TYPE Original Research PUBLISHED 15 February 2023 DOI 10.3389/fnhum.2023.1068216 The effects of white matter OPEN ACCESS hyperintensities on MEG power spectra in population with mild EDITED BY Magali González-Colaço Harmand, University Hospital of the Nuestra Señora de Candelaria, Spain REVIEWED BY cognitive impairment Roberto Nuevo, Rey Juan Carlos University, Spain Yoshihito Shigihara, Lucia Torres-Simon 1,2 *, Pablo Cuesta 1,3,4 , Hokuto Hospital, Japan Alberto del Cerro-Leon 1,2 , Brenda Chino 1,5 , Lucia H. Orozco 1 , *CORRESPONDENCE Elisabeth B. Marsh 6 , Pedro Gil 4,7 and Fernando Maestu 1,2,4 Lucia Torres-Simon 1 lucia.torres@ucm.es Center of Cognitive and Computational Neuroscience, Universidad Complutense de Madrid (UCM), Madrid, Spain, 2 Department of Experimental Psychology, Cognitive Processes and Speech Therapy, Universidad SPECIALTY SECTION Complutense de Madrid (UCM), Madrid, Spain, 3 Department of Radiology, Rehabilitation, and Physiotherapy, This article was submitted to Universidad Complutense de Madrid (UCM), Madrid, Spain, 4 Instituto de investigación Sanitaria San Carlos Brain Health and Clinical Neuroscience, (IdISSC), Madrid, Spain, 5 Institute of Neuroscience, Autonomous University of Barcelona (UAB), Barcelona, a section of the journal Spain, 6 Department of Neurology, The Johns Hopkins School of Medicine, Baltimore, MD, United States, Frontiers in Human Neuroscience 7 Department of Geriatric Medicine, Hospital Universitario San Carlos, Madrid, Spain RECEIVED 12 October 2022 ACCEPTED 23 January 2023 PUBLISHED 15 February 2023 Cerebrovascular disease is responsible for up to 20% of cases of dementia CITATION worldwide, but also it is a major comorbid contributor to the progression of other Torres-Simon L, Cuesta P, del Cerro-Leon A, neurodegenerative diseases, like Alzheimer’s disease. White matter hyperintensities Chino B, Orozco LH, Marsh EB, Gil P and Maestu F (2023) The effects of white matter (WMH) are the most prevalent imaging marker in cerebrovascular disease. The hyperintensities on MEG power spectra in presence and progression of WMH in the brain have been associated with general population with mild cognitive impairment. Front. Hum. Neurosci. 17:1068216. cognitive impairment and the risk to develop all types of dementia. The aim of doi: 10.3389/fnhum.2023.1068216 this piece of work is the assessment of brain functional differences in an MCI COPYRIGHT population based on the WMH volume. One-hundred and twenty-nine individuals © 2023 Torres-Simon, Cuesta, del Cerro-Leon, with mild cognitive impairment (MCI) underwent a neuropsychological evaluation, Chino, Orozco, Marsh, Gil and Maestu. This is an open-access article distributed under the MRI assessment (T1 and Flair), and MEG recordings (5 min of eyes closed resting terms of the Creative Commons Attribution state). Those participants were further classified into vascular MCI (vMCI; n = 61, License (CC BY). The use, distribution or mean age 75 ± 4 years, 35 females) or non-vascular MCI (nvMCI; n = 56, mean reproduction in other forums is permitted, provided the original author(s) and the age 72 ± 5 years, 36 females) according to their WMH total volume, assessed with copyright owner(s) are credited and that the an automatic detection toolbox, LST (SPM12). We used a completely data-driven original publication in this journal is cited, in accordance with accepted academic practice. approach to evaluate the differences in the power spectra between the groups. No use, distribution or reproduction is Interestingly, three clusters emerged: One cluster with widespread larger theta power permitted which does not comply with these and two clusters located in both temporal regions with smaller beta power for vMCI terms. compared to nvMCI. Those power signatures were also associated with cognitive performance and hippocampal volume. Early identification and classification of dementia pathogenesis is a crucially important goal for the search for more effective management approaches. These findings could help to understand and try to palliate the contribution of WMH to particular symptoms in mixed dementia progress. KEYWORDS cerebrovascular disease (CVD), small vessel cerebral disease, white matter hyperintensites, mild cognitive impairment, magnetoencephalography (MEG), power spectra analysis Frontiers in Human Neuroscience 01 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 Introduction electroencephalography (EEG) and magnetoencephalography (MEG) are included. While EEG is less expensive and more popular in a Due to high life expectancy, aged-related pathologies like clinical context, MEG provides a better disposition for the estimation dementia are rising. Alzheimer’s Disease (AD) is the most common of neural sources, a better signal to noise ratio for higher frequency subtype of dementia; however Vascular Dementia (VaD) or major bands and it is a more patient-friendly technique, and indeed it has vascular cognitive impairment (VCI) is responsible for up to 20% also already come into clinical use in some hospitals for the clinical of cases worldwide (Catindig et al., 2012; Rizzi et al., 2014; O’Brien practice (Hoshi et al., 2022). and Thomas, 2015; Kalaria, 2018). Besides pure VaD, cerebrovascular Furthermore, the study of electrophysiological spectral disease (CBVD) is also major comorbid contributor to the progression signatures, measured with MEG resting state recordings, has of other neurodegenerative diseases, observed to some extent in been well-established for early detection and prognosis in almost all patients with dementia or MCI (Khan et al., 2016), for neurodegenerative disorders (López-Sanz et al., 2019), especially example in 50%–90% of the patients diagnosed with AD (Santos for Alzheimer’s disease (López-Sanz et al., 2018; Nakamura et al., et al., 2017), and up to 50% of dementias worldwide (Wardlaw et al., 2018) and mild cognitive impairment (MCI; López-Sanz et al., 2019). Furthermore, the presence of cerebrovascular neuropathology 2016). According to the cerebrovascular factors they can also be in post-mortem studies has been related to an increment in the found in studies analyzing MEG spectral features for patients with risk of developing dementia even more than the evidence of AD stroke (Tecchio et al., 2006; Chu et al., 2015; Kielar et al., 2016; neuropathology; but concomitant pathologies skyrocket the risk Johnston et al., 2022), which is one of the major causes of major VCI, compared to people with no brain alterations, or even with evidence of even using ultrasonography measures to assess carotid blood flow exclusively Alzheimer or cerebrovascular-type lesions (Azarpazhooh (Matsumoto et al., 2021). Nevertheless, to the best of our knowledge, et al., 2018). no article has been published performing spectral analysis with MEG CBVD involves a spectrum of medical conditions or changes in older people diagnosed with mild or major VCI, and only a few in the cerebral and/or systemic vasculature with brain impact, can be found with EEG, as described in a recent systematic review including a huge number of pathologies and etiologies. Within performed by our research group (Torres-Simón et al., 2022). these vascular pathologies, cerebral small vessel disease (CSVD) In the present study, we assess MEG electrophysiological spectral is one of the most common accompaniments of aging, not only signatures that differentiate between participants clinically diagnosed in dementia patients but also in healthy populations. Specifically, with MCI, according to their cognitive profile, with and without white matter hyperintensities (WMH), seen as diffuse areas of evidence of WMH of vascular origin, resulting in two groups: vascular high signal intensity (hence, “hyperintense”) on T2-weighted or MCI (vMCI) and non-vascular MCI (nvMCI). Based on the results fluid-attenuated inversion recovery sequences (Alber et al., 2019), previously described for EEG studies, we hypothesize that patients are the most prevalent imaging marker on CSVD, accounting for with vMCI will display a spectral profile compatible with more 15% in the general population greater than 64 and more than pronounced slowness of brain activity (power increase of slow bands 90% in those older than the age of 80 (Drebette and Markus, and decrease of higher frequency bands), particularly in middle 2010; Kloppenborg et al., 2014). Although age seems to be closely temporal regions, where WMH predominantly appears in early stages related to WMH presence in the brain, the amount and progression of the disease. have been associated with a higher risk of stroke, dementia, and cognitive dysfunction (Drebette and Markus, 2010; Kloppenborg et al., 2014). Specifically, WMH has been associated with small Methods but robust effects on general cognitive impairment, attention, and executive functioning in patients with MCI and dementia (Mortamais The sample for the present study was recruited under the et al., 2014; Prins and Scheltens, 2015; Arvanitakis et al., 2016; Van framework of two national projects (PSI2009-14415-C03-01 and Den Berg et al., 2018; Lam et al., 2021). In this context, the high PSI2012-38375-C03-01) focused on research and early detection incidence and contribution of CBVD to cognitive impairment, added of dementia. Data were collected between 2010 and 2014 in to their treatable conditions, provide special importance to the early collaboration with three different clinical centers located in Madrid differentiation and prevention of the pathophysiological origin of (Spain): the Neurology Department in “Hospital Universitario cognitive decline in the old population (Livingston et al., 2017). In Clinico San Carlos,” the Center for Prevention of Cognitive addition, CBVD could help to rise more accurate prognosis and Impairment, and the Seniors Center of Chamartín District. The management of new therapeutic targets, with the final goal of slowing sample consisted in 129 participants who were native Spanish down the dementia progression. speakers diagnosed with mild cognitive impairment (MCI) according Electrophysiologic neuroimaging techniques are of special to the criteria established by Petersen (2004). Patients with interest for this early differentiation, as they bring useful information MCI showed cognitive impairments but did not fulfill the for assessing brain function and network dynamics revealing early criteria for a dementia diagnosis. To assess general cognitive and changes inaccessible to standard structural imaging techniques functional status, the following set of screening questionnaires or even to cognitive assessment. Neurophysiologic techniques, in were used: The Mini Mental State Examination (MMSE; Lobo contrast to the other functional neuroimaging techniques, provide et al., 1979), the Global Deterioration Scale (Reisberg et al., 1982), information about brain oscillations with a wide frequency range the Geriatric Depression Scale–Short Form (GDS; Yesavage et al., from delta to gamma, are capable to measure the neural activation 1982); the Functional Assessment Questionnaire (FAQ; Pfeffer directly instead of doing it by means of indirect measures such as et al., 1982), and the questionnaire for Instrumental Activities blood flow or metabolism and allow repeated measurements without of Daily Living (Lawton and Brody, 1969). Participants were any risk for the patients. Within the electrophysiological techniques, excluded if they had: (1) a history of psychiatric or neurological Frontiers in Human Neuroscience 02 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 disease; and (2) psychoactive drugs consumption or chronic use of quantity distribution among CI individuals aged 50–86 years (average medication, such as anxiolytics. In addition, alternative etiologies 67 ± 8). The obtained cut point corresponded to a WMH volume of cognitive decline were ruled out, such as B12 deficiency, poorly of 3,448 mm3 . controlled diabetes mellitus, thyroid problems, syphilis, or human immunodeficiency virus. All participants signed an informed consent. This study was MEG data acquisition approved by the “Hospital Universitario Clinico San Carlos” ethics committee and was performed in accordance with approved Five minutes of resting-state (closed-opened eyes) were recorded guidelines and regulations. with a Vectorview system (Elekta Neuromag) at the Center for Biomedical Technology (Madrid, Spain). All participants sat comfortably in a chair with their eyes closed. The arousal level Neuropsychological assessment of each subject was monitored with a video camera and checked via a conversation immediately following the measurement session. All patients underwent an additional neuropsychological MEG data were collected at a sampling frequency of 1,000 Hz and assessment to generate detailed cognitive profiles including Direct online band-pass filtered between 0.1 and 330 Hz. Each subject’s and Inverse Digit Span Test (Wechsler Memory Scale, WMS-III; head shape was defined relative to three anatomical locations (nasion Wechsler, 1997), Immediate and Delayed Recall (WMS-III; Wechsler, and bilateral preauricular points) using a 3D digitizer (Fastrak, 1997), and Phonemic and Semantic Fluency (Controlled Oral Word Polhemus, VT, USA) and the head motion was tracked through Association Test, COWAT; Benton and Hamsher, 1989). four head-position indicator (HPI) coils attached to the scalp. These HPI coils continuously monitored the participants’ head movements, while eye movements were monitored by a vertical Magnetic resonance recordings and analysis electrooculogram assembly (EOG) composed of a pair of bipolar electrodes. Raw recording data were first introduced to Maxfilter T1-weighted and T2-weighted 3D FLAIR MRI were obtained software (v 2.2, correlation threshold = 0.9, time window = 10 s) using a General Electric 1.5 Tesla magnetic resonance scanner, with to remove external noise using the temporal extension of the signal a high-resolution antenna and a homogenization PURE filter (Fast space separation method with movement compensation (Taulu and Spoiled Gradient Echo sequence). The following parameters were Simola, 2006). Then, magnetometers data (Garcés et al., 2017) used for T1-weighted imaging: repetition time (TR) = 11.2 ms, were automatically examined to detect ocular, muscle, and jump echo time (TE) = 4.2 ms, inversion time (TI) = 450 ms, Field Of artifacts using Fieldtrip software (Oostenveld et al., 2011), which View (FOV) = 25 cm, flip angle (FA) = 12◦ , 252 coronal slices (in- were visually confirmed by an MEG expert (initials). Twelve plain resolution: 256 × 256), voxel size: 0.98 × 0.98 × 1 mm3 and participants were dismissed due to bad quality MEG data, resulting acquisition time '8:00 min. T2-weighted 3D FLAIR images in a sample of 117 participants. The remaining artifact-free data were obtained with the following specifications: TR = 7,000 ms, was sectioned into 4-s segments. Independent component analysis- TE = 101 ms, TI = 2,112 ms, FOV = 24 cm, 252 coronal slices (in- based procedure (ICA) was applied to remove heart magnetic field plain resolution: 256 × 112), voxel size: 0.94 × 0.94 × 1.6 mm3 and artifacts and EOG components. Only those recordings with at acquisition time '4:57 min. least 20 clean segments (80 s of brain activity) were utilized in WMH segmentation (T1 and Flair) and lesion volume calculation subsequent analyses. were performed using an automatic computerized toolbox in SPM12, MEG clean time series were band-pass filtered (2 s padding) called Lesion Segmentation Tool (LST). We used the initial κ threshold between 2 and 45 Hz. Source reconstruction was carried out using a with default value of 0.3 as recommended by the developers (Schmidt regular grid of 1 cm spacing in the Montreal Neurological Institute et al., 2012), and subsequently the probabilistic threshold of 0.5, (MNI) template. The resulting model comprised 2,459 sources as it was the most accurate compared with the clinician’s manual homogeneously distributed across the brain. This model was segmentations. linearly transformed to each subject’s space. The leadfield was calculated using a single shell model (Nolte, 2003). Source time series were reconstructed using a Linearly Constrained WMH volume cut point Minimum Variance beamformer (Van Veen et al., 1997). The power spectrum of each grid’s node was computed by means As explained before in the introduction, a minimum load of a multitaper method (mtmfft) with discrete prolate spheroidal WMH is present in practically every individual above 60 years sequences (dpss) as windowing function and 1 Hz smoothing. old. Nevertheless, the amount and progression of this vascular For each node, relative power was calculated by normalizing by marker has been closely related with a higher risk of stroke and total power over the 2- to 30-Hz range. The source template dementia and also with cognitive dysfunction. In order to establish with 2,459 nodes in a 10 mm spacing grid was segmented a possible threshold from which WMH volume in the brain could into 78 regions of the Automated Anatomical Labeling atlas have clinical implications we have calculated a specificity-based cut (AAL, Tzourio-Mazoyer et al., 2002), excluding the cerebellum, point for the WMH damage quantity (Jack et al., 2017). For this basal ganglia, thalamus, and olfactory cortices. Those 78 regions purpose, we used the WMH segmentations of a preexisting dataset of interest included 1,202 of the original 2,459 nodes. Trials of 346 cognitively intact (CI) older individuals without evidence were averaged across participants ending up with a source- of vascular disease, according to the radiological report. This cut reconstructed power matrix of 1,202 nodes × 113 frequency point corresponded to the 95th percentile of the WMH damage steps × 117 participants. Frontiers in Human Neuroscience 03 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 Statistical procedure WMH volumes below the cut point (nvMCI group). Both groups were compared to assess possible differences in all demographic The assessment of significance between group power differences scores displayed in Table 1. Patients with vMCI showed significantly was addressed by relying on the cluster-based permutation test higher age (p < 0.001), Total gray matter volume (GM volume; (CBPT; Maris and Oostenveld, 2007; Zalesky et al., 2010). Power p = 0.014) and WMH volume (p ≪ 0.001). No significant differences differences were tested between groups for each pair of nodes between groups arise when cognitive performance and hippocampal using an ANCOVA test while adjusting for the effects of age and volumes were assessed. This last result is important since cognition total white matter volume. The resulting matrix of F-statistics (with and hippocampal atrophy are common markers of dementia status. the same dimension of original power matrix) was binarized by Descriptive data and statistical scores are depicted in Table 1. thresholding the matrix using a critical value computed with a p-value of 0.005. This binary matrix was split into two matrices attending for the sign of the differences between groups. Clusters Electrophysiological power differences were built by grouping spatially adjacent nodes that systematically between groups showed significant between-groups differences. All nodes within a cluster must exhibit the same sign of between-groups differences, The power spectra were used to determine the brain regions and indicating that the cluster was likely a functional unit. Only clusters frequency ranges with a significant difference between vMCI and involving at least 1% of the nodes (i.e., a minimum of 12 nodes) were nvMCI groups. The results showed three significant clusters. considered. Cluster-mass statistics were assessed through the sum of The first significant result (named θ cluster, CBPT p value = 0.036) all F-values across all nodes. To control for multiple comparisons, the emerged as a widespread cluster (see Figure 1A), whose brain entire analysis pipeline was repeated 10,000 times after shuffling the oscillatory activity in the frequency range [4.75–8.00 Hz] (Figure 1C) original group’s labels. At each repetition, the maximum statistic of differed between groups. The power of this cluster, that fell within the surrogate clusters was kept, creating a maximal null distribution the classical theta band range, was enhanced in the vMCI group that would ensure control of the family-wise error rate (FWER) at the when compared to the nvMCI group (see Figure 1B). The cluster size cluster level. Cluster-mass statistics on each cluster in the original data ranged between 20 nodes up to 300 nodes reaching the maximum size set was compared using the same measure in the randomized data. at the frequency of 7 Hz (see Figure 1D). The average F value, across The CBPT p-value represented the proportion of the permutation all nodes of the cluster, is depicted in Figure 1E. A detailed description distribution with cluster-mass statistic values greater or equal to the of the ROIs involved in the θ cluster can be found in Supplementary cluster-mass statistic value of the original data. Only those clusters Table 1. that survived the CBPT at p < 0.05 or below were considered for Besides from the θ cluster, two significant results showing a the subsequent analyses as potential “MEG markers.” As descriptive diminished power in the vMCI group when compared to the values for each significant cluster, we computed the average power individuals included in nvMCI, were found within the beta frequency (across all nodes that belong to the cluster). These values were used as band (Figures 2B and 3B). MEG marker values for the subsequent Spearman correlation analysis The first beta cluster (henceforth called β1, CBPT p with measures of cognitive performance and structural integrity. In value = 0.0336) involved left temporo-parietal regions of the addition, we computed between groups statistics for these averaged brain (see Figure 2A). It emerged in the frequency range [12.25 Hz to power values using ANCOVA with age and total white matter volume 19.50 Hz] (Figure 2C). The size of the cluster ranged between 20 and as covariates. Statistical analyses were carried out using Matlab 127 nodes, peaking at 16 Hz (Figure 2D). The maximum average F R2021b (Mathworks Inc). value, across all nodes of the cluster, was found at 15 Hz (Figure 2E). A detailed description of the ROIs involved in the β1 cluster can be found in Supplementary Table 2. Data availability The second beta cluster (henceforth called β2, CBPT p value = 0.0401) involved right temporo-parietal regions of the brain The data that support the findings of this study are available from (see Figure 3A). The β2 cluster contralaterally mirrored the result the corresponding author, upon request. All the algorithms used in of the β1 cluster since it showed a decreased beta power in the the present article are reported in the “Methods” section. patients with vMCI when compared to patients without vascular damage in the right hemisphere of the brain (Figure 3B). It was defined in the frequency range [11.75 Hz to 19.50 Hz] (Figure 3C). Results The size of the cluster fluctuated between 15 and 105 nodes, peaking at 15 Hz (Figure 3D). The maximum average F value, across all Segregation of patients with MCI based on nodes of the cluster, was found at 14 Hz (Figure 3E). A detailed their WMH volume description of the ROIs involved in the β2 cluster can be found in Supplementary Table 3. The aim of this piece of work is the assessment of brain functional differences in an MCI population based on the WMH volume, and their possible correlation with cognitive performance. Correlations between electrophysiological The original 117 patients with MCI were divided into two groups markers and brain health (VMCI vs. nVMCI) using their WMH volumes and the 3,448 mm3 cut point. Sixty-one patients diagnosed with MCI showed WMH In order to help the interpretation of the results, we carried volumes above the cut point (hence vMCI group) whereas 56 had out Spearman correlation test between the average power of each Frontiers in Human Neuroscience 04 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 TABLE 1 Participants’ demographics. vMCI (61) nvMCI (56) p-value F value η2 Sex (females) 35 36 0.46 Age 75.26 ± 4.19 72.13 ± 5.11 <0.001* 13.28* 0.104 Education (years) 9.08 ± 4.67 8.44 ± 4.14 0.21 1.60 0.026 MMSE 26.15 ± 2.68 27.09 ± 2.48 0.92 0.01 <0.001 Immediate Recall 15.13 ± 8.07 16.70 ± 8.92 0.89 0.02 <0.001 Delayed Recall 5.23 ± 6.41 6.42 ± 6.80 0.87 0.03 <0.001 Dígit Spam Forward 6.95 ± 1.74 6.54 ± 1.66 0.09 3.00 0.048 Dígit Spam Backward 4.03 ± 1.37 4.20 ± 1.54 0.55 0.36 0.006 Phonemic Fluency 8.28 ± 4.25 8.40 ± 3.77 0.54 0.38 0.006 Semantic Fluency 11.77 ± 2.94 12.48 ± 3.96 0.55 0.37 0.006 GDS 3.89 ± 2.89 4.45 ± 3.50 0.12 2.50 0.041 Apoe-ε4 27 20 0.45 WMH volume (10 ± 6) · 103 (1 ± 1) · 103 ≪0.001 60.17 0.505 WMH number of lesions 23 ± 13 12 ± 8 0.06 3.70 0.059 GM volume (531 ± 51) · 103 (527 ± 48) · 103 0.014 6.36 0.097 WM volume (389 ± 50) · 103 (395 ± 57) · 103 0.53* 0.40* 0.003 LHV 0.0022 ± 0.0003 0.0023 ± 0.0004 0.87 0.03 <0.001 RHV 0.0022 ± 0.0003 0.0024 ± 0.0004 0.80 0.07 0.001 Values are presented as mean ± SD. MMSE, mini-mental state examination; GDS, geriatric depression scale; WMH volume, white matter hyperintensities total volume; WMH number of lesions, white matter hyperintensities number of lesions; GM volume, total gray matter volume (mm3 ); WM volume, total white matter volume (mm3 ); L/RHV, left/right hippocampal volume. Medial temporal lobe volumes were normalized with respect to the overall intracranial volume to account for differences in head volume among participants. p-values correspond with ANCOVA with age and WM volume as covariates, t-test (for age and WM volume, marked with *), and Fisher (sex and ApoE-ε4) tests- F values account for ANCOVA analysis. For age and WM volume the values correspond to t values. η2 accounts for partial-Eta squared for all comparisons but for the case of age and WM volume l where the values correspond to η2 values. cluster and scores of brain health (neurophysiological assessment predisposition, or cognitive impairment, as they were controlled for and structural quantitative scores associated with gray matter our analysis. atrophy). The significant results are depicted in Table 2, and To the best of our knowledge, this is the first study using MEG the scores included in the analysis are those described in spectral analysis to compare MCI populations with and without Table 1. cerebrovascular damage, specifically assessing the presence of WMH. As it can be seen in Table 2, all three power markers correlated The lack of literature regarding electrophysiological spectral patterns significantly with neuropsychological performance and hippocampal measured with MEG related to VCI makes it impossible to directly integrity. Importantly, whilst the average power of the θ cluster compare our results with previous literature, but we can confirm that correlated negatively with cognitive and structural integrities, both our findings are in line with those described in EEG studies (Torres- β clusters showed a positive correlation. This indicates that brain Simón et al., 2022). It is worth mentioning that in many of these EEG health seems to be associated with decreased θ power and increased studies, the inclusion criteria for the VCI population are vague or β power. This association emerged clearly when assessing correlations broad. In fact, in most cases, it is not possible to differentiate between between all power markers and hippocampal volumes, delayed recall VCI subtypes or to assess the degree of comorbidity in MCI or and MMSE performance. These results emphasize the relevance dementia patients with mixed pathology. Few articles described clear of the MEG markers for early differentiation since both MCI inclusion and exclusion criteria regarding MRI standards, specifically groups did not show statistical differences in cognition nor in for WMH. Additionally, in most of these studies, structural vascular hippocampal volumes. damage or white matter damage of vascular origin was assessed using internationally validated clinical visual scales. These scales are the most commonly used method to detect and evaluate the Discussion severity of WMH, being the Fazekas scale (Fazekas et al., 1987), and the age-related white matter changes scale (ARWMC; Wahlund In this study, we compared the spectral profiles during resting et al., 2001) the most recommended visual rating (Wahlund et al., state, assessed with MEG, for two MCI populations differentiated 2017). Nevertheless, these scales do not report true quantitative according to their level of cerebrovascular damage, described as the information (i.e., lesion volume), which makes the evaluation of the total volume of white matter hyperintensities in the brain. As we progression difficult when the changes are not abrupt. In this context, hypothesized, our results demonstrated the emergence of increased a clinician’s manual segmentation is the most accurate method to power for slow bands, specifically a widespread theta band (5–9 Hz), measure lesion volume in white matter and it is considered as the gold and smaller power was seen for high-frequency bands, in particular standard (Commowick et al., 2018). Notwithstanding, segmenting in the beta band (12–20 Hz) in bilateral temporal regions. Importantly the WMH manually is extremely time-consuming, and it comprises these power spectral changes were independent of age, gender, genetic high inter-rater variability making its application in clinical and Frontiers in Human Neuroscience 05 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 FIGURE 1 Significant result found within the theta frequency range. (A) Description of the regions involved in the θ cluster. The cluster was found to be significant between 4.75 Hz and 8.00 Hz. In the figure can be seen the initial, final, and maximum extensions of the cluster. (B) Violin plots and boxplots graphics describing the individual values for the average power of the cluster in the significant frequency range. (C) Representation of the average spectral power across all significant nodes. The significant frequency region is marked with dashes lines. (D) Number of grid nodes that are part of the cluster at each frequency step (maximum extension was found at 7 Hz). (E) Minimum, maximum, and average F values across all nodes contained within the cluster at each frequency step. scientific practice unfeasible. For this reason, automatic segmentation et al., 2007; van Straaten et al., 2012); but interestingly, this increment tools are increasing the presence in this research field and allow was even greater for patients with vMCI than for patients with for direct comparison of WMH between groups. Due to its great nvMCI (Babiloni et al., 2004; Moretti et al., 2004, 2007; Gawel et al., balance between accuracy and time dedication, we decided to include 2009; Sheorajpanday et al., 2013; Quandt et al., 2020). Furthermore, this automatic analysis to classify as objectively as possible our this theta band power increment has been closely linked to general patients in the two experimental groups according to their level of cognitive impairment, assessed with MMSE (Moretti et al., 2007). We cerebrovascular damage for later comparison. are pleased to confirm this correlation in our sample. Additionally, we found a similar negative correlation between the power of the theta cluster and delayed recall scores. This result helps to confirm that the increased power found could be considered the reflection in the brain Increment in theta power function of poorer cognitive functioning, although non-significant, derived from the cerebrovascular damage in our patients, which could An increase in theta power during the resting state for vMCI help to the prognosis of worse memory decline in the future. compared to nvMCI has been broadly reported. Previous literature While the association between increased theta power, age, reports that the presence and severity of cerebrovascular damage is and cognitive impairment (Schomer and Lopes da Silva, 2012) is related to a higher increment of power in the theta band. Specifically, well established, our data suggest that evidence of cerebrovascular it has been reported that patients with mild cognitive impairment pathologies (i.e., WMH) exacerbates this relationship. The sharper and evidence of vascular damage experienced an augmentation in increment in slow frequency bands found for vMCI compared theta band power compared to healthy age-matched controls (Gawel to nvMCI with similar severity of cognitive impairment could be Frontiers in Human Neuroscience 06 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 FIGURE 2 Significant result found within the beta frequency range in the left temporo-parietal region. (A) Description of the regions involved in the β1 cluster. The cluster was found to be significant between 12.25 Hz and 19.50 Hz. In the figure can be seen the initial, final, and maximum extensions of the cluster. (B) Violin plots and boxplots graphics describing the individual values for the average power of the cluster in the significant frequency range. (C) Representation of the average spectral power across all significant nodes. The significant frequency region is marked with dashes lines. (D) Number of grid nodes that are part of the cluster at each frequency step (maximum extension was found at 16 Hz). (E) Minimum, maximum, and average F values across all nodes contained within the cluster at each frequency step. explained by pathological changes in the brain driven by WMH. hippocampal volumes. This correlation seems especially important to The referred hyperintense lesions are predominantly located in the explain the possible association between cognitive impairments with deep white matter of the brain and are presumably related to small periventricular WMH, which has been previously hypothesized to vessel disease, in which little ischemic events increase white matter be reflecting the disruption of cholinergic projections from the basal injuries triggering axonal damage and even diffuse demyelination forebrain to the cortex (Alber et al., 2019). (Jang et al., 2017; Venkat et al., 2017; Kalaria, 2018); increase in blood-brain barrier permeability, contributing to neurodegeneration, apoptosis, and functional disruption (Farrall and Wardlaw, 2009; Schreiber et al., 2013); dysregulation of neurotransmitter systems, Decrement in beta power such as the cholinergic system (Caruso et al., 2019); and alterations of cerebral blood flow and chronic hypoperfusion (Yang et al., 2002; The positive relationship between a reduction in beta band Tak et al., 2011). Brain white matter tracks are able to connect power in resting state and general cognitive functioning (MMSE, cortical and subcortical gray matter across a relatively large distance, delayed and immediate memory performance, semantic fluency, and as temporal, and frontal lobes whose interaction is tightly related attentional spam) has been previously reported. Specifically, prior to theta wave generation (Anderson et al., 2010). Consequently, studies show smaller beta (β) band power in VCI patients compared an increase in widespread theta could be explained as a cortical with HC, once again associated with greater evidence of cognitive oscillatory activity reflection of deep white matter injury. Going symptomatology (van Straaten et al., 2012; Al-Qazzaz et al., 2017). a step further in this hypothesis, in the present study, we found Unfortunately, there were no comparisons made to patients with that the theta band cluster that emerged was negatively associated nvMCI or AD. Additionally, previous studies obtained differences in with bilateral hippocampal volume. The higher WMH total volume, alpha band for vMCI compared with HC associated with increased the larger the theta band power in the cluster and the lower the severity of cerebrovascular damage and cognitive impairment. They Frontiers in Human Neuroscience 07 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 FIGURE 3 Significant result found within the beta frequency range in the right temporo-parietal region. (A) Description of the regions involved in the β2 cluster. The cluster was found to be significant between 11.75 Hz and 19.50 Hz. In the figure can be seen the initial, final, and maximum extensions of the cluster. (B) Violin plots and boxplots graphics describing the individual values for the average power of the cluster in the significant frequency range. (C) Representation of the average spectral power across all significant nodes. The significant frequency region is marked with dashes lines. (D) Number of grid nodes that are part of the cluster at each frequency step (maximum extension was found at 15 Hz). (E) Minimum, maximum, and average F values across all nodes contained within the cluster at each frequency step. TABLE 2 Significant correlation between power markers and brain health. θ cluster β1 cluster β2 cluster rho p val rho p val rho p val MMSE −0.306 0.001 0.269 0.004 0.288 0.002 Dígit Spam Backward 0.184 0.050 Immediate Recall 0.207 0.027 Delayed Recall −0.314 0.001 0.303 0.001 0.250 0.008 Semantic Fluency 0.188 0.049 LHV −0.261 0.005 0.225 0.016 0.210 0.024 RHV −0.282 0.002 0.246 0.008 0.281 0.002 Spearman correlation analyses between the average power of each corresponding cluster and brain health in the whole sample. All correlations were computed with age and total white matter volume as covariates. MMSE, mini-mental state examination; L/RHV, left/right hippocampal volume normalized by total intracranial volume. Phonological fluency was not included in the table because it did not significantly correlate with any MEG marker. observed decrement in the low-alpha (Babiloni et al., 2004; Moretti 2015). All these results are in line with our main hypothesis that et al., 2007; Sheorajpanday et al., 2013; Wu et al., 2014) and stated that a more pronounced slowing of spectral brain activity high-alpha (Moretti et al., 2004; Al-Qazzaz et al., 2017) power and could be related to the presence and progression of brain damage a slower peak alpha frequency (Moretti et al., 2004; Neto et al., of vascular origin. The heterogeneity in the alpha band definition Frontiers in Human Neuroscience 08 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 and naming could be distorting these results, meaning that similar EEG results evidence that electrophysiology signatures are not ready underlying electrophysiological phenomena have been described as yet to be included in the diagnosis criteria for cognitive impairment different findings. of vascular origin because they are not robust, repeatable, and reliable In contrast to the widespread presentation of the theta power enough to be used as clinical biomarkers. MEG quantitative analysis alteration, the decrement in beta is explicitly localized in bilateral has been previously stated as a precise, non-invasive tool with high temporal regions. We are pleased to see that our findings go in line temporal resolution that is able to reflect changes in the bioelectrical with the well-stated idea that slower oscillatory powers are primarily activity of the brain. This brings us the opportunity to study produced by the thalamo-cortical network, while faster oscillatory brain functioning disruption due to changes in synaptic potentials powers depend on local activities (Klimesch, 1999) Additionally, produced by vasculature alterations before structural changes and/or beta band functioning alteration has been widely reported in studies cognitive decline are evidenced. Nevertheless, there is a critical need focused on patients with ischemic strokes (Kulasingham et al., to accurately classify VCI in future electrophysiological research in 2022; Pusil et al., 2022). Interestingly, the clusters that emerged in order to facilitate scientific results sharing and aggregation. CBVD the present study for lower beta band power overlap the deepest underlying cognitive impairment of vascular origin involves a variety brain regions surrounding periventricular areas where WMH most of medical conditions, pathologies, and etiologies. In this sense, trying predominantly appears and is strongly related to the positive to look for biomarkers generalizable to all of them is practically relationship that we found with both beta temporal clusters with impossible. In this context, we encourage a reductionist approach right and left hippocampal volumes. As previously explained in oriented to find specific electrophysiological signatures related to this the introduction, WMH is the neuroimaging evidence of the small specific cerebrovascular damage (i.e., WMH) reducing the incidence and recurrent ischemic events underlying the cerebral small vessels of other possible pathological variables, trying to establish a strong disease (CSVD). In this context, our finding related to local beta and accurate baseline for VCI research. This approach could help power reduction hypothetically could explain some kind of relation to understand and try to palliate particular symptoms in mixed between beta band local functioning and brain ischemic events, but dementias, where WMH damage is a comorbid contributor to the this is something that needs further investigation. progression of other neurodegenerative diseases. The present study establishes a baseline for MEG future research in the VCI field, showing general electrophysiological spectral patterns, registered with Limitations MEG. These results demonstrate the utility of MEG signal analysis to augment structural imaging studies in the differentiation of MCI The lack of previous literature assessing those with vMCI using subtypes and to understand the effect of CBVD in brain function and MEG is one of the main concerns of the present study as it does cognition. not allow us to directly compare our results. This lack of replication should be considered for results generalization, but it also entails the beginning of a new and unexplored research field. Data availability statement Although all the patients included in the sample met the MCI diagnosis criteria (Petersen, 2004), the lack of neuropathological The raw data supporting the conclusions of this article will be markers of AD (i.e., CSF/PET amyloid or tau markers) potentially made available by the authors, without undue reservation. increases the heterogeneity of the MCI population and could diminish the degree of association between VCI and AD that we can state with our data. However, significant differences were found despite this limitation. Ethics statement Finally, recent meta-analysis found that WMH volume estimation was not comparable between studies due to the lack of standardization The studies involving human participants were reviewed in the definition of WMH and the high technical variability in and approved and this study was approved by the “Hospital assessment (Melazzini et al., 2021). Despite the lack of evidence for Universitario Clinico San Carlos” ethics committee and was WMH volume references intervals, the literature strongly supports performed in accordance with approved guidelines and regulations. the role of WMH as a biomarker of longstanding cerebrovascular The patients/participants provided their written informed consent to disease, and its implication in the pathophysiology of stroke and participate in this study. cognitive impairment or dementia (Chutinet and Rost, 2014). With the intention of advancing in the search for possible thresholds for this imaging marker we have calculated and detailed reported our Author contributions findings and procedure, based on a referenced method for defining biomarker cut points for brain aging (Jack et al., 2017), to facilitate LT-S: conceptualization, methodology, formal analysis, further replication. investigation, writing—original draft, writing—review and editing, visualization, project administration, and funding acquisition. PC: conceptualization, methodology, formal analysis, investigation, Conclusions writing—original draft, writing—review and editing, visualization and supervision. AC-L: methodology, formal analysis, investigation, The lack of literature on electrophysiological spectral patterns and writing—original draft. BC: writing—review and editing, related to patients with mild (referred to in this study as vMCI) or supervision and funding acquisition. LO: investigation, and major VCI with MEG, and the lack of consensus and replication on writing—original draft. EM: writing—review and editing, Frontiers in Human Neuroscience 09 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 supervision, and validation. PG: writing—review and editing, Conflict of interest supervision, validation and funding acquisition. FM: writing—review and editing, resources, supervision, validation and funding The authors declare that the research was conducted in the acquisition. All authors contributed to the article and approved absence of any commercial or financial relationships that could be the submitted version. construed as a potential conflict of interest. Funding Publisher’s note LT-S acknowledges the financial support of predoctoral All claims expressed in this article are solely those of the researchers grant from Universidad Complutense de Madrid authors and do not necessarily represent those of their affiliated (CT42/18-CT43/18) and co-funded by Santander bank. organizations, or those of the publisher, the editors and the reviewers. Additionally, BC acknowledges the financial support of the Any product that may be evaluated in this article, or claim that National Council of Science, Technology and Technological may be made by its manufacturer, is not guaranteed or endorsed by Innovation (CONCYTEC, Perú) through the National Fund for the publisher. Scientific and Technological Development (FONDECYT, Perú). Finally, all the authors acknowledge the financial support from the Fundación Canaria Instituto de Investigación Sanitaria de Canarias (FIISC). The authors declare that this study received Supplementary material co-funding from Santander Bank. The funder was not involved in the study design, collection, analysis, interpretation of data, The Supplementary Material for this article can be found online the writing of this article, or the decision to submit it for at: https://www.frontiersin.org/articles/10.3389/fnhum.2023.1068 publication. 216/full#supplementary-material. References Alber, J., Alladi, S., Bae, H. J., Barton, D. A., Beckett, L. A., Bell, J. M., et al. (2019). Drebette, S., and Markus, H. S. (2010). The clinical importance of white matter White matter hyperintensities in vascular contributions to cognitive impairment and hyperintensities on brain magnetic resonance systematic review and meta-analysis. BMJ dementia (VCID): knowledge gaps and opportunities. Alzheimers Dement. (N Y) 5, 341:c3666. doi: 10.1136/bmj.c3666 107–117. doi: 10.1016/j.trci.2019.02.001 Farrall, A. J., and Wardlaw, J. M. (2009). Blood—brain barrier: Ageing and Al-Qazzaz, N. K., Ali, S. H. B. M., Ahmad, S. A., Islam, M. S., and Escudero, J. (2017). microvascular disease–systematic review and meta-analysis. Neurobiol. Aging 30, Automatic artifact removal in EEG of normal and demented individuals using ICA-WT 337–352. doi: 10.1016/j.neurobiolaging.2007.07.015 during working memory tasks. Sensors (Basel) 17:1326. doi: 10.3390/s17061326 Fazekas, F., Chawluk, J. B., Alavi, A., Hurtig, H. I., Zimmerman, R. A. (1987). MR signal Anderson, K. L., Rajagovindan, R., Ghacibeh, G. A., Meador, K. J., and Ding, M. abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. Am. J. Roentgenol. 149, (2010). Theta oscillations mediate interaction between prefrontal cortex and medial 351–356. doi: 10.2214/ajr.149.2.351 temporal lobe in human memory. Cereb. Cortex 20, 1604–1612. doi: 10.1093/cercor/ Garcés, P., López-Sanz, D., Maestú, F., and Pereda, E. (2017). Choice of bhp223 magnetometers and gradiometers after signal space separation. Sensors (Basel) 17:2926. Arvanitakis, Z., Capuano, A. W., Leurgans, S. E., Bennett, D. A., and Schneider, J. A. doi: 10.3390/s17122926 (2016). Relation of cerebral vessel disease to Alzheimer’s disease dementia and cognitive Gawel, M., Zalewska, E., Szmidt-Sałkowska, E., and Kowalski, J. (2007). Does EEG function in elderly people: a cross-sectional study. Lancet Neurol. 15, 934–943. (visual and quantitative) reflect mental impairment in subcortical vascular dementia? doi: 10.1016/S1474-4422(16)30029-1 J. Neurol. Sci. 257, 11–16. doi: 10.1016/j.jns.2007.01.046 Azarpazhooh, M. R., Avan, A., Cipriano, L. E., Munoz, D. G., Sposato, L. A., and Gawel, M., Zalewska, E., Szmidt-Sałkowska, E., and Kowalski, J. (2009). The value of Hachinski, V. (2018). Concomitant vascular and neurodegenerative pathologies double quantitative EEG in differential diagnosis of Alzheimer’s disease and subcortical vascular the risk of dementia. Alzheimers Dement. 14, 148–156. doi: 10.1016/j.jalz.2017.07.755 dementia. J. Neurol. Sci. 283, 127–133. doi: 10.1016/j.jns.2009.02.332 Babiloni, C., Binetti, G., Cassetta, E., Cerboneschi, D., Dal Forno, G., Del Percio, C., Hoshi, H., Hirata, Y., Kobayashi, M., Sakamoto, Y., Fukasawa, K., Ichikawa, S., et al. et al. (2004). Mapping distributed sources of cortical rhythms in mild Alzheimer’s disease. (2022). Distinctive effects of executive dysfunction and loss of learning/memory abilities A multicentric EEG study. Neuroimage 22, 57–67. doi: 10.1016/j.neuroimage.2003. on resting-state brain activity. Sci. Rep. 12:3459. doi: 10.1038/s41598-022-07202-7 09.028 Jack, C. R., Wiste, H. J., Weigand, S. D., Therneau, T. M., Lowe, V. J., Knopman, D. S., Benton, A., and Hamsher, K. (1989). Multilingual Aphasia Examination. Iowa City, IA: et al. (2017). Defining imaging biomarker cut points for brain aging and Alzheimer’s AJA Associates. disease. Alzheimers Dement. 13, 205–216. doi: 10.1016/j.jalz.2016.08.005 Caruso, P., Signori, R., and Moretti, R. (2019). Small vessel disease to subcortical Jang, H., Kwon, H., Yang, J. J., Hong, J., Kim, Y., Kim, K. W., et al. (2017). Correlations dementia: a dynamic model, which interfaces aging, cholinergic dysregulation and the between gray matter and white matter degeneration in pure Alzheimer’s disease, pure neurovascular unit. Vasc. Health Risk Manage. 15, 259–281. doi: 10.2147/VHRM.S190470 subcortical vascular dementia and mixed dementia. Sci. Rep. 7:9541. doi: 10.1038/s41598- 017-10074-x Catindig, J. A. S., Venketasubramanian, N., Ikram, M. K., and Chen, C. (2012). Epidemiology of dementia in Asia: insights on prevalence, trends and novel risk factors. Johnston, P. R., McIntosh, A. R., and Meltzer, J. A. (2022). Spectral slowing in chronic J. Neurol. Sci. 321, 11–16. doi: 10.1016/j.jns.2012.07.023 stroke reflects abnormalities in both periodic and aperiodic neural dynamics. Neuroimage Clin. 37:103277.doi: 10.1016/j.nicl.2022.103277 Chu, R. K. O., Braun, A. R., and Meltzer, J. A. (2015). MEG-based detection and localization of perilesional dysfunction in chronic stroke. Neuroimage Clin. 8, 157–169. Kalaria, R. N. (2018). Neuropharmacology the pathology and pathophysiology of doi: 10.1016/j.nicl.2015.03.019 vascular dementia. Neuropharmacology 134, 226–239. doi: 10.1016/j.neuropharm.2017. 12.030 Chutinet, A., and Rost, N. S. (2014). White matter disease as a biomarker for long-term cerebrovascular disease and dementia. Curr. Treat. Options Cardiovasc. Med. 16:292. Khan, A., Kalaria, R. N., Corbett, A., and Ballard, C. (2016). Update on vascular doi: 10.1007/s11936-013-0292-z dementia. J. Geriatr. Psychiatry Neurol. 29, 281–301. doi: 10.1177/0891988716654987 Commowick, O., Istace, A., Kain, M., Laurent, B., Leray, F., Simon, M., et al. Kielar, A., Deschamps, T., Chu, R.K.O., Jokel, R., Khatamian, Y. B., Chen, J. J., et al. (2018). Objective evaluation of multiple sclerosis lesion segmentation using a data (2016). Identifying dysfunctional cortex: dissociable effects of stroke and aging on resting management and processing infrastructure. Sci. Rep. 8:13650. doi: 10.1038/s41598-018 state dynamics in MEG and fMRI. Front. Aging Neurosci. 8:40. doi: 10.3389/fnagi.2016. -31911-7 00040 Frontiers in Human Neuroscience 10 frontiersin.org Torres-Simon et al. 10.3389/fnhum.2023.1068216 Klimesch, W. (1999). EEG alpha and theta oscillations reflect cognitive and Reisberg, B., Ferris, S. H., de Leon, M. J., and Crook, T. (1982). The global deterioration memory performance: a review and analysis. Brain Res. Brain Res. Rev. 29, 169–195. scale for assessment of primary degenerative dementia. Am. J. Psychiatry 139, 1136–1139. doi: 10.1016/s0165-0173(98)00056-3 doi: 10.1176/ajp.139.9.1136 Kloppenborg, R. P., Nederkoorn, P. J., Geerlings, M. I., and van den Berg, E. (2014). Rizzi, L., Rosset, I., and Roriz-Cruz, M. (2014). Global epidemiology of dementia: Presence and progression of white matter hyperintensities and cognition: a meta-analysis. Alzheimer’s and vascular types. Biomed. Res. Int. 2014:908915. doi: 10.1155/2014/908915 Neurology 82, 2127–2138. doi: 10.1212/WNL.0000000000000505 Kulasingham, J. P., Brodbeck, C., Khan, S., Marsh, E. B., and Simon, J. Z. (2022). Santos, C. Y., Snyder, P. J., Wu, W. C., Zhang, M., Echeverria, A., and Alber, J. (2017). Bilaterally reduced rolandic beta band activity in minor stroke patients. Front. Neurol. Pathophysiologic relationship between Alzheimer’s disease, cerebrovascular disease and 13:819603. doi: 10.3389/fneur.2022.819603 cardiovascular risk: a review and synthesis. Alzheimers Dement. (Amst) 7, 69–87. Lam, S., Lipton, R. B., Harvey, D. J., Zammit, A. R., and Ezzati, A. (2021). White matter doi: 10.1016/j.dadm.2017.01.005 hyperintensities and cognition across different Alzheimer’s biomarker profiles. J. Am. Schmidt, P., Gaser, C., Arsic, M., Buck, D., Förschler, A., Berthele, A., et al. (2012). Geriatr. Soc. 69, 1906–1915. doi: 10.1111/jgs.17173 An automated tool for detection of FLAIR-hyperintense white-matter lesions in multiple Lawton, M. P., and Brody, E. M. (1969). Assessment of older people: self-maintaining sclerosis. Neuroimage 59, 3774–3783. doi: 10.1016/j.neuroimage.2011.11.032 and instrumental activities of daily living. Gerontologist 9, 179–186. doi: 10.1093/geront/9. Schomer, D. L., and Lopes da Silva, F. H. (2012). Niedermeyer’s Electroencephalography: 3_Part_1.179 Basic Principles, Clinical Applications And Related Fields. Lippincott Williams & Wilkins Livingston, G., Sommerlad, A., Orgeta, V., Costafreda, S. G., Huntley, J., Ames, D., (LWW). et al. (2017). Dementia prevention, intervention and care. Lancet 390, 2673–2734. Schreiber, S., Bueche, C. Z., Garz, C., and Braun, H. (2013). Blood brain barrier doi: 10.1016/S0140-6736(17)31363-6 breakdown as the starting point of cerebral small vessel disease? - New insights from a Lobo, A., Ezquerra, J., Gómez Burgada, F., Sala, J. M., and Seva Díaz, A. (1979). rat model. Exp. Transl. Stroke Med. 5:4. doi: 10.1186/2040-7378-5-4 El miniexamen, cognoscitivo (un “test” sencillo, práctico, para detectar alteraciones Sheorajpanday, R. V. A., Marien, P., Weeren, A. J. T. M., Nagels, G., Saerens, J., van intelectuales en pacientes médicos). Actas Luso-Espanolas Neurol. Psiquiatr. y Ciencias Putten, M. J. A. M., et al. (2013). EEG in silent small vessel disease: sLORETA mapping Afines 7, 189–202. reveals cortical sources of vascular cognitive impairment no dementia in the default mode network. J. Clin. Neurophysiol. 30, 178–187. doi: 10.1097/WNP.0b013e3182767d15 López-Sanz, D., Bruña, R., Delgado-Losada, M. L., López-Higes, R., Marcos- Dolado, A., Maestú, F., et al. (2019). Electrophysiological brain signatures for the Tak, S., Yoon, S. J., Jang, J., Yoo, K., Jeong, Y., and Ye, J. C. (2011). Quantitative classification of subjective cognitive decline: towards an individual detection in the analysis of hemodynamic and metabolic changes in subcortical vascular dementia preclinical stages of dementia. Alzheimers Res. Ther. 11:49. doi: 10.1186/s13195-019- using simultaneous near-infrared spectroscopy and fMRI measurements. Neuroimage 55, 0502-3 176–184. doi: 10.1016/j.neuroimage.2010.11.046 López-Sanz, D., Bruña, R., Garcés, P., Camara, C., Serrano, N., Rodríguez-Rojo, I. C., Taulu, S., and Simola, J. (2006). Spatiotemporal signal space separation method for et al. (2016). Alpha band disruption in the AD-continuum starts in the subjective rejecting nearby interference in MEG measurements. Phys. Med. Biol. 51, 1759–1768. cognitive decline stage: a MEG study. Sci. Rep. 6:37685. doi: 10.1038/srep37685 doi: 10.1088/0031-9155/51/7/008 López-Sanz, D., Serrano, N., and Maestú, F. (2018). The role of Tecchio, F., Zappasodi, F., Tombini, M., Oliviero, A., Pasqualetti, P., Vernieri, F., et al. magnetoencephalography in the early stages of Alzheimer’s disease. Front. Neurosci. (2006). Brain plasticity in recovery from stroke: An MEG assessment. Neuroimage 32, 12:572. doi: 10.3389/fnins.2018.00572 1326–1334. doi: 10.1016/j.neuroimage.2006.05.004 Maris, E., and Oostenveld, R. (2007). Nonparametric statistical testing of EEG- and Torres-Simón, L., Doval, S., Nebreda, A., Llinas, S. J., Marsh, E. B., and Maestú, F. MEG-data. J. Neurosci. Methods 164, 177–190. doi: 10.1016/j.jneumeth.2007.03.024 (2022). Understanding brain function in vascular cognitive impairment and dementia with EEG and MEG: a systematic review. Neuroimage Clin. 5:103040. doi: 10.1016/j.nicl. Matsumoto, T., Hoshi, H., Hirata, Y., Ichikawa, S., Fukasawa, K., Gonda, T., et al. (2021). 2022.103040 The association between carotid blood flow and resting-state brain activity in patients with cerebrovascular diseases. Sci. Rep. 11:15225. doi: 10.1038/s41598-021-94717-0 Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N., et al. (2002). Automated anatomical labeling of activations in SPM using Melazzini, L., Vitali, P., Olivieri, E., Bolchini, M., Zanardo, M., Savoldi, F., et al. (2021). a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage White matter hyperintensities quantification in healthy adults: a systematic review and 15, 273–289. doi: 10.1006/nimg.2001.0978 meta-analysis. J. Magn. Reson. Imaging 53, 1732–1743. doi: 10.1002/jmri.27479 Van Den Berg, E., Geerlings, M. I., Biessels, G. J., Nederkoorn, P. J., and Moretti, D. V., Babiloni, C., Binetti, G., Cassetta, E., Dal Forno, G., Ferreric, F., et al. Kloppenborg, R. P. (2018). White matter hyperintensities and cognition in mild cognitive (2004). Individual analysis of EEG frequency and band power in mild Alzheimer’s disease. impairment and Alzheimer’s disease: a domain-specific meta-analysis. J. Alzheimers Dis. Clin. Neurophysiol. 115, 299–308. doi: 10.1016/s1388-2457(03)00345-6 63, 515–527. doi: 10.3233/JAD-170573 Moretti, D. V., Miniussi, C., Frisoni, G., Zanetti, O., Binetti, G., Geroldi, C., et al. (2007). van Straaten, E.C.W., de Haan, W., de Waal, H., Scheltens, P., van der Flier, W. M., Vascular damage and EEG markers in subjects with mild cognitive impairment. Clin. Barkhof, F., et al. (2012). Disturbed oscillatory brain dynamics in subcortical ischemic Neurophysiol. 118, 1866–1876. doi: 10.1016/j.clinph.2007.05.009 vascular dementia. BMC Neurosci. 13:85. doi: 10.1186/1471-2202-13-85 Mortamais, M., Artero, S., and Ritchie, K. (2014). White matter hyperintensities as early Van Veen, B. D., van Drongelen, W., Yuchtman, M., and Suzuki, A. (1997). Localization and independent predictors of Alzheimer’s disease risk. J. Alzheimers Dis. 42, S393–S400. of brain electrical activity via linearly constrained minimum variance spatial filtering. doi: 10.3233/JAD-141473 IEEE Trans. Biomed. Eng. 44, 867–880. doi: 10.1109/10.623056 Nakamura, A., Cuesta, P., Fernández, A., Arahata, Y., Iwata, K., Kuratsubo, I., et al. Venkat, P., Chopp, M., Zacharek, A., Cui, C., Zhang, L., Li, Q., et al. (2017). White (2018). Electromagnetic signatures of the preclinical and prodromal stages of Alzheimer’s matter damage and glymphatic dysfunction in a model of vascular dementia in rats with disease. Brain 141, 1470–1485. doi: 10.1093/brain/awy044 no prior vascular pathologies. Neurobiol. Aging 50, 96–106. doi: 10.1016/j.neurobiolaging. Neto, E., Allen, E. A., Aurlien, H., Nordby, H., and Eichele, T. (2015). EEG spectral 2016.11.002 features discriminate between Alzheimer’s and vascular dementia. Front. Neurol. 6:25. Wahlund, L. O., Barkhof, F., Fazekas, F., Bronge, L., Augustin, M., Sjögren, M., et al. doi: 10.3389/fneur.2015.00025 (2001). A new rating scale for age-related white matter changes applicable to MRI and Nolte, G. (2003). The magnetic lead field theorem in the quasi-static approximation and CT. Stroke 32, 1318–1322. doi: 10.1161/01.str.32.6.1318 its use for magnetoencephalography forward calculation in realistic volume conductors. Wahlund, L. O., Westman, E., van Westen, D., Wallin, A., Shams, S., Cavallin, L., et al. Phys. Med. Biol. 48, 3637–3652. doi: 10.1088/0031-9155/48/22/002 (2017). Imaging biomarkers of dementia: recommended visual rating scales with teaching O’Brien, J. T., and Thomas, A. (2015). Vascular dementia. Lancet 386, 1698–1706. cases. Insights Imaging 8, 79–90. doi: 10.1007/s13244-016-0521-6 doi: 10.1016/S0140-6736(15)00463-8 Wardlaw, J. M., Smith, C., and Dichgans, M. (2019). Small vessel disease: mechanisms Oostenveld, R., Fries, P., Maris, E., and Schoffelen, J.-M. (2011). Fieldtrip: open source and clinical implications. Lancet Neurol. 18, 684–696. doi: 10.1016/S1474-4422(19) software for advanced analysis of MEG, EEG and invasive electrophysiological data. 30079-1 Comput. Intell. Neurosci. 2011:156869. doi: 10.1155/2011/156869 Wechsler, D. (1997). Wechsler Memory Scale, 3rd Edition. Pearson Petersen, R. C. (2004). Mild cognitive impairment as a diagnostic entity. J. Intern. Med. Wu, L., Chen, Y., and Zhou, J. (2014). A promising method to distinguish 256, 183–194. doi: 10.1111/j.1365-2796.2004.01388.x vascular dementia from Alzheimer’s disease with standardized low-resolution brain Pfeffer, R. I., Kurosaki, T. T., Harrah, C. H., Jr., Chance, J. M., and Filos, S. (1982). electromagnetic tomography and quantitative EEG. Clin. EEG Neurosci. 45, 152–157. Measurement of functional activities in older adults in the community. J. Gerontol. 37, doi: 10.1177/1550059413496779 323–329. doi: 10.1093/geronj/37.3.323 Yang, D. W., Kim, B. S., Park, J. K., Kim, S. Y., Kim, E. N., and Sohn, H. S. (2002). Prins, N. D., and Scheltens, P. (2015). White matter hyperintensities, Analysis of cerebral blood flow of subcortical vascular dementia with single photon cognitive impairment and dementia: An update. Nat. Rev. Neurol. 11, 157–165. emission computed tomography: adaptation of statistical parametric mapping. J. Neurol. doi: 10.1038/nrneurol.2015.10 Sci. 203–204, 199–205. doi: 10.1016/s0022-510x(02)00291-5 Pusil, S., Torres-Simon, L., Chino, B., López, M. E., Canuet, L., Bilbao, Á., et al. (2022). Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M., et al. (1982). Resting-state beta-band recovery network related to cognitive improvement after stroke. Development and validation of a geriatric depression screening scale: a preliminary Front. Neurol. 13:838170. doi: 10.3389/fneur.2022.838170 report. J. Psychiatr. Res. 17, 37–49. doi: 10.1016/0022-3956(82)90033-4 Quandt, F., Fischer, F., Schröder, J., Heinze, M., Lettow, I., Frey, B. M., et al. (2020). Zalesky, A., Fornito, A., and Bullmore, E. T. (2010). Network-based statistic: identifying Higher white matter hyperintensity lesion load is associated with reduced long-range differences in brain networks. Neuroimage 53, 1197–1207. doi: 10.1016/j.neuroimage. functional connectivity. Brain Commun. 2:fcaa111. doi: 10.1093/braincomms/fcaa111 2010.06.041 Frontiers in Human Neuroscience 11 frontiersin.org