Health-related behavioral changes following the use of psychedelics in naturalistic settings

Psychedelics, once widely stigmatised, are increasingly studied for therapeutic benefits. Earlier clinical and observational work has linked both classic psychedelics (for example LSD and psilocybin) and non-classic compounds (for example ketamine and MDMA) to sustained improvements in psychiatric symptoms and reductions in substance misuse; some trials reported high abstinence rates for tobacco and reductions in alcohol use. Emerging observational evidence also suggests broader effects on health-related behaviours such as diet, physical activity, and reduced incidence of cardiometabolic disease, and a few studies have reported associations between lifetime psychedelic use and lower odds of criminal or violent behaviour. However, whether therapeutic effects observed in controlled settings generalise to naturalistic use remains uncertain, and the behavioural impacts of psychedelics in real-world contexts are not well characterised. Teixeira and colleagues set out to address this gap using data from the Psychedelics and Wellness Study (PAWS). The present analysis examines associations between lifetime psychedelic use and self-reported changes in a range of health-related behaviours following psychedelic experiences, specifically eating habits, substance use (alcohol, cannabis, opiates, benzodiazepines, tobacco), and impulses related to aggression, criminality, and suicidality. The study also explored whether reported changes varied by specific psychedelic substance, dosing pattern (including microdosing), and frequency of use.

The study employed a cross-sectional, anonymous online survey asking participants to retrospectively report perceived changes in behaviours attributed to their past psychedelic use. Inclusion criteria required participants to be at least 18 years old and to have used one or more classic psychedelics (LSD, psilocybin, DMT, 5-MeO-DMT, ayahuasca, or mescaline) at least once in their lifetime. Data collection occurred between August 2019 and July 2023, and the final analytic sample comprised 2,510 eligible respondents who provided informed consent. The Western Institutional Review Board determined the study exempt from full review under applicable regulations. Recruitment relied on convenience and snowball methods, including online platforms, social media, email, flyers, word-of-mouth and in-person outreach; no additional screening criteria were applied beyond age and lifetime psychedelic use. Participants first reported demographic and psychedelic-use characteristics — for example preferred substance, lifetime number of uses, and history of microdosing — before completing the study instruments. Behavioural outcomes were measured using two study-specific instruments developed for PAWS: the 26-item Psychedelic Change Questionnaire (PCQ-26) and the 8-item Negative Consequences Inventory (NCI-8). The PCQ-26 includes an item asking respondents to rate change in eating habits on a 1 (very much improved) to 7 (very much worse) scale. The NCI-8 asks respondents to rate change in eight negative outcomes attributed to psychedelic use — including suicidal desire, criminal and aggressive impulses/behaviours, and misuse of alcohol, cigarettes, cannabis, benzodiazepines and opiates/opioids — using the same 1–7 response format. Participants also indicated preferred psychedelic substance and whether they had ever microdosed. The extracted text repeats aspects of recruitment and ethical oversight but provides limited detail on the statistical approach. Reported analyses compared proportions reporting improvement versus no change or worsening across behaviours, and examined associations between reported changes and characteristics of psychedelic use (lifetime frequency, microdosing, and preferred substance). The extracted text does not clearly report the full set of statistical tests, modelling strategies, or adjustment for potential confounders; the abstract and results mention significance tests (for example p < .001 for associations with frequency and microdosing) but detailed methods for these tests are not provided in the available extraction.

A substantial fraction of respondents attributed positive behavioural changes to their psychedelic experiences. Improvements were reported for all assessed behaviours, with the proportion reporting improvement ranging from 35.9% for cannabis use to 72.1% for aggressive impulse control. More than half of participants indicated positive change for alcohol misuse (66%) and suicidal ideation (68.9%). Approximately half reported healthier eating behaviour (49.4%), reduced smoking (48.5%), and reduced criminal impulses (48.2%). Reported reductions were also observed for opiate/opioid misuse (40.2%) and benzodiazepine misuse (36.8%). For behaviours with positive change, responses were concentrated in the stronger improvement categories ("much" or "very much" improved) rather than minimal improvement. Analyses relating behavioural change to characteristics of psychedelic use showed consistent patterns. Participants who reported positive changes tended to report a higher median lifetime number of psychedelic uses compared with those reporting no change or negative outcomes; except for cannabis misuse, these differences were statistically significant across behaviours as reported in the extracted text. Engagement in microdosing was associated with higher rates of self-reported positive change across all health behaviour domains examined, and the abstract notes that more frequent use or microdosing were associated with greater likelihood of endorsing positive changes (p < .001). Comparisons by substance showed one notable difference: psilocybin users were more likely than LSD users to report positive changes in eating behaviours. Other pairwise substance comparisons did not reach significance in the available extraction. The results section notes variability in sample size across specific items, reflecting that participants who had not engaged in a given behaviour (for example non-smokers) were not asked to rate change in that domain. The extracted text does not present confidence intervals, adjusted effect estimates, or full statistical tables in the provided content, and detailed subgroup sample sizes beyond the overall n = 2,510 are not clearly reported.

Teixeira and colleagues interpret the findings as extending PAWS evidence that naturalistic psychedelic use is frequently associated with self-reported improvements in a range of health-related behaviours. The authors emphasise notable reductions in alcohol and tobacco use, healthier eating patterns, and decreases in impulsive behaviours including aggression, criminal tendencies and suicidal ideation. They present these results as broadly consistent with prior observational and clinical research suggesting that psychedelic exposure can catalyse positive behavioural modifications beyond symptom relief for mental disorders. The discussion situates the findings alongside studies of ayahuasca and other psychedelic-using cohorts that reported higher physical activity, improved diet and lower alcohol consumption compared with the general population, and with research reporting lower rates of violent or antisocial behaviours among people reporting lifetime classic psychedelic use. The authors also draw attention to a potential substance-specific signal, noting that psilocybin was more strongly associated with improved eating behaviours than LSD, while emphasising that most substance comparisons did not show significant differences. Microdosing emerged as an important correlate of self-reported improvement; participants who ever microdosed were significantly more likely to endorse positive changes across domains. The authors acknowledge that microdosing research is nascent, citing preliminary evidence of short-term behavioural and neural effects in healthy adults, but they note that evidence for clinical benefits in patient populations remains lacking. Several limitations are acknowledged. The reliance on retrospective self-report introduces recall and social desirability biases, and the convenience online sample likely favours individuals with positive experiences, producing selection bias. The cross-sectional design precludes causal inference; the authors explicitly state that observed associations cannot confirm that psychedelic use caused the behavioural changes. Finally, they call for longitudinal and controlled studies to validate these findings and to clarify the role psychedelics might play in promoting healthier lifestyle behaviours at both individual and population levels. The authors present their conclusions cautiously, suggesting the results add to a growing body of real-world evidence but stopping short of asserting therapeutic causation.