Psychedelic Journal Club: How Do Psychedelics Work?

A commentary on the article “REBUS and the Anarchic Brain: Toward a Unified Model of the Brain Action of Psychedelics” by Robin Carhart-Harris and Karl Friston, published in Pharmacological Reviews.

With positive clinical trial results piling up1,2 and federal research money beginning to flow, psychedelic drugs are capturing interest in the scientific and medical communities like never before3. But right now the state of psychedelic medicine is a little like test-flying a plane without understanding aerodynamics. Psychedelics are treating entrenched mental illnesses and improving lives, but nobody really knows how they work.

Zoom into the molecular level and the picture seems clear. “Classical” psychedelics include LSD, psilocin and its metabolic precursor psilocybin (the main active components of psychedelic mushrooms), DMT, and mescaline. These compounds are also referred to as serotonergic psychedelics, because at the receptor level, their mechanism of action is thought to be mediated by their ability to bind to and activate excitatory serotonin receptors, particularly the 2A subtype. In human studies, pretreatment of research volunteers with the serotonin 2A receptor blocker ketanserin before LSD administration completely prevented the acute psychoactive effects normally associated with psychedelics4. Such research convincingly makes the case that serotonin 2A receptor activation is crucial to psychedelics’ profound psychoactive effects.

But serotonin 2A receptor activation isn’t the whole story. Zoom out from the receptor level just a little bit – to the level of neural circuits and neural networks – and our understanding becomes murky. A satisfying explanation of how psychedelics work needs to not only describe what the molecules do to receptors and individual neurons, but also how changes in the patterns of activity across populations of neurons are linked to perceptual and cognitive effects. This latter question remains a mystery.

… psychedelics, combined with psychotherapy, can effectively treat a variety of mental health disorders including major depressive disorder, post-traumatic stress disorder, and substance use disorder3.

Important clues come from recent work showing that psychedelics, combined with psychotherapy, can effectively treat a variety of mental health disorders including major depressive disorder, post-traumatic stress disorder, and substance use disorder3. The etiologies of these mental illnesses are unknown, but the symptoms of the individual disorders are divergent enough to suggest that they are categorically different conditions with distinct underlying brain mechanisms. How then could it be possible that psychedelics – which share a single mechanism of action – are broadly useful for treating such a diverse range of conditions?

An article in the journal Pharmacological Reviews, by Robin Carhart-Harris and Karl Friston, proposes an answer to this question with a novel theory of psychedelics’ mechanism of action5. The theory is called Relaxed Beliefs Under Psychedelics, or REBUS for short (the S in REBUS comes from “psychedelics,” no doubt making for a catchier acronym than REBUP).

The REBUS theory is that at a fundamental level, these disorders of mental health share a common factor. Namely, to suffer from depression, addiction, or PTSD involves being caught in loops of rumination and possessing a deeply ingrained negative worldview and self-concept.

Carhart-Harris and Friston propose that the reason psychedelics provide relief from mental illness is that their mechanism of action at the level of neural population activity is to reduce the precision weighting of cognitive priors. To understand what this means, it is important to first recognize that this proposal is situated in the theoretical context of predictive processing and the free energy principle, theories expounded by Friston to explain brain functioning in general.

According to Friston’s free energy principle, the brain is fundamentally a prediction machine whose function is to generate an internal model of the world and the self as an embodied agent moving through the world6. Predictive processing is a theory of neural computation by which the free energy principle is implemented in the brain: prediction signals generated by internal models of the environment and the self are issued to brain areas that receive sensory input and compared with arriving inputs to produce prediction errors that then update and improve the generative model7.

For example, if you were walking down a street in your neighborhood, your brain’s internal models of the neighborhood could generate a prediction that the familiar local park will be there as you turn the corner. If instead you saw the park was replaced with another house, your internal model would receive a big prediction error, and your brain would update its model of your neighborhood. The next day when you approach the same spot in your neighborhood, your brain’s internal model would generate a new prediction about what it would see as you turn the corner.

The REBUS model proposes that psychedelics diminish the precision of the prediction signals – also known as priors – entailed by the free energy principle, relieving incoming sensory information and all subsequent signals that ascend the processing hierarchy from the constraints of internal models.

How does this theory account for psychedelics’ efficacy in psychiatric care? According to Carhart-Harris and Friston, the rumination and pessimistic outlook common to addiction, depression, and PTSD are symptoms of pathological internal models. These harmful internal models are like deep channels that confine neural activity to negative domains, and by diminishing the power of the predictive signals generated by these models, psychedelics lower the walls of the channels and allow neural activity to attain new trajectories and to establish less harmful patterns.

…psychedelics lower the walls of the channels and allow neural activity to attain new trajectories and to establish less harmful patterns.

This novel theory is intriguing and promising, especially in how it answers an urgent question about how psychedelics work by bringing in logic from the more established frameworks of the free energy principle and predictive processing. But REBUS is still just a conceptual model, and needs to be tested experimentally.

The REBUS theory is unsatisfying in one major way: How does it account for the acute perceptual effects of psychedelics? A truly comprehensive theory of psychedelics’ mechanisms of action should explain all the effects of their administration, including the long-term changes to outlook and personality, but also the perceptual distortions, complex imagery, synesthesia, hallucinations, and the general altered state of consciousness that constitute the psychedelic experience. It seems counterintuitive that these acute effects could be the result of perceptual prediction signals being disempowered. Loosening the grip of internal models in favor of incoming sense data suggests that perception should become more veridical, not more fantastical.

REBUS should be developed to clarify why that intuition is wrong, or the theory could lose credibility. In its current state, REBUS can accommodate empirical findings that psychedelics reduce precision weighting in perceptual priors just as easily as it can accommodate findings that there is no effect of psychedelics on perceptual priors, and that the effects of psychedelics on predictive signals only occur in high-level cognitive areas. Psychedelic researchers testing REBUS should steer clear of a situation in which the theory can fit whatever phenomena exist and whatever evidence is turned up. Clarifying REBUS so it can produce testable, falsifiable hypotheses will be the next step in the development of this promising theory.

It’s too soon to say whether REBUS will prove as successful for guiding psychedelic medicine as aerodynamics is for keeping planes aloft. It’s one of multiple theories seeking to explain how psychedelics work, but in the end there will be just one that flies the farthest.

Link to journal article

Do you think psychedelics are revolutionary psychiatric medicines? Or are they just the next entry in a long line of overhyped and ultimately disappointing attempts to biomedicalize mental illness? Tell us in the comments below!


Written by Sean Noah
Illustrated by Sean Noah
Edited by Desi Nesheva and Arielle Hogan


Become a Patron!


1 Mitchell, J. M. et al. MDMA-assisted therapy for severe PTSD: a randomized, double-blind, placebo-controlled phase 3 study. Nat Med 27, 1025-1033, doi:10.1038/s41591-021-01336-3 (2021).

2 Carhart-Harris, R. et al. Trial of Psilocybin versus Escitalopram for Depression. N Engl J Med 384, 1402-1411, doi:10.1056/NEJMoa2032994 (2021).

3 Lieberman, J. A. Back to the Future — The Therapeutic Potential of Psychedelic Drugs. New England Journal of Medicine 384, 1460-1461, doi:10.1056/NEJMe2102835 (2021).

4 Preller, K. H. et al. Changes in global and thalamic brain connectivity in LSD-induced altered states of consciousness are attributable to the 5-HT2A receptor. Elife 7, doi:10.7554/eLife.35082 (2018).

5 Carhart-Harris, R. L. & Friston, K. J. REBUS and the Anarchic Brain: Toward a Unified Model of the Brain Action of Psychedelics. Pharmacol Rev 71, 316-344, doi:10.1124/pr.118.017160 (2019).

6 Friston, K. The free-energy principle: a unified brain theory? Nat Rev Neurosci 11, 127-138, doi:10.1038/nrn2787 (2010).

7 Bastos, A. M. et al. Canonical microcircuits for predictive coding. Neuron 76, 695-711, doi:10.1016/j.neuron.2012.10.038 (2012).

Sean Noah

Sean is a postdoctoral researcher at the UC Berkeley Center for the Science of Psychedelics. He studies the link between how psychedelics change neural activity in visual cortex and their effects on visual perception. He received his PhD from UC Davis, where he studied the neural mechanisms of visual attention.