution, not a diagnosis or an individual guarantee.
Immediate danger, risk of harmful action, or severe disorganization: seek urgent help. In France, call 15 or 112. For suicidal thoughts or concerns about a loved one, 3114 is available free of charge, day and night. Do not try to resolve an emergency with another meditation.
Discomfort, adverse effects, and worsening symptoms: three different things
Feeling bored, noticing restlessness, or experiencing an unpleasant emotion is not enough to establish psychological harm. A session can be unpleasant without leaving lasting consequences. You still have the right to stop: a well-being practice is not an endurance test.
Research distinguishes several dimensions. An experience can be unexpected and unpleasant, and may or may not affect functioning. A worsening of symptoms means an increase in symptoms that were already present. An adverse event describes what happens during a program, without proving on its own that the program caused it. Not all studies use these terms in the same way.
In their 2021 methodological paper, Britton and colleagues distinguish, among other things, unpleasantness, impact, and duration. This distinction avoids putting a brief emotion in the same category as a difficulty that persistently disrupts sleep or relationships.
Three questions help describe the situation:
- Intensity: does the experience remain tolerable, or is it becoming overwhelming?
- Duration: does it ease after stopping, or continue beyond the session?
- Impact: does it interfere with sleep, work, relationships, or the ability to feel safe?
These pointers are not a diagnostic test. You do not need to wait a specific length of time or reach a score before asking for help.
Why turning attention inward can become difficult
Mindfulness involves training attention to present experience. That experience may include worrying thoughts, sensations of panic, or distressing memories. Noticing them more does not necessarily mean regulating them more easily, especially at first.
This is a plausible psychological explanation, discussed in 2026 research on digital practices, not proof that a brain circuit has been damaged. The expected benefit and the difficulty may involve the same target: noticing more of what is happening within you.
The Toll, here, means the hidden cost of a solution: a practice presented as soothing can require emotional effort you did not anticipate. Marketing sells calm. The session itself does not guarantee the advertised emotional state.
For someone with anxiety, focusing on the breath can also turn into anxious monitoring of breathing. Switching your focus, keeping your eyes open, or ending the exercise are possible adjustments. They do not make every situation safe and do not replace care if symptoms persist.
What studies measure, and what they may miss
The 8.3% figure does not predict your personal risk
The systematic review by Farias and colleagues, published in 2020, brings together 83 studies on various forms of meditation and interventions in adults. The pooled estimate of adverse events is 8.3%, with a 95% confidence interval of approximately 5–12%.
This result combines very different practices, populations, durations, and methods. Estimates vary widely between experimental and observational studies. It is therefore not an “8.3% risk per session,” a proportion of permanent harm, or a causal frequency attributable to five minutes of mindfulness. Anxiety and depressive symptoms are among the reported experiences, but the data do not allow individual prediction.
Table of data collection methods
Before comparing two percentages, look at how they were obtained:
| Method | What it can detect | What it cannot establish on its own |
|---|---|---|
| Spontaneous reporting to the teacher | Difficulties the participant considers worth mentioning | No complaints does not demonstrate an absence of difficulties |
| General question about a negative experience | An overall assessment that is easy to collect | May miss specific symptoms or include discomfort without consequences |
| Detailed checklist or structured interview | Nature, duration, and impact of specific symptoms | A high number of responses does not mean an equal number of cases of serious harm |
| Scales before and after the program | Worsening or improvement in the symptoms measured | May miss a new symptom outside the scale or a temporary difficulty |
| Comparison with a control group | Differences in change over time on the outcomes studied | Does not remove limitations involving selection, follow-up, and detection of rare events |
This table summarizes distinctions made by Britton and colleagues and in the survey by Van Dam and colleagues published in 2025. Among 886 US adults with meditation experience, the latter found different frequencies depending on the question and tool used. Its cross-sectional, retrospective design limits causal attribution. The population does not specifically represent beginners in France.
Reassuring findings count too
An analysis of MBSR programs combines a cohort of 2,155 participants and data from three trials, with 156 MBSR participants and 118 on a waiting list. It finds no greater worsening in the MBSR group on the outcomes studied. This finding does not cover every possible effect. A methodological discussion published afterward highlights limitations involving missing data and the waitlist comparator, among other issues.
In 2026, an analysis of two app trials also distinguishes distressing experiences from their causality. In the second study, reports did not differ significantly between people who practiced the guided meditations and those who did not. However, this comparison based on actual use is not a randomization between “meditating” and “not meditating.” One author reports leading the organization behind the program: keep that in mind, without automatically dismissing the results.
Individual difficulties can therefore be observed without demonstrating an overall increase in risk on every outcome. And a reassuring average result does not justify ignoring someone whose condition is worsening.
AXA and Pourquoi Docteur: putting the contradiction in context
The AXA Prévention page dated 2018 states that no negative adverse effects have been reported in studies. This wording does not describe the full body of research available today. It does not distinguish clearly enough between what was not reported and what was actively investigated.
By contrast, Pourquoi Docteur reports intrusive memories and depersonalization, among other experiences. These descriptions deserve to be taken seriously, without becoming proof that all mindfulness is dangerous.
The disagreement does not force you to choose between promises and alarm. Check the date, population, practice, and definition of the event. “Nothing reported” is not enough to guarantee safety. “Distressing experience reported” is not enough to demonstrate harm caused by meditation.
Intensive retreats and brief exercises: different contexts
A retreat may combine long periods of practice, silence, changes in routine, and varying levels of supervision. A brief exercise at home does not reproduce that exposure. Serious problems have been reported, particularly in intensive settings, but these reports cannot be used to calculate a universal excess risk between the two formats.
Britton and colleagues also describe difficulties outside retreats. Brief does not therefore automatically mean well tolerated. There is no threshold in minutes that guarantees a session will not trigger any distress.
Before intensive practice, discuss any history of trauma or psychiatric disorders and your current symptoms with a qualified professional. This is not a blanket ban on meditation. The aim is to discuss the format, available support, and the option to stop, particularly if your condition is already unstable.
Anxiety and depersonalization: when to pause and seek help
The term depersonalization refers to a feeling of detachment from yourself. The term derealization refers to a feeling that your surroundings seem unreal. These descriptions do not provide a remote diagnosis or prove that meditation is the cause. If they worry you, persist, or disrupt your activities, they call for an assessment rather than an imposed spiritual interpretation.
The following chart offers cautious guidance, not a validated clinical tool:
| What you notice | Cautious response |
|---|---|
| Tolerable boredom, many thoughts, or mild restlessness, with no impact afterward | Adjust your posture, switch your focus, or stop. There is no obligation to continue |
| A clear increase in panic, an overwhelming traumatic memory, or a worrying feeling of disconnection | Stop the session. Reconnect with an external point of reference and a familiar activity if possible. Seek help if it does not ease |
| Recurring or persistent symptoms, poorer sleep, difficulties at work or in relationships | Pause self-directed practice and consult a doctor or psychologist before resuming |
| Severe confusion, loss of contact with reality, new hallucinations, or unusual excitement with very little sleep | Seek an urgent medical assessment. Immediate danger or inability to stay safe: 15 or 112 in France |
| Suicidal thoughts | Contact 3114 in France. Immediate danger, a plan to act, or means available: 15 or 112 |
The government guidance on psychiatric emergencies helps distinguish a need for support from an emergency. Chest pain, feeling seriously unwell, or an unusual physical symptom should not automatically be attributed to anxiety or meditation.
Decision tree
What should you do if a session becomes difficult?
Unable to stay safe: 15 or 112 in France.
Panic, worrying disconnection, or persistent symptoms: do not restart practice on your own.
Boredom and distractions do not prove harm. Continuing remains optional.
Resume with support, or choose something else
After symptoms worsen, the priority is to return to a manageable level of functioning and understand the symptoms. Do not deliberately expose yourself again to prove that meditation caused them. Resuming is not compulsory, and there is no validated universal timeline for doing so.
If you are considering resuming, discuss these points with a doctor or psychologist, then with the teacher involved:
- Which symptoms mean I should pause again, and who should I contact?
- Does the format suit my current condition and history?
- Should I avoid breath focus, body scans, or practicing with my eyes closed for now?
- How can I check the effects on sleep and the rest of the day, rather than judging only the session?
A teacher can adapt instructions. A healthcare professional assesses clinical symptoms. The title of instructor is no substitute for that expertise. Be wary of guidance that presents every deterioration as a necessary step or discourages access to care.
The printable HTML chart and its PDF version let you record the context, what happens afterward, and the planned contact person. They do not calculate a risk score.
For mild discomfort without warning signs, the mindfulness exercises adapted for beginners offer several points of focus. If you prefer to set meditation aside, a familiar activity, an ordinary walk, or a conversation with someone close to you can provide other ways to take a break, without any promise of treating a disorder.
Do not replace a poorly tolerated practice with a pile of products to “repair” your brain either. The guide to the limitations and risks of acetyl-L-carnitine for the brain · FR addresses that separate decision. It describes neither a treatment for meditation-related difficulties nor an equivalent risk.
The Verdict
Main sources
Sources accessed on October 17, 2026. AXA Prévention and Pourquoi Docteur are cited to examine their wording, not as primary medical evidence.
