Psilocybin dose-dependently causes delayed, transient headaches in healthy volunteers

Classic hallucinogens such as psilocybin produce psychoactive effects largely via 5-HT2A receptor agonism and have a long history of ritual and recreational use. Earlier clinical reports and case series have intermittently noted headache following administration of classic psychedelics, and some small studies and reviews suggested a dose-related occurrence of head pain after psilocybin, but these observations were generally retrospective, inconsistent, or not systematically characterised. At the same time, several tryptamine-based and indole compounds are used in acute and prophylactic headache management, and anecdotal reports and case series have raised the possibility that psilocybin or related compounds might both provoke and, paradoxically, treat certain primary headache syndromes such as cluster headache. Johnson and colleagues designed the present study to prospectively and systematically characterise headache as an adverse event after controlled psilocybin administration. The investigators aimed to examine whether headache incidence, timing, duration and severity were dose-dependent across a broad range of oral psilocybin doses in medically and psychiatrically healthy volunteers, and to explore relationships between headache and other measured drug effects and physiological responses.

Eighteen medically and psychiatrically healthy volunteers (10 female, 8 male) participated. Only one participant reported prior hallucinogen use more than 20 years earlier. Screening included self-reported headache history; four participants reported a history of headache. The study followed established safety guidelines for human hallucinogen research. A double-blind, within-subjects crossover design was used. Each volunteer completed five 8-hour sessions at roughly one-month intervals, receiving oral psilocybin at doses of 0 (placebo), 5, 10, 20 and 30 mg/70 kg. Participants were randomised to one of two dose sequences: doses either increased sequentially or decreased sequentially across sessions, with the single placebo session intermixed among the active doses. Study monitors met participants for preparatory rapport-building sessions before dosing and provided support during each drug session in a comfortable, living-room-like setting; participants were encouraged to remain reclined, wear an eye mask and listen to music. Analgesic medications were not permitted during the session, but over-the-counter (OTC) headache remedies were allowed afterwards. Headache assessment was prospective. Follow-up questioning occurred one or two days after 82% of sessions (SD 2.1; maximum 14 days), and at later visits three weeks and 14 months after sessions (though headache was not explicitly queried at those later timepoints). For sessions with reported headache, participants gave onset and end times (used to compute duration), indicated whether the headache was single or intermittent, rated severity on a three-point scale (1 = mild, 2 = moderate, 3 = severe), and reported any medication taken (all reported treatments were OTC acetaminophen, ibuprofen, or aspirin). Statistical analysis used repeated-measures regression (SAS PROC MIXED) with Dose (0, 5, 10, 20, 30 mg/70 kg) and Sequence (ascending, descending) as factors; alpha was set at 0.05. Tukey post hoc tests compared each active dose to placebo. Additional repeated-measures multiple-regression models examined whether headache (as a dichotomous covariate) related to other outcomes, including measures of mystical-type experience (Hood Mysticism Scale, Mystical Experience Questionnaire), drug-effect strength (intensity subscale of the Hallucinogen Rating Scale) and cardiovascular indices (peak and area under the curve, AUC, for systolic/diastolic blood pressure and heart rate).

Eighteen participants contributed data, although three sessions in which participants had an ongoing headache at the time of drug administration were excluded from headache analyses; those excluded sessions had headache onset a mean of 13.3 h before dosing and a mean total headache duration of 28.5 h. Overall, headache incidence increased with psilocybin dose. Repeated-measures regression showed a significant main effect of Dose on headache incidence (F(4,64) = 12.74, p < 0.0001), with no significant effects of Sequence or Dose × Sequence. Post hoc Tukey comparisons indicated that each psilocybin dose produced headaches in a significantly greater number of participants than placebo (all p < 0.03). Among participants reporting headache, duration increased with dose and all reported headaches had resolved by the time of headache data collection. Headache onset occurred a mean of 7.0 h after psilocybin administration, with relatively little variation in mean onset across doses though some individual headaches began substantially later. Severity ratings for reported headaches were restricted to mild or moderate; no participant rated any headache as severe at any dose. The percentage of participants who took OTC headache medication increased with dose among those who reported headache. Analyses did not find systematic relationships between participant demographics and headache occurrence. Including headache as a covariate in repeated-measures multiple-regression models did not produce significant associations with measures of mystical experience, hallucinogenic drug-effect strength (HRS intensity), or cardiovascular measures (peak or AUC for blood pressure and heart rate), suggesting that headache was not correlated with those assessed psychological or physiological variables in this sample.

Johnson and colleagues interpret the findings as showing that psilocybin commonly produces a delayed, transient headache in healthy volunteers and that incidence, duration and severity increase in a dose-dependent manner. The delayed timing and brief duration distinguish these headaches from immediate autonomic effects during acute intoxication. Because the study did not collect the additional diagnostic features required to confirm migraine (nausea, vomiting, photophobia/phonophobia, unilateral or pulsatile quality, exacerbation by activity), the investigators stopped short of asserting a specific headache subtype. To account for the observations, the authors review several potential mechanisms. They highlight a nitric oxide (NO) pathway: NO donors such as nitroglycerin produce a characteristic delayed headache 5–6 hours after exposure, and psilocybin’s time-course of headache onset is similar, suggesting psilocybin might trigger delayed NO-related processes (including inducible nitric oxide synthase expression). Alternative or complementary mechanisms considered include facilitation of cortical glutamate release, suppression and subsequent rebound of dorsal raphe serotonergic activity with downstream locus coeruleus effects on cerebral circulation, 5-HT2B-mediated meningeal mechanisms, and effects on inflammatory gene expression (for example via NF-κB and iNOS pathways). The authors also note that psilocin’s receptor profile includes affinity at multiple serotonin receptor subtypes (5-HT2A, 5-HT2C, 5-HT1A among others), and discuss how interactions at different receptor populations could promote or mitigate headache generation. Several lines of evidence discussed argue against exclusive mediation by 5-HT2A receptor activation: a small interaction study found similar headache incidence when psilocybin was pretreated with the 5-HT2A blocker ketanserin; LSD does not appear to produce the same delayed headaches in published reports; and typical 5-HT2A antagonists have not shown clear prophylactic efficacy for migraine. Conversely, small studies combining psilocybin with haloperidol suggested attenuation of post-psilocybin headache, offering other mechanistic clues. The authors acknowledge that available drug interaction studies are underpowered and use inconsistent headache measures, limiting firm conclusions. Key limitations acknowledged include the small sample size, the lack of detailed headache phenomenology needed for formal headache classification, and reliance on follow-up interviews rather than continuous monitoring. The investigators recommend more detailed headache scales and assessment of headache features in future work, screening for personal or family migraine history, and consideration of genetic analyses implicated in migraine pathogenesis. Finally, they conclude that because the observed headaches were transient, mild-to-moderate and non-disabling, predisposition to such headaches need not be an exclusion criterion in most future psilocybin research, though investigators might postpone dosing if an active headache could compromise study aims that depend on an unencumbered subjective experience.

The investigators conclude that psilocybin produces dose-dependent, delayed and transient headaches in healthy volunteers. They propose several candidate mechanisms—most notably delayed NO release—as plausible contributors, while acknowledging that glutamatergic facilitation, dorsal raphe–locus coeruleus interactions, 5-HT2B agonism, and inflammatory gene-expression changes may also play roles, potentially in combination. Given that reported headaches were neither severe nor disabling, the authors advise that such adverse events should not be a barrier to future psilocybin research, but recommend more detailed headache characterisation and consideration of participant migraine history and genetic factors in subsequent studies to disentangle mechanisms and potential therapeutic implications.