Emotional breakthrough and psychedelics: validation of the emotional breakthrough inventory

Psychedelic therapy is regaining attention as a treatment paradigm in mental health, and previous work has shown that qualities of the acute psychedelic experience help to determine longer-term psychological outcomes. Much prior research has emphasised mystical-type or "peak" experiences (commonly measured with the Mystical Experience Questionnaire, MEQ) and challenging experiences (measured with the Challenging Experience Questionnaire, CEQ) as important state predictors. The authors argue that these instruments do not clearly capture a related but distinct phenomenon: the overcoming or resolution of difficult emotions or memories during a psychedelic experience, which they term emotional breakthrough (EB). They note conceptual overlap between EB and historical notions of catharsis, and suggest EB may act as a positive predictor of subsequent mental-health improvements whereas CEQ scores have sometimes predicted worse outcomes. Roseman and colleagues set out to develop and validate a brief Emotional Breakthrough Inventory (EBI). The study aimed to assess the EBI's internal consistency, factor structure, discriminant validity versus the MEQ and CEQ, dose- and context-dependence, and its ability to predict changes in psychological well-being. The investigators tested five explicit hypotheses concerning dose-dependence, influence of therapeutic intention/setting, discriminability from existing scales, and predictive power for changes in well-being, both alone and combined with MEQ and CEQ measures. The work was conducted as a prospective, naturalistic online survey of self-selected psychedelic users rather than a controlled laboratory trial.

The study was a prospective, observational online survey advertised via social media and hosted at psychedelicsurvey.com. Ethical approval and written informed consent were obtained. Baseline data were collected one week before participants' planned psychedelic experience and included demographics, intentions for the session and the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS), which served as the primary dependent variable. Retrospective reports of the acute psychedelic experience were collected one day after use and comprised the newly developed EBI, the MEQ and the CEQ, plus self-reported drug type and an estimated dose expressed in LSD-equivalent categories. Long-term outcomes (WEMWBS) were collected two weeks after the experience. Most analyses used data from 379 participants; prediction of change in well-being focused on a subsample of 75 participants with relatively low baseline WEMWBS scores (⩽45) to reduce floor/ceiling issues. EBI item construction began with eight candidate items derived from prior trial interview content and clinical observation; two items were negatively keyed. Items were rated on a 0–100 visual analogue scale. Five experts reviewed the items for face validity. Other instruments used were the 30-item MEQ and the 26-item CEQ (scored 0–100 after rescaling), and the 14-item WEMWBS (scored over two weeks). Drug-type options included psilocybin, LSD, ayahuasca, DMT variants, salvia, mescaline, ibogaine and free-text responses. Dosing was categorised into five LSD-equivalent bands from low to extremely high. Participants also reported whether the setting was designed with a therapeutic objective and rated therapeutic intention and willingness to confront difficult emotions. Statistical analysis used SPSS 25 and Matlab 2017. Exploratory factor analysis via principal component analysis (PCA) determined the EBI factor structure and informed item retention. Internal consistency was assessed with Cronbach's alpha (a measure of how closely related items are within a scale). Discriminant validity was tested by entering all 62 items from the EBI, MEQ and CEQ into a single PCA to evaluate whether EB emerged as a distinct factor. Predictive validity was evaluated with Pearson correlations and multiple regression analyses predicting ΔWEMWBS (two-week change score), with the EBI, MEQ and CEQ entered as independent variables. An initial prediction analysis used the targeted low-well-being subsample (n=75) and a secondary exploratory regression used a larger sample with available follow-up WEMWBS (n=253). Collinearity diagnostics (tolerance values) were reported for regression predictors.

Exploratory factor analysis of the eight initial EBI items (n=379) supported a two-factor solution. The Kaiser-Meyer-Olkin measure was 0.877 and Bartlett's test was significant, indicating suitability for PCA. Factor 1 comprised six positively keyed items and explained 57.1% of the variance; Factor 2 contained the two negatively keyed items and explained 18.3%. The two negative items were therefore discarded and the six-item EBI used in subsequent analyses. Internal consistency for the six-item EBI was very high (Cronbach's alpha=0.932). The average of the six items correlated strongly with the factor score (r=0.996, p<0.0001) and was used as the total EBI score. Mean scores were reported as: EBI 43±31.5, MEQ 57±22.63, and CEQ 19.7±16.4. EBI scores showed expected dose- and context-dependence. Estimated drug dose correlated positively with EBI (Spearman r s =0.192, p=0.0001, n=379). Baseline therapeutic intention and willingness to confront difficult emotions correlated with higher EBI scores (r s =0.256 and r s =0.279, respectively, p<0.0001, n=351). Participants who reported their setting was designed with a therapeutic objective had higher EBI scores (47.9±30.1, n=144) than those who did not (31.6±29.1, n=168); this difference was highly significant (p<0.0001) with a medium effect size (Cohen's d=0.55). Discriminant validity testing used PCA on all 62 items from EBI, MEQ and CEQ (KMO=0.941, Bartlett's test significant). A three-factor solution emerged: a first factor loaded heavily on MEQ items (with some EBI loading), a second factor loaded on CEQ items, and a third factor loaded on EBI items alone. The first three components explained 31.3%, 15.4% and 6.1% of the variance, respectively. Correlation analysis showed the EBI correlated strongly with MEQ dimensions and with the grief subscale of the CEQ, but MEQ subscales correlated more strongly with each other than with EBI, supporting discriminability. Predictive validity analyses focused on the low-baseline well-being subsample (n=75). For these participants, WEMWBS increased from 38.8±5.3 at baseline to 47.9±7.2 at two weeks (paired t-test significant). EBI scores correlated with change in well-being (ΔWEMWBS) at r=0.294, p=0.005 (one-tailed). A multiple regression with ΔWEMWBS as the dependent variable and EBI, MEQ and CEQ as predictors produced a significant model (r=0.45, p=0.0005, one-tailed). Standardised beta coefficients were all significant and in the predicted directions: βEBI=0.29, p=0.017; βMEQ=0.24, p=0.038; βCEQ=-0.35, p=0.002 (one-tailed). Collinearity tolerances were acceptable (0.631, 0.634, 0.814), indicating the predictors provided relatively independent information. An exploratory regression in a larger sample with two-week WEMWBS available (n=253) yielded a marginally significant overall model (r=0.153, p=0.059, one-tailed). In this group WEMWBS increased from 49.5±8.6 at baseline to 52.7±7.3 at two weeks. In the larger sample the standardised betas were smaller and marginal: βEBI=0.115, p=0.063; βMEQ=0.075, p=0.15; βCEQ=-0.086, p=0.01 (one-tailed). The authors attribute the reduced effect size in the larger sample to a ceiling effect at baseline WEMWBS. Overall, the combined model of EBI, MEQ and CEQ accounted for close to 20% of variance in well-being change in the targeted low-baseline sample.

Roseman and colleagues interpret their findings as supporting the EBI as a reliable, concise measure of a distinct component of the acute psychedelic experience that they label emotional breakthrough. The six-item EBI demonstrated strong internal consistency, dose- and context-dependence, and discriminant validity versus established measures of mystical-type and challenging experiences. Consistent with the authors' hypotheses, higher EBI scores were associated with greater increases in psychological well-being two weeks after a psychedelic experience, and a model combining EBI, MEQ and CEQ performed better at predicting well-being change than any single instrument alone. The combined model explained nearly 20% of variance in the targeted low-baseline subsample, which the authors consider notable given the non-clinical, largely well sample and the uncontrolled naturalistic design. The investigators emphasise the importance of context: therapeutic intention and a therapeutic setting were associated with stronger EB, reinforcing prior suggestions that contextual factors (for example music, empathic support and therapeutic alliance) are key mediators of psychedelic-assisted interventions. They situate EB alongside other candidate state and contextual predictors such as psychological insight and integration, and argue for broader measurement of these constructs in future work. The discussion also notes the potential neurobiological relevance of state measures, briefly referencing prior findings that brain entropy increases during psychedelics correlate with later psychological change. Three limitations highlighted by the authors are the sample composition (mostly healthy, experienced psychedelic users, limiting clinical generalisability), reliance on self-reported drug identity and dose in uncontrolled settings, and limited convergent-validity analyses owing to lack of existing measures closely aligned to EB. The authors recommend applying the EBI in clinical trials and controlled settings, comparing it with qualitative interview outcomes, and examining other dependent variables beyond general well-being (for example disorder-specific symptom change). They conclude that EB is an important component of psychedelic-assisted therapy that should be routinely measured alongside peak and challenging experiences to better quantify psychological mechanisms that accompany pharmacological effects.