Psychedelics triggered mania in up to 30% of naturalistic bipolar cases, rarely in trials

As serotonergic psychedelics move further into clinical development for depression and other conditions, one specific safety concern has persisted: that these drugs could trigger mania or hypomania, particularly in people with bipolar spectrum vulnerability who are typically excluded from clinical trials but not from real-world use. A systematic review and meta-analysis published this month in Molecular Psychiatry is among the first to quantify that risk across the available evidence [s1].

The design

Researchers systematically searched human studies examining manic or hypomanic symptoms following exposure to serotonergic psychedelics — psilocybin, LSD, mescaline, and DMT/ayahuasca — or MDMA, with the search protocol registered in advance (CRD420251160656) and databases searched through January 26, 2026 [s1]. Eligible study designs included randomized and non-randomized clinical studies, registry-based cohorts, cross-sectional surveys, and longitudinal observational studies [s1]. Risk of bias was assessed using standard tools appropriate to each study design [s1]. Twenty-three studies met inclusion criteria, with four contributing to formal meta-analysis [s1].

What it found

Rates of psychedelic-associated dysphoria, euphoria, hypomania, or mania varied enormously by context: as low as 5.8% in controlled trials of psilocybin-assisted psychotherapy for major depressive disorder, up to 30% in naturalistic studies of individuals who already had bipolar disorder [s1]. When manic symptoms did occur, they were typically acute and self-limited rather than persistent [s1]. Observational studies identified specific groups at higher risk: people with bipolar I disorder specifically, those with a family history of bipolar disorder, people with polysubstance use, and those using psychedelics in unsupervised or illegal settings [s1].

The meta-analysis of registry-based cohorts examining longer-term diagnostic transition — people later being diagnosed with bipolar disorder following psychedelic exposure — found a pooled prevalence of 4% (95% CI 2–8%, based on 7,478 people across the pooled studies, I² = 32.1%, indicating moderate but not severe inconsistency across the pooled studies) [s1]. Notably, the review found "little evidence for a hallucinogen-specific signal" in that transition-to-bipolar-disorder finding [s1] — meaning the observed 4% rate wasn't clearly higher than what would be expected from other risk factors independent of psychedelic exposure itself.

Why the gap between controlled and naturalistic settings is the headline finding

The roughly fivefold difference between rates in controlled psilocybin-assisted psychotherapy trials (5.8%) and naturalistic studies of people who already have bipolar disorder (up to 30%) is the review's most clinically actionable finding, because it isolates what's actually driving risk: not psychedelics generally, but psychedelic exposure specifically in people with existing bipolar vulnerability, especially outside the careful screening, dosing, and monitored setting that clinical trials use. Controlled trials routinely screen out and exclude people with bipolar I disorder or significant bipolar spectrum symptoms — which is precisely why their reported rates run so much lower than naturalistic studies that include exactly the population controlled trials exclude.

What the diagnostic-transition finding does and doesn't mean

The 4% rate of later bipolar disorder diagnosis following psychedelic exposure, with "little evidence for a hallucinogen-specific signal" [s1], is a reassuring finding on its own terms — it suggests psychedelic exposure isn't clearly driving new bipolar disorder diagnoses beyond what background rates or other risk factors would predict. But this is pooled registry data across different studies and populations, and the review's own framing treats this as a distinct, more uncertain question from the acute mania/hypomania risk — the two findings shouldn't be conflated as saying the same thing.

What this doesn't establish

This is a systematic review pooling 23 quite different studies — controlled trials, naturalistic surveys, and registry cohorts — using different definitions, populations, and follow-up windows, with only four contributing to formal meta-analysis; the rest are synthesized narratively rather than statistically combined. That heterogeneity limits how precisely any single risk estimate here should be treated as universally applicable. The review's authors explicitly note that long-term outcomes and the effects of repeated psychedelic exposure remain insufficiently studied [s1].

What the authors conclude

The review's overall characterization: serotonergic psychedelics carry "a low but clinically meaningful relative risk" of transient mood-related symptoms specifically in susceptible individuals, while appearing "relatively safe" within controlled clinical settings [s1] — a summary that argues for continued careful screening in clinical contexts while acknowledging real risk exists outside them, particularly for people with bipolar spectrum vulnerability using psychedelics without supervision.

What to watch

Longitudinal research following people with and without bipolar spectrum vulnerability after psychedelic exposure, and whether clinical trial screening protocols evolve as more data on this risk accumulates. This article is not medical advice.

Sources

1. Psychedelic-induced hypomania and mania: a systematic review and meta-analysis — Molecular Psychiatry, 29 May 2026