Who Is Actually Starting to Use Psychedelics? The NSDUH Data Is Surprising
- NSDUH analysis: 1,005,421 respondents (2002–2019) and 173,808 (2021–2023) — first-time psychedelic use has increased only modestly overall (0.71% to 0.79%), but the demographic story is more striking.
- New psychedelic use among 12–17 year-olds *decreased* (OR 0.96 per year) over the 2002–2019 period; among adults aged 65+ it *increased* significantly (OR 1.56).
- LSD new use increased meaningfully (OR 1.08/year, 95% CI 1.07–1.09) in contrast to psilocybin and MDMA, which showed no comparable trend.
- Despite state-level decriminalization and therapeutic legalization in Oregon (2020) and Colorado (2022), 2021–2023 data showed no significant change in overall new psychedelic use rates — the clinical pipeline and recreational pipeline remain separated.
The political story about psychedelics focuses on regulatory milestones — FDA breakthrough designations, state ballot measures, MAPS's MDMA-for-PTSD application. But the epidemiological story, based on a nationally representative survey of over a million Americans across two decades, tells a more nuanced and clinically relevant picture. The question for practitioners isn't whether psychedelics are being studied — it's who is actually using them, and whether clinical access and recreational trends are moving in lockstep.
What the Trends Show
Between 2002 and 2019, first-time psychedelic use remained largely stable at around 0.71% of the US population per year. The 2021–2023 window showed a slight increase to 0.79%, but this was not statistically significant — suggesting that Oregon and Colorado's legalization moves had not yet substantially altered population-level initiation rates.
The substance-specific and age-specific findings are more clinically informative. LSD showed a consistent year-over-year increase across the full period, while psilocybin — despite being the current focus of therapeutic research and decriminalization campaigns — showed no comparable trend. This counterintuitive finding likely reflects that psilocybin's cultural moment is relatively recent, while LSD has had a continuous presence in drug culture since the 1960s.
The demographic shift is the most clinically significant finding. Adolescent new use is declining — the "gateway drug" concern that historically drove drug policy is not supported by recent data. Meanwhile, adults over 65 are initiating psychedelic use at increasing rates. This is a population that practitioners may not expect to encounter in discussions about hallucinogens, but the data says otherwise. This age group is also the one most likely to be taking medications with potential pharmacokinetic interactions (anticoagulants, SSRIs, cardiac medications) and most likely to have subclinical cardiac or vascular conditions.
For Your Practice
The immediate clinical implication is screening. If you work with older adults — particularly in palliative, geriatric psychiatry, or late-life depression contexts — consider adding psychedelic use to your substance use history, including psilocybin mushrooms, which are increasingly available through informal channels regardless of regulatory status. The older patient who doesn't mention "drug use" may still be microdosing psilocybin for depression or anxiety.
The second implication is about the therapeutic pipeline itself. The gap between state-level legalization and population-level initiation suggests that even favorable regulation doesn't immediately translate to widespread new use. Therapeutic psilocybin services in Oregon and Colorado are proceeding slowly due to cost, limited facilitator training capacity, and lack of insurance coverage. The clinical moment is arriving, but more slowly than the media narrative suggests.
The fastest-growing new psychedelic users in the US are over 65 — the generation that once outlawed these substances is now quietly experimenting with them.
NSDUH methodology changed in 2020, creating a data gap; 2021–2023 estimates may not be directly comparable to 2002–2019. Self-reported drug use is systematically underreported. State-specific analyses are not available from the published data.