Is psychedelic use associated with cancer?: Interrogating a half-century-old claim using contemporary population-level data

This analysis of survey data (n=210.000) finds no association between psychedelics use and the risk of cancer or leukaemia (hematologic cancer). This counteracts alarmist reporting from the late 60s, implying that psychedelic use could lead to cancer (through chromosomal damage). The study is the first to use population-level data to dispel this myth.

Abstract

“Background: In 1967, concerns about the carcinogenic potential of psychedelics arose after a study reported chromosomal damage in human leukocytes following in vitro lysergic acid (LSD) exposure. Worries were further heightened by subsequent reports of leukemia and other cancers in LSD users. Additional investigations of psychedelics’ effects on chromosomes were published over the next decade, with the majority suggesting these concerns were unfounded. However, the relationship between psychedelics and cancer has been explored only minimally from an epidemiological perspective.

Aims: To determine whether associations exist between psychedelic use and either lifetime cancer or hematologic cancer diagnoses.

Methods: We analyzed data from adult participants in the 2015–2019 administrations of the National Survey on Drug Use and Health for associations between lifetime use of psychedelics and lifetime diagnosis of either any cancer or hematologic cancer.

Results: We identified no associations between lifetime psychedelic use and either lifetime cancer diagnosis or hematologic cancer diagnosis. Subanalyses of lifetime lysergamide, phenethylamine, and tryptamine use also revealed no associations with lifetime cancer or hematologic cancer diagnosis.

Conclusions: While laboratory studies and case reports from the 1960s and 1970s generated concerns about psychedelics’ carcinogenic potential, this analysis of recent epidemiological data does not support an association between psychedelic use and development of cancer in general or hematologic cancer. Important study limitations to consider include a lack of information about psychedelic dosage, number of lifetime psychedelic exposures, and the temporal relationship between psychedelic use and cancer diagnosis.

Authors: Brian S. Barnett, Kathleen Ziegler, Rick Doblin & Andrew D. Carlo

Summary

Please also take into account this correction from May 2025:

While preparing for a subsequent unrelated analysis using data from the National Survey on Drug Use and Health (NSDUH), the authors realized that they had input the complex survey design variables incorrectly for the analysis in this published manuscript. Specifically, the NSDUH nesting variables (VEREP and VESTR) were incorporated incorrectly into the R program survey object (as part of the survey package). When the authors realized this error had been made, they did a thorough review of all analyses. In Tables 1, S1 (supplement), S2 (supplement), S3 (supplement), and 2, the authors found that the error impacted the weighted standard errors and p-values, but not the weighted percentages. In Table 3, they found that the error impacted the model standard errors, confidence intervals, and p-values, but not b coefficient estimates. This coding error did not impact any of the unweighted calculations. The authors are confident that the identified coding errors in their survey object do not change the interpretation, direction, or magnitude of the findings in their investigation. During this thorough review, however, the authors did note additional errors unrelated to the coding error above, which have since been corrected and are summarized below (though these additional errors also do not change the interpretation, direction, or magnitude of the findings in this study):
1.	
Although this is not an error, the p-values have been modified in accordance with the journal formatting style in the tables.
2.	
In Table 1, the incorrect variable for lifetime stimulant use was used. Instead of “stmnmlif” (which asks participants whether they have ever taken a stimulant medication NOT prescribed by a physician), the authors used “stmanylif” (which asks participants whether they have ever taken a stimulant). This impacted the weighted and unweighted percentage calculations, in addition to significance testing. The authors confirmed that this error was not made in the statistical models.
3.	
In Table 1, the weighted percentage for classic psychedelic users between 50-64 years old was input as 28.2 when it should have been 28.1.
4.	
In Table 1, the coding error in the survey object led to incorrect standard errors and p-values for BMI and “Age when First Drank an Alcoholic Beverage.” However, the weighted percentages were correct in the original version and the standard error and p-value changes the authors made do not alter the interpretation of the original findings.
5.	
In the Supplementary Material, tables S1, S2 and S3, the incorrect variables for lifetime stimulant use were used. Instead of “stmnmlif” (which asks participants whether they have ever taken a stimulant medication NOT prescribed by a physician), the authors used “stmanylif” (which asks participants whether they have ever taken a stimulant). In the same tables, similar errors were made with sedatives (“sedanylif” was input instead of “sednmlif”) and tranquilizers (“trqanylif” was input instead of “trqnmlif”). This impacted the weighted and unweighted percentage calculations, in addition to significance testing. The authors confirmed that these errors were not made in the statistical models.
6.	
In the Supplementary Material, tables S1, S2 and S3 the coding error in the survey object led to incorrect standard errors and p-values for BMI and Age when First Drank an Alcoholic Beverage. However, the weighted percentages were correct in the original version and the standard error and p-value changes made do not alter the interpretation the original findings.
7.	
In Table 2, the authors re-conducted all analyses. The weighted and unweighted percentages were all found to be correct in the original version. P-values did change in the weighted analyses, but fundamental statistical significance did not change for any of the weighted percentage comparisons. Also, in Table 2, the authors found a data input error in the lifetime phenethylamine use percentage among participants with a history of any cancer – the number was input as 5.0 when it should have been 4.9.
8.	
In Table 3, the authors re-conducted all analyses. The model b coefficient estimates were found to be correct in the original version. Standard errors, confidence intervals and p-values for the estimates did change, but fundamental statistical significance did not change for any of the b coefficient estimates of interest. Therefore, the interpretation of the model findings did not change after the authors corrected the coding error described above. Corrected model outputs for all four models are in the corrected supplement as Tables S4, S5, S6, and S7.
Two subject experts and the Journal Editors confirmed that the changes to the article alter the results but do not change the conclusions.

Introduction

Lysergic acid (LSD) appeared to cause chromosomal breaks and rearrangements in cultured human leukocytes, but another study found that the number of chromosomal abnormalities in leukocytes from LSD users was not significantly higher than the number found in control leukocyte cultures.

Results from a number of studies investigating the mutagenic, carcinogenic, and teratogenic potential of LSD, N,N-dimethyltryptamine (DMT), mescaline, and psilocybin were published over the next decade. Some studies reported that chromosomes purportedly damaged by LSD resembled those found in leukocytes of patients with chronic myeloid leukemia, and others reported that chromosome aberrations appearing similar to those occurring in Fanconi anemia and other conditions associated with high rates of leukemia and other cancers.

In 1967, concerns about the carcinogenic potential of psychedelics arose. However, subsequent studies suggested these concerns were unfounded, and the relationship between psychedelics and cancer has only been explored minimally from an epidemiological perspective.

In vitro studies reporting associations between LSD exposure and chromosomal damage were criticized for using high concentrations of LSD and long exposure times, and for using impure LSD by non-medical users. Furthermore, some studies reporting elevated rates of chromosomal damage by LSD were eventually retracted.

In 1971, a review of nine in vitro studies of LSD and 21 in vivo studies found that there was no dose-response relation and that the magnitude of chromosomal damage caused by LSD was similar to that caused by many commonly used substances.

The majority of studies refuting the original findings on LSD-associated chromosomal damage tapered off in the 1970s, and in 1977 Maimon Cohen concluded that such results were not seriously anticipated.

Despite the bulk of evidence weighing against carcinogenic risks associated with psychedelic exposure, concerns still linger in the public consciousness and the medical literature. There has been little contemporary research on the carcinogenic potential of psychedelics.

Given the recent resurgence of interest in psychedelics’ therapeutic potential, and the possibility that psychedelics could return to clinical use in coming years, it is important to clarify the relationship between lifetime use of psychedelics and subsequent development of cancer.

Study data and population

The NSDUH is a nationally representative survey conducted by the Substance Abuse and Mental Health Services Administration. It is designed to gather information on patterns and trends in alcohol and other substance use, assess the consequences of substance use, and identify groups at high risk for substance use.

Participants are paid $30 to participate, and the survey had national interview response rates of 64.92% (weighted) and 66.62% (unweighted) in 2019. The survey uses audio computer-assisted self-interviewing (ACASI) for questions about drug use and other sensitive topics, and uses a sample-weighted design to ensure results accurately represent the general population.

We limited our study population to NSDUH adult participants (18 years and older) and combined data from survey years 2015 to 2019 to increase sample size.

Variables and variable selection rationale

The NSDUH ACASI program asks participants about lifetime diagnoses of cancer and other conditions, and then provides a list of different types of cancers for participants to select from. Psychedelic use was assessed by asking participants if they had ever used LSD, MDMA, mescaline, DMT, peyote, psilocybin, or San Pedro cactus. The use of AMT, DMT, and 5-MeO-DIPT was also assessed by asking if they had ever used any other hallucinogens.

We analyzed classifications of psychedelics based on chemical structure separately, and generated variables for lifetime use of tryptamine, lysergamide, and phenethylamine psychedelics.

Our logistic regression models included lifetime use of psychedelics, age, sex, race/ethnicity, body mass index, BMI, educational attainment, income, marital status, rural versus urban location of residence, lifetime marijuana use, lifetime heroin and prescription opioid use, and lifetime inhalant use.

We included variables assessing health insurance status, risk taking behavior, lifetime use of tobacco and alcohol, as well as lifetime use of psychedelics, in our models to assess the association between lifetime use of these substances and cancer diagnosis.

Data analysis

We used descriptive statistics to compare the percentages of respondents who had a history of using psychedelics with those who did not.

We specified four logistic regression models to examine the associations between lifetime psychedelic use and lifetime diagnoses of any cancer or hematologic cancer. The statistical significance was set at p 0.05.

Results

Among 210,021 unique NSDUH respondents, 37,222 (17.72%) reported lifetime psychedelic use. Lifetime psychedelic users were more likely to be male, non-Hispanic White, non-Hispanic Native American/Alaska Native, non-Hispanic multiracial, highly educated, higher income, uninsured, lower BMI, and residents of metropolitan areas.

Lifetime tryptamine psychedelic users were more likely to be male, between 26 and 64, non-Hispanic White, non-Hispanic Native American/Alaska Native, non-Hispanic multiracial, highly educated, higher income, uninsured, lower BMI, residents of metropolitan areas, and never married.

Lifetime lysergamide psychedelic users were more likely to be between 35 and 64, male, non-Hispanic White, non-Hispanic multi-racial, uninsured, lower BMI, residents of small metropolitan areas, never married, divorced/separated, and more frequently engaged in risky behavior.

Lifetime phenethylamine psychedelic users were more likely to be male, between ages 18 and 49, non-Hispanic White, non-Hispanic Native American/Alaska Native, non-Hispanic multiracial, highly educated, uninsured, lower BMI, and residents of large metropolitan areas.

Regarding cancer prevalence among lifetime psychedelic users, 5.3% of lifetime psychedelic users reported a lifetime history of any type of cancer compared to 6.5% of non-users, and 0.4% of lifetime psychedelic users reported a lifetime history of hematologic cancer compared to 0.5% of non-lifetime users.

Multivariable logistic regression analyses revealed no associations between lifetime psychedelic use and lifetime cancer diagnosis, or lifetime hematologic cancer diagnosis, or lifetime tryptamine, lysergamide, or phenethylamine psychedelic use and lifetime cancer diagnosis.

All models adjusted for age, gender, race/ethnicity, education, income, marital status, history of risky behavior, lifetime history of cocaine use, lifetime history of stimulant use, lifetime history of sedative use, lifetime history of tranquilizer use, lifetime history of heroin use, lifetime history of pain killer use.

Discussion

This study examined the relationships between lifetime psychedelic use and cancer diagnoses in the general population. It found that lifetime psychedelic users had higher frequencies of several characteristics associated with cancer development, including male sex, non-Hispanic white race/ethnicity, never married or divorced marital status, and risky behavior. Psychedelic users have a higher prevalence of protective factors for cancer than non-psychedelic users, including younger age, lower BMI, higher educational attainment, and higher income level.

In this population-level study involving 210,021 unique adults, lifetime use of psychedelics was not associated with lifetime diagnosis of cancer in general or hematologic cancer. However, lifetime use of lysergamide, phenethylamine, and tryptamine psychedelics was associated with statistically significant decreases in prevalence of lifetime cancer diagnoses.

We found no association between lifetime hematological cancer diagnosis and lifetime LSD use, while a 1997 case – control study found an association between lifetime LSD use and non-Hodgkin’s lymphoma in men. However, the relationship between LSD use and non-Hodgkin’s lymphoma was limited to dose-response.

We found no association between lifetime classic psychedelic use and past year cancer diagnosis, nor did we find any association between lifetime classic psychedelic use and lifetime hematological cancer diagnosis.

As investigations into clinical applications of psychedelics grow and psychedelic use in non-research settings increases, more detailed epidemiological data about psychedelic use is needed to understand the physiological safety of psychedelics.

Strengths and limitations

The primary strength of this study is its use of data from multiple administrations of a large, rigorously conducted nationally representative survey sample. Its limitations include its reliance upon retrospective self-report of survey respondents and its focus on the civilian non-institutionalized population of the United States.

We did not find an association between psychedelic use and cancer diagnosis in a study of active duty military members, prisoners, and people in addiction treatment centers. However, we cannot exclude the possibility that individuals who first used psychedelics after a cancer diagnosis could have been included. Psychedelic users who initiated use after cancer diagnosis would have likely been excluded from the sample, but increasing media coverage of therapeutic applications of psychedelic-assisted therapy in people with cancer could have led to more people using psychedelics. The study has several limitations, including the fact that lifetime psychedelic users were more likely to be uninsured and that some participants may have used impure psychedelics.

Conclusions

In the late 1960s, concerns were raised that LSD and other psychedelics might induce chromosomal damage, possibly leading to leukemia and other cancers. However, epidemiological data on substance use have shown no association between psychedelic use and cancer diagnosis.