Personality traits explain the relationship between psychedelic use and less depression in a comparative study

Data were collected via an online survey advertised in Swedish psychedelic-interest forums in February 2023. Inclusion criteria for the user sample were at least one classical psychedelic experience in the last year, age 16 or older, and native fluency in Swedish. Over a two-week recruitment window the researchers obtained 400 psychedelic users; the sample size was power-calculated to detect between-group differences at Cohen's d = 0.20. The average number of lifetime psychedelic experiences in the user sample was M = 6.9 (Md = 7.0; SD = 4.4). Mean participant age was 34.5 years (SD = 8.5), range 16–63; 68% were male and 32% female. A control group of 400 sex- and age-matched Swedish speakers without psychedelic experience was recruited from a survey panel drawn via the national personal address register (SPAR). Ethical approval was obtained from the Swedish ethical review authority (Dr: 2022-06810-02) and informed consent was collected online. The survey included standard self-report instruments: the PHQ-9 for depressive symptoms (9 items; Cronbach's α = .89 in this sample), the Insomnia Severity Index (ISI), a shortened 20-item Adolescent Dissociative Experience Scale (A-DES; α = .93), and a 30-item IPIP-NEO instrument to assess the Big Five personality traits (internal consistencies reported as N α = .88, E α = .81, O α = .78, A α = .75, C α = .78). Problematic substance use was screened with AUDIT (alcohol) and DUDIT (drugs). Group comparisons were performed with independent t-tests (using Levene's test to decide variance assumptions) and Cohen's d effect sizes; the authors treated d < 0.20 as trivial and d > 0.30 as of interest. The pre-specified significance threshold was p < .01. To test whether personality traits could account for mental health differences, analysis of covariance (ANCOVA) was used with the Big Five traits entered as covariates. The authors also report supplying a Pearson correlation matrix in supplemental materials and note that analyses were implemented in Jamovi v2.3.28.

Comparisons between the 400 psychedelic users and 400 matched non-users showed several notable differences. On primary mental health screens, psychedelic users scored lower on depressive symptoms measured by the PHQ-9 (effect size d = -0.29) and slightly lower on the ISI for insomnia symptoms; problematic alcohol use (AUDIT) and dissociation (A-DES) did not differ significantly between groups. By contrast, the DUDIT drug-use score was substantially higher among psychedelic users (d = 1.27), indicating greater reported drug use overall. Personality comparisons revealed robust between-group differences across the Big Five. The largest effect was for openness (d = 1.72), with users scoring much higher than non-users; agreeableness was also higher (d = 0.88), and users showed higher extraversion and conscientiousness and lower neuroticism (specific d values for the latter traits were reported but the extracted text emphasises openness and agreeableness as particularly large effects). Demographically, sex and age matching was successful; however, psychedelic users rated their current individual socioeconomic status (SES) higher (d = 0.42) while reporting slightly lower family home SES. When the researchers adjusted for Big Five personality traits using ANCOVA, the between-group difference in PHQ-9 scores became non-significant (ANCOVA reported F = 3.7, p ≈ .05, η² ≈ .00 when controlling for the Big Five). Neuroticism accounted for the largest share of PHQ-9 variance (F(1,767) = 211.5, η² = .23) and adjusting for personality reduced the PHQ-9 explained variance from about 2% to approximately 0% (as reported). For drug use (DUDIT), controlling for personality reduced the explained variance from 29% to 20%, but the group difference in DUDIT remained statistically significant (F(1,767) = 186.1, p < .001). The authors note that personality covariates thus explained most or all of the depression difference and around one third of the drug-use difference between groups.

Sjöström and colleagues interpret their findings as showing that, in this Swedish sample, recreational psychedelic users report slightly lower depressive symptoms but substantially higher overall drug use, and that the apparent depression advantage is largely accounted for by differences in baseline personality traits. They emphasise that openness was markedly higher among users (an effect size close to two standard deviations), with higher agreeableness and extraversion and lower neuroticism also observed. Controlling for the Big Five removed the statistical difference in depressive symptoms and reduced the drug-use difference, suggesting that personality is an important confounder when comparing mental health across users and non-users. The authors discuss two non-exclusive explanations: individuals with particular personality profiles (for example high openness or low neuroticism) may be more likely to seek psychedelic experiences, or psychedelic experiences might induce changes in personality (for example increases in openness) that then relate to mental health. They note that prior clinical and naturalistic studies have reported antidepressant effects of psychedelics and longer-term personality change, but emphasise that their cross-sectional data cannot resolve causality. The elevated DUDIT scores among users are discussed as potentially reflecting the instrument's focus on illegal drug use regardless of drug type; participants may have reported psychedelic use rather than dependence-inducing substances, which could confound interpretation. The lack of higher dissociation or alcohol problems in users is highlighted, and the authors propose that neuroticism may modulate vulnerability to sustained dissociation. Key limitations the study team acknowledges include the cross-sectional design that precludes causal inference, self-selection and volunteer bias inherent in forum-based recruitment, and the use of brief screening instruments rather than clinical diagnostic interviews. They also note that psychedelic users rated current SES higher than controls, which may relate to personality and mental health differences. The authors recommend including personality measures in future psychedelic research to disentangle selection effects from potential causal effects of psychedelic experiences.

The separate Conclusion restates the main takeaways: recreational psychedelic users in this national Swedish sample reported marginally better mental health on screening measures, but this apparent advantage appears to be at least partly attributable to differences in personality traits — particularly higher openness. The authors argue for routine inclusion of personality assessment in psychedelic research to better understand how individual differences interact with psychedelic experiences and reported mental health outcomes.