Resting-state functional connectivity alterations after mild COVID-19 infection Authors: Anna Szente, Ákos Arató, Husamalddin Ali Alhour, Gábor Perlaki, Gergely Orsi, Szilvia Anett Nagy, Eszter Áfra, Barnabás Dudás, Dalma Tényi, Péter Barsi, Norbert Kovács, József Janszky, Gergely Darnai Published: 2025 ABSTRACT AbstractThe long-term effects of coronavirus disease 2019 (COVID-19) are more pronounced after severe infection, although years after the pandemic was declared, long-term neurological complications after mild infection have begun to unfold. The most common complaints include fatigue and cognitive disturbances. Brain imaging techniques have also revealed structural and functional alterations. In the present study, we included 40 subjects who recovered after mild COVID infection and 36 healthy control subjects. We administered mental health–related questionnaires assessing fatigue and sleep disturbances, performed neuropsychological tests and analysed resting-state functional connectivity using multivariate pattern analysis. Our results indicate no cognitive dysfunction in the post-infection group, but alterations in functional connectivity manifested as mixed hyper- and hypoconnectivity compared to the control group. Altered resting-state functional connectivity might be an adaptive neural response to structural alterations, as seen in viral infections in general. Our results suggest that despite the lack of any neurological complaints or neurocognitive signs, the presence of increased and decreased functional connectivity might be a part of natural recovery from mild COVID-19 infection or the sign of ongoing post-COVID neurological sequele.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-025-12273-8.Keywords: COVID-19, fMRI, MVPA, Resting-state, Resting-state functional connectivity Introduction Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) caused a sudden, rapid increase in coronavirus disease 2019 (COVID-19), which eventually led the World Health Organization (WHO) to declare the COVID-19 pandemic in March 2020, causing millions of deaths [1]. While individual differences occur, most patients present initial respiratory symptoms, such as cough or in more severe cases, shortness of breath, accompanied by headache, chills and fever [2]. Although all age groups are affected, severe cases mostly appear in comorbid, elderly patients [3], and children seemed to be spared from critical conditions [4]. At the beginning of the COVID-19 pandemic, the involvement of the nervous system was not considered to be a common phenomenon: only severe cases presented altered consciousness, encephalopathy, ischaemic stroke and encephalitis [5, 6]. Four years since the declaration of the COVID-19 pandemic, it has become clear that the risk of neurological complications is noticeable even 12 months after the infection. A wide range of symptoms can develop in nonhospitalised individuals too: anosmia or dysgeusia [7], cognitive and memory disorders (“brain fog” [8]), sensory alterations affecting younger people, and severe, chronic conditions, such as cerebrovascular or musculoskeletal diseases [9]. Based on the common residual symptoms, the definition of long-COVID was introduced: symptoms 3 months after the infection lasting for at least 2 months without any alternative explanation [10]. The background of long-COVID remains unclear. A possible explanation might be persistent brain tissue damage, but because mild, nonhospitalised cases may develop long-lasting symptoms, unresolved inflammation could also be the culprit [11, 12]. According to a study of 223 recovered individuals, white matter alterations were present without cognitive deficits, which also points to chronic neuroinflammation [13]. With severe cases of COVID-19, distinctive patterns of parenchymal lesions have been described based on magnetic resonance imaging (MRI). According to the abovementioned study, the neurobiological background remains unclear, but an immunological explanation might be assumed, because only one patient had the virus in his cerebrospinal fluid (CSF) [14]. In mild cases, olfactory bulb involvement and cranial nerve alterations might be more characteristic [15]. Structural brain imaging has been employed to explore decreased cortical thickness even after mild COVID-19 infection [16]. Regarding functional MRI measurements, alterations in functional connectivity (FC) affecting olfactory networks have been reported [17, 18]; however, long-term measurements of nonhospitalised, mild cases are often overlooked. Resting-state functional connectivity (RSFC) can be interpreted as the connection between spatially distributed brain regions involved in the same function [19], and it provides valuable insights, biomarkers or guidance for treatment in various clinical conditions [20]. It can be conducted in 10 to 15 min, and although taken at rest, can be used to indicate functional connections during tasks as well [21]. In this study, we aimed to explore RSFC differences between individuals recovered after mild COVID-19 infection and a control group by using multivariate pattern analysis (MVPA). MVPA is a sensitive method to analyse FC patterns; contrary to univariate analyses, it explores FC globally and retrieves data from the activity patterns [22]. We also conducted neuropsychological tests and administered mental health–related questionnaires to compare cognitive measurements between the two groups. Methods Participants Participants were recruited via a Google Form shared on social media platforms. A telephone interview was conducted to inform applicants about the research and the possible contraindications of MRI (for more detailed information, see Supplementary File 1.). Seventy-six participants met the inclusion criteria, aged between 18 and 40 years. There were 40 subjects in the post-COVID group, whose initial infection took place in the first and second waves of the pandemic, and 36 in the control group. All subjects in the post-COVID group presented a positive polymerase chain reaction (PCR) or fast antigen test result. The additional inclusion criteria were mild symptoms (no need of hospitalisation, and no pneumonia or dyspnoea) and no acute or chronic health problems at the time of the MRI scan. According to self-reports, the control subjects had not had COVID-19 or showed COVID-19 or flu-like symptoms since the beginning of the pandemic. All participants were right-handed and reported no history of neurological or psychiatric diseases. Neither the post-COVID, nor the control group received any treatment or supplements during our study. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethical Committee of the National Public Health Center (6843-5/2021/EÜIG). All subjects were informed and gave written informed consent. Questionnaires A list of the most common symptoms during the infection was used to determine mild COVID-19-related issues. The intensity of each symptom was described on a scale from 1 (‘none’) to 10 (‘worst’) [23]. The Fatigue Impact Scale (FIS) was used to examine the presence of fatigue symptoms in daily activities. It consists of 40 items divided into three subscales: the cognitive, physical and psychosocial domains. Each item is answered on a scale from 0 (‘no problem’) to 4 (‘an extreme problem’) [24]. Complications related to the loss of smell were explored using the Questionnaire of Olfactory Disorders Negative Statements (QOD-NS). Here, the modified 17-item version was used. Each item refers to olfactory function–related negative experiences, with scores ranging from 0 (‘disagree’) to 3 (‘agree’) [25]. The Athens Insomnia Scale (AIS) was used to evaluate sleep quality. This self-reported questionnaire consists of eight items, which are scored from 0 (‘no problem at all’) to 3 (‘very serious problem’), to explore sleep-quality, well-being and daytime sleepiness [26]. The presence of anxiety and depression was measured by Spielberger State-Trait Anxiety Scale [27] and the Beck Depression Inventory [28]. Handedness was assessed using the Edinburgh Handedness Inventory [29]. Additionally, the elapsed time between the first positive COVID-19 test and the MRI measurement was specified. Neuropsychological tests The following tests were used for neuropsychological assessment: the Rey–Osterrieth Complex Figure Test is used to explore visual perception and memory by copying and recalling a certain figure [30]. Digit Span (DS) investigates the limits of working memory [31]. Verbal learning was measured using the Rey Auditory Verbal Learning Test (AVLT) [32], and visuospatial working memory was assessed with the Corsi Block Tapping Test (CBT) [33]. Completing all these tests, provided enough time for the Rey-Osterrieth Complex Figure Test recall as well [34]. Data acquisition MRI examinations were performed on a 3T MRI scanner (Magnetom PrismaFit Siemens Healthcare, Erlangen, Germany), with a 64-channel head/neck coil at the Diagnostic Centre of Pécs. During the RS fMRI measurements, the participants were instructed to keep their eyes open and to focus on a white cross presented at the centre of the screen on a black background using MRI-compatible LCD screen (BOLDscreen 24 LCD for fMRI, Cambridge Research Systems Ltd., Rochester, United Kingdom). Functional images were acquired with the following parameters: repetition time (TR)/echo time (TE) = 3000/30 ms; flip angle (FA) = 83°; field of view (FOV) = 210 × 210 mm2; 70 × 70 matrix; 44 axial slices with a thickness of 3 mm; 2040 Hz/pixel receiver bandwidth. A sagittal three-dimensional T1-weighted magnetisation-prepared rapid gradient-echo (MPRAGE) sequence was obtained for registration procedure for each subject. The parameters were: TR/TI/TE = 2530/1100/3.37 ms; FA = 7°; 176 slices; slice thickness = 1 mm; FOV = 256 × 256 mm2; matrix size = 256 × 256, 200 Hz/pixel receiver bandwidth. Functional magnetic resonance imaging (fMRI) data preprocessing SPM’s CONN Toolbox (version 22.a), a MATLAB-based platform, was used to perform preprocessing steps, denoising and FC analysis [35, 36]. The preprocessing steps included realignment, unwarp (with fieldmaps for susceptibility distortion correction), slice-timing correction and normalisation. Functional data were smoothed with a full-width half maximum (FWHM) of 8 mm. Acquisitions with framewise displacement above 0.9 mm or global blood-oxygenation-dependent (BOLD) signal changes above 5 standard deviations were flagged as outliers using Artificial Detection Tools (ART). During denoising, these potential outlier scans were regressed out (scrubbing). The intention was to exclude subjects if their number of volumes flagged as outliers divided by the total number of volumes was greater than 25%; however, this did not occur in our analysis. Moreover, the following confounding effects were also regressed out from the BOLD signal: five CompCor noise components from the white matter timeseries, five CompCor noise components from the CSF timeseries, motion parameters and their first order derivatives (12 factors), session effects and their first order derivatives (2 factors), and linear trends (2 factors). Lastly, bandpass frequency filtering of the BOLD timeseries between 0.008 and 0.09 Hz and visual check of the outputs were carried out [36]. MVPA MVPA was used to estimate whole-brain RSFC. This data-driven approach computes the FC map between each individual seed voxel and the rest of the brain and is capable of decoding the complex neuronal background of conditions [37] in a diverse range of topics such as fMRI patterns in pain perception [38], FC in depression [39] or electroencephalogram (EEG) patterns [40]. With this approach, it was possible to determine whether the FC pattern for the whole brain was different in subjects after mild COVID-19 infection compared with the control group. Because the connection between each voxel and the rest of the brain is analysed, to discover maximal intersubject heterogeneity, dimensionality reduction is needed, which was 64 for this study (the default setting). During group-level analysis, the number of eigenpattern scores (which characterises FC of each voxel) was defined a priori with a ratio of 1:10 (the number of resulting components: the number of subjects), which resulted in seven components. This ratio seems to be sensitive enough to detect large effects of FC, but it is also capable of finding marginal effects [41]. The resulting MVPA components showed between-subject variability of FC patterns between the two groups (post-COVID vs. control) that were used as seeds in the subsequent seed-to-voxel connectivity analysis [41]. To identify potential seeds, statistical analyses were performed on each of the 7 components using a general linear model (GLM), where the post-COVID > control between-subject contrast was used as regressor of interest, and age and sex were included as covariates. The results were thresholded using a combination of a cluster-forming p < 0.001 voxel-level threshold, and a familywise error–corrected p < 0.05 cluster-size threshold. Seed-to-voxel analysis Using the MVPA seeds, in the next step, seed-to-voxel maps were estimated, characterizing the spatial pattern of FC with each of the seed areas. FC strength was estimated separately for each seed area and target voxel, modelling the association between their BOLD signal timeseries. Group-level analyses were performed using a GLM, where the post-COVID > control contrast was used as regressor of interest, while age and sex were included as covariates again. Thus, SBC analysis showed group differences of brain regions with increased or decreased FC between the MVPA seed and the rest of the brain while controlling for the effects of age and sex. The results were thresholded using a combination of a cluster-forming p < 0.001 voxel-level threshold (except for two seeds with robust results, for which a threshold of p < 0.0001 was used), and a familywise error–corrected p < 0.05 cluster-size threshold. The clusters that consisted of fewer than 100 voxels were excluded. Statistical analysis Questionnaire scores and neuropsychological test results were compared between the two groups using the Mann–Whitney U-test. Regarding group-level SBC results of the post-COVID group, FC values for clusters showing differences between post-COVID and control group were exported to SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA) for further analysis to discover the possible effect of the questionnaire scores and neuropsychological tests, using Spearman correlation analysis. Demographic characteristics of the two groups were compared by Fisher’s exact test (sex distribution) and the Mann–Whitney U-test (age). Results Demographic characteristics, questionnaire scores and neuropsychological tests There was no significant difference in age and sex between the post-COVID and the control group (Table 1), while Table 2 shows the intensity of experienced symptoms during acute infection for the COVID-19 group. Table 1Demographic characteristicsGroupp-valuePost-COVID groupControl groupSexFemalen = 26Malen = 14Femalen = 22Malen = 140.813aAgec24 (22–27)26 (26-32.25)26 (24–28)26 (24–27)0.942ba Fisher’s exact testb Mann-Whitney U-testc Values represent median and interquartile range Demographic characteristics Female n = 26 Male n = 14 Female n = 22 Male n = 14 a Fisher’s exact test b Mann-Whitney U-test c Values represent median and interquartile range Table 2Intensity of commonly experienced symptoms during the infection in the post-COVID groupCOVID symptomMedian (interquartile range)Febrile sense4 (1–6)Chill4 (1-6.25)Cough4 (2-5.25)Sputum2 (1–3)Sore throat2 (1-4.25)Runny nose2 (1–5)Nasal stuffiness3 (1-5.25)Loss of smell10 (8–10)Chest pain2 (1-4.25)Chest discomfort2 (1–4)Shortness of breath1 (1–3)Loss of taste8 (2.5–10)Loss of appetite4 (1–7)Epigastric soreness1 (1–2)Nausea1 (1–1)Vomiting1 (1–1)Abdominal pain1 (1–1)Diarrhea1 (1–2)Constipation1 (1–1)Headache6 (4.75-8)Muscle ache5 (3–7)Arthralgia3 (1–7)General weakness7 (5–10) Intensity of commonly experienced symptoms during the infection in the post-COVID group Table 3 shows the comparison between the post-COVID and control groups in the mental health–related questionnaire scores and neuropsychological test results. There was no significant difference in cognitive performance nor in the questionnaire scores except for the QOD-NS. Table 3Comparison of questionnaire and neuropsychological test scores between post-COVID and control groupsQuestionnairepost-COVID group (n = 40)median (interquartile range)Control group (n = 36)median (interquartile range)Mann-whitney U-test p-valuesSTAI-T41 (35.5–45)40 (35–46)0.708STAI-S37 (33–42)35 (33-40.5)0.243BDI7 (3–13)4 (1–10)0.140FIS Cognitive7 (3.75-14)8 (4-12.25)0.790FIS Physical5 (2-8.5)7 (2–10)0.782FIS Social8 (3–17)9 (6-14.25)0.688FIS Total18 (11-38.25)24 (12-37.5)0.712QOD-NS6 (3-12.25)1 (12-37.5)< 0.001*AIS10 (2–7)4 (2-5.25)0.206Digit-Span forward7 (6–8)7 (6–7)0.552Digit-Span backward6 (5–6)5 (5–6)0.230Corsi forward6 (5–7)6 (6–7)0.267Corsi backward6 (6-6.25)6 (5–7)0.777AVLT total learning58 (51–64)60 (51-63.25)0.647Rey copy36 (36–36)36 (36–36)0.079Rey memory25 (22–29)25 (22–29)0.896*denotes significant results Comparison of questionnaire and neuropsychological test scores between post-COVID and control groups *denotes significant results The median number of days between the onset of a positive COVID-19 test and MRI measurement was 202 days (interquartile range [IQR] = 117.75–257.25 days, min = 79, max = 607. Study enrolment period ran from January 2021 and April 2022. All our participants in the post-COVID group were infected in the first and second epidemic waves of COVID-19. Results of the MVPA analysis Table 4; Fig. 1 shows that MVPA analysis uncovered six significant seeds showing differences in RSFC between the post-COVID and control groups. Table 4Seeds of MVPA analysisMVPA seedsPeak MNI coordinatesAnatomical regionVoxels per clusterpFWEDirection of correlationMVPA seed a-24, 78, -18Left fusiform gyrus, left lingual gyrus1990.0001post-COVID > ControlMVPA seed b-56, -46, -16Left inferior temporal gyrus, left middle temporal gyrus750.013Control > post-COVIDMVPA seed c6, 26, 30Right paracingulate gyrus910.000737Control > post-COVIDMVPA seed d30, 12, 12Right insular cortex, right central opercular cortex670.006286Control > post-COVIDMVPA seed e-18, 30, 50Left superior frontal gyrus1120.000024post-COVID > ControlMVPA seed f52, -36, 26Right supramarginal gyrus430.031458post-COVID > ControlpFWE refers to family-wise error corrected p-value Seeds of MVPA analysis pFWE refers to family-wise error corrected p-value Seeds of MVPA analysis. Color bars indicate Z-scores (red-yellow indicate post-COVID > Control while blue-light blue indicate Control > post-COVID). Axial slices are shown in radiological convention Results of the seed-to-voxel analysis Seed-to-voxel connectivity analysis revealed that four out of the six MVPA seeds (a, b, c and d) showed significant FC differences with certain brain regions (Table 5; Fig. 2) between the groups. There was no correlation between the FC values and the questionnaire scores and neuropsychological test results. Table 5Seed-to-voxel analysisMVPA seedTargetPeak MNI coordinatesCluster sizepFWEDirectionMVPA seed aBilateral caudate nucleus, left nucleus accumbens, bilateral thalamus4, 0, 8472< 0.000001post-COVID > ControlLeft middle temporal gyrus, left superior temporal gyrus-60, -24, -62630.000047post-COVID > ControlLeft angular gyrus and left lateral occipital cortex-48, -58, 502010.000283post-COVID > ControlLeft frontal pole and left superior frontal gyrus-12, 42, 361760.000615post-COVID > ControlRight frontal pole and right superior frontal gyrus12, 26, 421570.001139post-COVID > ControlLeft central opercular and left insular cortex-36, 2, 141320.002662Control > post-COVIDMVPA seed bLeft lingual gyrus-14, -70, -123020.003666post-COVID > ControlRight insular cortex, right temporal pole36, 8, -42630.007673post-COVID > ControlAnterior cingular gyrus, bilateral paracingular gyrus-6, 8, 422050.02459post-COVID > ControlAnterior cingular gyrus4, -12, 421750.04659post-COVID > ControlMVPA seed cLeft superior frontal gyrus, left precentral gyrus-18, -6, 502430.01410post-COVID > ControlLeft superior parietal gyrus, left lateral occipital cortex-20, -64, 682080.02791post-COVID > ControlMVPA seed dLeft lingual gyrus, left occipital fusiform gyrus-14, -76, -102460.000218Control > post-COVIDRight lingual gyrus, right temporal occipital fusiform cortex22, -46, -101320.001973Control > post-COVIDLeft occipital pole and left lateral occipital cortex-18, -88, 21300.002124Control > post-COVID Seed-to-voxel analysis Seed-to-voxel analysis. Color bars indicate Z-scores (red-yellow indicate post-COVID > Control, while blue-light blue indicate Control > post-COVID). Axial slices are shown in radiological convention Discussion In 2020, the WHO declared the COVID-19 pandemic, and nervous system involvement has been shown to unfold during the acute phase of COVID-19 infection as well as chronically after infection [42]. The exact pathomechanism is still unknown, but it has been hypothesised to involve an increase in pro-inflammatory cytokines, a retrograde neuronal pathway through the olfactory bulb or direct invasion of the nerves [43]. According to a recent study, blood–brain barrier disruption and inflammation might be behind cognitive dysfunction (brain fog) in patients with long COVID [44]. Our main findings indicate several alterations in resting-state functional connectivity. We used MVPA to analyse the pattern of RSFC in the whole brain and our results indicated the presence of hyper- and hypoconnectivity after COVID-19 infection. These findings are novel given that a previous study comparing severe, moderate and mild cases of COVID-19 found these kinds of changes only after severe infection [45]. There were similar results reported when using MVPA in COVID-19 survivors with subjective cognitive complaints [46], although based on our results, FC alterations can be present without cognitive symptoms. Viral infections other than COVID-19 can also be associated with FC alterations: patients with human immunodeficiency virus (HIV) infection who are on antiretroviral therapy presented increased and decreased FC without neuropsychological symptoms [47]. Extensive hyperconnectivity was found in the subacute phase after Zika virus infection, although those participants presented neurological symptoms [48]. These results indicate a similar neural response to viral infections. Enhanced FC could be interpreted as a compensatory mechanism to preserve essential functions after brain injuries or structural brain impairments [49]. Therefore, one of the possible explanations might be that the lack of neuropsychological deterioration is the result of the hyperconnectivity patterns found in the post-COVID group. The arrangement of hyperconnectivity patterns might indicate the involvement of higher-level organisation networks: specific central and subcortical, interconnected brain regions integrate information between other networks, playing a role in global FC [50]. Based on our analysis, several central and subcortical structures are affected, including the superior frontal, superior parietal and lingual gyri, and the thalamus. The involvement of these regions has been suggested to affect neural communication on a global level [51]. We identified hypoconnectivity in several brain networks, which is expected after neurological injury, mixed with enhanced connectivity. This pattern might depend on the stage of progression, including decreased FC during the subacute phase of infection. Hence, coexistent hyper- and hypoconnectivity might be a part of a natural recovery [52], as it was found in other conditions. This pattern was observed in chronic cases of inflammatory conditions, such as Systemic Lupus Erythematosus (SLE). SLE patients with and without neuropsychiatric symptoms presented altered FC connectivity, and a possible explanation could be the response to neuronal damage, and thus, compensatory amendments [53]. As recent study suggested that RSFC might be a biomarker in optimizing medication in Alzheimer’s Disease [54], this might be another aspect of interpreting altered FC results. In this study, we examined individuals after mild COVID-19 infection with questionnaires as well and compared the results with a control group. Given that previous studies have shown cognitive impairments even after mild COVID-19 infection [55], we expected to find deficits in neuropsychological tests; however, there was no difference in performance between the two groups. The main reason behind the lack of significant results might be the use of a neuropsychological test battery in our study. Here, we used neuropsychological tests designed to assess clinical populations and validated on neurological cohorts; therefore, these tools are less sensitive to minor impairments. On the other hand, Dressing et al. [56] found only minor impairments on cognitive testing. Moreover, there might be continuous improvement regarding cognitive functions after COVID-19 infection, and thus cognitive symptoms in the past cannot be excluded [57]. We also hypothesised that after mild COVID-19 infection, individuals experience sleep difficulties and fatigue. However, our findings do not support this hypothesis. This might be due to our sample characteristics: our sample involved mostly university students, and education seems to be a protective factor [58]. Not surprisingly, there was a significant deterioration in the quality of life associated with olfactory impairments in the post-COVID group, as denoted by the QOD-NS. Although olfactory disorders also improve over time, in a small percentage of individuals, these symptoms persist. Despite the exact mechanism is also unclear, olfactory symptoms might predict future neurodegenerative conditions [59]. However, we found no cognitive deterioration associated with olfactory symptoms. All our participants in the post-COVID group were infected during the first and second epidemic waves, which are associated with the ‘wild-type’ strain and emerging variants, such as the ‘alpha’ variant of COVID-19 [60]. There are several limitations to our study. Although the recruitment took place during the first and second waves of the COVID-19 pandemic, asymptomatic, but previously infected individuals could be involved in the control group. The post-COVID group provided a positive PCR test, but the control group included participants who only stated that they had not experienced any previous COVID-19 symptoms. This could be avoided using blood tests to exclude any previous infections. Second, a longitudinal follow-up measurement could provide additional information to interpret the FC alterations, to evaluate whether the reported FC alterations are still present. Besides that, our results apply to the initial variants of COVID-19, other variants have emerged since then. Comparing different variants from different epidemic waves, would widen our knowledge on viral infections. The median days between the MRI measurement and the COVID-19 infection was 202 days; an earlier assessment of RSFC might yield different results, therefore we are not able to present data on the acute phase. Based on the neuropsychological results, cognitive deterioration after COVID-19 infection cannot be excluded completely, because more sensitive tests might show different scores. Besides that, our battery could have included a wider range of neuropsychological tests (e.g. AVLT Delayed Recall). Lastly, it is difficult to compare the results of FC among studies due to the heterogeneity of the methods [49]. Conclusion Although numerous researchers have investigated the long-term effects of COVID-19 infection, its background is still unclear, especially in cases after mild infection. In this study, we did not find any difference in mental health–related questionnaires nor neuropsychological tests, but we found the mixed presence of hyper- and hypoconnectivity based on RSFC. Although previous studies investigated functional connections after COVID-19, in our case, MVPA allowed us to evaluate whole-brain RSFC in a data-driven way [41]. Therefore, this approach might be capable of shedding light on a global postinfectious condition. RS brain network function might be indicative of the neurological prognosis, even in the presence of severe structural lesions [61]. Based on our results, altered FC might be part of physiological recovery. Although these conclusions should be interpreted with caution, taking the abovementioned limitations into account, they offer directions for further studies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Author contributions ASz. wrote the main manuscript text. GP, GO, SZAN and GD designed the MRI measurements. ÁA, HAA, EÁ and BD conducted the MRI measurements. ASz, ÁA, HAA and GD analysed the data. ASz, GD, GP, GO and JJ contributed to the interpretation of data. JJ, NK, DT and PB revised the manuscript. Funding TKP2021-EGA-13 has been implemented with the support provided from the National Research, Development and Innovation Fund of Hungary, financed under the TKP2021-EGA funding scheme. Project no. TKP2021-EGA-16 has been implemented with the support provided from the National Research, Development and Innovation Fund of Hungary, financed under the TKP2021-EGA funding scheme. Project no. RRF-2.3.1-21-2022-00011 has been implemented with the support from the National Laboratory of Translational Neuroscience, Adult Nervous System Disorders. J.J. was supported by National Research, Development and Innovation Fund of Hungary (NKFIH-K142479). SAN was supported by the János Bolyai Research Scholarship of the Hungarian Academy of Sciences, EKÖP-24-4-II-PTE-250 University Research Fellowship Program of the National Research, Development and Innovation Office. Data availability Data supporting these findings are available from the corresponding author on reasonable request. Declarations Human ethics and consent to participate The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethical Committee of the National Public Health Center (6843-5/2021/EÜIG). All subjects were informed and gave written informed consent. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement Data supporting these findings are available from the corresponding author on reasonable request.