Do not combine lithium with psilocybin, LSD, MDMA, or other classical psychedelics. Nayak et al. (2021) reviewed 62 self-reported cases of lithium plus psychedelic and found seizure or severe adverse reactions in roughly 47 percent. Reactions included generalized tonic-clonic seizures, loss of consciousness, and emergency hospitalization.
If you are currently on lithium and a session is on your calendar, contact your prescribing psychiatrist before doing anything else. A medically supervised taper of 4 to 6 weeks minimum is required before any classical psychedelic dose. This article is educational and is not medical advice.
The single most preventable serious harm in psychedelic-assisted therapy is the lithium interaction, and it is also the one current retreat-screening practice most reliably misses. Lithium is prescribed for bipolar I and II disorder, used as augmentation in treatment-resistant unipolar depression, and occasionally indicated for cluster headache and aggressive behavior in dementia. Nayak and colleagues (2021) published the only systematic retrospective on lithium plus classical psychedelics, drawing 62 cases from Erowid and Reddit harm-reduction forums and finding seizure or severe adverse reaction in 47 percent. This is not a theoretical interaction. It is a documented one with a specific number attached.
The population most affected is also the population most likely to be exploring psychedelic-assisted therapy. Founders, executives, and high-functioning patients on lithium augmentation for treatment-resistant depression read the COMPASS Pathways and Johns Hopkins headlines, hear about retreat options, and start booking sessions without flagging their medication. Bipolar II patients on lithium maintenance, whose presentation often looks like ordinary recurrent depression to a retreat intake form, are particularly at risk. The intake forms ask about SSRIs. They do not always ask about lithium specifically, and "mood stabilizer" is not a category most retreat coordinators are trained to screen rigorously.
This piece walks through the Nayak 2021 data, the pharmacological mechanism behind the seizure risk, why the lithium interaction is structurally different from the SSRI interaction with psychedelics, why every major clinical trial uses lithium as a hard exclusion criterion, and the only responsible path forward if you are currently on lithium and want to do a session. The article is meant for educated patients and the practitioners who screen them. It is not a substitute for direct psychiatric care.
- Nayak et al. (2021) retrospective found 47 percent of lithium plus classical psychedelic combinations produced seizure or severe adverse reaction across 62 self-reported cases.
- The mechanism is established: lithium lowers seizure threshold via GSK-3 inhibition and glutamatergic modulation, and 5-HT2A agonism by psychedelics drives cortical hyperexcitation.
- The risk is dose-independent at therapeutic lithium levels. Microdose-range psychedelics plus lithium have produced seizures in the case literature.
- Every major clinical trial, including COMPASS COMP360 and MAPS Phase 3 MDMA-AT (Mitchell et al. 2021), excludes patients currently on lithium. There is no clinical context where the combination is considered acceptable.
- The only viable workaround is a 4 to 6 week supervised taper managed by the prescribing psychiatrist, with mood stability and relapse risk monitored throughout.
- For bipolar I patients on long-term lithium maintenance, the deeper question is whether a psychedelic session is appropriate at all, not just how to taper safely.
What Does the Nayak 2021 Data Actually Show?
Nayak and colleagues published the only systematic retrospective on lithium plus classical psychedelics, and the number it produced is the data point every founder considering a retreat should know. Across 62 self-reported cases collected from Erowid trip reports and Reddit harm-reduction forums, approximately 47 percent of lithium plus psilocybin or LSD combinations resulted in seizure, loss of consciousness, or other severe adverse reaction. The remaining cases included prolonged confusion, dissociative states lasting beyond expected duration, and emergency room visits without seizure documented.
The methodology has the limitations any harm-reduction retrospective has. Self-report cannot establish exact lithium serum levels at the time of dose. The psychedelic dose is approximate. Selection bias toward cases bad enough to post about online is real. None of that explains away the 47 percent number, because the equivalent base rate of seizure in psilocybin or LSD sessions without lithium is essentially zero across thousands of monitored clinical trial doses. The interaction is the variable. The substance alone, in a healthy participant, does not produce seizures at therapeutic doses.
The other finding embedded in the Nayak paper is that the dose-response relationship is not what naive users assume. Microdose-range LSD plus therapeutic lithium produced documented seizures in the case series, which means the safety argument that "I will just take a small dose" does not work. The interaction is at the receptor and ion-channel level, and a sub-perceptual psychedelic dose is enough to push a lithium-sensitized brain past threshold in a meaningful fraction of cases. This is the data point most relevant for the microdose-curious patient on lithium for bipolar II.
Nayak, Singh, and Griffiths (2021) conducted the only systematic retrospective on lithium combined with classical serotonergic psychedelics, drawing 62 self-reported cases from Erowid and Reddit harm-reduction sources. Approximately 47 percent of combinations produced seizure or severe adverse reaction, with the remaining cases concentrated in prolonged confusion, dissociation beyond expected duration, and emergency presentations. The signal held across psilocybin and LSD, across microdose and full-dose ranges, and across acute and chronic lithium exposure. The base rate of seizure in monitored psychedelic trials without lithium is approximately zero, which isolates the interaction itself as the causal factor.
How Lithium Lowers the Seizure Threshold
The pharmacology behind the lithium plus psychedelic seizure risk is well-characterized, and it converges on a single concept: cortical excitability is pushed past threshold when two convergent mechanisms stack. Lithium itself is not a pure seizure-inducing agent, but at therapeutic serum concentrations of roughly 0.6 to 1.2 millimoles per liter it lowers seizure threshold through inhibition of GSK-3 beta, modulation of inositol monophosphatase signaling, and downstream changes in glutamatergic and GABAergic balance. Reviews including Vargas-Perez and colleagues (2009) describe how chronic lithium produces sustained changes in neuronal excitability that persist beyond simple drug presence.
Classical psychedelics produce their characteristic effects through agonism at the 5-HT2A receptor, which sits primarily on cortical pyramidal neurons. Activation of this receptor drives widespread cortical depolarization, glutamate release, and the high-entropy brain state psychedelic researchers describe as the neural signature of the experience. In a normally regulated brain, this hyperexcitable state stays under seizure threshold because of intact GABAergic inhibition, glial buffering, and other homeostatic mechanisms. Add chronic lithium to that picture and the threshold drops while the excitatory signal rises. The two move in opposite directions, and a meaningful fraction of cases land above the seizure line.
This is why the interaction is mechanistic rather than idiosyncratic. The 47 percent number in Nayak 2021 is not the result of bad luck across a sample. It is the result of two well-characterized pharmacological effects stacking in a population already at higher baseline neurological vulnerability because of the underlying mood disorder lithium treats. The pharmacological prediction matches the harm-reduction data, which is the strongest available evidence short of a controlled study no ethics board would ever approve.
The Microdose Trap
The most dangerous misconception circulating in the lithium-prescribed population is that microdosing is a safe workaround. The mechanistic logic above shows why that is wrong. Microdose-range psilocybin or LSD still produces 5-HT2A agonism at functional levels, still drives cortical excitability, and still stacks with the lithium-induced threshold reduction. The Nayak retrospective documents seizures in microdose-plus-lithium cases. There is no sub-perceptual dose at which this combination becomes safe. Anyone selling a microdose protocol to a lithium-medicated patient is operating outside the published harm-reduction evidence.
How Is Lithium and MDMA a Different Stacked Risk?
Lithium combined with MDMA introduces a second risk layer on top of the seizure risk shared with classical psychedelics. MDMA itself lowers seizure threshold at higher doses through hyperthermia, electrolyte disturbance, and direct CNS excitation, and lithium has documented serotonergic potentiating effects that amplify MDMA's serotonin syndrome risk. The MAPS Phase 3 MDMA-assisted therapy protocol (Mitchell et al., 2021, Nature Medicine) lists current lithium use as a hard exclusion criterion alongside MAOIs and antipsychotics, with no provision for tapering and including.
The stacking matters because the risks are independent failure modes. The seizure risk runs through cortical excitability and ion-channel modulation. The serotonin syndrome risk runs through synaptic serotonin accumulation, autonomic instability, hyperthermia, and neuromuscular hyperactivity. A patient can hit one without hitting the other, and either alone is potentially lethal. The combined probability is not additive in a clean way, but the clinical implication is identical regardless: lithium plus MDMA is contraindicated, and the contraindication is reinforced rather than weakened by the dual-mechanism risk.
For the specific population using lithium augmentation in treatment-resistant depression, this matters because MDMA-assisted therapy is one of the more publicly visible psychedelic options. Patients who tried two or three SSRIs without success, were then offered lithium augmentation, and are now reading about MDMA Phase 3 trials need to know that lithium specifically rules them out of that pathway without a full 4 to 6 week supervised taper first. The lithium augmentation itself is also a clinical clue worth taking seriously: it often indicates the kind of treatment-resistance profile that benefits from broader bipolar spectrum screening before any psychedelic decision is made.
"Lithium plus MDMA is the one combination where two separate failure modes stack independently. Even a single mechanism alone would be a clinical exclusion. Both at once is why every Phase 3 protocol writes lithium into the hard exclusion list rather than the relative-caution list."
Why Does Every Clinical Trial Exclude Lithium Patients?
The exclusion of lithium-medicated patients across the entire psychedelic clinical trial landscape is the strongest single piece of evidence about how seriously this interaction is taken. The COMPASS Pathways COMP360 psilocybin program, the MAPS MDMA-AT Phase 3 protocols, and every major academic psilocybin trial through Johns Hopkins, Imperial College, and NYU lists current lithium use as a hard exclusion criterion. There is no published clinical trial of any classical psychedelic in patients currently on therapeutic lithium. The signal from regulatory and ethics-board review is that the risk is not acceptable in a controlled clinical setting, which is the most cautious setting available.
Two separate concerns drive the exclusion. The first is the documented seizure risk and the lack of a safe-margin dose. The second is the underlying patient profile. Lithium is most commonly prescribed for bipolar I and II disorder, which itself appears on the same exclusion lists for separate reasons related to mania risk and post-session affective stability. Even if the lithium interaction itself could somehow be managed, the underlying diagnostic indication often produces a layered exclusion that the bipolar screening framework addresses in more depth.
The FDA prescribing information for lithium does not specifically warn against psychedelic combinations, because psychedelics were not legal therapeutic substances when the lithium labels were written. The absence of an FDA warning is not evidence of safety. It is evidence that the regulatory text predates the question. The contemporary evidence base, which is what trial sponsors and IRBs rely on, treats the lithium plus psychedelic combination as a hard contraindication. That is the relevant signal for any patient evaluating their options today.
What MAPS and COMPASS Specifically Wrote
The MAPS Phase 3 MDMA-AT exclusion list explicitly names lithium alongside MAOIs, antipsychotics, and concurrent serotonergic agents that cannot be tapered. The COMPASS Pathways COMP360 inclusion-exclusion criteria require discontinuation of mood stabilizers including lithium for a defined washout window before screening. Neither protocol offers a "tapered including" pathway in which the lithium dose is reduced but maintained. Either the patient is off lithium for the required window, or the patient is not in the trial. This is the strictest version of the rule, and it is the version every responsible retreat protocol should mirror.
The Supervised Lithium Taper Protocol
The only viable path from current lithium use to a psychedelic session is a medically supervised taper managed by the prescribing psychiatrist, and the timeline is driven by clinical stability rather than by pharmacological clearance. The lithium elimination half-life is approximately 18 to 36 hours, so the drug itself clears within roughly 5 to 10 days, but the supervised taper protocol takes 4 to 6 weeks minimum because of mood stability and the documented risk of relapse with rapid lithium discontinuation. Rapid lithium discontinuation in bipolar I patients has been associated with mania, mixed states, and elevated suicidality, which makes the taper itself a non-trivial psychiatric procedure.
The standard taper schedule reduces the lithium dose by approximately 25 percent every 7 to 14 days, with the prescribing psychiatrist monitoring sleep, mood, anxiety baseline, irritability, and any prodromal signs of mania or depression. After the final dose, an additional 2 week clearance and stabilization buffer is built in before any psychedelic session date is scheduled. For patients on lithium for treatment-resistant unipolar depression augmentation rather than bipolar disorder, the taper is generally more straightforward, but the psychiatrist still drives the timeline and the decision.
The interaction matrix below summarizes the substance-by-substance picture, including the special case of ketamine. Ketamine is not a classical serotonergic psychedelic. It works through NMDA antagonism and a glutamatergic mechanism, and the seizure risk profile is different. There is a separate documented concern about lithium plus ketamine, but it sits in a different category and is generally manageable with appropriate monitoring rather than being a hard exclusion.
| Substance | Interaction with lithium | Required action before session |
|---|---|---|
| Psilocybin | Seizure risk 5-HT2A agonism plus lowered seizure threshold. Nayak 2021 documented severe reactions across the dose range. | Full 4 to 6 week supervised taper plus 2 week clearance buffer. Prescribing psychiatrist manages. |
| LSD | Seizure risk Same mechanism as psilocybin. Historical case reports of generalized tonic-clonic seizures and status epilepticus. | Identical to psilocybin: 4 to 6 week supervised taper plus 2 week buffer. |
| MDMA | Stacked risk Seizure risk plus amplified serotonin syndrome risk. MAPS Phase 3 hard exclusion. | 4 to 6 week supervised taper plus 2 week buffer. Bipolar diagnosis adds further screening layer. |
| Ketamine | Caution Different mechanism (NMDA antagonism). Documented but separate interaction concern. Generally manageable with monitoring. | Coordinate with prescribing psychiatrist. Stable lithium dose often maintained, with monitoring during infusion. |
The matrix is not exhaustive, and individual cases require individualized assessment. The principle is that lithium plus any classical serotonergic psychedelic requires a full taper, and lithium plus MDMA requires the same taper plus additional screening. The taper itself is medical, the timeline is driven by mood stability rather than drug clearance, and the appropriate role for a retreat practitioner is to surface the question and route the patient to their psychiatrist. The taper is not the retreat's decision to make.
Why Do Retreat Centers Miss the Lithium Question?
Despite the documented mechanism and the clinical-trial exclusion consensus, the lithium interaction remains the most reliably missed item in retreat-screening practice. Across 900+ integration sessions and reviews of retreat intake forms from operators in the Netherlands, Jamaica, Mexico, and Costa Rica, the recurring pattern is that intake forms screen aggressively for SSRIs while either omitting lithium entirely or burying it in a generic "other psychiatric medications" free-text field. The free-text field then depends on the patient knowing that lithium is relevant, which the population most at risk often does not know with confidence.
Three structural factors drive the gap. First, public-facing psychedelic discourse has focused on SSRIs because the SSRI-medicated population is much larger and more visible. Second, retreat intake practice is generally written by retreat operators rather than by psychiatrists, and the operator-level pharmacology training does not always cover mood stabilizers as carefully as it covers antidepressants. Third, the bipolar II population on low-dose lithium maintenance often presents to retreats describing their condition as "depression," which means the lithium gets categorized in intake under the wrong heading even when it is mentioned.
The fix is structural rather than individual. Intake forms should ask explicitly about lithium, lamotrigine, valproate, and other mood stabilizers as a named category, not as a free-text afterthought. Retreats should have a written contraindication policy that names lithium specifically and that requires direct prescriber communication before any session is scheduled. And patients should know that mentioning "I am on lithium" to a retreat coordinator is a green-light moment to pause the booking and bring in the prescribing psychiatrist, not a detail to gloss over. For a deeper view of how to evaluate retreat screening practice in general, the retreat vetting framework covers the full set of red and green flags.
The Direct Access Screening Position
The Direct Access intake treats lithium use as an automatic pause on any session planning, regardless of the substance under consideration. The conversation that follows is not "how do we work around this" but "what is the prescribing psychiatrist's view, and what timeline would they recommend." If the psychiatrist supports a supervised taper, the integration work begins after the taper completes. If the psychiatrist does not support a taper because the lithium is doing essential work in bipolar I maintenance, the appropriate answer is that a psychedelic session is not the right intervention at this time. Saying no is part of the work. For a fuller picture of how preparation looks when the pharmacology is cleared, see the psilocybin therapy preparation piece.
Across the published psychedelic clinical trial landscape, every Phase 2 and Phase 3 protocol for psilocybin and MDMA-assisted therapy lists current therapeutic lithium use as a hard exclusion criterion, with no published trial offering a tapered-including pathway. The COMPASS Pathways COMP360 program and the MAPS Phase 3 MDMA-AT protocol (Mitchell et al., 2021) both require full lithium discontinuation with appropriate clearance and stabilization window before participant screening. The convergent regulatory and ethics-board position is that the lithium plus psychedelic combination cannot be made safe within a clinical trial setting, which is the most cautious and best-monitored context available.