Mental health of a self-selected sample of psychedelic users and self-medication practices with psychedelics

Mason and colleagues situate the study within renewed scientific interest in classical psychedelics (for example, psilocybin, LSD, ayahuasca) and related entactogens (MDMA) as potential treatments for a range of psychiatric conditions, including treatment‑resistant depression, PTSD and substance-use disorders. They note that although standard therapies (psychotherapy, antidepressants, anxiolytics) benefit many patients, a substantial minority remain symptomatic, motivating exploration of novel therapeutic agents. Population surveys have reported that lifetime use of psychedelics is not consistently associated with worse mental health and in some large samples has been linked to lower odds of psychological distress and suicidality. This background raises the question whether people who have used psychedelics attempt to self‑medicate psychiatric or physical illness with these substances. The present study therefore aimed to estimate the prevalence of self‑reported psychiatric disorders in a self‑selected sample of psychedelic users and to assess whether those with diagnosed disorders sought treatments outside conventional medical advice, specifically whether they used psychedelics to treat symptoms. The investigators also compared self‑rated effectiveness of standard treatments offered by medical professionals with self‑administered psychedelic treatments, and hypothesised that psychedelic users with a diagnosed psychopathology would be likely to self‑medicate with psychedelics.

The study used an online cross‑sectional questionnaire distributed via several websites and forums between May and July 2017. Eligibility required participants to be aged 18 years or older. The survey was implemented in Qualtrics; 4,892 individuals began the survey and 1,967 (40%) provided consent, met the age criterion and completed the questionnaire. Reported completion time averaged 96 minutes, though it varied with the number of substances a respondent had used. Collected measures included demographics (age, gender, continent, highest education), lifetime history of use of multiple substances (alcohol, nicotine, cannabis, MDMA, psilocybin, LSD, ayahuasca, and novel psychoactive substances), and detailed follow‑up questions on frequency, age of first use, recency, combinations with other substances, adverse effects, typical setting and motives for use. Mental and physical health questions comprised binary screening items (for example, "Have you ever suffered from a mental disorder?"), and, where affirmative, an open‑ended field to list disorders. The investigators subsequently mapped reported mental disorders to 20 DSM‑5 categories. Participants who reported a disorder were asked whether it had been diagnosed by a medical professional, whether any treatment had been offered (medication, therapy, or other), whether they had sought treatments outside medical recommendation, and whether they had ever used psychedelics to treat the condition. Treatment effectiveness was assessed using three 100‑mm visual analogue scales (VAS): perceived whether the treatment "worked," whether symptoms had remitted sufficiently to restore daily functioning, and whether quality of life (QOL) improved. VAS scores were summarised as means and standard deviations and were also categorised into quartiles (Q1: 0–25%; Q2: 26–50%; Q3: 51–75%; Q4: 76–100%). To compare standard treatments offered by medical professionals with self‑administered psychedelic treatments, the authors used binary logistic regression on the three effectiveness questions, contrasting the highest quartile (Q4) versus the lowest quartile (Q1). Reported outputs included odds ratios (ORs) with 95% confidence intervals where available, and significance was set at p ≤ .05. Analyses were performed in SPSS v24.0. Ethical approval was obtained from the relevant institutional committee and participants provided informed consent.

The analysed sample comprised 1,967 respondents with a mean age of 25.9 years (SD 8.7; maximum age 71). Most participants were male (N = 1,549; 79%; mean age 25.2, SD 8.4), 20% were female (N = 392; mean age 28.1, SD 9.1) and 1% (N = 26) identified as "other." Educational attainment was relatively high: 57% reported university education and 12% postgraduate degrees. Participants reported origins spanning five continents, principally America (60%) and Europe (33%). Lifetime drug‑use frequencies (presented in a table in the paper) indicated that the three most commonly used substances were cannabis (94%), alcohol (91%) and LSD (79%). The least commonly reported were ayahuasca (8%), cannabinoid novel psychoactive substances (8%) and stimulant NPS (13%). The majority of respondents rated their current mental health favourably on VAS, with a sample mean of 74.8% (SD 23.7); physical health mean was 74.4% (SD 20.9). Over half of respondents rated mental and physical health in the top quartile (Q4). Regarding mental disorders, 900 respondents (46%) indicated they had ever suffered or were currently suffering from a mental disorder; of these, 77% (N = 698) reported a professional diagnosis. Among diagnosed cases, 99% (N = 691) reported being offered some form of treatment: medication was reported in 81% (N = 558) of diagnosed cases, therapy in 71% (N = 494), and 8% (N = 53) reported another type of offer. When asked whether they searched for treatments outside medical recommendations, 77% (N = 540) of diagnosed respondents answered affirmatively; 81% (N = 436 of those who sought alternatives) said they had used or were using psychedelics to treat the disorder. Self‑rated effectiveness differed markedly between standard treatments and self‑administered psychedelics. For treatments offered by a medical professional, mean VAS scores (±SD) were approximately 49% (±32.3) for "did it work," 48% (±34.0) for "symptoms disappeared sufficiently," and 54% (±35.9) for "QOL improved." For self‑administered psychedelic treatment, means were higher: 81% (±20.3) for "did it work," 72% (±25.3) for "symptoms disappeared," and 87% (±19.3) for "QOL improved." Logistic regression contrasting Q4 versus Q1 indicated significantly greater odds of reporting high effectiveness for psychedelic self‑treatment compared with offered treatments on the three questions: OR for "did it work" = 41.93 (p < .001; 95% CI not fully provided in the extracted text), OR for "symptoms disappear" = 4.24 (p < .001; 95% CI [2.94, 6.13]), and OR for "QOL improved" = 1.56 (p = .01; 95% CI not fully reported). On familial mental‑health measures, 39% of respondents reported a first‑degree relative diagnosed with a mental disorder and 11% reported a first‑degree relative who had committed suicide. With respect to physical health, 21% (N = 421) reported ever having a physical disorder; 88% of these were diagnosed by a medical professional and 85% were offered treatment (medication in 83% of diagnosed cases, therapy 19%, other 28%—percentages indicate multiple categories could apply). About half (52%) of those diagnosed with a physical disorder sought treatment outside medical recommendations and 28% reported using psychedelics to treat the physical condition. Mean VAS ratings for standard medical treatments of physical disorders were 70% (±30.6) for "did it work," 66% (±33.8) for "symptoms disappeared," and 69% (±33.4) for "QOL improved." Only the "symptoms disappeared" outcome was entered into regression for physical disorders due to small cell counts for the other items; that OR was 0.89 and not statistically significant (p = .81).

Mason and colleagues interpret their findings as indicating a relatively high lifetime prevalence of self‑reported psychiatric disorders among this self‑selected sample of psychedelic users: 46% reported ever having a mental disorder, with depression and anxiety the most common. The authors note that these prevalence figures appear higher than typical general‑population lifetime rates cited in epidemiological studies, though they acknowledge heterogeneity across surveys and age ranges. They emphasise that the most prevalent disorders in the general population (depression and anxiety) were also dominant in the current sample. A central finding highlighted is that 77% of participants with a diagnosed mental disorder had searched for treatments outside medical recommendation and that 81% of those reported having used psychedelics as a self‑treatment. Self‑administered psychedelic treatments were rated substantially more effective on average than treatments offered by medical professionals, and logistic regression indicated higher odds of reporting maximal effectiveness for psychedelics on the three effectiveness measures. The authors relate these self‑reports to previous case series and survey work suggesting beneficial effects of ayahuasca and other psychedelics and to population studies that have not found associations between lifetime psychedelic use and worse mental health and, in some cases, suggest reduced suicidality. Several important limitations are acknowledged. The sample was self‑selected and the survey relied on retrospective, self‑rated effectiveness measures, which may introduce bias. The cross‑sectional design and the way questions were asked did not distinguish clearly between current versus past disorders or establish temporal sequencing between disorder onset and psychedelic use; therefore causality cannot be inferred. The wide range of reported disorders and the diversity of psychedelics used made it difficult to examine drug‑specific or disorder‑specific associations. The authors also note that many respondents reported combining drugs, complicating attribution of effects to a single substance. They highlight unresolved questions for future research, including whether psychedelics can be effective without concomitant psychotherapy, how "set and setting" and intentions influence outcomes, and whether group administration produces different effects than individual use. Mason and colleagues conclude by stating that, in this sample, some psychedelic users do self‑medicate and perceive greater benefit from psychedelics than from standard offered treatments, and that these findings support further clinical investigation of psychedelics as potential treatments for psychiatric disorders.