The study matters because meditation training has exploded far beyond monasteries and meditation centers. It is now embedded in psychotherapy clinics, hospitals, schools, and corporate wellness programs, and surveys suggest tens of millions of adults practice some form of it. Alongside that growth, a serious scientific conversation has emerged about unintended harm. Researchers have documented meditation-related experiences ranging from time-limited anxiety and traumatic re-experiencing to, in rarer cases, lasting problems such as executive dysfunction, delusional beliefs, hallucinations, depression, and insomnia. Just as potent psychological or pharmacological therapies can produce adverse effects, the argument goes, so can meditation, and clinicians need to know who is most at risk.

Until now, the evidence pointing to mental health symptoms as a risk factor came largely from retrospective, cross-sectional surveys. In those studies, meditators with substantial prior experience, often more than one hundred lifetime hours, reported on past symptoms and past distressing experiences, sometimes months or years after the fact. Such designs suffer from well-known recall biases, and, critically, they lack a comparison group of non-meditators. Without knowing the base rate of distressing experiences in daily life, it is impossible to say whether difficult experiences during meditation are caused by the practice or simply reflect what people with elevated symptoms experience anyway. A person prone to anxiety may report intense anxiety during a retreat, but they may also report intense anxiety during an ordinary week at home.

A handful of prospective studies tried to address the recall problem by measuring symptoms before mindfulness-based interventions and tracking outcomes afterward, with mixed results. But those studies typically involved moderate practice doses, from zero to forty-five minutes a day over four to eight weeks, and were designed as treatments tailored to people with mental health difficulties. None included a non-meditation control group. The highest-intensity common format, the residential retreat, with six to eleven hours of daily practice in a secluded setting, remained comparatively underexamined, even though prior work suggests most distressing or impairing meditation-related experiences begin during or shortly after retreats.

The new study, a secondary analysis of a trial preregistered at ClinicalTrials.gov in December 2020, tackled these limitations with a prospective matched-control design. Eighty-nine participants enrolled in one of five six-day residential insight (Vipassana) retreats run by the Tovana Israel Insight Meditation Society, while forty-four controls were drawn from the same meditation community and matched on age and lifetime meditation experience. The retreat schedule was demanding: at least ten roughly forty-five-minute sessions of sitting and walking meditation per day, totaling a minimum of seven and a half hours of formal practice, embedded in silence, alongside loving-kindness sessions, mindful movement, and mindful work periods. Teachers gave daily talks and small-group interviews, and participants progressively shifted from focused-attention practice to broader, deconstructive awareness practices.

Before the retreats began, all participants completed the Patient Health Questionnaire-9 for depression and the Overall Anxiety Severity and Impairment Scale for anxiety. During the retreat, or an equivalent six-day window for controls, they filled out the Peak Meditative Experience Scale, reporting on thirty-two unusual, intense experiences and rating each one’s hedonic tone from extremely unpleasant to extremely pleasant. Nine items, including intense anxiety, flashbacks, and overwhelming sadness, had previously been identified as predominantly unpleasant. Two weeks after the retreat, participants rated both the positive and negative impact of each endorsed unpleasant experience, allowing the researchers to compute a net post-retreat impact score. The sample showed wide variability in symptoms, though most participants fell below clinical screening cutoffs, and roughly two-thirds had more than one hundred lifetime meditation hours.

The statistical strategy centered on bootstrapped moderation models testing whether baseline symptoms interacted with group membership in predicting unpleasant experiences. The hypothesis was straightforward: higher symptoms should predict more distressing experiences among retreat participants relative to controls. Instead, the interaction for depression symptoms was significant in the opposite direction. Retreat participants scoring one standard deviation below the mean on depression endorsed significantly more unpleasant peak experiences than matched controls with similar symptoms, while those at or above the mean did not differ from controls. The anxiety results followed the same pattern, with a marginally significant interaction and the same reversal among low-symptom participants. Sensitivity analyses examining the valence, frequency, and duration of unpleasant experiences found no evidence that elevated symptoms increased risk on any dimension, and the findings held after controlling for lifetime meditation hours, retreat days, and gender.

The two-week follow-up added a further twist. The relationship between pre-retreat anxiety and the net impact of distressing experiences was best described by a quadratic curve: participants with both low and high baseline anxiety reported that their unpleasant retreat experiences had, on balance, a more positive than negative impact two weeks later, while those with moderate anxiety showed a net zero difference. Depression symptoms showed no significant relationship with post-retreat impact at all. The authors suggest several interpretations. Low-symptom participants may have been more resilient to begin with, and resilience is linked to evaluating difficult experiences as opportunities for growth. High-symptom participants, meanwhile, may have benefited most, consistent with evidence that people with greater baseline vulnerability often gain more from mindfulness-based interventions. Intensive practice in a supportive setting, with teachers available and instructions emphasizing non-judgment and acceptance, may even function as a form of stress inoculation, teaching practitioners to approach anxiety rather than flee from it.

Perhaps the most consequential implication concerns causation. The study found that greater depression and anxiety symptoms were associated with more unpleasant experiences in both groups, meditators and controls alike. That means the correlations reported in earlier retrospective studies may be real and replicable, but attributable not to meditation itself. Distressing experiences among people with elevated symptoms may simply be part of the background noise of daily life, and intensive practice neither amplifies nor uniquely generates them, at least in this retreat format. The authors argue this demonstrates the necessity of no-meditation control conditions for anyone hoping to attribute distressing experiences or adverse effects to meditation training, a methodological point with echoes of long-standing debates about distinguishing treatment effects from base-rate phenomena in psychotherapy research.

The study has limits the authors are careful to acknowledge. A fully randomized active-control design would provide stronger causal evidence than matching, the sample was not well powered to detect small interaction effects, and most participants fell below clinical thresholds for depression and anxiety, so the findings may not generalize to severely symptomatic clinical populations. The researchers also call for future work on symptoms beyond depression and anxiety, including posttraumatic stress and psychotic symptoms, on longer and more intensive retreat formats, and on replication across retreat centers, traditions, and cultures. Still, if the results hold up in larger and more diverse samples, the clinical message is striking: depression and anxiety symptoms alone may not constitute contraindications to intensive meditation retreat training, and for some practitioners, the difficult moments on the cushion may ultimately be the ones that help the most.

Subject of Research: Whether pre-existing depression and anxiety symptoms predict distressing experiences during intensive mindfulness meditation retreat training

Article Title: Mental Health Symptoms Do Not Increase Risk for Distressing Experiences During Intensive Mindfulness Meditation Training: Evidence from a Prospective Matched-Control Intervention Study

Article References: Mental Health Symptoms Do Not Increase Risk for Distressing Experiences During Intensive Mindfulness Meditation Training: Evidence from a Prospective Matched-Control Intervention Study. (n.d.). https://doi.org/10.1007/s12671-026-02982-9

Image Credits: AI Generated

DOI: 10.1007/s12671-026-02982-9

Keywords: mindfulness, meditation retreats, adverse effects, depression, anxiety, matched-control study, insight meditation, mental health risk factors, Vipassana, clinical psychology, contemplative science, preregistered research

The problem the researchers set out to solve is surprisingly basic. Mindfulness research has exploded over the past two decades, with hundreds of trials showing benefits for mental health and well-being in both clinical and non-clinical populations. Yet the overwhelming majority of studies have measured practice quantity, such as how many minutes or sessions participants completed, while treating practice quality as an unmeasured black box. Existing mindfulness questionnaires capture related but different constructs, including trait mindfulness, which reflects a person’s general disposition, and state mindfulness, which captures momentary awareness during practice. Others assess adherence to practice schedules or mechanisms like equanimity. None, according to the authors, had ever been built to assess the quality of the practices themselves, even though evidence suggests that quality, not just duration, may drive outcomes, particularly for novice meditators who have not yet learned to hold attention steadily.

Drawing on Buddhist psychology and modern clinical practice, the team focused on the two practices they describe as the fundamental pillars of mindfulness meditation: concentration, known in the Pali tradition as śamatha, and insight, known as vipassanā. In śamatha practice, the meditator rests attention on a single object, typically the sensation of breath at the nostrils, training the executive functions of sustaining attention, inhibiting the habitual pull toward thoughts, and shifting attention back without frustration. In vipassanā practice, the meditator opens awareness to whatever arises, observing body sensations, emotions and thoughts as impermanent, impersonal phenomena. The researchers deliberately excluded loving-kindness practice from the initial instrument, noting that classical taxonomies classify it as an affective-directed method distinct from the cognitive-directed śamatha and vipassanā practices taught in the earliest Buddhist sources.

The MMPQS was designed to be completed immediately after a meditation session, capturing the practice while it remains fresh. Each of the two scales began with ten items rated on a seven-point Likert scale ranging from not at all true to completely true. Item development drew on traditional meditation manuals and the clinical teaching sequence of MiCBT, and items were refined through an iterative consensus process among experienced meditation teachers on the research team. Concentration items probe the ability to stay focused and to refocus quickly after distraction, with examples such as being able to return to the breath after thoughts or images intervene. Insight items assess the capacity to feel a range of sensations in the body, to perceive experience with objectivity and distance, and to avoid reacting to physical, emotional or mental events, mapping directly onto the interoceptive awareness, equanimity and impermanence components of vipassanā.

What distinguishes this study methodologically is its use of Rasch analysis, a psychometric approach that many readers will find both technical and quietly revolutionary. Traditional scoring of questionnaires assumes that Likert-scale responses behave like interval data, so that the difference between scores of 25 and 26 means the same as the difference between 29 and 30. In reality, ordinal ratings rarely satisfy this assumption, introducing measurement error into every downstream analysis. Rasch models explicitly acknowledge the ordinal nature of the data and provide a mathematically principled conversion from raw summed scores to interval-level estimates on a logit scale. The researchers provide conversion tables so that clinicians and researchers can transform raw MMPQS scores without additional software, enabling precise detection of change over time and meaningful comparisons across groups and studies.

The Rasch analysis itself was an iterative winnowing process. Data came from 368 clinical participants at two psychology clinic sites, all diagnosed with at least one comorbid mental health condition and most seeking help for anxiety disorders. Participants completed the Concentration scale after concentration practices and the Insight scale after insight practices on two measurement occasions, with combined samples of 414 and 313 observations respectively. On the Concentration scale, items measuring things like feeling that time passed quickly or feeling motivated and engaged proved to be double-barreled, vague or peripheral to the construct, while two nearly identical refocusing items created local dependency. Removing these and combining two remaining dependent item pairs as testlets yielded a clean six-item scale with high reliability, person separation index values between 0.89 and 0.94, and strict unidimensionality.

The Insight scale presented a different pattern of problems. Its initial items all fit the Rasch model, but the scale failed tests of unidimensionality, and one item showed differential item functioning by age, meaning it behaved differently for younger and older respondents. Through successive rounds of analysis, redundant and double-barreled items were removed until a six-item version emerged that preserved full coverage of the insight construct: two items for interoceptive awareness, two for equanimity, and two for awareness of impermanence and non-self. Neither final scale showed differential item functioning for age, gender, prior meditation experience or time of measurement, indicating that the instruments perform equivalently across these groups and that changes in scores over the course of treatment reflect genuine improvements in practice quality rather than shifting item behavior.

Validity testing delivered results that the authors describe as consistent with their hypotheses. Both scales showed small to moderate correlations in the expected directions with measures of equanimity, depression, anxiety, stress, awareness, kindness, life satisfaction and self-efficacy for meditation, supporting construct and divergent validity. More striking were the predictive findings. Concentration practice quality significantly predicted lower levels of depression, anxiety and stress, though not life satisfaction. Insight practice quality predicted reduced symptoms and, uniquely, increased life satisfaction. The authors suggest the difference may reflect the sequential structure of the therapy: concentration training begins in week two and is practiced for only a week or two before insight practices are introduced, possibly too brief a window for patients to feel meaningful gains in life satisfaction so early in treatment. Scores on both scales also rose significantly from the first to the second measurement occasion, consistent with the idea that practice quality improves with continued training.

For clinicians and teachers, the appeal of the MMPQS lies in its practicality. Because the scales are short and administered immediately after practice, they can serve as feedback tools, helping novice meditators understand the skills their practice is meant to cultivate, tracking their progress session by session, and flagging specific difficulties, whether sustaining attention or holding equanimity in the face of unpleasant sensations, that may warrant extra guidance. A participant scoring 25 on the raw Concentration scale, for example, converts to an interval score of 22.11, while a score of 30 converts to 25.10, with equal intervals between converted scores now representing genuine equal units of the underlying ability rather than arbitrary ordinal steps.

The authors are careful about the limits of their findings. The validation sample was culturally homogeneous, predominantly White Australian adults treated within a single therapy program, and the iterative, partly data-driven item selection could have capitalized on chance characteristics of this particular sample, so confirmation in independent samples is needed before the six-item solutions are treated as final. Detailed practice frequency and duration data were not collected, and the temporal lag between concentration and insight training introduces a confound into the predictive analyses. Future work will extend the MMPQS to non-clinical populations, culturally diverse samples, other mindfulness-based programs, and eventually a loving-kindness subscale. Still, the study marks a turning point: for the first time, researchers can measure not just whether people meditate, but whether they meditate well, opening a path toward understanding precisely how the quality of attention, equanimity and insight translates into relief from suffering.

Subject of Research: Development and validation of scales measuring the quality of mindfulness meditation practices using Rasch analysis

Article Title: Development and Validation of the Mindfulness Meditation Practice Quality Scales (MMPQS) in a Clinical Population Using Rasch Analysis

Article References: Strohmaier, S., Cayoun, B. A., Shires, A. G., Medvedev, O. N., & Krägeloh, C. U. (2026). Development and Validation of the Mindfulness Meditation Practice Quality Scales (MMPQS) in a Clinical Population Using Rasch Analysis. Mindfulness. https://doi.org/10.1007/s12671-026-02990-9

Image Credits: AI Generated

DOI: 10.1007/s12671-026-02990-9

Keywords: mindfulness, meditation, Rasch analysis, psychometrics, MMPQS, clinical psychology, Mindfulness-integrated Cognitive Behavior Therapy, concentration meditation, insight meditation, mental health, measurement, equanimity