Our study was performed in a total of N = 66 participants that were assigned to a group of MDD patients and a group of adults without MDD diagnosis. The MDD group contained a total of N = 38 patients (23 females, 15 males, Mean_age = 40.82 years ±2.09 S.E., Age-range: 19 to 65 years) with a medical diagnosis of MDD, comprising a current or previous major depressive episode in accordance to DSM-IV-TR (American Psychiatric Association, 2000) criteria. Only patients without any previous diagnosis or high screening score of a psychotic disorder, eating disorder, learning disorder, autistic spectrum disorder, or any other medical or physical conditions, affecting the central nervous system (e.g., multiple sclerosis, Parkinson’s disease, dementia) were included. Data from N = 35 MDD patients were collected as part of the Cognitive Function and Mood Study (CoFaM-Study; Baune and Air, 2016), and were tested within the Burghof Clinic (Rinteln, Germany). Additional data from N = 3 MDD patients were acquired in the AMEOS Clinical Center (Hildesheim, Germany). From the MDD group, a total of N = 4 patients had to be excluded due to poor quality of EEG recordings (e.g., lack of artifact-free intervals or less than 65% of critical segments remained after artifact correction), resulting in N = 34 MDD patients (20 females, 14 males, Mean_age = 41.62 years ±2.19 S.E., Age-range: 21 to 65 years), which remained for further analyses.

The comparison group contained a total of N = 28 adults without a diagnosis of MDD (all females, Mean_age = 21.29 years ±0.57 S.E., Age-range: 19 to 34 years). These participants were tested at the University of Hildesheim (Hildesheim, Germany).

Before signing an informed consent form, all participants were provided with a detailed explanation (written and in person) of study procedures and measures. The study was conducted according to the guidelines of the Declaration of Helsinki, and was approved by the Human Research Ethics Committee at the University of Adelaide (Approval Number: H-160-2011), the Royal Adelaide Hospital Human Research Ethics (Approval Number: 111230), and the ethics committee of the FB-1 (Educational and Social Sciences) at University of Hildesheim (Reference Number: 223/2022).

For the analyses of serum/plasma levels of biomarkers, samples of peripheral blood were collected always between 8:30 a.m. and 11:30 a.m. (within less than 3 min to avoid the initiation of pituitary stress responses; Vahl et al., 2005). Blood samples were centrifuged (1800–2,122 × g, 10–15 min) and frozen at least at −20° in a refrigerator (monitored by a central alert system 24 h/day and 7 days/week). Using an enzyme immune assay (EIA; Dedimed GmbH Europarc Lab, Kleinmachnow, Germany, MVZ DIAMEDIS, Sennestadt, Germany), BoDV-1-specific CICs, generally considered to reflect a positive (i.e., an activated) status of BoDV-1 infection, were detected (using a cut-off value of 0.1). Using flow cytometry and the BD Cytometric Bead Array (CBA) Human Inflammatory Cytokines Kit (Becton, Dickinson and Company BD Biosciences, San Jose, USA), lithium heparin plasma from MDD patients of the Burghof Clinic was analyzed according to concentrations of pro-inflammatory cytokines IL-6 and IL-8. The analysis was conducted by the Lab Division of Molecular Neurobiology of Mental Health (Münster, Germany). For participants tested in Hildesheim, concentrations of IL-6 and IL-8 were detected from their serum using a chemiluminescent immune assay (CLIA; using an IL-8 detection limit of 5.00 pg/ml and an IL-6 detection limit of 3.50 pg/ml). This analysis was conducted by the MVZ Laboratory Limbach (Lehrte, Germany).

Resting-state EEG activity (with eyes closed) was continuously recorded from passive (GVB-geliMED GmbH, Bad Segeberg, Germany; EEG recordings in Rinteln) or active AgCl electrodes (actiCAP; Brain Products GmbH, Gilching, Germany; EEG recordings in Hildesheim) using a NeuroPrax (NeuroConn, Ilmenau, Germany; EEG recordings in Rinteln) or BrainAmp DC amplifier (Brain Products GmbH, Gilching, Germany; EEG recordings in Hildesheim). Dependent on the amplifier, sampling rate was either 512 or 1,000 Hz (down-sampled to 512 Hz). Brain electrical activity was recorded from electrode positions Fp1, Fp2, F7, F3, Fz, F4, F8, T3, C3, Cz, C4, T4, T5, P3, Pz, P4, T6, O1, and O2 (EEG recordings in Hildesheim additionally employed FC5, FC1, FC2, FC6, CP5, CP1, CP2, CP6, and Oz). All electrode positions were defined according to the international 10–20 system. Electrode impedances were kept below 5 kΩ for passive electrodes, and below 20 kΩ for active electrodes. As reference, linked ears (A1, A2; EEG recordings in Rinteln), or FCz (re-referenced to mastoid positions TP9 and TP10; EEG recordings in Hildesheim) were used. AFz was used as ground. For EEG recordings conducted in Rinteln, no electrooculographic (EOG) activity was recorded, but EEG recordings in Hildesheim employed two additional EOG electrodes, which were placed below and near the outer canthus of the right eye. EEG resting-state recordings were conducted in a semi-darkened dim and quiet room. All participants were seated comfortably on a chair with neck support. EEG-activity was recorded for a total duration of 10 min (EEG recordings in Hildesheim) or 20 min (EEG recordings in Rinteln). All participants were informed to close their eyes and to remain relaxed, yet alert. During the recording, participants from the Burghof Clinic were additionally instructed to open their eyes for a short duration of 5 s after 3, 6, 15 and 18 min of recording, to hyperventilate after 10 min of recording (for a total duration of 2 min), and to relax afterwards. However, these sections of the EEG were not analyzed for this study.

All EEG raw data were analyzed offline using the BrainVision Analyzer 2.0 (Brain Products, Gilching, Germany). After down-sampling and re-referencing, EEG data were bandpass filtered using a low cut-off value (order 8) of 0.1 Hz (Mohammadi et al., 2015), and a high cut-off value (order 4) of 40 Hz (Neuner et al., 2014). In addition, a 50 Hz-Notch filter had been applied. Using manual raw data inspection, EEG data were visually inspected to mark artifacts and identify artifact-free intervals for the further analysis. Due to missing EOG channels in recordings from the Burghof Clinic, all ocular artifacts were detected using an Independent Component Analysis (ICA). Signal components, which were topographically associated to ocular activity, were subtracted from the EEG signal (Neuner et al., 2014). Afterwards, an artifact-free interval of at least 2 min (placed between 1 and 8 min of EEG recording) was selected for further analysis. EEG data from the chosen interval were segmented into 2-s epochs (50% overlap) and DC-and baseline-corrected, before artifacts with amplitudes over ±75μV (related to muscle tension, tremor, or drowsiness) were marked as bad intervals (Mohammadi et al., 2015). Participants with less than 65% of remaining artifact-free segments were excluded from further analysis. The mean proportion of segments remaining for the further analysis was 91.50% ± 1.05 S.E. (range: 65 to 99%). A Fast Fourier Transformation (FFT; symmetric Hanning window of 5%; according to Mohammadi et al., 2015) was employed to extract relative power values (i.e., absolute values per 0.5 Hz bin normalized with respect to the frequency range of 0.5 to 34 Hz). FFT-transformed segments (using a bin-width of 0.5 Hz) were averaged for each channel, before individual frequency band values were exported. In accordance to both the guideline of the International Federation of Clinical Neuropsychology (Nuwer et al., 1999) and Neuner et al. (2014), frequency bands were defined as followed: delta (0.5–3.5 Hz), theta (4–7 Hz), alpha (7.5–12 Hz), and beta (13–34 Hz). For detailed statistical analyses, mean relative power values were extracted for the alpha band at fronto-central (Fz), fronto-medial (F3 and F4), centro-central (Cz), centro-medial (C3 and C4), as well as for parieto-central (Pz) and parieto-medial (P3 and P4) electrode positions. In addition, a spatially averaged value across 19 electrodes, which were employed for all EEG recordings, was computed for each participant.

All statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS; Version 28 & 29 for Mac, IBM, New York, USA). Descriptive statistics were calculated and presented as mean and standard error (S.E.). Pearson’s Chi-Square (χ²) tests were used to examine categorical data. For comparative analyses, repeated measurement analysis of variances (ANOVAs) with post-hoc (Bonferroni-corrected) pairwise comparisons were employed. Where appropriate, Greenhouse–Geisser corrected p-values were applied. Pearson’s correlation coefficient r was used to determine possible correlations between variables. For all analyses, the significance level was set to p < 0.05.

Based on the analyses of blood samples, MDD patients and participants from the comparison group were divided into subgroups of BoDV-1 positive and BoDV-1 negative participants. In the clinical group, N = 14 MDD patients (41.2%; 9 females, 5 males, Mean_age = 40.14 years ±3.90 S.E., Age-range: 21 to 58 years, 12 participants on medication) were identified as BoDV-1 positive (indicating an activated status of BoDV-1 infection), whereas N = 20 patients (58.8%; 11 females, 9 males, Mean_age = 42.65 years ±2.59 S.E., Age-range: 23 to 65 years, 15 participants on medication) were tested as BoDV-1 negative (indicating no BoDV-1 infection or a non-activated status of BoDV-1 infection). In the group of adults without MDD diagnosis, N = 14 participants (50%; all females, Mean_age = 20.00 years ±0.30 S.E., Age-range: 19 to 23 years) were identified as BoDV-1 positive, whereas N = 14 participants (50%; all females, Mean_age = 22.57 years ±1.00 S.E., Age-range: 19 to 34 years) were tested as BoDV-1 negative. Most important, no significant association between the status of BoDV-1 infection (positive vs. negative) and the clinical diagnosis of MDD (yes vs. no) could be observed (p > 0.05). Regarding to sex, no differences could be observed within each study group for BoDV-1 positive vs. BoDV-1 negative subgroups (p > 0.05). However, comparisons between MDD patients vs. adults without MDD diagnosis revealed a significant sex difference for BoDV-1 positive subgroups [χ²(1) = 6.09, p < 0.05, Cramer’s ν = 0.47] and BoDV-1 negative subgroups [χ²(1) = 8.57, p < 0.01, Cramer’s ν = 0.50]. Regarding to age, between-groups comparisons revealed BoDV-1 positive participants without MDD diagnosis to differ from BoDV-1 negative participants without MDD diagnosis [t(26) = 2.47, p < 0.05, Cohen’s d = 0.93]. Such an age difference was not observed for clinical subgroups (p > 0.05). In addition, comparisons between MDD patients vs. adults without MDD diagnosis revealed a significant age difference for BoDV-1 positive subgroups [t(26) = 5.16, p < 0.001, Cohen’s d = 1.95] and BoDV-1 negative subgroups [t(32) = 6.24, p < 0.001, Cohen’s d = 2.18]. Finally, clinical subgroups did not differ with respect to their proportion of participants using psychotropic drugs (p > 0.05).

For all subgroups, the means of spatially averaged values of relative EEG power (including all EEG frequencies from 0.5 to 34 Hz) are summarized in Figure 1. As it was expected for recordings of resting-state EEG in a closed-eyes condition, a significant increase in power of the EEG alpha frequency band (7.5–12 Hz) was observed for all study groups.

To further test for relative power-differences in the EEG alpha band, a 2 × 2 × 2 × 3 repeated measurement ANOVA with the between-subjects factors “Group” (MDD patients vs. adults without MDD diagnosis) and “Status of BoDV-1 infection” (positive vs. negative), as well as the within-subjects factors “Hemisphere” (left vs. right hemispheric electrodes) and “Electrode position” (frontal, central, and parietal) was performed. The analysis revealed a significant main effect of “Electrode position” [F(1.40, 81.48) = 17.05, p < 0.001, η²_p = 0.23], as well as a significant interaction between the factors “Group” and “Electrode position” [F(1.40, 81.48) = 45.65, p < 0.001, η²_p = 0.44; see Figure 2]. All remaining main effects and interactions proved to be non-significant (p > 0.05). As this ANOVA did not reveal any hemisphere-specific effects, a 2 × 2 × 3 repeated measurement ANOVA was computed considering only midline “electrode positions” (Fz, Cz, and Pz) as a within-subjects factor, and “Group” (MDD patients vs. adults without MDD diagnosis) as well as “Status of BoDV-1 infection” (positive vs. negative) as between-subjects factors. This analysis confirmed a significant interaction between the factors “Group” and “Electrode position” [F(1.36, 79.08) = 39.10, p < 0.001, η²_p = 0.40; see Figure 2]. All remaining interactions and main effects proved to be non-significant (p > 0.05).

For MDD patients in comparison to adults without MDD diagnosis, post-hoc comparisons showed a significant increase in mean relative EEG alpha power at midline electrode positions Fz (Mean_MDD = 3.115 ± 0.241 S.E., Mean_{without MDD} = 2.155 ± 0.262 S.E., |Mean|_Diff = 0.959, p < 0.01), as well as a significant decrease at electrode position Pz (Mean_MDD = 2.380 ± 0.219 S.E., Mean_{without MDD} = 3.385 ± 0.238 S.E., |Mean|_Diff = 1.005, p < 0.01), but no mean relative EEG alpha power difference at Cz (Mean_MDD = 2.873 ± 0.228 S.E., Mean_{without MDD} = 2.588 ± 0.247 S.E., |Mean|_Diff = 0.285, p > 0.05). More precisely, adults without MDD diagnosis showed a significantly decreased mean relative EEG alpha power at electrode position Fz in comparison to Cz (|Mean|_Diff = 0.432, p < 0.001) and Pz (|Mean|_Diff = 1.230, p < 0.001), as well as a significant decrease at electrode position Cz compared to Pz (|Mean|_Diff = 0.797, p < 0.001). In contrast, MDD patients showed a reversed activation pattern with a significantly increased mean relative EEG alpha power at electrode position Fz in comparison to Cz (|Mean|_Diff = 0.242, p < 0.05) and Pz (|Mean|_Diff = 0.735, p < 0.001), as well as a significant increase at electrode position Cz compared to Pz (|Mean|_Diff = 0.493, p < 0.01).

To summarize, a posterior/parietal to anterior/frontal decrease in mean relative EEG alpha power was observed in adults without MDD diagnosis, whereas a reversed pattern with an increase in EEG alpha power from posterior/parietal to anterior/frontal electrode positions was observed in MDD patients. With respect to BoDV-1 status or hemispheric differences, no significant differences in EEG resting-state alpha activity could be observed.

Because of detection limits, IL-8 concentrations <5.00 pg/ml were coded as 4.99 pg/ml, and IL-6 values <3.50 pg/ml were coded as 3.49 pg/ml in all participants. To examine possible differences between concentrations of pro-inflammatory IL-6 or IL-8 with respect to MDD diagnosis and BoDV-1 status, two separate 2 × 2 ANOVAs with the factors “Group” (MDD patients vs. adults without MDD diagnosis) and “Status of BoDV-1 infection” (positive vs. negative) were conducted. For mean concentrations of IL-6, the analysis revealed no significant main effects, and no interaction (p > 0.05; for descriptive data see Table 1). However, for mean concentrations of IL-8, the analysis revealed a weak, but significant “Group” and “Status of BoDV-1 infection” interaction [F(1, 58) = 5.97, p < 0.05, η²_p = 0.09], with higher IL-8 concentrations in BoDV-1 negative MDD patients compared to BoDV-1 positive MDD patients (|Mean|_Diff = 3.34, p < 0.01) or BoDV-1 negative adults without MDD diagnosis (|Mean|_Diff = 3.43, p < 0.01; see Table 1). However, no difference was observed between BoDV-1 positive and negative participants within the subgroup of adults without MDD diagnosis (|Mean|_Diff = 1.15, p > 0.05), as well as between MDD patients and adults without MDD diagnosis within the subgroup of BoDV-1 positive participants (|Mean|_Diff = 1.06, p > 0.05).

Figure 3 illustrates the variability of mean relative fronto-central EEG alpha power in dependence on concentrations of IL-6 and IL-8 for the different subgroups of our study. For participants without MDD diagnosis, IL-6 concentrations were almost always observed below or near the detection limit (min to max = 3.49 to 4.86 pg/ml; see Table 1). With respect to IL-8 concentrations, subgroups of adults without MDD diagnosis showed IL-8 values to predominately vary between 4.99 and 10.00 pg/ml. In general, concentrations of pro-inflammatory cytokines in participants without MDD diagnosis were – irrespectively of their BoDV-1 status – accompanied by rather low values (<5%) and low variability in fronto-central EEG alpha power (see Figure 3 and Table 1).

The observed result pattern was quite different in MDD patients. Comparable to adults without MDD diagnosis, both subgroups of MDD patients revealed the majority of IL-6 concentrations to lie below or near the detection limit, but demonstrated a slightly higher variability in mean relative fronto-central EEG alpha power (see Figure 3 and Table 1). In BoDV-1 positive MDD patients, the majority of IL-8 concentrations were observed to vary – comparable to IL-8 concentrations in the comparison group – between 4.99 and 10.00 pg/ml, but mean relative fronto-central EEG alpha power showed a high variability of data with relative power values of up to 7% (see Figure 3 and Table 1). However, in the group of BoDV-1 negative MDD patients, participants showed an increased variability of IL-8 concentrations with values ranging from 4.99 to 21.20 pg/ml, showing a concomitant high variability in mean relative fronto-central EEG alpha power with values of up to 6% (Figure 3 and Table 1). Correlation analyses confirmed these observations and revealed a weak, but significant negative correlation between mean IL-8 concentrations and mean relative fronto-central EEG alpha power for BoDV-1 negative MDD patients (r = −0.46, p < 0.05), and no significant correlations for any of the other study groups (p > 0.05).

We observed a similar proportion of participants with a BoDV-1 positive and BoDV-1 negative status of infection (determined through BoDV-1 specific CICs). Differing sensitivities in detection methods impair the comparability of prevalence data reported in different studies (Frank et al., 2022). However, among all methods, BoDV-1 specific CICs, as they were determined in our study, belong to the most frequently used markers of BoDV-1 virus infection, and are considered to reasonably explain infection profiles of dormant and activated states (Bode et al., 2001; Liu et al., 2015). Contrary to previously published studies conducted before the pandemic coronavirus disease 2019 (COVID-19), our results were unable to confirm that MDD patients are overall more likely affected by BoDV-1 infection than healthy individuals without any psychiatric diagnosis (Dietrich et al., 1998; Ludwig and Bode, 2009; Mazaheri-Tehrani et al., 2014; Wang et al., 2014). Our study was conducted during and shortly after the COVID-19 pandemic (i.e., data of adults without MDD diagnosis were assessed shortly after the pandemic, whereas data of most MDD patients, with the exception of two individuals, were collected before the onset of the COVID-19 pandemic). A possible activation of the BoDV-1 infection by the pandemic might explain, why we observed rates of 41.2% for BoDV-1 positive MDD patients and of 50% for adults without MDD diagnosis, which are much higher than many previously reported estimations showing a rather low prevalence of BoDV-1 positive cases, especially for activated states (Ludwig and Bode, 2009; Niller et al., 2020). However, the observed high rate of CICs in BoDV-1 positive MDD patients of our study fits well to previous studies, reporting higher detection rates of BoDV-1 specific RNA transcripts in lymphocytes, increased serum antibody titers against BoDV-1 specific nucleoprotein (p40) and phosphoprotein (p24), and higher BoDV-1 specific CICs in MDD patients (Bode et al., 1995, 2001; Ikuta et al., 2002; Miranda et al., 2006; Sauder et al., 1996). For example, Mazaheri-Tehrani et al. (2014) investigated diagnosed psychiatric patients and reported a high infection (CIC) prevalence of 40.4%, which perfectly matched to our results in MDD patients. On the other hand, for healthy controls, Mazaheri-Tehrani et al. (2014) reported a prevalence of 29.5%, which was rather lower than what we observed for adults without MDD diagnosis. Liu et al. (2015) addressed the risk of BoDV-1 infection in health care settings in China, and found 21.8% of the hospital staff (with the highest prevalence among psychiatry and oncology personnel) to be CIC positive, which exceeded both the prevalence of 18.2% detected in MDD outpatients, and of 11.1% detected in healthy controls. Finally, Bode et al. (2020) reported a prevalence of 57.9% for BoDV-1 infected psychiatric patients in Germany, but again only of 24% for healthy carriers. Interestingly, these authors also reported a prevalence of 37.3% for healthy carriers in the Czech Republic. To summarize, previous studies confirmed a high BoDV-1 infection (CIC) prevalence in psychiatric patients that we also confirmed in our study. However, our study revealed a prevalence of BoDV-1 infection of 50% for healthy carriers, which is much higher than the prevalence rates between 10 and 40%, which were reported for healthy individuals in most of the previous studies. The high prevalence rate observed in our study might reflect an effect of infection with the severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) and of COVID-19 on the activation of BoDV-1 infections in healthy carriers. In addition, our observed prevalence rate in healthy participants argues against zoonotic reservoirs as important for the subclinical spread of the BoDV-1 virus (Bode et al., 2020, 2022; Cain and Ly, 2023; Lipkin et al., 2001). Epidemiological surveys are required to further clarify the role of COVID-19 on BoDV-1 infection in humans.

Previously, non-invasive EEG studies using ERPs suggest BoDV-1 infection to be associated with neurophysiological and behavioral changes (Dietrich, 2011; Dietrich and Bode, 2008; Dietrich et al., 1998, 2000, 2005, 2020; Gosztonyi, 2008; Gosztonyi et al., 2020; Marty et al., 2022; Zhang et al., 2022). However, in our study, focusing on closed-eyes EEG resting-state alpha activity, we were unable to observe any BoDV-1-dependent changes, which are probably more likely to occur in eyes-opened conditions, where the sensory systems actively process environmental stimuli. In our study, we decided to record EEG activity in an eyes-closed task-free condition to avoid any confounding effects of external stimuli, instructions, or task execution. In addition, resting state seems more self-relevant than standard cognitive tasks and drive subjects to direct their attention away from the (bottom-up) processing of the environment, thus permitting (top-down) processing of the internal mental context (Fingelkurts and Fingelkurts, 2015). The EEG alpha oscillation is commonly thought to arise primarily from the occipital and parietal parts of the cortex (where usually the highest alpha power can be observed), and reflects primarily cortical dendritic activity synchronized across a large part of the visual cortex (Kellaway, 2003; Liu et al., 2012; Tatum et al., 2006; Zou et al., 2009). Furthermore, the thalamus and thalamocortical neurons have been proposed to play an active role in generating and modulating posterior alpha rhythmic activity (Liu et al., 2012). Simultaneously recorded functional magnetic resonance imaging (fMRI) and EEG measurements conducted during rest revealed a negative correlation between posterior alpha-EEG and pulvinar BOLD-fMRI, and a positive correlation between posterior alpha-EEG and anterior as well as medial dorsal (thalamic) nuclei, which appear functionally connected to the cingulate cortex (Liu et al., 2012), forming a cingulo-insular-thalamic network, comprising the dorsal anterior cingulate cortex, anterior insula, anterior prefrontal cortex, and the thalamus (Sadaghiani et al., 2010). This network is proposed (I) to mediate cortical arousal (Saper et al., 2005), (II) to maintain tonic internalized alertness (Coste and Kleinschmidt, 2016; Sadaghiani et al., 2010), (III) to correlate with processes of introspection, internal mentation, or integration of stimulus-independent thought processes (Bowman et al., 2017), and (IV) to correlate with fluctuations in involuntary attention (Dosenbach et al., 2008; Portas et al., 1998). Liu et al. (2012) speculated that while the pulvinar may be intimately involved in the generation of the alpha rhythm in the occipital cortex, such rhythmic activities may be under central modulatory control (Steriade, 1993) effectuated through the anterior and medial dorsal parts of the thalamus. Thus, an increase in posterior alpha (PA) power during rest suggests an (probably primarily pulvinar-mediated) inhibition of external awareness of the environment and a suppression of potentially distracting information (Foxe and Snyder, 2011), whereas a lower PA power may indicate a higher level of cortical arousal and outward vigilance, which may be primarily under central modulatory control of anterior and medial dorsal nuclei (Hopman et al., 2020; Kirschfeld, 2005; Liu et al., 2012). In line with this notion, simultaneously recorded fMRI and EEG measurements conducted during rest showed PA power to be positively correlated with activity in neural regions associated with introspection and rumination (Bowman et al., 2017), and negatively correlated with regions associated with vigilance and visual processing (Laufs et al., 2003). Directing visuospatial attention to a given visual field consistently results in a power suppression of alpha activity over parieto-occipital areas contralateral to the attended visual field (Arana et al., 2022; Händel et al., 2011), showing that modulations of EEG alpha activity correlate with the direction of visuospatial attention. A reduction in the EEG alpha power over parieto-occipital areas had also been reported in children with Attention Deficit Hyperactivity Disorder (ADHD) and frontocortical hypoactivity (Bozhilova et al., 2022; Lenartowicz et al., 2018).

As it was expected from previous studies, we observed a posterior to anterior decrease in EEG alpha power, and – compared to MDD patients – a higher PA power in healthy adults without MDD diagnosis. These data are indicative of a predominance of processes of stimulus-independent and spontaneous self-referential or introspective thought processes in healthy participants in the eyes-closed resting state condition (Bowman et al., 2017). Compared to healthy participants, PA power was significantly reduced in MDD patients, while EEG resting-state alpha activity became significantly increased at anterior/frontal electrode sites (i.e., a reversed topographic activity pattern, compared to adults without MDD diagnosis). This decrease in PA power and increase in anterior alpha (AA) power (representing a frontal hypoactivation; Baxter, 1989; Jaworska et al., 2012) might be indicative of an increased (bottom-up driven) external awareness and outward vigilance in MDD patients during rest, accompanied by a decrease in interoceptive awareness and (top-down driven) impairments to regulate and integrate self-referential processes (Liu et al., 2012; Veer et al., 2010). In line with this notion, decreases in resting-state functional connectivity of the anterior insula (and the related network) have been reported for medication-free MDD patients without comorbidity (Veer et al., 2010).

Beside from modulations in PA power, several clinical studies reported an increase in the asymmetry of AA power (indicative of a left anterior hypoactivation) in acute depressed or remitted psychiatric patients, which was associated with an underactivation of the behavioral approach system, and a bias towards negative-valenced stimuli and an activation of the withdrawal-related behavioral system (Coan and Allen, 2004; Debener et al., 2000; Henriques and Davidson, 1991; Jaworska et al., 2012; Knott et al., 2001; Olbrich and Arns, 2013; Pollock and Schneider, 1990). However, our results were not able to confirm any hemispheric asymmetries in resting-state alpha activity, either for frontal, central or parietal electrode positions.

With regard to the analysis of pro-inflammatory cytokines IL-6 and IL-8, our data are in line with overall inconclusive findings regarding concentrations and effects of pro-inflammatory cytokines in depression (Himmerich et al., 2019; Köhler et al., 2017; Lamers et al., 2020; Tsai, 2021). Contrary to previous findings (Dowlati et al., 2010; Himmerich et al., 2019; Howren et al., 2009; Köhler et al., 2017), our results do not support the assumption of an immune activation in either individuals with or without MDD diagnosis. Most important, the observed concentrations of IL-6 and IL-8 in BoDV-1 positive participants were highly comparable to those observed in BoDV-1 negative adults without MDD diagnosis. Compared to adults without MDD diagnosis, MDD patients failed to demonstrate elevated levels of IL-6, which were previously reported to be associated with the severity of depression (Miller et al., 2009). In our sample (compared to healthy participants), IL-8 was only slightly elevated in BoDV-1 negative, but not in BoDV-1 positive MDD patients. Previous studies (e.g., Hannestad et al., 2011; Iwata et al., 2013) showed that levels of proinflammatory cytokines can be downregulated by psychotropic drugs (e.g., antidepressants), but in our study, BoDV-1 negative and positive MDD patients received comparable medical treatment. Furthermore, serum levels of proinflammatory cytokines could be influenced by demographic factors such as age (Himmerich et al., 2019; Janelidze et al., 2015) and sex (Birur et al., 2017; Kruse et al., 2021). However, in our study, BoDV-1 positive and negative MDD patients did neither differ in age nor sex. One explanation for slightly elevated IL-8 concentrations in BoDV-1 negative, but not in BoDV-1 positive MDD patients might be, that an activation of BoDV-1 in positive MDD patients may bypass the activation of the innate immune system and the release of pro-inflammatory and anti-inflammatory cytokines (Jordan and Lipkin, 2001; Reuter et al., 2010; Zhai et al., 2013). Previous studies have shown that IL-8 might affect neurons and their excitability including the modulation of neurotransmitter release or spontaneous synaptic activity (Giovannelli et al., 1998; Puma et al., 2001). Although the observed negative correlation between IL-8 levels and fronto-central EEG resting-state alpha activity is consistent with the notion that IL-8 may exert neuromodulatory effects (Rostène et al., 2007; Stuart et al., 2015), this finding still needs to be interpreted with caution, and has to be confirmed in future, more systemic studies. These studies should employ larger sample sizes and a combination of non-invasive EEG and invasive measurements, including consideration of a wide range of immunological parameters, endocrine variables, and bioenergetic changes in the intracellular network dynamics of neuron’s mitochondria (Hitzler et al., 2019; Karabatsiakis and Schönfeldt-Lecuona, 2020). The use of invasive methods highlights the overall organic character of both BoDV-1 infections and MDD-related somatic vulnerability, which physiologically affects more than neurons in the brain. For example, BoDV-1-specific changes in the EEG of MDD patients depend on complex changes in inflammatory and hormonal processes that may alter the mitochondrial adenosine triphosphate (ATP) production, which is the main driver of cellular activity, including neuronal activity (Karabatsiakis et al., 2014; Boeck et al., 2018). Invasive measurements may facilitate the identification of potentially important correlations with biological variables outside the central nervous system (CNS), which help to elucidate the precise biological mechanisms leading to observable changes in non-invasive EEG measurements.