A small, exploratory study suggests that people who report experiencing mediumistic trances appear to show distinct patterns of brain connectivity compared to individuals engaged in prayer. The researchers propose that the brain activity recorded during these episodes suggests an altered state of consciousness that does not resemble typical sleep or psychiatric conditions. The findings were published in the journal Psychiatry Research: Neuroimaging.
Mediumship describes a practice where a person subjectively believes they are acting as a vessel for communication from a source existing outside physical reality. Many individuals report these experiences as a sudden shift in their awareness, allowing them to hear or see alleged spiritual entities. While personal accounts of such events are common in many cultures, attempting to study the physiological process through modern neuroscience represents a relatively new and highly complex frontier.
Previous neuroimaging studies have observed basic brain activity in religious practitioners by looking at localized hotspots. However, researchers want to understand exactly what happens in the wider brain network when a person voluntarily enters what they describe as a trance state. Instead of looking at isolated regions, they are trying to identify how different areas collaborate.
Lead author Jessica Placido, a researcher at the Bahiana School of Medicine and Public Health in Brazil, investigated this phenomenon alongside colleagues from several Brazilian universities. They specifically examined brain functional networks. This analytical approach treats the human brain as a highly interconnected system.
Traditional electroencephalography, or EEG, measures the brain’s electrical output through sensors placed directly on the scalp. Older experiments usually measured the raw power of brainwaves at specific, isolated locations. Placido and her team opted for a dynamic network method to see how these signals relate to one another.
Functional network analysis maps how different parts of the brain synchronize their activity over time. The researchers applied a mathematical framework called graph theory to track this behavior. In this system, the electrical readout from each scalp sensor represents a distinct node in a larger web.
The connections between these nodes are called edges. By constantly mapping the nodes and edges, the scientists create an evolving graph of human cognition. If two distant regions show the exact same sequence of electrical changes at the same time, they establish a functional connection.
The team conducted a small study using electroencephalography data collected from just 28 women in Campo Grande, Brazil. The researchers recruited participants from local centers practicing Spiritism. Spiritism provides a cultural context where channeling spirits is a voluntary practice, allowing the researchers an opportunity to scan subjects in a controlled environment.
Half of the women in the selection pool were experienced practitioners who claimed to have regularly channeled spirits for at least five years. The other 14 women served as a control cohort. These individuals attended the same religious centers and shared identical spiritual viewpoints, but they did not act as channels themselves.
The researchers confirmed this division by asking participants to complete a self-reported questionnaire called the Inventory of Anomalous Experiences. This survey tracks how often people claim to encounter unusual anomalies, like out-of-body experiences, unexplainable intuition, or seeing spirits. The medium group reported a much greater frequency of anomalous experiences compared to the control group.
To start the experiment, both groups sat quietly with their eyes closed for seven minutes in a dimly lit room. This initial phase established a baseline measurement for their resting brain activity. Following the baseline, the two groups participated in fundamentally different mental and physical tasks.
The medium group proceeded with an alleged psychophonic trance process, meaning they actively spoke out loud as a spirit supposedly communicated through their vocal cords. Meanwhile, the control group participated by listening to the session while silently praying. The researchers recorded the electrical patterns on their scalps throughout the concurrent session.
With the gathered data, the team mapped out multiple frequency bands. The human brain produces several types of electrical waves depending on a person’s mental state. They tracked delta waves, which are slow and typically linked to restorative rest, and theta waves, which are slightly faster and associated with deep relaxation.
The scientists also tracked alpha waves, which happen when a person is awake but mentally calm, and beta waves, which relate to active focused thinking. To build the network graphs, the software hunted for matching sequences of electrical spikes and dips between the different sensors. This technique allowed the scientists to quantify how long different brain areas remained in sync.
When comparing the trance phase to the baseline rest phase, the mediums displayed elevated brain connectivity in the delta and theta frequency ranges. Their brain regions synchronized more frequently and maintained those connections over longer periods. No such connectivity spikes occurred in the control group.
An increase in both delta and theta brainwaves typically signals the transition from wakeful awareness to early sleep. However, the subjects within the medium group were actively speaking and sitting upright during the recording. While the researchers propose that a mediumistic trance operates as a unique altered state of consciousness entirely separate from rest, it is also highly likely that physical actions, such as speaking and altered breathing, influenced these specific readings.
The scientists tracked an additional metric to determine whether the brain communicated across its two halves or kept activity confined locally. During the trance, the mediums experienced a change in their alpha waves. The left hemisphere altered its behavior, establishing an equal balance of connections within its own side and across to the right hemisphere.
The left side of the human brain generally handles language formulation, linguistic comprehension, and spoken speech. Because the mediums were actively vocalizing messages during the session, the left hemisphere logically demonstrated an altered state of activity. Yet, silent prayer also heavily utilizes these same brain regions.
Scientific literature suggests that praying involves internal dialogue and social narratives, which activate similar cortical networks. Despite the overlapping reliance on language networks for praying, the control group did not show the same robust shifts in the left hemisphere. The researchers suspect the detected brain patterns relate exclusively to the nature of the trance itself, though the sheer difference in vocalization remains a prominent confounder.
The non-medium control group only showed a networking change in their beta waves. During the prayer session, the likelihood of one brain area dominating the rest of the network decreased. Their brain networks became more universally synchronized while they prayed.
The researchers acknowledge a few major limitations to the experimental procedure. Because this was a small study of only 28 individuals, the observed connectivity patterns represent a highly localized sample that requires rigorous future replication. Most notably, the specific tasks performed by the two groups varied drastically in their physical output.
The mediums spoke out loud during the trance, while the controls prayed entirely in silence. A physical difference like vocalizing a sentence introduces muscle movements and breathing changes that can severely disrupt and influence the raw electrical activity detected by an EEG on the scalp. The scientists recommend that future projects ask the control group to perform an improvised speaking task to mirror the physical actions of a channeled message, which would help rule out false signals.
Additionally, the portable recording equipment possessed a sampling limit that prevented the extraction of the gamma frequency band. Gamma waves are the fastest electrical signals generated by the brain and often deal with intense awareness. Upgraded medical capabilities would allow future investigators to monitor this missing frequency.
Despite these significant confounding factors, the researchers assert that the functional network approach offers a productive way to map religious experiences. The brain network dynamics observed in the mediums do not align with patterns associated with severe psychiatric diagnoses like schizophrenia. Based on the preliminary mapping data, the researchers conclude that alleged mediumship alters brain synchronization in a non-pathological manner, even if the exact physiological triggers of these alterations remain up for debate.
The study, “Brain functional network dynamics in women with alleged mediumship: a controlled study,” was authored by Jessica Placido, Marco Aurelio Vinhosa Bastos Jr., Paulo Roberto Haidamus de Oliveira Bastos, Thaise G.L. de O. Toutain, and Jose Garcia Vivas Miranda. It was published in February 2026.