Pharmacology of Psychedelics: Receptors, Pharmacokinetics, Dose and Interactions
Psychedelic-Assisted Medicine: Science and Practice
Learning objectives for this lesson:
- Define affinity, efficacy, agonism, partial agonism, antagonism and receptor occupancy, and apply each term to a named psychedelic.
- Explain how blockade studies with ketanserin support the conclusion that serotonin 5-HT2A receptor activation produces the classic psychedelic state in people.
- Calculate receptor occupancy from a plasma psilocin concentration, and interpret what the saturation of that curve means for dose.
- Contrast the molecular actions of LSD, MDMA, ketamine and ibogaine, and state which claims about each rest on human data and which rest on structural, cell or animal work.
- Describe the conversion of psilocybin to psilocin, and explain why DMT is inactive when swallowed alone and how the beta-carbolines in ayahuasca make an oral preparation active.
- Compare the onset, peak and offset of six agents measured in controlled human studies, relate those profiles to the length of a supervised session, and weigh the evidence on route of administration and on fixed compared with weight-adjusted dosing.
- Distinguish tolerance, physical dependence and compulsive use, summarize the tolerance and self-administration evidence for classic psychedelics, MDMA and ketamine, and evaluate why compulsive use of classic psychedelics is rare.
- Identify the pharmacokinetic, pharmacodynamic and additive interactions that a medication review screens for, naming the evidence behind each and what it implies for screening decisions.
- Evaluate what user-reported interaction resources can and cannot establish, and describe how a clinician would use a client's account of them.
This course was developed by Dr. Kiffer G. Card.
Receptor Pharmacology: What Each Agent Binds
Introduction
HSCI 180 Lesson 3 introduced the two halves of pharmacology: pharmacokinetics, which describes what the body does to a drug, and pharmacodynamics, which describes what a drug does to the body. That lesson built the general machinery of receptors, agonists, antagonists and dose-response curves. This lesson applies the machinery to one family of compounds whose members act through strikingly different molecular routes while producing experiences that practitioners often group together. Lesson 1 placed psilocybin, lysergic acid diethylamide (LSD), N,N-dimethyltryptamine (DMT), mescaline, MDMA, ketamine and ibogaine in classes. This section asks what each one actually binds, and how researchers know.
Three terms carry most of the weight. Affinity is how tightly a drug attaches to a receptor, usually reported as a dissociation constant or an inhibition constant, where a smaller number means tighter binding. Efficacy is how much biological signal the drug produces once it is attached. Receptor occupancy is the proportion of a receptor population bound by the drug at a given moment, which depends on the concentration of drug at the receptor and on its affinity. A drug can have high affinity and no efficacy, in which case it occupies receptors while producing no signal of its own, and that combination defines an antagonist. A drug with high affinity and low efficacy is a partial agonist. Figure 3.1 sets out the four molecular routes that this section examines.
Learning Objectives
- 1.1 Define affinity, efficacy, agonism, partial agonism, antagonism and receptor occupancy, and apply each term to a named psychedelic.
- 2.1 Explain how blockade studies with ketanserin support the conclusion that serotonin 5-HT2A receptor activation produces the classic psychedelic state in people.
- 3.1 Calculate receptor occupancy from a plasma psilocin concentration using the single-site binding model fitted by Madsen and colleagues, and interpret what saturation means for dose.
- 4.1 Describe the binding kinetics of lysergic acid diethylamide and explain why structural findings in cells are weaker evidence about people than blockade studies.
- 4.2 Describe how MDMA acts through monoamine transporters, and identify the human pretreatment evidence that this is its route of action.
- 4.3 Contrast the molecular actions of ketamine and ibogaine with those of the classic psychedelics, and state which claims rest on human data and which rest on cell or animal work.
1.1 The Vocabulary of Receptor Pharmacology
Precision about these terms matters because claims about psychedelics are often stated loosely. A drug described as acting "on serotonin" could be an agonist at one receptor subtype, an antagonist at another, a reuptake inhibitor at a transporter, or a releaser that reverses a transporter, and each of those produces a different clinical picture. The flip cards below define the terms used through the rest of the lesson.
With these terms in place, the question becomes which of them describes each agent, and what evidence supports each attribution.
2.1 Serotonin 5-HT2A Receptors and the Blockade Evidence
The consensus position, stated in Nichols's (2016) review in Pharmacological Reviews, is that classic psychedelics are agonists or partial agonists at serotonin 5-HT2A receptors, with particular importance for the receptors on the apical dendrites of pyramidal cells in layer five of the cortex. Binding studies contribute to that claim, and most of its strength comes from a simple experimental logic: give people a drug that blocks the receptor, then give the psychedelic, and see whether the experience still happens.
The box below sets out the three human studies that carry most of this argument. Their designs differ in an instructive way. The first two block the receptor before the psychedelic is given, which tests whether 5-HT2A activation is necessary for the state to begin. The third gives the blocker after the psychedelic has taken effect, which tests whether continued occupancy is necessary for the state to continue.
Three blockade studies in people
Vollenweider et al. (1998), NeuroReport. In healthy volunteers, the psychotomimetic effects of psilocybin were blocked in a dose-dependent way by the 5-HT2A antagonist ketanserin and by the atypical antipsychotic risperidone, while the dopamine antagonist haloperidol increased them. The authors described this as the first evidence in people that the psilocybin state follows 5-HT2A receptor activation independently of dopamine stimulation.
Preller et al. (2017), Current Biology. Twenty-two participants received LSD with and without ketanserin pretreatment while undergoing functional magnetic resonance imaging. General subjective LSD effects were fully blocked by ketanserin, and the drug also prevented the attribution of personal relevance to previously meaningless stimuli.
Becker et al. (2023), International Journal of Neuropsychopharmacology. Twenty-four healthy participants received ketanserin or placebo one hour after one hundred micrograms of LSD. Ketanserin shortened subjective effects from 8.5 hours to 3.5 hours, reversed alterations of mind including visual changes and ego dissolution, and reduced cardiovascular effects and pupil dilation, without altering the pharmacokinetics of LSD. The authors concluded that ketanserin can serve as a planned or rescue option to shorten the LSD response.
These three studies are examples of a wider pattern. A systematic review of interactions involving classic psychedelics reported that several drugs blocking this receptor, among them ketanserin, trazodone, risperidone and chlorpromazine, reduced the effects of LSD and psilocybin across the studies it collected, so this step in the mechanism is generally well supported (Halman et al., 2024).
The third study matters for practice as well as for theory. A session that has become unmanageable cannot be undone by waiting for metabolism alone, and Becker et al. (2023) show a pharmacological route to shortening it in a research setting. Lesson 6 develops the management of difficult experiences and medical events, including the medications kept available in trials.
Table 3.1 summarizes the primary molecular targets of the agents this course returns to, with the human or laboratory evidence behind each entry.
Table 3.1. Primary molecular targets of six agents, with the kind of evidence behind each attribution.
| Agent | Primary molecular action | Principal evidence |
|---|---|---|
| Psilocybin (through psilocin) | Agonist at serotonin 5-HT2A receptors; also binds 5-HT1A, 5-HT2B and 5-HT2C receptors | Blockade by ketanserin and risperidone in people (Vollenweider et al., 1998); occupancy measured by positron emission tomography (Madsen et al., 2019) |
| LSD | Agonist at 5-HT2A receptors, with activity at other serotonin receptors and at dopamine receptors | Blockade and reversal by ketanserin in people (Preller et al., 2017; Becker et al., 2023); crystal structure at the 5-HT2B receptor (Wacker et al., 2017) |
| DMT (in ayahuasca and by injection) | Agonist at 5-HT2A receptors; inactive by mouth without monoamine oxidase inhibition | Human dose-response studies (Strassman et al., 1994); oral activity explained by beta-carboline inhibition of monoamine oxidase (McKenna et al., 1984) |
| Mescaline | Agonist at 5-HT2A receptors; a phenethylamine by chemical family, unlike the tryptamines above | 5-HT2A agonism shared with the class (Nichols, 2016); effects matched to those of LSD and psilocybin in a controlled comparison in people (Ley et al., 2023) |
| MDMA | Enters nerve terminals through monoamine transporters and reverses them, releasing serotonin, noradrenaline and dopamine; raises circulating oxytocin | Human pharmacology review (de la Torre et al., 2004); transporter blockers reduce its effects in people (Liechti et al., 2000; Hysek et al., 2012); oxytocin measured after dosing (Holze et al., 2020) |
| Ketamine | Open-channel blocker at NMDA glutamate receptors, with lower-affinity actions at other targets and active metabolites | Pharmacological review of ketamine and its metabolites (Zanos et al., 2018) |
| Ibogaine and noribogaine | Interacts with transporters, opioid receptors, sigma receptors, glutamate receptors and nicotinic receptors at low micromolar concentrations | Review of ibogaine pharmacology and cardiac effects (Koenig & Hilber, 2015); human pharmacokinetics (Glue et al., 2015) |
The table conceals one important difference. For psilocybin and LSD, the link from receptor to experience has been tested directly in people. For ibogaine, the list of targets comes from binding and electrophysiology work, and a 2021 review of the iboga alkaloids describes its mechanism as poorly defined, with weak affinity for a large number of receptors (Iyer et al., 2021).
3-D Model 3.1 lets you compare where several of the targets in Table 3.1 are found in the human brain, using group-average maps from positron emission tomography (PET) for the 5-HT2A and 5-HT1A receptors, the serotonin transporter, the NMDA receptor and the mu-opioid receptor.
3.1 From Plasma Concentration to Receptor Occupancy
Blockade studies support the conclusion that the receptor is necessary. A separate question is how much of the receptor population a given dose occupies. Madsen et al. (2019) answered this in eight healthy volunteers using positron emission tomography with the 5-HT2A agonist radioligand carbon-11 Cimbi-36. Each participant was scanned at baseline and once or twice more on the same day after a single oral dose of psilocybin between 3 and 30 mg, with blood samples and intensity ratings taken during the scans.
Occupancy reached 43 to 72 percent at the first post-dose scan. The relationship between plasma psilocin concentration and neocortical occupancy followed a single-site binding model, shown in Equation 3.1, with a maximum attainable occupancy of 76.6 percent (95 percent confidence interval 67.3 to 88.0) and a half-maximal concentration of 1.95 micrograms per litre (95 percent confidence interval 1.17 to 3.15). Subjective intensity, rated from 0 to 10, followed the same form with a maximum of 10.8 and a half-maximal concentration of 4.5 micrograms per litre.
Worked Example 3.1: Occupancy at a measured concentration
Holze et al. (2023) reported a mean maximal psilocin concentration of 17 nanograms per millilitre after a 25 mg oral dose of psilocybin, which is the same as 17 micrograms per litre, the unit used by Madsen et al. (2019). Substituting into Equation 3.1 with the Madsen parameters gives occupancy = (76.6 × 17) ÷ (1.95 + 17) = 1302.2 ÷ 18.95 = 68.7 percent. At the 15 mg concentration of 11 micrograms per litre, the same calculation gives (76.6 × 11) ÷ (1.95 + 11) = 842.6 ÷ 12.95 = 65.1 percent. A concentration more than one and a half times higher raises occupancy by less than four percentage points, because the curve is already close to its plateau.
(1) Use Calculator 3.1 with the preset for the 15 mg concentration and record the occupancy and the predicted intensity. (2) Switch to the 30 mg preset and record both again. (3) Set the concentration to 1.95 micrograms per litre, the half-maximal value, and confirm that occupancy is half of the maximum. (4) Using the three results, write one sentence explaining why a doubling of the plasma concentration in the upper range changes occupancy far less than a doubling in the lower range.
(1) At 11 micrograms per litre the model gives 65.1 percent occupancy and a predicted intensity of 7.7 out of 10. (2) At 21 micrograms per litre it gives 70.1 percent and 8.9 out of 10. (3) At 1.95 micrograms per litre occupancy is 38.3 percent, half of the 76.6 percent maximum, which is what the half-maximal concentration means. (4) The model saturates: once most receptors are occupied, additional drug has few free receptors left to bind, so the same proportional rise in concentration buys progressively less occupancy. The intensity curve fitted by Madsen et al. (2019) has a higher half-maximal concentration of 4.5 micrograms per litre, so predicted intensity is still rising across the range in which occupancy has nearly levelled off.
Two cautions belong with this model. It was fitted in eight people, and the authors reported wide variation between individuals in the concentration reached from a given dose. The radioligand is an agonist, which may bind preferentially to receptors in a high-affinity state, so the occupancy estimate is tied to the method. The finding that the model would not fit well with maximum occupancy fixed at 100 percent is discussed by the authors as a limitation with several possible explanations, including receptor internalization.
4.1 LSD and the Shape of a Long Experience
LSD acts at the same receptor as psilocin, yet its effects last several hours longer (8.2 hours compared with 4.9 hours in the direct comparison by Ley et al., 2023). Part of the explanation is pharmacokinetic and appears in Section 2. Another part concerns how the molecule sits in the receptor. Wacker et al. (2017) solved the crystal structure of LSD bound to the human 5-HT2B receptor and reported that LSD dissociates exceptionally slowly from both that receptor and the 5-HT2A receptor. Molecular dynamics simulations suggested that a lid formed by the second extracellular loop closes over the binding pocket, and a mutation predicted to make the lid more mobile accelerated binding and release while selectively reducing one branch of downstream signalling.
This is careful structural work, and it is work in crystals, cell lines and simulations. It supports a mechanism for slow release from the receptor; it does not by itself demonstrate that slow release explains the duration of the experience in people. Becker et al. (2023) provide the human counterpart: when ketanserin displaces LSD from the receptor one hour after dosing, the experience shortens from 8.5 hours to 3.5 hours while the plasma concentration of LSD follows its usual course. Taken together, the two findings support the view that the experience lasts as long as the receptor stays occupied. Lesson 4 develops what happens downstream of the receptor in brain networks and plasticity.
4.2 MDMA: Releasing Monoamines Through the Transporters
MDMA belongs to the same chemical family as mescaline, the phenethylamines, yet its main action is different in kind. De la Torre et al. (2004) describe MDMA as a potent releaser and reuptake inhibitor at the presynaptic transporters for serotonin, dopamine and noradrenaline, acting through the membrane transporters and vesicular storage systems. The practical consequence is that MDMA raises the amount of messenger in the synapse, with serotonin release predominating, and that its effects depend on the transporter being available.
Human studies confirm that dependence. Liechti et al. (2000) found that the serotonin reuptake inhibitor citalopram attenuated the psychological effects of MDMA. Hysek et al. (2012) reported that pretreatment with duloxetine, which inhibits both serotonin and noradrenaline transporters, prevented MDMA-induced elevations in well-being and emotional excitation in a placebo-controlled laboratory study. Blocking the door prevents the drug from using it, which is strong evidence that the door is the route of action. Section 4 returns to this finding, because the same mechanism explains why people taking serotonergic antidepressants report blunted responses.
MDMA also raises circulating oxytocin. Holze et al. (2020) compared LSD, MDMA and d-amphetamine in the same participants and found that MDMA, and neither of the other two, increased plasma oxytocin concentrations. Later studies from the same laboratory found that LSD and mescaline also raised circulating oxytocin (Ley et al., 2023) and that MDMA raised it more than LSD did (Straumann et al., 2023), so the contrast between the agents is one of degree. Oxytocin is often invoked to explain the feelings of closeness that distinguish MDMA from classic psychedelics, and the association is real. The step from circulating oxytocin to the subjective state rests on weaker ground: Kuypers et al. (2014) reported no evidence that MDMA-induced enhancement of emotional empathy was related to peripheral oxytocin levels. Lesson 9 examines what this pharmacology was used for in the MDMA trials for post-traumatic stress disorder.
4.3 Ketamine and Ibogaine: Two Kinds of Complexity
The last two agents complicate the picture in different ways. Ketamine has a well-defined primary target and a complicated metabolism. Ibogaine has many targets at once.
Zanos et al. (2018), in a review in Pharmacological Reviews, describe ketamine as a racemic mixture of two forms, S-ketamine and R-ketamine, in clinical use since 1970. Its anaesthetic and analgesic properties are generally attributed to direct inhibition of NMDA glutamate receptors, which it blocks from inside the open channel. The review lists lower-affinity actions at gamma-aminobutyric acid, dopamine, serotonin, sigma, opioid and cholinergic targets, and at voltage-gated sodium and hyperpolarization-activated channels, and it emphasizes that ketamine is rapidly and extensively metabolized to norketamine, dehydronorketamine, hydroxyketamine and hydroxynorketamine. The authors argue that these metabolites may have broader clinical relevance than was previously considered. The esketamine product used in depression treatment contains the S form alone. Health Canada first authorized it in May 2020 as a nasal spray for adults with major depressive disorder who have not responded adequately to at least two courses of antidepressant treatment in the current episode, used together with a selective serotonin reuptake inhibitor or a serotonin and noradrenaline reuptake inhibitor; Health Canada's regulatory decision summary records a new Notice of Compliance issued on 10 January 2023, after a Federal Court of Appeal decision on data protection. Lesson 8 covers the trials and the regulatory decisions.
Koenig and Hilber (2015) describe ibogaine as an indole alkaloid from the root bark of Tabernanthe iboga whose psychoactive mechanisms are poorly understood because its pharmacology is complex. Reported interactions at low micromolar concentrations include neurotransmitter transporters, opioid receptors, sigma receptors, glutamate receptors and nicotinic receptors. Long-lasting effects after ibogaine are attributed to its long-lived active metabolite, noribogaine. Glue et al. (2015) describe the conversion of ibogaine to noribogaine as mediated primarily by the liver enzyme CYP2D6, and showed in healthy volunteers that pretreatment with the CYP2D6 inhibitor paroxetine roughly doubled exposure to the combined active moiety. Section 4 returns to the cardiac consequences of this pharmacology, and Lesson 10 covers the use of ibogaine in opioid use disorder.
A single well-defined target supports a straightforward experiment: block the target, and see whether the effect disappears. A molecule acting at many targets at once does not permit that design easily, because blocking any one target leaves the others in place. This is one reason that mechanistic claims about ibogaine remain weaker than mechanistic claims about psilocybin, even though ibogaine has been used in treatment settings for decades. Lesson 4 returns to the general problem of grading mechanism claims by the kind of evidence behind them.
3-D Model 3.2 shows ketamine’s primary target, the NMDA receptor, and lets you step through how glutamate and glycine open its channel, how magnesium blocks the channel at rest and where ketamine lodges inside the open channel.
Section 2 turns from where these drugs act to when they act: how a swallowed dose becomes an active molecule, how long it stays, and how those facts shape the structure of a supervised session.
Key Takeaways
- 1.1 Affinity is how tightly a drug attaches to a receptor, efficacy is how much signal it produces once bound, and receptor occupancy is the share of a receptor population bound at a given moment. An agonist binds and activates, a partial agonist activates with a smaller maximum signal, and an antagonist such as ketanserin binds without activating. Psilocin is an agonist at serotonin 5-HT2A receptors, and several classic psychedelics behave as partial agonists there.
- 2.1 Classic psychedelics are agonists or partial agonists at serotonin 5-HT2A receptors, and the human blockade studies with ketanserin support the conclusion that this activation produces the state. Ketanserin given before psilocybin or LSD blocked the effects (Vollenweider et al., 1998; Preller et al., 2017). Ketanserin given one hour after LSD shortened subjective effects from 8.5 hours to 3.5 hours without changing the pharmacokinetics of LSD (Becker et al., 2023), which shows that continued receptor occupancy sustains the experience.
- 3.1 In the model fitted by Madsen et al. (2019), occupancy equals the maximum attainable occupancy of 76.6 percent multiplied by the plasma psilocin concentration, divided by the sum of the half-maximal concentration of 1.95 micrograms per litre and the concentration. Because the curve saturates, raising the concentration above about ten micrograms per litre changes occupancy very little: 11 micrograms per litre gives 65.1 percent and 17 gives 68.7 percent. The model was fitted in eight people with an agonist radioligand, so its estimates are tied to that sample and method.
- 4.1 LSD acts at the same receptor as psilocin but lasts several hours longer. Crystal, cell and simulation work shows that it dissociates very slowly under a lid formed by an extracellular loop (Wacker et al., 2017), which supports a mechanism without demonstrating it in people. The human evidence that the experience lasts as long as the receptor stays occupied comes from the ketanserin reversal study.
- 4.2 MDMA enters nerve terminals through monoamine transporters and reverses them, releasing mainly serotonin. Human pretreatment studies in which citalopram or duloxetine reduced or prevented its effects show that the transporter is its route of action. MDMA also raises circulating oxytocin, although the step from oxytocin to the subjective state rests on weaker ground.
- 4.3 Ketamine blocks the open channel of the NMDA glutamate receptor and is extensively metabolized to compounds that may themselves have clinical relevance. Ibogaine and its metabolite noribogaine interact with several target families at low micromolar concentrations. Because a 2021 review describes its mechanism as poorly defined, mechanistic claims about ibogaine remain weaker than those about psilocybin.
Reflection
A hypothetical colleague argues that nobody can say which receptor produces the psychedelic state, because psilocin binds several serotonin receptor subtypes. Three human studies bear on the argument. Vollenweider et al. (1998) found that the psychotomimetic effects of psilocybin were blocked in a dose-dependent way by the 5-HT2A antagonist ketanserin and by risperidone, while the dopamine antagonist haloperidol increased them. Preller et al. (2017) gave 22 participants LSD with and without ketanserin pretreatment and found that general subjective effects were fully blocked by ketanserin. Becker et al. (2023) gave ketanserin or placebo to 24 participants one hour after one hundred micrograms of LSD, and found that subjective effects shortened from 8.5 hours to 3.5 hours while the pharmacokinetics of LSD were unchanged. In 150 to 250 words, explain what these three designs together establish, identify which design supports the claim that continued receptor occupancy is needed for the state to continue, and state one limit that applies to all three.
The three studies answer different questions. Vollenweider et al. (1998) and Preller et al. (2017) blocked the receptor before the drug was given, which tests whether 5-HT2A activation is needed for the state to begin. In both, the state largely failed to appear. The contrast with haloperidol in the first study adds a further element, because a dopamine antagonist increased the effects, which argues against a dopaminergic explanation. Becker et al. (2023) is the design that speaks to continuation: ketanserin was given one hour after LSD, when the experience was under way, and the experience shortened by five hours while plasma LSD followed its usual course. Because the concentration was unchanged and the experience ended early, the state depends on continued occupancy of the receptor by LSD.
A limit common to all three is that ketanserin is not perfectly selective, so blockade of a related receptor could contribute. The samples are also small and consist of healthy volunteers, and each study measures subjective report without a clinical outcome. The evidence supports a strong claim about the receptor and a much weaker one about therapeutic benefit, which Lessons 8 to 10 address separately.
Minimum 20 characters required.
Question 1: Becker et al. (2023) gave ketanserin to participants one hour after one hundred micrograms of LSD. What did they report?
Question 2: In the model of Madsen et al. (2019), why does raising plasma psilocin from 11 to 21 micrograms per litre change occupancy by only about five percentage points?
Question 3: What is the principal molecular action of MDMA?
Question 4: What does the crystal structure reported by Wacker et al. (2017) establish?
Pharmacokinetics, Onset and Duration: Fitting an Agent to a Session
Introduction
HSCI 180 Lesson 3 described absorption, distribution, metabolism and elimination, and showed how the half-life of a drug sets the time course of its effects. Those ideas do the organizing work here. A supervised psychedelic session is built around a time course: how long until the person feels anything, when the effects peak, when they can safely be discharged, and how many staff hours the whole process takes. Section 1 asked what each agent binds. This section asks when, and for how long, enough of it is present to matter.
Two features of this drug family make the pharmacokinetics unusually consequential. The first is that one commonly used agent is a prodrug, inactive until the body converts it. The second is that a traditional preparation, ayahuasca, works only because a second group of compounds blocks the enzyme that would otherwise destroy the active one. The section takes these in turn before comparing the resulting time courses across agents, routes and doses.
Learning Objectives
- 5.1 Explain why psilocybin is a prodrug and describe the conversion to psilocin, using the measured plasma values from human pharmacokinetic studies.
- 5.2 Explain why DMT is inactive when swallowed on its own, and how the beta-carbolines in ayahuasca make an oral preparation active.
- 6.1 Compare the onset, peak and offset of six agents measured in controlled human studies, and relate those profiles to the length of a supervised session.
- 6.2 Describe how the route of administration changes the time course of DMT and ketamine, and what that allows a research team to control.
- 6.3 Weigh the evidence on fixed dosing compared with weight-adjusted dosing of psilocybin in trials, and state what the evidence does not settle.
5.1 Psilocybin Is a Prodrug: The Conversion to Psilocin
Swallowed psilocybin is dephosphorylated to psilocin, and psilocin is the molecule that occupies the receptor. The evidence that this conversion is essentially complete is direct. Brown et al. (2017) gave escalating oral doses of psilocybin to twelve healthy adults under monitoring and assayed blood and urine with liquid chromatography and tandem mass spectrometry. No psilocybin was found in plasma or urine. Renal clearance of intact psilocin accounted for less than two percent of total clearance, the pharmacokinetics of psilocin were linear across the range of doses studied, and the elimination half-life of psilocin was three hours with a standard deviation of 1.1 hours. Figure 3.2 shows the sequence.
Holze et al. (2023) characterized the time course of fixed doses in a larger sample, giving 15, 25 and 30 mg of psilocybin to 28, 23 and 28 healthy participants and measuring unconjugated psilocin alongside subjective ratings for 24 hours. Mean maximal concentrations were 11, 17 and 21 nanograms per millilitre, maximal concentrations were reached after an average of two hours, and elimination half-lives of free psilocin were 1.8, 1.4 and 1.8 hours. Mean durations of subjective effects were 5.6, 5.5 and 6.4 hours. Becker et al. (2022) reported the same short half-life of 1.8 hours for free psilocin, with a range of 1.1 to 2.2 hours. These values are shorter than the three hours reported by Brown et al. (2017), and Table 3.2 gives a third value, 2.3 hours (Ley et al., 2023). The studies differ in dose (weight-adjusted in Brown et al., fixed in the others) and in pharmacokinetic modelling, Holze et al. measured unconjugated (free) psilocin, and Brown et al. noted an extended elimination phase in some participants that suggests psilocin being released again from its glucuronide, so a range of roughly 1.5 to 3 hours is a fairer summary than any single figure. A systematic review of psilocin pharmacokinetics found the parameters consistent across studies, with terminal half-lives from 1.23 to 4.72 hours, and describes psilocybin as a prodrug that is rapidly absorbed and converted to psilocin (Otto et al., 2025).
Two numbers in that paragraph look inconsistent. The half-life of free psilocin is under two hours, and the experience lasts between five and six hours. The resolution is the shape of the occupancy curve from Section 1. Because occupancy saturates, a concentration well below the peak still occupies a large share of receptors, so the experience persists while the concentration falls through two or more half-lives. In the data of Madsen et al. (2019), intensity ratings were more closely associated with occupancy than with the plasma psilocin concentration.
3-D Model 3.3 follows psilocin to its site of action at a cortical synapse, where it activates 5-HT2A receptors and the signalling they trigger inside the neuron, and its final steps contrast the route that MDMA takes through the serotonin transporter.
5.2 Ayahuasca: Making an Oral Dose of DMT Possible
DMT swallowed on its own produces no psychedelic effect, because monoamine oxidase in the gut wall and liver breaks it down before it reaches the circulation. McKenna et al. (1984) analyzed the constituents of eight ayahuasca samples from Peru, measured the beta-carbolines harmine, harmaline and tetrahydroharmine in Banisteriopsis caapi and DMT in the admixture plants, and screened the beta-carbolines for monoamine oxidase inhibition in a laboratory assay. They reported that the beta-carbolines are highly active reversible inhibitors of the enzyme and proposed that this protects DMT from deamination and makes it orally active. Figure 3.3 sets out the mechanism.
Human studies confirmed the time course. Callaway et al. (1999) measured all four alkaloids in the tea and in the plasma of fifteen volunteers who drank a standardized preparation in a ceremonial context, and reported an average time to maximum DMT concentration of 107.5 minutes with a standard deviation of 32.5 minutes, with peak times of psychoactivity coinciding with high alkaloid concentrations. Riba et al. (2003) gave encapsulated freeze-dried ayahuasca at two doses to eighteen volunteers in a double-blind placebo-controlled trial, and found maximal DMT concentrations of 12.14 and 17.44 nanograms per millilitre, a median time to maximum concentration of 1.5 hours at both doses, and subjective effects peaking between 1.5 and 2 hours. Riba et al. (2001) had earlier reported that psychological effects were first noted after 30 to 60 minutes, peaked between 60 and 120 minutes, and were resolved by 240 minutes.
Riba et al. (2003) also found that harmine plasma levels were negligible and that deaminated monoamine metabolite levels in urine were not decreased, which suggests that the enzyme inhibition is predominantly peripheral, in the gut and liver. That detail matters for Section 4, where the interaction between monoamine oxidase inhibition and other serotonergic drugs is examined.
6.1 Onset, Peak and Offset Across Six Agents
Comparing time courses across agents requires studies that measure them the same way. The Basel group has published a series of placebo-controlled crossover studies that define onset, peak and offset from individual effect-time curves for the rating "any drug effect", using a threshold of ten percent of each participant's maximum response. Table 3.2 assembles the values, adding ayahuasca and ketamine from studies using their own measures. Interactive Tool 3.1 then plots the same values so that they can be compared against a planned session length.
Table 3.2. Onset, peak, offset and plasma half-life for six agents, as measured in controlled human studies.
| Agent and dose studied | Onset | Peak effect | Offset | Plasma half-life | Source |
|---|---|---|---|---|---|
| Psilocybin 20 mg by mouth | 0.5 h | 2.1 h | 5.3 h | 2.3 h (psilocin) | Ley et al. (2023), n = 32 |
| LSD 100 micrograms by mouth | 0.4 h | 2.3 h | 8.6 h | 3.5 h | Ley et al. (2023), n = 32 |
| Mescaline 500 mg by mouth | 0.9 h | 3.4 h | 12.0 h | 3.6 h | Ley et al. (2023), n = 16 |
| MDMA 100 mg by mouth | 0.5 h | 1.9 h | 4.9 h | 8.3 h | Straumann et al. (2023), n = 24 |
| Ayahuasca by mouth, 0.5 to 1.0 mg DMT per kg | 0.5 to 1 h | 1.5 to 2 h | About 4 h | Peak DMT at 1.5 h | Riba et al. (2001, 2003), n = 6 and 18 |
| Ketamine 0.5 mg per kg, intravenous over 40 minutes | Minutes | About 0.7 h | Within 1 h of the end of the infusion | 186 minutes at analgesic doses | Zarate et al. (2006); Włodarczyk et al. (2021); Clements and Nimmo (1981) |
Three patterns deserve attention. Psilocybin and LSD begin at almost the same time, and LSD lasts roughly three hours longer, which is a staffing difference as much as a pharmacological one. Mescaline has both the slowest onset and the longest duration, and Ley et al. (2023) attribute its longer action to its later time to maximal plasma concentration, since its elimination half-life is close to that of LSD. MDMA has the shortest duration of the four oral agents measured by the Basel group, consistent with the pooled analysis by Vizeli and Liechti (2017), who reported a duration of subjective effects of 4.2 hours with a standard deviation of 1.3 hours across 166 healthy participants given 75 or 125 mg in nine studies.
Matching a session length to the time course of each agent
Every bar shows measured values from the studies in Table 3.2: the bar begins at onset, the marker shows the time of peak effect, and the bar ends at offset. Move the session slider to set how many hours of supervised time are planned after dosing, then read which agents finish inside that window. The last slider adds the preparation and discharge time that surrounds the acute effects, so that the whole clinic day can be compared.
Values are the measured means in Table 3.2. The session line is a planning input chosen by the reader, and no study prescribes it.
(1) Select all six agents and set the supervised time to six hours. Which agents are still within their acute window when that time ends, and by how long? (2) Set the supervised time to nine hours and record which agents now finish inside the window. (3) Keep the supervised time at nine hours and set the added preparation and discharge time to two hours; record the total clinic day. (4) Using the numbers you recorded, write two sentences explaining why a service that offers LSD and mescaline sessions would need a different staffing model from one that offers psilocybin or ketamine sessions.
(1) At six hours, psilocybin (offset 5.3 hours), MDMA (4.9 hours) and ayahuasca (about 4 hours) finish inside the window, and ketamine finishes far earlier. LSD runs 2.6 hours beyond it and mescaline 6 hours beyond it. (2) At nine hours, LSD also finishes inside the window, with mescaline still running three hours beyond. (3) Nine hours of supervised time plus two hours of preparation and discharge gives an eleven-hour clinic day. (4) The staffing answer follows from the offsets: a mescaline session reaches a twelve-hour acute window before any preparation or discharge time is added, so it exceeds an ordinary eight-hour shift for a two-person team, while a ketamine session ends within two hours of the start of the infusion. Lesson 7 develops the structure of the session itself.
6.2 Route of Administration: What Intravenous Dosing Buys
Two of the agents in this lesson are given intravenously in research, and both show how the route changes what a team can control. Vogt et al. (2023) tested five intravenous DMT regimens in 27 healthy participants: placebo, two continuous infusion rates, and two combinations of a loading bolus with an infusion. The bolus doses produced very intense effects that peaked within the first two minutes and decreased within ten to fifteen minutes. The infusions without a bolus produced slowly rising effects that reached a plateau after 30 minutes and stayed at that level until the infusion stopped at 90 minutes, after which all drug effects subsided completely within fifteen minutes. The plasma data show a short early elimination half-life of 5.0 to 5.8 minutes followed by a later phase of 14 to 16 minutes.
The authors also reported that subjective effects were stable from 30 to 90 minutes despite further rising plasma concentrations, which they interpreted as acute tolerance to continued infusion. Section 3 returns to tolerance. For present purposes the point is a practical one: an intravenous infusion allows the intensity and the length of the state to be set and then ended, which an oral capsule does not allow. Both bolus conditions produced more negative subjective effects and more anxiety than the infusions, so the fastest route was also the least comfortable.
Ketamine in depression trials is given by a different intravenous schedule. Zarate et al. (2006) infused 0.5 mg per kilogram over 40 minutes in a randomized placebo-controlled crossover study in eighteen people with treatment-resistant major depression, with ratings at 40, 80, 110 and 230 minutes and at one, two, three and seven days. Antidepressant improvement reached significance within 110 minutes and remained significant through the following week. Włodarczyk et al. (2021), using the same dose and route in an observational study, measured dissociative symptoms with the Clinician-Administered Dissociative States Scale before the infusion, at 40 minutes and at one hour, and reported that all scores declined to an absent level within one hour after the infusion in every patient at every session. The acute dissociative state and the antidepressant effect therefore separate in time, which is part of why Lesson 8 treats the mechanism of ketamine's antidepressant action as an open question.
6.3 Dose, Body Weight and the Evidence on Fixed Dosing
Early psilocybin trials adjusted the dose to body weight, usually expressed as milligrams per 70 kilograms. As of October 2026, many recent trials, including the registration trials for depression, use fixed doses. Three lines of evidence support the change, and the box below sets them out.
Fixed doses compared with weight-adjusted doses of psilocybin
Clearance does not track weight. Brown et al. (2017) found that variation in psilocin clearance was not predicted by body weight in their twelve participants, and simulated that an oral dose of 25 mg would approximate the exposure produced by 0.3 mg per kilogram.
Response does not track weight. Garcia-Romeu et al. (2021) pooled ten previous studies, with 288 participants in total, and examined acute subjective effects in those given 20 mg per 70 kg, those given 30 mg per 70 kg, and those whose weight-adjusted dose was about 25 mg. Across body weights from 49 to 113 kilograms, no significant associations were found between subjective effects and body weight or sex. The authors concluded that the convenience and lower cost of fixed dosing outweigh any advantage of weight adjustment.
Fixed doses behave predictably. Holze et al. (2023) found dose-proportional pharmacokinetics for 15, 25 and 30 mg, dose-dependent intensity and duration of subjective effects, and no influence of body weight on pharmacokinetics or response.
What this evidence settles is narrow. It concerns psilocybin, in the dose range used in trials, with acute subjective effects as the outcome, in samples that were mostly healthy volunteers or trial participants meeting narrow eligibility criteria. It does not establish that fixed dosing is appropriate for other agents, that acute subjective effects are the right outcome on which to base a dosing decision, or that the result would hold at the extremes of body size, since the pooled range stopped at 113 kilograms. Garcia-Romeu et al. (2021) described their analyses as post hoc, which is a further reason for caution. Research on dosing is evolving, so the current practice may change.
The dose-response relationship itself is clearer. Holze et al. (2023) reported maximal effects of 58, 73 and 80 percent of the scale for 15, 25 and 30 mg, and durations of 5.6, 5.5 and 6.4 hours. Ley et al. (2023) found that 300 mg of mescaline produced lower peak effects than 500 mg of mescaline, 100 micrograms of LSD and 20 mg of psilocybin, which were similar to one another. The authors concluded that differences in the altered states produced by these three substances are dose-dependent, and that the substances themselves did not produce qualitatively different states at equivalent doses. Lesson 5 examines what participants experience within those states, and how the strength of the experience relates to later outcomes.
Section 3 turns to what happens when these agents are taken more than once: tolerance within and across drugs, and the evidence on whether they produce compulsive use.
Key Takeaways
- 5.1 Psilocybin is a prodrug: swallowed psilocybin is dephosphorylated to psilocin, the molecule that occupies the receptor, and Brown et al. (2017) found no psilocybin in plasma or urine. After fixed oral doses, psilocin peaked at an average of about two hours with a half-life under two hours, yet the experience lasted five to six hours, because the saturating occupancy curve keeps a large share of receptors bound while the concentration falls.
- 5.2 DMT swallowed alone is inactive because monoamine oxidase in the gut wall and liver breaks it down before it reaches the circulation. The beta-carbolines in ayahuasca reversibly inhibit that enzyme, mainly in the gut and liver, so oral DMT becomes active, with peak concentrations at about 1.5 to 2 hours.
- 6.1 Measured onsets, peaks and offsets differ enough across agents that the choice of agent determines the length of the clinic day, from under two hours for an intravenous ketamine session to twelve hours for a high mescaline dose. Psilocybin and LSD begin at almost the same time, and LSD lasts roughly three hours longer, which is a staffing difference as much as a pharmacological one.
- 6.2 An intravenous infusion allows the intensity and length of the state to be set and then ended, which an oral capsule does not. Infused DMT held a plateau and subsided within fifteen minutes of stopping, while bolus doses peaked within two minutes and produced more anxiety (Vogt et al., 2023). Ketamine infused over 40 minutes produces dissociation that resolves within an hour of the end of the infusion, while its antidepressant effect lasts days.
- 6.3 Body weight did not predict psilocin clearance or acute subjective effects in the studies examined, which supports fixed dosing of psilocybin within the trial dose range. The evidence is narrow: it concerns psilocybin, acute subjective effects as the outcome, mostly healthy or narrowly selected samples, body weights up to 113 kilograms and, in the pooled study, post hoc analyses.
Reflection
A hypothetical outpatient service is planning which agents it could offer if an authorized route were available, and asks you to advise on scheduling. The measured values are these. Psilocybin 20 mg by mouth: onset 0.5 hours, peak 2.1 hours, offset 5.3 hours (Ley et al., 2023). LSD 100 micrograms by mouth: onset 0.4 hours, peak 2.3 hours, offset 8.6 hours (Ley et al., 2023). Mescaline 500 mg by mouth: onset 0.9 hours, peak 3.4 hours, offset 12.0 hours (Ley et al., 2023). Ketamine 0.5 mg per kilogram given intravenously over 40 minutes: dissociative scores measured at 40 minutes had declined to an absent level within one hour of the infusion in every patient (Włodarczyk et al., 2021). In 150 to 250 words, compare what each agent would require of the service in staff hours, explain why mescaline lasts longer than LSD even though their plasma half-lives are similar, and state two things these numbers cannot tell the service.
The staffing demand follows the offset. A ketamine session is finished within about two hours of starting the infusion, so it can sit inside an ordinary clinic half-day. Psilocybin needs roughly five and a half hours of supervised time after dosing, and preparation and discharge push the day towards seven or eight hours. LSD needs about three hours more than psilocybin, and mescaline reaches twelve hours before any preparation time is added, which exceeds an ordinary eight-hour shift.
Mescaline outlasts LSD because its effects peak later. Ley et al. (2023) reported similar elimination half-lives of about 3.5 hours for both, and attributed the longer mescaline response to its later time to maximal plasma concentration and a broader peak, with a comparable rate of decline.
The numbers cannot tell the service how long an individual will take, because the ranges around these means are wide, including an offset range of 7.9 to 22 hours for the higher mescaline dose. They also say nothing about what support is needed during those hours, which depends on the therapeutic model covered in Lesson 7, or about the aftercare period, during which a person remains affected after formal discharge.
Minimum 20 characters required.
Question 1: Brown et al. (2017) found no psilocybin in the plasma or urine of participants given oral psilocybin. What does that show?
Question 2: Why is DMT inactive when swallowed on its own, and active in ayahuasca?
Question 3: Ley et al. (2023) found a longer response to mescaline than to LSD, although both had plasma half-lives of about 3.5 hours. What explains this?
Question 4: What does the evidence on fixed compared with weight-adjusted dosing of psilocybin support?
Tolerance, Cross-Tolerance and Dependence Potential
Introduction
Section 2 followed a single dose from administration to offset. This section asks what happens when these agents are taken more than once: whether the response shrinks within a session or across days, what tolerance shared between drugs reveals about their target, and whether the agents produce dependence or compulsive use.
Learning Objectives
- 7.1 Distinguish tolerance, physical dependence and compulsive use, and apply each term correctly to psychedelics.
- 7.2 Describe the evidence for acute tolerance within a single session, contrasting a single oral dose with a continuous infusion.
- 7.3 Describe the evidence for tolerance across repeated daily doses, and the limits of that evidence.
- 7.4 Explain what cross-tolerance between LSD and psilocybin implies about their shared receptor target.
- 7.5 Summarize the self-administration evidence for classic psychedelics and contrast it with the evidence for MDMA and ketamine.
- 7.6 Evaluate the pharmacological, experiential and contextual explanations offered for the low rate of compulsive use of classic psychedelics, and state the limits of each.
7.1 Tolerance, Dependence and Abuse Potential
Three words are often used as if they were one. Tolerance means that a repeated dose produces a smaller effect, so that a larger dose would be needed to produce the original effect. Physical dependence means that the body has adapted to regular exposure, so that stopping produces a withdrawal syndrome. Compulsive use, the behavioural core of addiction, means continued use that the person finds difficult to control and that persists despite harm. HSCI 180 Lesson 8 developed the neurobiology behind these distinctions, and HSCI 402 develops the treatment of substance use disorders. This section applies the distinctions to the agents in this course, because they separate here in an unusual pattern for the classic psychedelics: tolerance develops quickly, a 2026 review of abuse potential reports finding no accounts of withdrawal after these drugs (Heal et al., 2026), and compulsive use appears to be uncommon, with a different picture for MDMA and ketamine. Each part of that summary rests on a different kind of study, and the sections below name the evidence and how firm it is.
Heal et al. (2026) make a point worth holding onto through this section. In regulatory assessments of abuse potential, the term that drug regulators and scheduling laws use for the likelihood that a drug will be taken outside medical use for its rewarding effects, physical dependence on its own is not considered a signal, because it is a physiological adaptation to repeated exposure that also occurs with many drugs that have no abuse potential, including selective serotonin reuptake inhibitors. The question that matters for scheduling and for clinical risk is whether a drug functions as a reinforcer, which is a question about behaviour.
7.2 Acute Tolerance Within a Single Session
Within one session, the oral classic psychedelics show little evidence of acute tolerance. Liechti (2017), reviewing the modern LSD studies, reports that after a single dose the pharmacodynamic effects lasted as long as LSD was present in the body, with no evidence of acute tolerance to the effects of LSD. The picture from Section 1 explains why: occupancy tracks concentration, and the experience tracks occupancy.
Continuous intravenous infusion is different. Vogt et al. (2023) found that the subjective effects of DMT were stable from 30 to 90 minutes of infusion even though plasma concentrations continued to rise, which the authors interpreted as acute tolerance to continued infusion. The distinction is useful: a drug that is given once and then cleared rarely has the chance to show acute tolerance, while a drug held at a steady concentration for an hour does.
7.3 Tolerance Across Repeated Doses
Tolerance to classic psychedelics appears quickly when doses are repeated on consecutive days. Liechti (2017) notes that tolerance has been reported with repeated daily LSD administration over three to seven days, citing work from the 1950s. Isbell et al. (1961) gave each drug repeatedly over one to two weeks to small groups of participants and found that tolerance developed to both, to a lesser degree for psilocybin than for LSD, and that tolerance to either drug carried over to the other, which is called cross-tolerance. Figure 3.4 sets out the pattern described in these reports.
Interactive Tool 3.2 lets this pattern be explored by varying the interval between doses and the speed of recovery. The simulator is a teaching device: the shape it draws follows the published descriptions, and the numbers it produces are illustrative, because those descriptions give the direction and rough speed of the pattern rather than a measured recovery curve.
How dosing interval shapes tolerance and recovery (illustrative)
Each bar is the predicted response to one dose, as a percentage of the response to the first dose. The model assumes that every dose adds a fixed amount of tolerance and that tolerance decays with the recovery half-time you set. Change the interval between doses to see how spacing changes the response to later doses.
Illustrative model. It reproduces the direction and rough speed of the pattern reported by Isbell et al. (1961) and reviewed by Liechti (2017); the values are not measurements.
(1) Choose the daily dosing preset and record the predicted response to the fourth dose. (2) Choose the weekly preset and record the same value. (3) Keep the weekly interval and set the recovery half-time to its longest setting of ten days, then record the response to the fourth dose again. (4) Write two sentences on what the third result implies for a research team deciding how far apart to place two dosing sessions, and on why the uncertainty in the recovery half-time matters for that decision.
(1) With daily dosing and a three-day recovery half-time, the model gives 25 percent of the first response by the fourth dose. (2) With weekly dosing it gives 84 percent. (3) With a ten-day recovery half-time and weekly dosing it gives 45 percent. (4) The safe interval therefore depends on a quantity that the published descriptions give only roughly, so a team choosing an interval is relying on an estimate. If recovery is slower than the estimate assumes, a second session at the same dose would produce a weaker experience, and that weaker response would be hard to distinguish from a person simply responding differently on the day.
Two kinds of tolerance
Pharmacological tolerance is an adaptation at the drug's target or in its metabolism, and it is what the studies of acute and repeated-dose tolerance described above measure. A second kind of tolerance is learned. HSCI 180 Lesson 3 described conditioned tolerance, in which the body prepares for a drug in the setting where it has been taken before, and expectancy effects, in which what a person anticipates changes what they report. Both operate alongside receptor-level adaptation whenever a drug is given more than once in the same place by the same team.
Separating the two requires designs that vary the setting and the expectation, which most of the tolerance literature for this drug class did not do. Lesson 5 examines expectancy and blinding in detail, including the problem that participants usually know what they have received. A reported decline in response across sessions can therefore reflect receptor adaptation, reduced novelty, or a changed relationship with the research team, and the older studies cannot distinguish these.
The tolerance evidence also bears on repeated low-dose regimens. Madsen et al. (2019) observed that microdosing research would concern doses producing much lower occupancy than those in their study. How fully receptor occupancy recovers between the doses of such a schedule is not reported by the studies reviewed here, and a regimen that repeats every few days sits in the interval that those studies describe least well. Lesson 5 covers the self-blinding microdosing study and what it was able to establish about effects at these doses.
7.4 Cross-Tolerance and What It Shows
Cross-tolerance is pharmacologically informative. If tolerance to LSD reduces the response to psilocybin, as Isbell et al. (1961) reported, the two drugs are probably acting through the same mechanism, and an adaptation at that mechanism explains both. This is a behavioural argument for a shared target that was made before the receptor itself was identified, and it agrees with the blockade evidence in Section 1.
3-D Model 3.4 shows LSD and psilocin in the same binding pocket of the 5-HT2A receptor, the shared target to which the cross-tolerance findings point, and its earlier steps show the lid over LSD described in Section 1.
Cross-tolerance also has practical consequences. People who describe recent repeated use of one classic psychedelic may respond less to another, which matters when a clinician is interpreting a client's account of an experience that seemed unexpectedly weak. Trials also ask about recent use: the phase 3 trial of MDMA-assisted therapy for PTSD known as MAPP1, for example, excluded people who had used ecstasy in the six months before the first experimental session (ClinicalTrials.gov, NCT03537014).
7.5 Dependence Potential: What the Self-Administration Evidence Shows
The standard laboratory test of whether a drug functions as a reinforcer is intravenous self-administration, in which an animal can press a lever to receive an infusion. Drugs with high abuse potential maintain high rates of responding. Heal et al. (2026) summarize the classic psychedelic literature as a series of self-administration experiments in which none of the compounds tested had shown efficacy, and they offer two candidate explanations: animals may not experience the sensory and perceptual changes as rewarding, or rapid pharmacological tolerance may extinguish responding before a stable level of self-administration can be reached. They conclude that most, if not all, conventional 5-HT2A agonist psychedelics will probably not serve as positive reinforcers in this model.
Liechti (2017) states the same conclusion from the clinical side: LSD and other classic psychedelics are misused but are not addictive substances leading to compulsive drug taking, withdrawal or self-administration in animals. Johnson et al. (2018), assessing psilocybin against the eight factors of the United States Controlled Substances Act, describe limited reinforcing effects supporting marginal and transient self-administration in non-human animals, note that a few lifetime use occasions is the usual pattern among people who use psilocybin-containing mushrooms, and judge the scope of use and associated harms to be low compared with the drugs that serve as reference points for high abuse potential. Their conclusion was that Schedule IV placement may be appropriate if a psilocybin medicine were approved, which is a claim about scheduling under United States law. Lesson 11 covers the Canadian routes and schedules.
MDMA and ketamine sit differently, and Table 3.3 sets out the contrast.
Table 3.3. Tolerance and dependence indicators across the agents in this course, with the evidence behind each entry.
| Agent | Tolerance with repeated dosing | Self-administration in animals | Pattern of use reported in people |
|---|---|---|---|
| Psilocybin | Develops with repeated daily doses, to a lesser degree than for LSD; cross-tolerance with LSD (Isbell et al., 1961) | Marginal and transient (Johnson et al., 2018; Heal et al., 2026) | A few lifetime occasions is the usual pattern (Johnson et al., 2018) |
| LSD | Reported with daily dosing over three to seven days (Liechti, 2017) | Not maintained (Liechti, 2017; Heal et al., 2026) | Misused, without compulsive use or withdrawal (Liechti, 2017) |
| DMT given by infusion | Acute tolerance during a continuous infusion (Vogt et al., 2023) | Not described in the sources reviewed here | Not characterized in the sources reviewed here |
| MDMA | Chronic tolerance to reinforcing effects reported with long-term self-administration (Fantegrossi, 2008) | Maintained in rhesus monkeys (Fantegrossi, 2008) | Widely used recreationally (de la Torre et al., 2004); acute effects shorter than those of LSD (Vizeli and Liechti, 2017) |
| Ketamine | Dissociative scores fell across repeated infusions in one observational series (Włodarczyk et al., 2021) | Maintained by intravenous infusions in rhesus monkeys (Moreton et al., 1977) and rats (De Luca and Badiani, 2011) | Recreational use and harms from prolonged exposure are documented (Zanos et al., 2018) |
| Ibogaine | Not described in the controlled human studies reviewed here | Not described in the sources reviewed here | Used in single administrations in treatment settings (Koenig & Hilber, 2015) |
The row for ketamine deserves a qualification. Włodarczyk et al. (2021) measured dissociative symptoms across eight infusions in an observational study, and the only statistically significant difference between measurements was between the second and the eighth infusion, so the decline across sessions is a modest finding from a small uncontrolled series. Lesson 6 covers the bladder toxicity associated with frequent ketamine use, and Lesson 8 covers the private ketamine clinics where repeated dosing has become common.
The flip cards below name the four kinds of evidence used in this section, since each supports a different strength of claim.
7.6 Why Compulsive Use Is Rare, and What That Explanation Misses
Three explanations are usually offered for the low rate of compulsive use of classic psychedelics, and each is partial.
Tolerance arrives quickly enough that daily use produces little effect, so the pattern of escalating daily use seen with stimulants or opioids has no pharmacological pay-off. Heal et al. (2026) list rapid tolerance as one candidate explanation for the failure of self-administration studies. The limit of this explanation is that tolerance also recovers within days, so it blocks daily use without blocking weekly use.
The states are intense, sometimes difficult, and demand several hours of withdrawal from ordinary activity. Lesson 5 covers challenging experiences and how often they occur. Several hours of altered consciousness is a poor fit with the brief, repeatable reward that characterizes compulsive stimulant use. The limit of this explanation is that it describes an average and says nothing about the minority who do use frequently.
Availability, price and social setting shape the pattern of use for every drug. Classic psychedelics have been scheduled for decades, and most use happens in planned, occasional settings. The limit of this explanation is that it predicts that patterns would change if access changed, which is one of the open questions in the policy debate that Lesson 11 examines.
Two cautions close this section. First, a low dependence potential is not a general safety claim. Lesson 6 covers physical and psychiatric risks, which are the risks that screening is designed to manage, and those risks exist whether or not a drug produces compulsive use. Second, the evidence here comes from animal models that may not capture what humans find reinforcing, from survey patterns in populations with limited access, and from older human studies that would not meet current standards. Heal et al. (2026) describe the evaluation of abuse potential for this drug class as unresolved in several respects, in a field that they describe as moving very quickly. Research here is evolving, so confident summaries deserve caution.
Section 4 turns to the interactions that arise when these agents meet the medications a person is already taking, and to how that question is handled in screening.
Key Takeaways
- 7.1 Tolerance means that a repeated dose produces a smaller effect, physical dependence means that stopping produces a withdrawal syndrome, and compulsive use means continued use that is difficult to control and persists despite harm. In regulatory assessments of abuse potential, physical dependence on its own is not considered a signal, because it also occurs with drugs such as selective serotonin reuptake inhibitors; the question that matters is whether a drug functions as a reinforcer.
- 7.2 After a single oral dose of LSD, effects lasted as long as the drug was present, with no evidence of acute tolerance, because occupancy tracks concentration and the experience tracks occupancy. During a continuous DMT infusion, subjective effects held steady from 30 to 90 minutes while plasma concentrations kept rising, which Vogt et al. (2023) interpreted as acute tolerance.
- 7.3 Classic psychedelics produce rapid tolerance when doses are repeated on consecutive days, reported for LSD over three to seven days and to a lesser degree for psilocybin (Isbell et al., 1961), while spacing doses allows the response to recover. The studies described here did not measure the full recovery curve in people, and the older studies cannot separate receptor adaptation from conditioned tolerance, expectancy or reduced novelty.
- 7.4 Tolerance to LSD carried over to psilocybin and the reverse, which implies that the two act through the same mechanism. This behavioural argument for a shared target was made before the receptor was identified and agrees with the blockade evidence. In practice, recent repeated use of one classic psychedelic may blunt the response to another.
- 7.5 In the animal experiments reviewed by Heal et al. (2026), classic psychedelics did not maintain intravenous self-administration, and Johnson et al. (2018) describe a few lifetime occasions as the usual human pattern of psilocybin use. MDMA is self-administered by rhesus monkeys, and ketamine maintains self-administration in rhesus monkeys and rats. A null result in the animal model may reflect the drug or the limits of the model.
- 7.6 Rapid tolerance, the intense and demanding nature of the experience, and limited access each help explain why compulsive use of classic psychedelics is rare, and each is partial: tolerance recovers within days, the experiential account describes an average, and the contextual account predicts change if access changes. A low dependence potential is a claim about reinforcement, and it carries no information about cardiovascular, psychiatric or situational risk.
Reflection
A hypothetical news article states that psychedelics are non-addictive and therefore safe. You have the following evidence. Heal et al. (2026) report that intravenous self-administration studies of classic psychedelics in animals have not shown efficacy, and offer two explanations: the sensory changes may not be experienced as rewarding, or rapid tolerance may extinguish responding. Johnson et al. (2018) describe limited reinforcing effects, marginal and transient self-administration in animals, and a few lifetime use occasions as the usual pattern among people who use psilocybin-containing mushrooms. Liechti (2017) notes tolerance with repeated daily LSD dosing over three to seven days, and Isbell et al. (1961) demonstrated cross-tolerance between LSD and psilocybin. Fantegrossi (2008) reports that MDMA is self-administered by rhesus monkeys. In 150 to 250 words, write a correction of the article's claim that separates the dependence evidence from the safety claim, states what the animal evidence can and cannot support, and explains why MDMA belongs in a different category.
The article conflates two separate claims. The evidence on dependence is reasonably consistent: in the animal experiments that Heal et al. (2026) reviewed, classic psychedelics did not maintain self-administration, tolerance develops quickly enough to make daily use unproductive, and Johnson et al. (2018) describe a few occasions in a lifetime as the usual pattern of human use. That evidence supports a statement that these drugs rarely produce compulsive use. It does not support a statement about safety, which concerns different outcomes, including cardiovascular effects, psychiatric risk and the harms of unsupervised use, and which Lesson 6 examines.
The animal evidence has a particular weakness. A negative result in a self-administration study can mean that the drug is not reinforcing, or that the model cannot detect reinforcement of this kind, and Heal et al. (2026) name rapid tolerance as a possible reason that responding is extinguished before a stable pattern appears. Negative findings in a model therefore constrain the claim less firmly than positive findings would.
MDMA belongs in a separate category because it is self-administered by rhesus monkeys, which places it with drugs that function as reinforcers. Grouping it with the classic psychedelics under one heading about addiction potential obscures a real pharmacological difference.
Minimum 20 characters required.
Question 1: Why is physical dependence on its own not treated as a signal of abuse potential in regulatory assessment?
Question 2: Isbell et al. (1961) demonstrated cross-tolerance between LSD and psilocybin. What does that finding support?
Question 3: In the intravenous self-administration experiments reviewed by Heal et al. (2026), classic psychedelics did not maintain responding. How should that be interpreted?
Question 4: Vogt et al. (2023) found that DMT effects were stable from 30 to 90 minutes while plasma concentrations kept rising. What did they conclude?
Drug Interactions and Medication Screening
Content note
This section describes deaths reported after ibogaine use and after a combination of MDMA with an antidepressant, and it discusses seizures and cardiac arrhythmias. If you would like support while reading, the 9-8-8 Suicide Crisis Helpline is available across Canada by call or text at 9-8-8, and the Hope for Wellness Help Line serves Indigenous people across Canada at 1-855-242-3310. Both were verified on 17 October 2026.
Introduction
Many people who are considered for psychedelic-assisted therapy are already taking medication, often for the condition that brought them to the referral. Screening therefore involves a medication review before anything else is decided, and that review needs a working model of how interactions arise. The section first sets out that model, then works through the specific combinations that appear in trial protocols and in clinical guidance, and ends with the resources that clients themselves consult.
A boundary is worth stating at the outset. This section is about clinical screening and about reading the research literature. It does not tell any reader what to take or what to combine. Lesson 6 develops screening and eligibility as a whole, Lesson 7 covers the preparation phase in which the medication conversation happens, and Lesson 11 covers the legal routes through which an authorized session can take place in Canada.
Learning Objectives
- 8.1 Distinguish pharmacokinetic, pharmacodynamic and additive physiological interactions, and give a psychedelic example of each.
- 8.2 Summarize the evidence on serotonergic antidepressants and the acute response to psilocybin and MDMA, naming the design behind each finding.
- 8.3 Explain how monoamine oxidase inhibition, including the inhibition built into ayahuasca, creates a risk of serotonin toxicity, and why a medication review asks about more than antidepressants.
- 8.4 Explain the reported association between lithium and seizures, and state what an analysis of online reports can and cannot establish.
- 8.5 Describe how inhibition of the liver enzyme CYP2D6 changes exposure to ibogaine, MDMA and LSD.
- 8.6 Describe the cardiac screening questions raised by ibogaine, including QT prolongation, electrolyte levels and the role of CYP2D6.
- 8.7 Explain why the evidence does not settle whether to taper an antidepressant before a session, and describe the process through which that decision is made.
- 9.1 Describe what user-facing interaction resources publish, what they state about their own limits, and how a clinician would use a client's account of them.
8.1 Three Mechanisms of Interaction
Figure 3.5 sets out the three mechanisms that account for the interactions discussed in this section. A pharmacokinetic interaction changes how much of a drug reaches the circulation or how long it stays there. A pharmacodynamic interaction arises when another drug acts at the same receptor or transporter, blunting or amplifying the effect. An additive physiological interaction arises when two drugs push the same body system in the same direction, for example serotonin activity, the QT interval or blood pressure.
8.2 Serotonergic Antidepressants and a Blunted Response
A frequent screening question in this field concerns antidepressants, because many referred patients take one. Three kinds of study address it.
The first is a controlled pretreatment experiment. Becker et al. (2022) used a double-blind placebo-controlled crossover design in which healthy participants took escitalopram at 10 mg daily for seven days and then 20 mg daily for seven days, including the day of dosing, before receiving 25 mg of psilocybin. Escitalopram pretreatment had no relevant effect on the positive mood effects of psilocybin and significantly reduced bad drug effects, anxiety, adverse cardiovascular effects and other adverse effects. It did not alter the pharmacokinetics of psilocin. The authors described the finding as a first indication that psilocybin can be taken during escitalopram treatment without losing its positive effects.
The second is a survey of people who used psychedelics outside the laboratory. Barbut Siva et al. (2024) compared people taking a serotonin reuptake inhibitor with people who were not, and found that those not taking one reported stronger mystical experiences by 18.2 percent, stronger challenging experiences by 50.9 percent and stronger emotional breakthrough experiences by 31.9 percent, with no difference in visual effects. Both groups showed similar improvements in well-being and depressive symptoms afterwards. A survey of this kind cannot separate the drug interaction from the differences between people who are and are not taking antidepressants.
The third is a clinical trial that deliberately kept the antidepressant in place. Goodwin et al. (2023) gave a single 25 mg dose of the synthetic psilocybin formulation COMP360 with psychological support to nineteen participants with treatment-resistant depression who continued a selective serotonin reuptake inhibitor, in an open-label study with three weeks of follow-up. The mean change in the Montgomery-Asberg Depression Rating Scale total score at week three was a reduction of 14.9 points, with a 95 percent confidence interval from 20.7 to 9.2, and response and remission were each evident in eight of the nineteen participants. Twelve participants had a treatment-emergent adverse event, most of them mild and resolving on the day of onset, with no serious events and no indication of increased suicidal ideation or behaviour. The design was open-label with no comparison group, so it can show feasibility, and a test of efficacy would require a comparison group.
Taken together, the three designs answer different questions and none of them settles the clinical one. A systematic review of interactions involving classic psychedelics describes understanding of these interactions as limited and calls for further research, so this is an area in which the evidence is still developing (Halman et al., 2024).
For MDMA the direction of the evidence is clearer, because MDMA depends on the transporter that the antidepressant occupies. Liechti et al. (2000) found that the serotonin reuptake inhibitor citalopram attenuated the psychological effects of MDMA. Hysek et al. (2012) found that duloxetine, which inhibits both the serotonin and the noradrenaline transporters, prevented MDMA-induced elevations in well-being and emotional excitation. Blocking the transporter blocks the drug's route of action, which is the mechanism described in Section 1.
3-D Model 3.5 shows the transporter that the antidepressant occupies, with the selective serotonin reuptake inhibitor paroxetine bound at its central site, and compares the actions of ibogaine and MDMA at the same protein.
8.3 Monoamine Oxidase Inhibition and Serotonin Toxicity
Serotonin toxicity, also called serotonin syndrome, is the interaction with the most serious consequences in this field. Foong et al. (2018), writing for Canadian family physicians, describe it as a drug-induced condition caused by excess serotonin at synapses, presenting with a combination of neuromuscular, autonomic and mental status features. Most cases involve two drugs that raise serotonin by different mechanisms, and the authors identify the combination of a monoamine oxidase inhibitor with a serotonin reuptake inhibitor as the most concerning. Their list of serotonin reuptake inhibitors includes the selective serotonin reuptake inhibitors, the serotonin and noradrenaline reuptake inhibitors, tramadol, certain tricyclic antidepressants, dextromethorphan and St John's wort. They also note that tramadol, methadone, meperidine and fentanyl are weak serotonin reuptake inhibitors that can contribute, while morphine, codeine, oxycodone and buprenorphine do not.
Two features of this drug class make the question concrete. Ayahuasca contains reversible monoamine oxidase inhibitors as an essential part of its pharmacology, as Section 2 described. Callaway and Grob (1998) set out the resulting concern in print: the interaction between the harmala alkaloids in ayahuasca and selective serotonin reuptake inhibitors may induce a serotonin syndrome with potentially grave outcome. MDMA releases serotonin directly. Vuori et al. (2003) described four deaths after ingestion of MDMA with moclobemide, a reversible monoamine oxidase inhibitor, and concluded that the probable cause of death in each case was serotonin syndrome from the interaction. None of the four had been prescribed moclobemide, which led the authors to call for warnings against attempts to potentiate the effects of illicit drugs.
These reports help explain why trial protocols have required participants to stop psychiatric medications before dosing (the MAPP1 trial required a medically supervised discontinuation lasting at least five half-lives plus one week; Mitchell et al., 2021), and why a medication review asks about more than antidepressants. A person may be taking tramadol for pain, dextromethorphan in a cough preparation, or St John's wort bought without a prescription, and a review that asks only about psychiatric medication will miss all three.
8.4 Lithium: A Signal From an Unusual Data Source
A 2021 review of psilocybin and bipolar depression noted that people with bipolar disorder had been excluded from modern clinical trials, out of concern about activating mania or worsening the illness, which leaves little trial evidence on the risks of psychedelics alongside a mood stabilizer (Gard et al., 2021). Nayak et al. (2021) approached the gap differently, analyzing reports of classic psychedelics taken with mood stabilizers posted on three websites.
What the lithium analysis found and what it can support
The finding. Of 62 reports describing a classic psychedelic taken with lithium, 47 percent involved seizures and a further 18 percent described a challenging experience, while 39 percent involved medical attention. Of 34 lamotrigine reports, none described a seizure, and 65 percent of the lamotrigine reports were judged to have left the psychedelic experience unaffected, compared with 8 percent of the lithium reports. The authors concluded provisionally that psychedelic use may pose a significant seizure risk for patients taking lithium.
What the design can support. The reports are self-selected accounts written by people who chose to post them, with no denominator, no verification of what was taken, and a plausible bias towards posting dramatic outcomes. The analysis cannot estimate how often a seizure occurs per exposure.
Why it still matters. The comparison with lamotrigine is internal to the same data source and the same posting behaviour, which makes the contrast harder to explain by reporting bias alone. Given how little trial evidence there is on this combination, a signal of this size is enough to shape screening, and the authors state the limits plainly. Two of the authors are employed by Erowid Center, which operates one of the data sources, and the paper discloses this.
The lithium case is a useful example of how evidence of different strengths is combined in practice. A clinician does not need a randomized trial to decide to ask about lithium, and a researcher should still describe the evidence as weak.
8.5 Enzymes, Exposure and the CYP2D6 Problem
Pharmacokinetic interactions change how much drug is present. Three examples from this course's agents show the range.
Straumann et al. (2023) co-administered 100 mg of MDMA with 100 micrograms of LSD in 24 healthy participants. The peak plasma concentration of LSD rose from 1.9 to 2.1 nanograms per millilitre, the elimination half-life rose from 3.9 to 5.2 hours, and subjective effects lasted 9.9 hours with the combination compared with 8.4 hours for LSD alone. The authors attribute the change to inhibition of the enzyme CYP2D6 by MDMA, and note that patients taking antidepressants that inhibit CYP2D6, including fluoxetine, paroxetine, duloxetine and bupropion, could show similarly higher LSD concentrations.
Glue et al. (2015) gave a single 20 mg oral dose of ibogaine to 21 healthy volunteers who had been pretreated for six days with placebo or with the CYP2D6 inhibitor paroxetine. In the placebo group, ibogaine was rapidly converted to noribogaine, with median peak noribogaine concentrations at four hours. In the paroxetine group, ibogaine itself was absorbed substantially, remained detectable to 72 hours, and had an elimination half-life of 10.2 hours. Exposure to the combined active moiety was about twice as high. The authors suggested that genotyping patients and at least halving the intended dose in CYP2D6 poor metabolizers may be prudent.
Schmid et al. (2015) studied bupropion with MDMA in 16 healthy participants. Bupropion reduced the heart rate response and the rise in plasma noradrenaline, and at the same time increased plasma MDMA concentrations and prolonged its subjective effects, while MDMA increased plasma bupropion concentrations. An interaction can therefore reduce one effect and increase another at the same time, which is why a screening decision cannot be summarized as a single direction of risk.
The common thread is CYP2D6, a liver enzyme with well-known genetic variation that HSCI 180 Lesson 3 introduced through the example of codeine. People who metabolize poorly through this enzyme, and people taking drugs that inhibit it, reach higher concentrations from the same dose.
8.6 Cardiac Risk and Ibogaine
Ibogaine raises a different kind of interaction question, because the concern there is additive physiology as well as altered exposure. In cell studies, Koenig et al. (2014) reported that therapeutic concentrations of ibogaine reduce currents through human ether-a-go-go-related gene potassium channels expressed in a cultured cell line, which provides a mechanism by which the drug may generate life-threatening cardiac arrhythmias. The picture in heart cells is less simple. In guinea pig ventricular cardiomyocytes the same group found that ibogaine did not prolong the action potential and shortened it at higher concentrations, which they attributed to the drug also blocking L-type calcium channels, an effect they expect to count for little in human cells at the concentrations reached in treatment (Koenig & Hilber, 2015). Koenig and Hilber (2015) reviewed the case reports of long QT intervals, arrhythmias and sudden cardiac death associated with ibogaine, and set out the cardiac risk in detail. Among the risk factors they ask clinicians to clarify before ibogaine is given are abnormal electrolyte levels, pre-existing cardiovascular disease, drug interactions and genetic variation in drug metabolism. They single out methadone, which is associated with QT prolongation, has a long half-life of 15 to 55 hours and also inhibits CYP2D6.
Alper et al. (2012) reviewed all available autopsy, toxicological and investigative reports for the known fatalities outside West Central Africa that were temporally related to ibogaine use between 1990 and 2008. Nineteen people, fifteen men and four women aged 24 to 54, died within 1.5 to 76 hours of taking ibogaine. The clinical and postmortem evidence did not suggest a characteristic syndrome of neurotoxicity. In 12 of the 14 cases with adequate postmortem data, advanced pre-existing medical conditions, mainly cardiovascular, or one or more other substances explained or contributed to the death. The authors also identified seizures associated with withdrawal from alcohol and benzodiazepines, and the uninformed use of ethnopharmacological forms of ibogaine, as apparent risk factors.
Three screening implications follow directly. Other medications that prolong the QT interval add to the risk, so a full medication list matters. Low potassium was found in every case report in their review in which it was tested, and low magnesium in half of the reports, so potassium and magnesium levels matter (Koenig & Hilber, 2015). Because CYP2D6 determines exposure, as described under 8.5 Enzymes, Exposure and the CYP2D6 Problem, a CYP2D6-inhibiting antidepressant and ibogaine together raise concentrations of the drug whose cardiac risk is at issue. Lesson 6 develops cardiac screening and monitoring, and Lesson 10 covers the evidence for ibogaine in opioid use disorder, where these risks are weighed against a serious condition.
Table 3.4. Interaction classes most often addressed in screening, with the mechanism, the reported effect and the source.
| Combination | Mechanism | Reported effect | Source |
|---|---|---|---|
| Escitalopram with psilocybin | Pharmacodynamic, at serotonin receptors and transporters | Positive mood effects preserved; bad drug effects, anxiety and cardiovascular effects reduced; psilocin kinetics unchanged | Becker et al. (2022), controlled crossover, healthy volunteers |
| Ongoing serotonin reuptake inhibitor with a classic psychedelic | Pharmacodynamic | Weaker mystical, challenging and emotional breakthrough experiences reported; visual effects unchanged | Barbut Siva et al. (2024), survey |
| Citalopram or duloxetine with MDMA | Pharmacodynamic, transporter blockade | Psychological effects of MDMA attenuated or prevented | Liechti et al. (2000); Hysek et al. (2012) |
| Monoamine oxidase inhibitor with a serotonin releaser or reuptake inhibitor | Additive serotonergic | Serotonin toxicity; four deaths reported after MDMA with moclobemide | Foong et al. (2018); Vuori et al. (2003) |
| Ayahuasca with a serotonin reuptake inhibitor | Additive serotonergic, with monoamine oxidase inhibition built into the preparation | Potential for serotonin syndrome with grave outcome | Callaway and Grob (1998) |
| Lithium with a classic psychedelic | Not established | Seizures in 47 percent of 62 online reports; none in 34 lamotrigine reports | Nayak et al. (2021), online reports |
| CYP2D6 inhibitor with ibogaine | Pharmacokinetic | About twofold higher exposure to the active moiety | Glue et al. (2015), controlled study |
| CYP2D6-inhibiting antidepressant (bupropion; fluoxetine, paroxetine or duloxetine) with MDMA or LSD | Pharmacokinetic, with pharmacodynamic effects | Bupropion raised plasma MDMA, prolonged its subjective effects and reduced the heart rate response; higher LSD concentrations expected with CYP2D6 inhibitors, which the authors infer rather than test | Schmid et al. (2015), controlled crossover; Straumann et al. (2023) |
| QT-prolonging medication with ibogaine | Additive cardiac | Mechanism for arrhythmia through potassium channel inhibition (cell studies); deaths reviewed | Koenig et al. (2014); Koenig and Hilber (2015); Alper et al. (2012) |
8.7 Tapering Decisions
If an antidepressant blunts the response, one option is to stop it before the session. That decision has two sides, and the evidence does not resolve it. On one side, Barbut Siva et al. (2024) found weaker experiences among people taking a reuptake inhibitor, and the strength of the acute experience has been associated with outcome in several trials, which Lesson 5 examines. On the other side, Becker et al. (2022) found the positive effects of psilocybin preserved under escitalopram, Goodwin et al. (2023) reported substantial symptom change in people who continued their antidepressant, and stopping an antidepressant risks a return of the symptoms that brought the person to treatment, during the weeks when they are preparing for an intense experience.
What follows for practice is a process, and the evidence supports no single rule. The decision belongs to the prescriber who manages the medication, in consultation with the person and the team. It requires a plan for monitoring symptoms during any taper, an agreed point at which the taper would be abandoned, and a realistic account for the person of what is known and unknown. Lesson 6 develops stabilization and eligibility, and Lesson 7 covers how this conversation fits into preparation.
9.1 Interaction Resources Written for People Who Use Drugs
Clients arrive having read something. As of October 2026, two resources appear most often. TripSit publishes a combination chart and an online tool that return a category for a pair of substances, and the tool states on its own page that it is a quick-reference guide and not medical advice, that additional research must always be done, and that people should not mix drugs when in doubt. Its wiki page on drug combinations opens with a warning that the material is for educational purposes, that the authors do not endorse any of the combinations, and that the page will always be a work in progress. PsychonautWiki describes itself as an open encyclopedia of psychonautics, and its pages assemble user-contributed descriptions of effects, durations and combinations. Both were consulted on 17 October 2026.
The course's position on these sources, set out in Lesson 1, applies here. They document what people who use these drugs report, and they are useful for exactly that: understanding the vocabulary a client uses, and knowing what guidance a client has already encountered. They are not evidence of safety or efficacy, their entries are not systematically reviewed, and a category label on a chart carries no information about dose, individual health conditions or the specific products involved.
Nayak et al. (2021) show a third possibility. Treated as data about reports, with the sampling limits stated, user-contributed accounts can generate a signal about a combination that no trial will study, which can then shape screening. The difference lies in how the source is handled: as a dataset with known biases in a peer-reviewed analysis, as against a lookup table treated as a clinical answer. The practical consequence for a clinician is that a client's printout is a starting point for a medication review, and the review itself needs the prescription list, the health history and, where relevant, the prescriber.
Interactive Tool 3.3 puts these pieces together. Four hypothetical referrals are described, each with a medication list. For each medication, choose the screening implication that the evidence in this section supports, then check your answers.
Medication screening exercise
Each referral below is hypothetical and is not drawn from any real service or patient. Select the implication that fits each medication, then check your answers. Feedback names the evidence from this section. The exercise asks what a screening review would flag for discussion, and no part of it tells anyone what to take.
This lesson has moved from the receptor to the time course to the medication list. Section 1 established that the classic psychedelic state follows 5-HT2A receptor activation, and that MDMA, ketamine and ibogaine reach their effects by other routes. Section 2 showed how a swallowed dose becomes an active molecule and how long each agent occupies a session. Section 3 separated tolerance, dependence and compulsive use. Section 4 turned the pharmacology into the questions a screening review asks. Lesson 4 takes the next step, from the receptor to the brain networks and the plasticity that have been proposed to explain lasting change.
A hypothetical 54-year-old person with treatment-resistant depression has been referred for consideration of psilocybin-assisted therapy through an authorized route in Canada, the Special Access Program, which since 1 October 2026 has operated under the Controlled Substances Regulations (SOR/2025-242); Lesson 11 describes this route and its history, from the section 56 exemptions that began in August 2020, with applications supported by TheraPsil, to the January 2022 amendments that reopened the program to psilocybin. The person takes paroxetine 20 mg daily, lithium for a long-standing mood disorder diagnosis, tramadol for back pain, and an over-the-counter cough preparation containing dextromethorphan during the winter. Using Table 3.4 and the evidence set out under 8.2 to 8.6, write a short screening note of about 150 words that (a) names each medication and the mechanism by which it could interact, (b) states which single item would most change the team's decision and why, and (c) names the evidence behind that judgement and its strength. Do not recommend a dose or a combination; the note is a list of what the team and the prescriber would need to discuss.
Key Takeaways
- 8.1 A medication review looks for three mechanisms: a pharmacokinetic interaction changes how much drug reaches the circulation or how long it stays, a pharmacodynamic interaction occurs at the same receptor or transporter, and an additive interaction pushes the same physiological system, such as serotonin activity, heart rhythm or blood pressure, in the same direction. CYP2D6 inhibition with ibogaine, transporter blockade with MDMA, and QT prolongation with ibogaine are examples of each.
- 8.2 In a controlled crossover study, escitalopram left the positive mood effects of psilocybin intact, reduced adverse effects and did not change psilocin kinetics (Becker et al., 2022). A survey found weaker mystical, challenging and emotional breakthrough experiences among people taking a reuptake inhibitor (Barbut Siva et al., 2024), and an open-label trial in people who continued a selective serotonin reuptake inhibitor showed substantial symptom change without a comparison group (Goodwin et al., 2023). For MDMA the evidence is clearer, because citalopram and duloxetine block the transporter through which it acts.
- 8.3 Serotonin toxicity is the interaction with the most serious consequences, and a monoamine oxidase inhibitor combined with a serotonin reuptake inhibitor or releaser is the most concerning combination. Ayahuasca carries monoamine oxidase inhibition in its own pharmacology, and four deaths were reported after MDMA was taken with moclobemide. Tramadol, dextromethorphan and St John's wort also add serotonergic load, so a review that asks only about psychiatric medication will miss them.
- 8.4 In an analysis of online reports, 47 percent of 62 accounts of a classic psychedelic taken with lithium involved seizures, compared with none of 34 lamotrigine reports. Self-selected reports have no denominator and cannot estimate risk per exposure, but the internal comparison with lamotrigine makes the signal harder to explain by reporting bias alone. Given how little trial evidence there is on this combination, the signal is enough to shape screening, while the evidence is still described as weak.
- 8.5 CYP2D6 inhibition raises exposure: paroxetine pretreatment roughly doubled exposure to the active moiety of ibogaine, MDMA raised LSD concentrations and lengthened its effects, and bupropion raised plasma MDMA while reducing the heart rate response. People who metabolize poorly through CYP2D6, and people taking drugs that inhibit it, reach higher concentrations from the same dose.
- 8.6 Ibogaine inhibits cardiac potassium channels in cell studies, which provides a mechanism for QT prolongation and arrhythmia, and in a review of 19 deaths, pre-existing illness, mainly cardiovascular, or other substances explained or contributed to most of the deaths with adequate data. Screening therefore asks about other QT-prolonging medications such as methadone, about potassium and magnesium levels, and about CYP2D6 inhibitors that raise exposure.
- 8.7 Stopping an antidepressant before a session may strengthen the acute response, but it risks a return of the symptoms that brought the person to treatment, and the evidence does not resolve the trade-off. The decision belongs to the prescriber, in consultation with the person and the team, with a plan for monitoring symptoms and an agreed point at which a taper would be abandoned.
- 9.1 Resources such as TripSit and PsychonautWiki document what people who use drugs report and state their own limits; they help a clinician understand a client's vocabulary and prior guidance, and they are not evidence of safety or efficacy. User reports can generate a signal when analyzed as data with stated biases, and a client's printout is a starting point for a medication review rather than a clinical answer.
Reflection
A hypothetical 46-year-old person has been referred for consideration of psilocybin-assisted therapy through an authorized route in Canada. Their medication list is: escitalopram 20 mg daily for depression; lithium, prescribed for several years; tramadol 50 mg as needed for back pain; and a cough preparation containing dextromethorphan used in winter. The relevant evidence is as follows. Becker et al. (2022) found that escitalopram pretreatment preserved the positive effects of psilocybin and reduced anxiety and cardiovascular effects, with no change in psilocin kinetics. Nayak et al. (2021) found seizures described in 47 percent of 62 online reports of a classic psychedelic taken with lithium, and in none of 34 lamotrigine reports. Foong et al. (2018) list tramadol and dextromethorphan among drugs that raise serotonin and identify a monoamine oxidase inhibitor combined with a serotonin reuptake inhibitor as the most concerning combination. In 150 to 250 words, set out what the screening review would flag for each medication, say which item would most change the team's decision, and state how strong the evidence behind that judgement is.
Escitalopram raises a question about the strength of the acute response and about the decision to continue or taper, which belongs to the prescriber. Becker et al. (2022) found positive effects preserved and adverse effects reduced under escitalopram, so continuing the medication is a defensible option and the main risk of tapering is a return of symptoms before the session.
Lithium is the item that would most change the decision. The association with seizures reported by Nayak et al. (2021) is large and the internal comparison with lamotrigine, drawn from the same posting behaviour, is difficult to explain by reporting bias alone. The evidence is nonetheless weak in form: self-selected online reports with no denominator, no verification of what was taken, and authors who disclose an affiliation with one of the data sources. A team would treat it as grounds for careful discussion with the prescriber and for considering a change in mood stabilizer well before any session.
Tramadol and dextromethorphan each add serotonergic load and are easy to miss in a history, since one is taken as needed and the other is bought without a prescription. Both belong on the list that the prescriber reviews for cumulative serotonergic risk.
Minimum 20 characters required.
Question 1: According to Foong et al. (2018), which combination is the most concerning cause of serotonin toxicity?
Question 2: Nayak et al. (2021) found seizures in 47 percent of 62 lithium reports and in none of 34 lamotrigine reports. What can that analysis support?
Question 3: Glue et al. (2015) pretreated volunteers with the CYP2D6 inhibitor paroxetine before a single 20 mg dose of ibogaine. What did they find?
Question 4: How should a practitioner treat a client's printout from a user-facing drug combination chart?
Final Assessment
Bringing It All Together
This lesson followed four questions about the same group of drugs: what they bind, how long they last, what happens when they are repeated, and what they do alongside other medication. The answers separate the agents more sharply than their shared label suggests. Classic psychedelics produce their state through serotonin 5-HT2A receptors, a conclusion that rests on blockade studies in people as well as on binding data. MDMA works through monoamine transporters, ketamine through the NMDA receptor channel, and ibogaine through several target families at once, which is why mechanistic claims about ibogaine remain the weakest of the group.
The time course turned out to matter as much as the target. A swallowed dose of psilocybin is inactive until it is converted to psilocin, oral DMT is inactive until an enzyme is inhibited, and the length of a session is set by the time to peak concentration as much as by the half-life. Those facts translate directly into staffing, monitoring and the structure of a clinic day, which Lesson 7 develops.
The final section turned pharmacology into practice. A medication review asks three questions: whether another drug changes exposure, whether it occupies the same target, and whether it pushes the same physiological system in the same direction. The evidence behind the answers ranges from controlled crossover studies to analyses of self-selected online reports, and naming the strength of each is part of using it well.
Key Takeaways from this lesson
- Affinity, efficacy and receptor occupancy are separate properties: an agonist such as psilocin binds and activates the receptor, a partial agonist produces a smaller maximum signal, and an antagonist such as ketanserin binds without activating.
- Classic psychedelics are agonists or partial agonists at serotonin 5-HT2A receptors, and the human blockade studies with ketanserin support the conclusion that this activation produces the state, which a 2024 systematic review found consistent across studies. Ketanserin given one hour after LSD shortened subjective effects from 8.5 hours to 3.5 hours without changing the pharmacokinetics of LSD, which shows that continued receptor occupancy sustains the experience.
- Receptor occupancy follows a saturating curve, so raising the plasma psilocin concentration above about ten micrograms per litre changes occupancy very little.
- MDMA works through monoamine transporters, ketamine through the NMDA receptor channel, and ibogaine through several target families at once. The slow release of LSD from its receptor rests on structural work in cells, and mechanistic claims about ibogaine, which rest on binding and electrophysiology work, remain the weakest of the group.
- Psilocybin is a prodrug, and human studies find no psilocybin in plasma or urine because it is converted to psilocin before reaching the general circulation. Oral DMT is active only when monoamine oxidase is inhibited, which the beta-carbolines in ayahuasca accomplish mainly in the gut and liver.
- Measured onsets, peaks and offsets differ enough across agents that the choice of agent determines the length of the clinic day, from under two hours for an intravenous ketamine session to twelve hours for a high mescaline dose. An intravenous infusion lets a team set and end the state, and body weight did not predict psilocin clearance or acute subjective effects in the studies examined, which supports fixed dosing of psilocybin within the trial dose range.
- Classic psychedelics produce rapid tolerance with daily dosing and cross-tolerance between agents, and in the self-administration experiments covered by a 2026 review they had not maintained responding, while MDMA is self-administered by rhesus monkeys. A low dependence potential is a claim about reinforcement, and it carries no information about cardiovascular, psychiatric or situational risk.
- Medication screening looks for three mechanisms: altered exposure through liver enzymes, competition at the same receptor or transporter, and additive effects on serotonin, heart rhythm or blood pressure. The evidence behind each ranges from controlled crossover studies to analyses of self-selected online reports.
- User-facing interaction resources document what people who use drugs report and are not evidence of safety or efficacy; a client's account of them is a starting point for a medication review, and only a peer-reviewed analysis with stated biases turns such reports into a usable signal.
Core Concepts Reviewed
Section 1: affinity, efficacy, agonism and partial agonism, receptor occupancy, the ketanserin blockade designs, LSD binding kinetics, transporter reversal by MDMA, NMDA channel blockade by ketamine, and the multiple targets of ibogaine and noribogaine.
Section 2: prodrug conversion of psilocybin to psilocin, monoamine oxidase inhibition and oral DMT, onset, peak, offset and half-life across six agents, the effect of route of administration, and the evidence on fixed compared with weight-adjusted dosing.
Section 3: tolerance, acute tolerance, cross-tolerance, physical dependence, compulsive use, intravenous self-administration as a model, and the limits of negative findings in that model.
Section 4: pharmacokinetic, pharmacodynamic and additive interactions, serotonin toxicity and monoamine oxidase inhibition, the lithium seizure signal, CYP2D6 and exposure, QT prolongation with ibogaine, tapering decisions, and the standing of user-reported interaction resources.
The reflection below asks you to explain the pharmacology of this lesson to a general audience, using the findings you have worked with.
Reflection
A hypothetical community education evening asks you to explain, in plain language, why two people given the same 25 mg dose of psilocybin can have very different experiences. Draw on the whole lesson. The facts available to you are these: Madsen et al. (2019) fitted a single-site binding model in which occupancy rises steeply at low plasma psilocin concentrations and flattens above about ten micrograms per litre, with a maximum of 76.6 percent and a half-maximal concentration of 1.95 micrograms per litre, and reported wide variation between individuals in the concentration reached from a given dose. Holze et al. (2023) reported mean maximal concentrations of 11, 17 and 21 nanograms per millilitre for 15, 25 and 30 mg, with effects lasting 5.5 to 6.4 hours and no influence of body weight. Becker et al. (2022) found that escitalopram pretreatment reduced anxiety and adverse effects without changing psilocin kinetics. Barbut Siva et al. (2024) found weaker mystical and challenging experiences among people taking a serotonin reuptake inhibitor. In 150 to 250 words, give three reasons for the variation that are grounded in these findings, and name one source of variation that pharmacology alone does not explain.
The first reason is pharmacokinetic. The same swallowed dose does not produce the same plasma concentration in every person. Madsen et al. (2019) found wide variation in the concentration reached from a given dose, and body weight does not account for it, since Holze et al. (2023) found no influence of weight on either kinetics or response.
The second reason is the shape of the occupancy curve. Below about ten micrograms per litre the curve is steep, so two people whose concentrations differ modestly can occupy quite different shares of the receptor population, while above that range the curve flattens and further differences matter less. Where a person falls on that curve shapes the intensity they report.
The third reason is other medication. Becker et al. (2022) showed that an antidepressant taken beforehand changed the character of the response, reducing anxiety and adverse effects, and Barbut Siva et al. (2024) found weaker mystical and challenging experiences among people taking a reuptake inhibitor.
What pharmacology does not explain is the contribution of the person and the setting: expectation, preparation, the relationship with the people present and the room itself. Lesson 5 examines that evidence, and it accounts for differences that no concentration measurement predicts.
Minimum 30 characters required.
Final Knowledge Assessment
Question 1: Which statement best describes the relationship between receptor occupancy and the subjective intensity of psilocybin?
Question 2: A hypothetical service wants sessions that finish within seven hours of dosing. Using the measured offsets, which agent falls outside that window?
Question 3: Why does a psilocybin experience last five to six hours when the half-life of free psilocin is under two hours?
Question 4: Which pair correctly matches an agent with its principal molecular action?
Question 5: What do the findings of Preller et al. (2017) and Becker et al. (2023) together establish about the 5-HT2A receptor?
Question 6: A hypothetical person taking lithium asks about psilocybin. Which response reflects the evidence most accurately?
Question 7: Which finding most directly supports the claim that MDMA acts through monoamine transporters in people?
Question 8: Which statement about tolerance to classic psychedelics is supported by the evidence in this lesson?
Question 9: A hypothetical trial team is deciding whether participants should stop an antidepressant before a psilocybin session. Which summary is accurate?
Question 10: Which screening question follows most directly from the pharmacology of ibogaine?
Question 11: What does the evidence on body weight and psilocybin dosing show?
Question 12: Which statement about intravenous DMT is supported by Vogt et al. (2023)?
Question 13: Why is a negative result in an animal self-administration study weaker evidence than a positive result?
Question 14: A hypothetical client brings a printout from a user-edited wiki listing drug combinations. Which use of that document is appropriate?
Question 15: Which sequence orders four agents by the duration of their acute effects, from shortest to longest, using the measured values?
Glossary: Key Terms, People & Studies
📚 Reference page, available throughout the lesson
The terms, measures and researchers that this lesson uses, with short definitions for reference while you read.