Event‐Related Spectral Perturbation, Inter Trial Coherence, and Functional Connectivity in motor execution: A comparative EEG study of old and young subjects

Abstract Introduction The motor‐related bioelectric brain activity of healthy young and old subjects was studied to understand the effect of aging on motor execution. A visually cued finger tapping movement paradigm and high‐density EEG were used to examine the time and frequency characteristics. Methods Twenty‐two young and 22 healthy elderly adults participated in the study. Repeated trials of left and right index finger movements were recorded with a 128‐channel EEG. Event‐Related Spectral Perturbation (ERSP), Inter Trial Coherence (ITC), and Functional Connectivity were computed and compared between the age groups. Results An age‐dependent theta and alpha band ERSP decrease was observed over the frontal–midline area. Decrease of beta band ERSP was found over the ipsilateral central–parietal regions. Significant ITC differences were found in the delta and theta bands between old and young subjects over the contralateral parietal–occipital areas. The spatial extent of increased ITC values was larger in old subjects. The movement execution of older subjects showed higher global efficiency in the delta and theta bands, and higher local efficiency and node strengths in the delta, theta, alpha, and beta bands. Conclusion As functional compensation of aging, elderly motor networks involve more nonmotor, parietal–occipital, and frontal areas, with higher global and local efficiency, node strength. ERSP and ITC changes seem to be sensitive and complementary biomarkers of age‐related motor execution.


INTRODUCTION
Due to the increasing life expectancy across the world, the proportion of the adult population aged over 60 is rapidly increasing (Ageing-Demographics, WHO Dataset, 2022). Beside the age-related decline in cognitive and memory functions, deteriorating motor performance piqued the interest of neuroscientists (Seidler et al., 2010). Motor dysfunction involves the central and peripheral nervous systems (Borzuola et al., 2020). Deteriorating motor performance in elderly persons manifests itself as slowing of movement (Krampe, 2002), decrease of coordination ability (Seidler et al., 2002), or as increased variability of movement executions (Cooke et al., 1989).
Aging-related structural remodeling and selective gray matter volume reduction mostly affect changes in prefrontal, fusiform, inferior temporal and superior parietal cortices (Raz et al., 1997). Cortical volume reduction is the result of shrinkage or loss of large neurons (Haug & Eggers, 1991;Terry et al., 1987). During normal aging neuroglia cells also undergo numerous changes just as white matter integrity (Bennett & Madden, 2014;Pannese, 2021).
PET and fMRI studies of motor tasks showed overactivation of the prefrontal and sensorimotor regions for older subjects.
These additionally activated brain areas can compensate for performance level despite the structural deterioration in the aging brain (Calautti et al., 2001;Heuninckx et al., 2005Heuninckx et al., , 2008Seidler et al., 2010;Ward & Frackowiak, 2003). A meta-analysis of 40 functional brain-imaging studies (39 fMRI/1 PET) indicates that aging-related motor control involves not only motor cortex but also posterior brain areas such as the occipital-temporal cortex (Zapparoli et al., 2022). In addition, in a recent review of resting state PET and fMRI studies, data from eleven network measures were summarized (Deery et al., 2023) and concluded that the older brain is less modular, more integrated, and less efficient.
Oscillations in the brain regions provide a fundamental coordinating mechanism. Based on Buzsáki and Draguhn (2004), it is realistic to assume that the amplitude and/or phase synchronicity of the oscillations measured at each EEG electrode is related to the event-related activity of the underlying cortical neurons. Connectivity information of different brain regions can provide insights into large-scale neuronal communication and may illustrate how neural pathways are transformed and operate (Bullmore & Sporns, 2009). EEG has much higher temporal resolution than fMRI and thus has a potential to provide a detailed temporal view and show the dynamism of task-related connectivity network changes.
We assume that EEG could confirm the previous fMRI observations further in spatial, spectral, and phase coherence domains monitored by high-density and high-temporal-resolution EEG. Using the finger tap movement (button press) paradigm, as we did before in poststroke condition (Gyulai et al., 2021), could expand our understanding of electrophysiology of brain aging.
Inter Trial Coherence (ITC) is the measure of the phase synchronization of single-trial oscillations relative to a time-locking event, with a value between 0 and 1 (Makeig et al., 2004). Movement-related ITC or phase-locking index in the delta-theta frequency band is a ubiquitous movement-related signal and independent phenomenon of movement initiation (Popovych et al., 2016). Only a few and controversial studies have addressed the phenomena of delta-theta phase-locking value changes of movement-related potentials in the aging brain, contradicting in whether movement-related phase locking is an age-dependent or -independent phenomenon (Liu et al., 2017;Rosjat et al., 2018Rosjat et al., , 2021. Studying the organization of Functional Connectivity networks, which represent the synchronization of physiological signals originating from different brain regions, is a further method in the analysis of the activated motor cortex. Brain networks are organized in a highly efficient manner and can be characterized by graph theoretical metrics such as node degree, global and local efficiency (Bullmore & Sporns, 2009 (1) to find differences between young and elderly motor execution-related brain activity; (2) to characterize age-related differences in cortical networking (3) to determine whether there are, if any, differences in bioelectric activity between young and elderly which can be used as biomarker(s) in aging; and (4) to provide data for future EEG studies, where age could be a factor in the data interpretation.

Participants
Twenty-two young (10 females, mean age: 23.6 years, SD = 3.1) and 22 elderly adults (9 females, mean age: 67.6 years, SD = 10.9) took part in this study. Clinically healthy young and elderly subjects were included in the study, who had no previous history of neurological or psychiatric problems. They were normotensive and the physical examinations were negative. They were engaged in normal daily activities.
All subjects were right-handed according to the Edinburgh Handedness Inventory (Oldfield, 1971). The study was approved by the Ethics Committee of the National Mental, Neurological and Neurosurgical Institute. All subjects provided written informed consent before the study.

The experimental paradigm
Visually cued finger tapping movements were executed in the experiment. The visual cue, a black square of size 0.8 visual degree, was presented on a 22" Samsung (1680 × 1050 pixels) LCD display unit over a middle gray background. The gray level intensity of the square was adjusted linearly in a black-middle gray-black sequence in a cyclic manner with a cycle time between 7-13 s, chosen randomly for each trial (Gyulai et al., 2021). Subjects were seated in front of the display (viewing distance: 70 cm) with supported elbows, and were instructed to press a button with their right or left index finger on a custom-made feedback panel when the contrast between the cue square and the background became zero (i.e., square indistinguishable from the background). The experiment was executed in blocks of 50 finger presses.
Each block was performed using only one hand. Young subjects executed six blocks (3 blocks for each hand). Elderly subjects executed four blocks (2 blocks for each hand). Left and right hand blocks were executed alternately; the left or right hand side to start with was selected randomly by dice throwing. The possible pattern of block execution hence was either L-R-L-R or R-L-R-L.

EEG recording and data processing
EEG was recorded using a Biosemi Active Two EEG device (Biosemi B.V.) with 128 electrodes placed on the scalp according to the Biosemi ABC layout (see Figure 1 for layout and electrode labels). Data were digitized at a 2048 Hz sampling rate with 24-bit A/D conversion.
Trigger events marking the stimulus presentation and response key presses were recorded via the standard trigger port of the Biosemi EEG device.
Data processing was performed by custom MATLAB (The Math-Works Inc., Natick) scripts using EEG analysis toolboxes. Data were high-pass and low-pass filtered at 0.5 Hz and 70 Hz frequencies, respectively, by 4th-order Butterworth filters, followed by a 50 Hz notch filter (Q = 45 quality factor) eliminating power line noise. All filtering steps were performed as zero-phase filtering using the filtfilt() MATLAB function. EEG artifacts (blinks, eye movements, muscle noise, noisy channels) were removed using a six-stage artifact elimination process based on Independent Component Analysis (Delorme et al., 2007;Onton et al., 2006;Weiss et al., 2016). Independent components representing identified by taking into account the results of MARA 2011 (Winkler et al., 2011), FASTER (Nolan et al., 2010), and ADJUST (Mognon et al., 2011) EEGLAB plug-ins, then removed before reconstructing the EEG signals. After artifact elimination, occasional bad channels were interpolated in the EEGLAB toolbox using spherical spline interpolation.
After the data preprocessing steps, 113 (SD = 17.1) and 88.3 (SD = 5.4) trials for young and old volunteers, respectively, were kept for each hand (left and right index finger movements) for further data processing. EEG data were then segmented into [-4500 to 2500 ms] response-locked epochs using the button press event for trial synchronization, and downsampled to 256 Hz sampling rate. To reduce volume conduction effects, scalp current source density estimate was computed using the spherical spline Laplacian method (Perrin et al., 1989) in the Scalp Current Density Toolbox (Kayser & Tenke, 2006) with default parameters (unit sphere radius; the maximum degree of Legendre polynomials: 10; spline flexibility: m = 4; smoothing constant:

Event-Related Spectral Perturbation and Inter Trial Coherence calculations
Event-Related Spectral Perturbation (ERSP) was calculated from the Morlet wavelet-transformed data for all channels according to the single-trial gain model method of Grandchamp and Delorme (2011).
First, the baseline-corrected single-trial power values, P % k are computed for each frequency-time point pairs as where ′ B (f, k) is the mean baseline spectral estimate for trial k at frequency f and is defined as .
Inter Trial Coherence (ITC) is a descriptive statistical measure characterizing the circular variance of event-related phase information or in other words, the phase consistency across trials (Makeig et al., 2004;Tallon-Baudry et al., 1996). It is defined by the magnitude of the vector average of the oscillatory phases at every point of the time-frequencychannel domain across the trials. A 0 value represents random phase distribution, whereas a value 1 represents identical phase values in all trials. ITC is calculated from the complex Morlet wavelet-transformed epoch time-frequency information with the circ_r() function of the CircStat MATLAB Toolbox as described in (Berens, 2009).
In a previous study (Gyulai et al., 2021), we showed that baselinerelated ERSP and ITC values are at their maximum in the 300 ms interval centered around the finger movement onset, in connection with the finger tap execution. Consequently, ERSP and ITC values of this 300 ms interval were used for comparing the motor execution of old and young subjects.

F I G U R E 1
The 128-channel Biosemi ABC electrode layout and the electrode labelling convention. In brackets, the corresponding electrode in the international 10-10 system.

Functional connectivity calculation
Sensor-space Functional Connectivity was calculated based on the weighted debiased Phase Lag Index (dwPLI) (Vinck et al., 2011) using custom scripts in the FieldTrip toolbox (Oostenveld et al., 2010). dwPLI characterizes phase synchronization of oscillations across trials and is less sensitive to zero-phase connectivity, that is, volume conduction generated spurious connections, than other phase-based connectivity metrics, such as the phase-locking value (PLV). Connectivity was computed for each EEG frequency band in the sensor space for a 300 ms time window centered at the time of the finger tap press. Sensor space was used to avoid potential source localization error induced inaccuracies and performing MRI scans of the subjects. The resulting weighted connectivity association matrices in which each entry represents the connection strength between two electrodes were then subsequently processed with the Brain Connectivity Toolbox (Rubinov & Sporns, 2010) to compute connectivity graph measures such as global and local efficiency and node strength (Bullmore & Sporns, 2009;Chennu et al., 2014;Edmunds et al., 2019).
Global efficiency represents the integration property of a network; the efficiency with which different areas can transfer information to one another. This metric is the inverse of the average shortest path length from one node to all other nodes. Local efficiency, on the other hand, represents the local information processing capability of the network, the "strength" of connectivity of a local node neighborhood or cluster. Node degree (or node strength in a weighted network) represents the number (or sum of weights) of the connections of a given node.

Statistical analysis
The ERSP and ITC values covering the -150 to 150 ms time and 0.5-70 Hz frequency interval were subjected to statistical analysis within subjects (baseline-related ERSP and ITC change) and between groups (comparing old and young subjects by ERSP, ITC). To control the multiple comparison problem, the cluster-based permutation method (Maris & Oostenveld, 2007) was used in the FieldTrip MAT-LAB toolbox (Oostenveld et al., 2010). Paired t-tests were computed by the ft_statfun_depsamplesT() FieldTrip function. The cluster alpha level was set to a 0.001 threshold value to minimize false positives.
The lower than usual threshold was selected based on suggestions for cluster-based statistical analysis (Eklund et al., 2016). Spatial adjacency was defined by combining distance and triangulation methods from the ft_prepare_neighbours() FieldTrip function (Oostenveld et al., 2010). The threshold of distance was set to 0.4. The significance of the clusters was estimated based on 10,000 Monte Carlo randomizations. Clusters with p-value > .05 were rejected. The same clusterbased permutation statistical analysis was used to find differences between old and young subjects. However, in these test, the paired ttests were calculated by the ft_statfun_indepsamplesT() FieldTrip function. All other parameters were identical to the ones described above.
The statistical analysis of old versus young Functional Connectivity was performed in EEGLAB using the statcond() function. Since dwPLI relies on epoch-based phase synchronization, for each subject we obtained one association matrix for each EEG frequency band, averaged from the 300 ms target interval dwPLI values, from which we computed the node strength and local efficiency values for each electrode, and one global efficiency value per subject, respectively. These values were then tested for group-level mean differences with the nonparametric permutation test by the function statcond().

ERSP in young adults
In right (dominant hand) finger movement, delta band ERSP maximum was found on the left hemisphere at electrodes D14, D17-19, D28 (see Figure 1 for electrode layout and labels). The frontal electrode D19 correlates with electrode C3 of the international 10-10 system that is considered to represent the left sensorimotor area (its cortical projection is mainly over the postcentral gyrus (69%) and-to a lesser extent-the precentral gyrus (19%)). Theta ERSP showed significant increase in the active left hemisphere with a characteristic doublepeak pattern. The anterior peak area was located midline, including electrode C23 (which is equivalent to FCz in the 10-10 system), and had larger magnitude than the second peak. Significant decrease in the alpha band ERSP was found over the entire posterior area of the brain.

ERSP differences of old and young subjects
In right index finger movements, no significant delta band ERSP difference was found between old and young subjects. In the theta frequency band, however, elderly subjects had significantly smaller ERSP values than young ones. Theta band ERSP differences were found over the column 2). No significant differences were found in the gamma band.

ITC of young subjects
In right index finger movements, the maximum of delta band ITC was In left index finger movements, delta band ITC maximum was found over the right central electrodes B18-23. In the 10-10 system, electrode B20 corresponds to C2, whose main cortical projection is the precentral gyrus. Theta band ITC also showed a characteristic doublepeak pattern over the active right hemisphere. The smaller, anterior theta band ITC peak was observed over the medial-frontal electrodes, while the greater posterior peak was located at the parietal electrodes.

ITC of elderly subjects
In right (dominant hand) finger movements, the maximum of the delta band ITC was detected at the left frontal-midline (premotor) and parietal areas. In the theta band, ITC had a characteristic double-peak pattern over the left hemisphere, similar to the young subjects. The posterior theta band ITC peak area covered electrodes D26, 29, A6-7.
The location of the significant alpha band ITC cluster corresponded to the posterior theta ITC peak (p = .0001, SD = 0.0001) (Figure 4 column 2).
In left index finger movements, the delta band ITC peak area was located over the right frontal-central-parietal electrodes B3-4, B17, B23, and B32. The anterior peak of the theta band ITC area covered more frontal electrodes. As in the right finger movement, the location of the significant alpha band ITC cluster corresponded to the posterior theta ITC peak (positive cluster p = .0001, SD = 0.0001) ( Figure 5, column 2). No significant differences were found in the beta and gamma bands for either finger.

Comparison of ITC between young and old
In right index finger movements, we found significant difference for the event-related delta and theta band ITC between old and young subjects. We found that elderly subjects had a significantly more extended ITC area in front of and behind the left central-parietal motor region Significant ITC differences were found in the delta and theta bands between old and young subjects. We found that the spatial extent of increased ITC values was larger in old subjects.

Connectivity results
The sensor-space Functional Connectivity association matrices were computed for the delta, theta, alpha, and beta bands (left and right index finger movements, respectively) based on the debiased weighted Phase Lag Index (dwPLI) connectivity method. The association matrices are shown in Figure 6. The matrices are weighted, not binary, to enable comparison of connection weight distribution differences between old and young subjects in each frequency band. For easier comparison of the old and young subject association matrices, identical weight value ranges are used for each band.
We found significant differences (p < .01) in each frequency band between the old and young group node strength values, showing higher mean node strength for old subjects. Figure 7 shows the distribution of In older subjects, the execution-related delta and theta band ITC for both hands affects the contralateral parietal-occipital areas. In the right finger movement execution, additionally, significant delta band ITC appears in the left prefrontal-premotor area. An age-dependent theta and alpha ERSP decrease was observed for both hands over the frontalmidline area, including SMA in the center. In the right finger movement execution of older adults, theta, alpha, and beta ERSP decrease was found over the right prefrontal area. Old subjects also showed significant alpha and beta ERSP changes in right finger tap execution over the right sensorimotor area. Elderly subjects showed significant beta ERSP F I G U R E 9 Old and young subject sensor-space Functional Connectivity graphs of the right finger tap task in the delta (top) and theta (bottom) frequency bands. The networks of the old subjects show higher participation of the occipital-parietal areas, whereas young subjects displayed increased connections in the frontal areas.
change over the ipsilateral parietal regions during the finger movement executions with both hands. The global and local efficiency, node strength were higher by both right and left button press in the delta, theta, alpha, and beta frequencies in older adults.
The finger tapping movement execution is a frequently used paradigm (Gerloff et al., 1998;Körmendi et al., 2021;Popovych et al., 2016;Urbano et al., 1996) to study the organization of the motor execution and corresponding decision-making processes. Attention, decision-making, motor execution, and controlling motions are associated with theta oscillations (Cavanagh et al., 2012). In the present study, we targeted exclusively the execution phase of the finger movement. It could be assumed that the synchronization and temporal organization of pyramidal, principal cells corresponding to motor execution are regulated by theta oscillations of interneurons. Cortical theta activity is suggested to be a slower variant of the thalamocortical alpha rhythms (Buzsáki et al., 2013). Hippocampal theta is linked to voluntary movement in rats (Vanderwolf, 1969). Recently, in human intracranial EEG experiments, Ramayya et al. (2021) have found that theta oscillations increased near the areas of movement-related neural populations during movements.
In previous PET and fMRI studies, overactivation involving the prefrontal and higher-level sensorimotor regions was found in elderly subjects (Calautti et al., 2001;Heuninckx et al., 2005Heuninckx et al., , 2008Seidler et al., 2010;Ward & Frackowiak, 2003). In the delta-theta bands, both for right and left index finger movements, elderly subjects showed a more extended significant ITC activity over the parietaloccipital regions compared to young subjects. Our finding challenges the observation of Liu's group, which concluded that no phase-locking differences appear in the delta-theta frequency bands between old and young subjects during visually cued or self-initiated finger tapping tests (Liu et al., 2017). They stated that phase locking in the delta-theta frequencies is a general, age-independent phenomenon underlying the execution of simple finger movements, but the extent of synchronization between motor areas (i.e., interregional phase-locking value) significantly differed depending on age: multiple additional intra-and interhemispheric connections were found in older subjects (Rosjat et al., 2018). Our study showed significant age-related increase over the parietal-occipital and the left prefrontal areas. Our finding is in agreement with fMRI observations. In elderly subjects, increased motor-related fMRI activation was observed in the superior temporal gyrus, supramarginal gyrus, secondary somatosensory area, ipsilateral precuneus, presupplementary motor area, predorsal premotor area, dorsolateral prefrontal cortex, as well (Heuninckx et al., 2005). A novel meta-analysis of 40 functional brain-imaging studies indicates that aging-related motor control involves not only motoric brain regions but also posterior areas such as the occipital-temporal cortex (left calcarine fissure and left superior occipital gyrus, the right occipital-temporal cortex (Zapparoli et al., 2022). Calautti et al. (2001) found overactivation for older subjects in the superior frontal cortex (premotor-prefrontal junction) ipsilateral to the moving fingers.
Our study found extended contralateral premotor-prefrontal delta band ITC area in right finger movements. During bimanual finger tapping employing EEG-based dynamic causal modeling, elderly subjects showed significant couplings between the left prefrontal cortex and the left lateral premotor area (Loehrer et al., 2016). As a novelty in our analysis, we confirmed the fMRI findings with the delta-theta F I G U R E 1 0 Old and young subject sensor-space Functional Connectivity graphs of the left finger tap task in the delta (top) and theta (bottom) frequency bands. The networks of the old subjects show more extended network connections to frontal areas.
ITC calculation, the increase of posterior executive motor network in elderly subjects. ITC carries information on the temporal pattern of neural activity of a given neural population under the same condition (Popovych et al., 2016). Higher ITC suggests a higher involvement of cortical areas in the given task. A dynamic graph-based approach using EEG phase locking in the low frequencies (2-7 Hz) indicated different neural information processing in older subjects; their networks displayed an overall increased connectivity, especially in motor-related electrodes (Rosjat et al., 2021). Our results imply a growing active area working with delta-theta frequencies in older subjects. The higher global efficiency of our older subjects supports the observation of Rosjat's group.
MEG studies of visuospatial processing found that neural responses in the theta and lower alpha range (4-10 Hz) correlate with the chrono-logical age of the prefrontal and motor cortices (Wiesman & Wilson, 2019). In a lifespan EEG study, the theta Inter Trial Coherence in the medial-frontal cortex (electrode FCz) increased from childhood to early adulthood and started to decrease from early adulthood to old age (Papenberg et al., 2013). We found that the motor executionrelated theta band ITC increased over the parietal-occipital regions in older adults, but not in the midline frontal areas (electrode FCz).
In a motor-related theta activity EEG study, strong aging-related theta power suppression was observed at the medial-frontal-central (FCz) region (Yordanova et al., 2020). The study suggests that this altered regulation in older subjects is due to the suppression of the medial-frontal integrating mechanism. Emerging evidence suggests that midfrontal theta oscillations are involved in cyclically orchestrating brain computations, which is more likely to be related to response execution during the tasks than to conflict processing (Duprez et al., 2020). Oscillatory midfrontal theta dynamics during reactive control mostly reflect motor-related adjustments and the theta power is predictive of motor slowing (Kaiser & Schütz-Bosbach, 2021). SMA-M1 connectivity was found to be reduced in older adults with TMS (Green et al., 2018;Rurak et al., 2021) during right finger tapping.
Alpha sources in posterior brain regions were found to significantly decrease with aging (Babiloni et al., 2006). During an auditory discrimination motor task registered with 12 electrodes, widespread lower alpha amplitudes were detected in elderly subjects during cognitive and motor tasks (Dushanova & Christov, 2014).
We found significant changes in alpha ERSP in elderly subjects in the frontal-midline/SMA regions during finger tap executions by both hands. Alpha band ERSP decrease above the SMA region is a novel finding and could be interpreted as an overactivation mechanism of the aging motor network.
Sensorimotor alpha and beta rhythm changes may reflect different neural trajectories in aging (Schmiedt-Fehr et al., 2016). Sensorimotor beta desynchronization is associated with motor preparation, execution, and imagination (Neuper et al., 2006;Pfurtscheller & Lopes da Silva, 1999). In a dominant hand grip MEG study, the increasing age was associated with increased movement-related beta desynchronization only in the ipsilateral M1 region (Rossiter et al., 2014). In an fMRI hand grip study, reduced ipsilateral M1 deactivation was observed in older subjects at both hands (Ward et al., 2008). During right hand grips, TMS and fMRI observations suggested that the ipsilateral cortical motor areas, in particular ipsilateral M1, play a central role in maintaining performance levels with aging through increasingly facilitatory corticocortical influences (Boudrias et al., 2012). Beta band activity in sensorimotor and parietal cortex are important for accurate motor performance (Chung et al., 2017). Our beta band ERSP results confirm MEG and fMRI studies showing age-related changes over ipsilateral regions. In addition, theta, alpha, and beta band ERSP values could be used as biomarkers of aging. Delta and theta band ITC phase synchrony could also be a further age-related marker.
The aging brain shows that decreased efficiency of individual functional areas and that these changes are related to changes in connectivity and the small-world architecture (modular structure) of the brain networks (Goh, 2011;Rakesh et al., 2020). Graph theory provides a suitable framework to study the integration (global efficiency), and segregation (local efficiency) properties of the connectivity network. EEG-based connectivity analysis of the aging brain. Wang et al. (2018) performed an EEG Functional Connectivity analysis of an audiovisual integration task and found increased global and local efficiency and degree in older subjects. They hypothesized that increase in efficiency metrics during a cognitive task indicate the presence or activation of more inter and intra module connections, representing less efficient execution and an increased cognitive demand in the aging brain. Our connectivity results reinforce these findings, as we have found significant increase in node strength as well as in global and local efficiency in the old group. This supports the idea that the aging brain requires stronger cooperation of different functional modules during the execution of a given task. Our findings suggest that both neural connectivity and the organization of these connections are important markers of processing efficiency and can be used in the future to characterize the aging process of the brain. EEG seems to be an ideal, low-cost, and easyto-use technology for future longitudinal ITC and connectivity studies to identify quantitative aging measures.

CONCLUSIONS
This study investigated aging effects in motor execution comparing old and young subjects. 128-channel high-density EEG measurements were performed using a visually cued finger tapping movement experimental paradigm. Event-Related Spectral Perturbation, Inter Trial Coherence, and Functional Connectivity were computed from the measurement data and analyzed in different EEG frequency bands.
We found significant changes in delta and theta ITC over extended parietal-occipital areas. Alpha and beta band ERSP decreased in elderly subjects over the frontal-midline and the ipsiparietal areas, respectively. ITC and Functional Connectivity network analysis confirmed that in old subjects a more extended network of cortical activation areas is involved in motor execution. These differences could be interpreted as bioelectric markers of aging. This approach may initiate further studies, where age could be a factor in the data interpretation in healthy and diseased conditions.

CONFLICT OF INTEREST STATEMENT
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.