What psychedelics do in the brain according to new international fMRI research

In recent years, more and more studies on psychedelics and the brain have been appearing, but the results do not always neatly follow the same direction. That is precisely why this new publication in Nature Medicine Interesting. The researchers combined 11 separate resting-state fMRI datasets of five psychedelics, including psilocybin, LSD, mescaline, DMT, and ayahuasca, to see which brain changes recur most consistently. Their central conclusion is remarkably nuanced: psychedelics do not seem to simply cause all brain networks to fall apart, but primarily lead to increased communication between networks that are normally more strongly separated.

That is an important difference. In the older explanation, it is sometimes stated that psychedelics make the brain chaotic. This study shows instead that a brain state arises in which certain boundaries between networks temporarily become more flexible. In particular, networks involved in abstract thinking, self-reflection, and information integration become more strongly coupled to visual and bodily systems. According to the authors, this is the most robust, reproducible signature visible across different substances and datasets.

More communication between self-networks and sensory systems

The strongest finding of this mega-analysis is that psychedelics increase the functional coupling between transmodal networks and sensory or motor networks. By transmodal networks, the researchers mean, among others, parts of the default mode network and the frontoparietal network. These are systems often associated with self-reflection, meaning-making, and higher information integration. The sensory and motor systems, on the other hand, comprise networks for seeing, feeling, moving, and attention to the environment.

That pattern aligns well with what people often describe during a psychedelic session. Thoughts can feel more visual. Emotions are not only thought but also felt physically. Music seems to penetrate more deeply. Symbolism can be experienced more tangibly or directly. The study does not prove that these subjective experiences are caused by a single network mechanism, but it does provide a strong neurobiological framework for the idea that different layers of experience temporarily become more intertwined.

The authors also describe this as a kind of flattening of the normal processing hierarchy in the brain. In the normal state, there is more separation between concrete sensory input and abstract self-related processing. Under psychedelics, that separation seems to become less rigid. This aligns with models in which psychedelics allow more room for new associations, different perspectives, and less rigid information processing.

Less strong evidence for general network disintegration

A second important conclusion is perhaps even more interesting, because it contradicts a popular idea from earlier studies. Individual studies often spoke of a broad decline in within-network connectivity, meaning less cohesion within existing networks. In this large Bayesian analysis, the authors found only limited and selective evidence for this. Many of these effects were weak to moderate, varied by medium and by network, and frequently overlapped with zero when uncertainty was properly accounted for.

Visual and somatomotor subnetworks, in particular, still showed relatively consistent declines. For many other networks, such as parts of the default mode network and the frontoparietal network, the evidence was less conclusive. This means that the popular summary “psychedelics cause brain networks to fall apart” is too crude. This study points to targeted reconfiguration rather than general dysregulation.

This is relevant in practice. It suggests that the psychedelic state consists not only of disorganization but also of new forms of integration. It is precisely this combination of less rigid fixed patterns and increased communication between previously separate systems that can explain why insights are sometimes experienced as new, vivid, or deeply felt.

Subcortical areas also participate, especially the striatum.

The study examined not only the major cortical networks but also subcortical areas. There, the involvement of the caudate and putamen stood out in particular. In the analysis, these regions showed clearer linkages with sensory and, to some extent, transmodal networks. According to the authors, this striatal pattern is precisely one of the most consistent subcortical findings.

That is interesting, because the striatum is involved in the linkage between perception, behavior, context, and response selection. In other words, psychedelics appear to influence not only reflection and experience, but also systems that help determine how input is weighted and how the body, attention, and behavior become attuned to it. The thalamus, which receives much emphasis in some theories, emerged less strongly and less consistently in this mega-analysis than was often assumed.

Differences between psilocybin, LSD, DMT, and ayahuasca

Psilocybin and LSD showed strikingly similar patterns in this analysis. This supports the idea that these substances share many mechanisms at the level of large brain networks. This does not mean that the session experience is identical, but rather that their acute influence on network organization overlaps to a significant extent.

Qualitatively, DMT showed the strongest disruptions of network linkages, but certainty regarding this was more limited due to the small sample size and greater individual variation. Ayahuasca, on the other hand, deviated the most from the other substances and showed a profile that was more difficult to interpret. The authors cite two logical explanations for this: ayahuasca is pharmacologically more complex than pure DMT, and the dataset was very small. This makes this study strong for general mechanistic lines, but less suitable for establishing strict rankings between substances.

What this might mean for guided psychedelic sessions

This study is not a treatment study and therefore does not directly prove which form of psychedelic therapy works best. The participants were healthy adults and the primary outcome was brain connectivity during the acute state. Nevertheless, the findings are relevant to how guided sessions can be understood.

If psychedelics indeed lead to increased communication between networks for self-reflection, sensory processing, and bodily experience, then it becomes clear why set and setting are so important. Music, safety, intention, the space, the relationship with the facilitator, and the way bodily signals are received can all have a greater influence when the brain temporarily functions in a more open and less hierarchically organized manner. That is a reasoned implication of this study, not a directly tested treatment outcome, but it aligns well with the practice of careful preparation and guidance.

For people with fixed thought patterns, persistent self-narratives, or emotional rigidity, this also offers an interesting perspective. If the usual separation between thinking, feeling, and perceiving temporarily becomes more flexible, space may open up for new interpretations and more profound insights. Not because the brain simply becomes dysregulated, but because it temporarily integrates information in a different way. The study thus primarily supports a model of reorganization and loosening, not of pure disintegration.

Why this study is strong, but also has limitations

The strength of this article lies in its design. The researchers combined 11 datasets, used a uniform preprocessing pipeline, and applied a Bayesian hierarchical model. As a result, they were able to examine not only which effects were visible on average, but also which effects are likely to be truly robust across studies and resources. This makes the study methodologically significant.

At the same time, there are limitations. The datasets differed in dosage, route of administration, time of scanning, MRI scanner strength, study design, and sample size. Some studies lacked a placebo or a fixed sequence. Moreover, some results proved sensitive to analytical choices, such as the presence or absence of global signal regression. The authors are transparent about this. Their conclusion is therefore not that every detail is now definitively established, but rather that, despite all these differences, a clear core is still visible.

Conclusion

This international mega-analysis shows that psychedelics likely primarily alter communication between major brain networks. Instead of a simple story of network breakdown, we see a picture of temporary reconfiguration. Self-related and abstract systems become more strongly intertwined with visual, bodily, and sensory networks. Additionally, the caudate and putamen, in particular, demonstrate a prominent role. Intra-network degradation may also exist, but appears less broad and less robust than previous standalone studies sometimes suggested.

For guided psychedelic sessions That is an interesting thought. Perhaps the power of psychedelics lies not only in breaking free from fixed patterns, but also in temporarily enabling new connections between feeling, perception, meaning, and self-experience. That is precisely why the circumstances in which such a session takes place remain so important.