The research was carried out in cities in the northeastern region of Brazil. The recruitment of participants was performed using the snowball sampling strategy, a technique in which a participant indicates another who meets the necessary requirement to participate. This is a non-probability sampling method that can be useful when the researcher is not able to known the number of persons with this ability in the population. The study staff did not known the participants identify during sociodemographic questionnaire applying. They only received their identity in order to complete the cognitive testing.

Nineteen psychophonic mediums who regularly attended Spiritist centers were included. These subjects reported having the ability to communication when deceased “beings” communicate directly via speaking. After initial selection, the subjects were divided into two groups. The first was considered to be the more “experienced” group (MEG) with 11 subjects, while the second was considered to be the less “experienced” group (LEG) with 8 subjects. Experience was defined by the length of time they had practiced mediumship (weekly or monthly). For the MEG, more than 10 years and a minimum of 1 trance activity per week were taken into account. For the LEG, 1–5 years of practice and up to 4 practices of the phenomenon per month were accepted. The period between 6 and 9 years was considered intermediate (the intermediate group).

Subjects who were absent from the activities developed in the Spiritist centers for more than 1 year were excluded, as well as those with decompensated psychiatric disorders (depression, schizophrenia), reports of sleep disorders (such as parasomnias, REM sleep behavioral disorder), a diagnosis of epilepsy according to the criteria of the International League Against Epilepsy (ILAE), subjects on psychoactive medications, reports of drug use or abuse, and subjects in the intermediate group.

A comparison of neuropsychological performance between the MEG and LEG was performed to analyze the differences in cortical processing. In previous studies with neuroimaging in meditation using the mindfulness technique, different patterns of cortical activity and higher speed in executive processing have been observed depending on the years that subjects had previously practiced meditation (Falcone and Jerram, 2018).

The data from the sociodemographic questionnaire were collected using Google forms; the participant’s identity and any form of identification were concealed. The following data were requested: date of birth, gender, level of education, profession/occupation, ethnicity, and marital status.

For the purposes of inclusive criteria, they were asked about the ability to enter into a mediumistic trance, especially in laboratory and scientific research environments; the period when they started practicing mediumship; what type of mediumship they practiced (psychophonic, psychographic, clairvoyant); how often they entered into a mediumistic trance, weekly and/or monthly; and how the participant characterized their state of consciousness during mediumistic communication (conscious, semiconscious, unconscious).

To verify the exclusion criteria, they were asked about the existence of previous diseases; if they continuously and/or daily used psychiatric medications; if they had ever had a seizure or sleep disorder; and if the time of mediumistic practice corresponded to the intermediate period of 6–9 years.

Due to the COVID-19 pandemic, we sought to preserve the participants’ health by maintaining face-to-face contact only for the application of neuropsychological tests. Thus, some instruments were adapted to an online format via Google forms: The informed consent form, the Beck Depression Inventory (BDI), the Self-Report Questionnaire (SRQ), and the Trauma History Questionnaire (THQ) were used exclusively for research purposes. After verifying the inclusion criteria, neuropsychological tests were applied, and cognitive functions were evaluated.

The participants had a mean age of 45.1 ± 9.8. The mean age at which the manifestation of mediumship was perceived was 21.4 ± 12.9 years. The sample was 68% female and 32% male.

There were 8 individuals in the LEG, representing 42% of the sample, and 11 in the MEG, denoting 58%. Regarding the level of consciousness during the mediumistic trance state, 58% were conscious, 32% were semiconscious, and 11% were unconscious. Those who reported being unconscious belonged to the MEG group.

As for level of education, 37% of the participants had up to 12 years of schooling, and 63% had 12 or more years of schooling. Regarding marital status, 37% were in a stable relationship, and 63% lived alone or with relatives.

The SRQ showed an incidence of common mental disorders in 21% of the sample (p = 0.008, chi-square test), all belonging to the LEG, according to Table 1.

Using the BDI, major depression was ruled out in 100% of the subjects. On the THQ, 100% of the sample mentioned having experienced at least one traumatic event, 10% had only experienced trauma related to crime events, 26% had only experienced trauma in general and during disasters, 31% had experienced situations related to crime and trauma in general, and 31% had experienced trauma related to physical and/or sexual violence. Those who experienced physical/sexual violence had also experienced all other events, as shown in Table 2.

In the evaluation of cognitive functions, on the BCSB, 100% of the subjects exhibited scores in the average and/or above average range at the end of the test, which evaluates late memory. The same outcome was found on the digit span subtest of the WASI; 100% of the subjects performed above or equal to the average found among Brazilian individuals. On the cubes subtest, 90% of the subjects scored higher than or equal to the average. On the FAS (verbal fluency) test, only 42% of the subjects scored higher than or equal to the Brazilian average (31% in the MEG and 10% in the LEG), as seen in Table 3.

On the FDT test, in the evaluation of the automatic processes that involve reading and counting, 53% showed percentiles equal to or higher than the average, of which 64% belonged to the MEG. In the controlled choice process, 68% had percentiles equal to or higher than the average; of these subjects, 64% belonged to the MEG. When the FDT subtest named controlled choice process was performed, there was statistical significance related to the age of perception of mediumship, with p = 0.038—chi-square test (Table 4).

Regarding performance in controlled process alternation, 63% had a score higher than or equal to the average; of these subjects, 45% belonged to the MEG. In the evaluation of inhibitory control, 58% had a score higher than or equal to the average, and 45% belonged to the MEG. On the cognitive function of flexibility, 74% had a score higher than or equal to the mean, and 64% belonged to the MEG.

In a comparative study between mediums and individuals with dissociative identity disorder, mediums differed from persons with dissociative identity disorder, indicating better social adjustment, lower prevalence of mental disorders, lower use of mental health services, no use of antipsychotics, a lower prevalence of histories of physical or sexual abuse in childhood, and sleepwalking (Moreira-Almeida et al., 2008). As a result, mediumship stood out, demonstrating better mental health and social adjustment. Other research compared spiritualist mental mediums to non-medium spiritualists. It suggested that mediums had better psychological wellbeing and reported lower psychological distress (Roxburgh and Roe, 2011). Another study with 3,023 participants suggested that individuals who claimed mediumship had higher dissociation scores than non-claimants, but neither group exceeded the threshold for pathology (Wahbeh and Radin, 2017).

According to Bastos et al. (2018), during mediums’ training, prior to communicating with dead individuals, one of the most important skills to be acquired is control over the manifestations of possession. There is a role of learning this control of the possession phenomenon, which focuses on reading and education, as to the principles of the spiritism and behavioral adjustment to cultural rules (Espirito Santo, 2010). This process reflects broad acquisition and application of aspects of executive processing, with inhibitory control over the possession phenomenon, as well as cognitive flexibility for changes in relation to the trance state and readjustment to the real environment.

Spirit possession cannot be considered pathological in relation to mental health, except when it causes clinically significant discomfort and impairment in social and occupational functioning (van Duijl et al., 2010). However, in this study, a difference was found in the prevalence of common mental disorders only in the LEG (i.e., in the group with less time of regular study and systematic activities aimed at controlling mediumship). It is inferred that the regularity of practices in the cultural and religious aspects of this practice may favor the concept of spirit possession as a phenomenon and not a disorder.

Another study suggested that performance on the verbal fluency test was significantly different among participants with intermittent dissociative phenomena when compared to transient dissociative phenomena (Al-Adawi et al., 2019). Such a finding did not corroborate the current study’s outcomes. Although 58% of the participants scored below average on the FAS test, there was no statistical significance between the groups. However, a below-average performance only on the FAS test—which was not observed in the digit span test and in all FDT subtests—makes executive dysfunction unlikely, even when stratifying the groups into MEG and LEG. It is hypothesized that the cognitive impairment observed in the study by Al-Adawi et al. (2019) may be related to the thought content experienced during possession phenomena.

Brain areas associated with the preservation and learning of memory, attention, and executive functioning tend to be dysfunctional in people with dissociative phenomena (Cima et al., 2001). Dissociation can be characterized by subtle deficits in neuropsychological performance, such as attention impairment, and some cognitive phenomena associated with dissociation seem to be dependent on the emotional or attentional context (Giesbrecht et al., 2008). However, in general, the mediums in this study exhibited scores within the Brazilian average on neuropsychological tests for the assessment of long-term memory, working memory, visuo-constructive ability, and planning. It is necessary to differentiate dissociative disorder from dissociative phenomena related to cultural experiences. Cognitive health also seems to follow this trend.

The fast, efficient reproduction of a series of elements on the FDT test indicates not only the presence of focused attention, but also the ability for automatization, learning, resistance to fatigue, and inhibitory control (Broverman et al., 1966). In the choice process, several variables are verified at the same time: attention, phonology, semantics, control, and inhibition of one response and activation of another. In our study, there was statistical significance between the age the medium began to exhibit mediumistic abilities and their scores within the choice process (a lower age of onset was associated with a higher score in this FDT subtest). We are unaware of studies that address such a comparison. There is probably a greater cortical susceptibility to the dissociative mediumistic trance phenomenon in some subjects, or the learning process and cortical specialization are facilitated by the early onset of mediumistic recognition. The phenomenon of neuronal pruning occurs until the age of 21, and is fundamental in the specialization of the functionality of individuals and preparation for the cognitive challenges that the individual will come across (Holguera and Desplan, 2018).

Mediumistic trance can be understood as a form of mediation in which an individual voluntarily access degrees of states of conscious. Both strategies (meditation and trance) may allow the subjects to access their ego-conscious (Wahbeh et al., 2019). Some studies showed an improvement in cortical activity by meditation using the mindfulness technique (Tanaka et al., 2014; Taren et al., 2017; Wang et al., 2020). Functional MRI neuroimaging studies have shown a change in the pattern of cortical activation between less experienced and more experienced mediumship practitioners. While human insula activation is observed in less experienced meditation practitioners, more experienced meditation practitioners activate areas of the right medial frontal gyrus, anterior cingulate gyrus, globus pallidus, and putamen (Falcone and Jerram, 2018).

In Brazil, the study conducted by Peres et al. (2012) with psychographic mediums (communication with deceased beings through writing) showed that among more experienced mediums, who wrote while in the trance state, cerebral blood flow was consistently lower in the regions responsible for complex writing compared to the control condition. During the mediumistic trance, there was low activity in the brain areas responsible for complex writing in the MEG. In this PET study, the more experienced mediums reported deeper trance states in contrast to the less experienced mediums. Cortical neuroimaging changes can be further detailed by thorough neuropsychological evaluation. In the present study, there was a tendency for better performance on tests that assessed subgroups of frontal tasks, suggesting better executive processing associated with the practice of dissociative, mediumistic trance phenomena.