Contents lists available at ScienceDirect

Neuropsychologia

journal homepage: www.elsevier.com/locate/neuropsychologia

Auditory priming improves neural synchronization in auditory-motorentrainment

Jewel E. Crastaa,b,⁎, Michael H. Thautc, Charles W. Andersond, Patricia L. Daviesa,e,William J. Gavine

a Department of Occupational Therapy, Colorado State University, Fort Collins, CO 80523, USAb Kennedy Krieger Institute, Baltimore, MD 21205, USAc Faculty of Music, Music and Health Science Research Center, University of Toronto, Toronto, Ontario, Canada M5S 1A1d Department of Computer Science, Colorado State University, Fort Collins, CO 80523, USAe Department of Molecular, Cellular & Integrative Neuroscience, Colorado State University, Fort Collins, CO 80523, USA

A R T I C L E I N F O

Keywords:Auditory-motor entrainmentSynchronizationEEG

A B S T R A C T

Neurophysiological research has shown that auditory and motor systems interact during movement to rhythmicauditory stimuli through a process called entrainment. This study explores the neural oscillations underlyingauditory-motor entrainment using electroencephalography. Forty young adults were randomly assigned to oneof two control conditions, an auditory-only condition or a motor-only condition, prior to a rhythmic auditory-motor synchronization condition (referred to as combined condition). Participants assigned to the auditory-onlycondition auditory-first group) listened to 400 trials of auditory stimuli presented every 800ms, while those in themotor-only condition (motor-first group) were asked to tap rhythmically every 800ms without any externalstimuli. Following their control condition, all participants completed an auditory-motor combined condition thatrequired tapping along with auditory stimuli every 800ms. As expected, the neural processes for the combinedcondition for each group were different compared to their respective control condition. Time-frequency analysisof total power at an electrode site on the left central scalp (C3) indicated that the neural oscillations elicited byauditory stimuli, especially in the beta and gamma range, drove the auditory-motor entrainment. For thecombined condition, the auditory-first group had significantly lower evoked power for a region of interest re-presenting sensorimotor processing (4–20 Hz) and less total power in a region associated with anticipation andpredictive timing (13–16 Hz) than the motor-first group. Thus, the auditory-only condition served as a primingfacilitator of the neural processes in the combined condition, more so than the motor-only condition. Resultssuggest that even brief periods of rhythmic training of the auditory system leads to neural efficiency facilitatingthe motor system during the process of entrainment. These findings have implications for interventions usingrhythmic auditory stimulation.

1. Introduction

The auditory and motor systems of the brain interact duringmovement to rhythmic auditory stimuli through the process of en-trainment (Molinari et al., 2003). Entrainment or sensorimotor syn-chronization is described as a form of behavior wherein an action istemporally coordinated with a predictable external stimulus (Pressing,1999). Rhythmic synchronization between the auditory and motorsystem underlies and facilitates many functional everyday behaviors.For example, studies examining rhythm in speech suggest that dis-rupting the natural rhythm of a sentence hampers intelligibility (Peelle

and Davis, 2012). Moreover, researchers have found that rhythm inspeech and other sounds provide a predictive cue to the time of onset ofthe subsequent stimuli (Giraud and Poeppel, 2012). These predictionsoccur via stimulus-driven entrainment or phase-locking of neural os-cillations, which modulate neuronal excitability and generate maximalsensitivity during the subsequent critical time period (Giraud andPoeppel, 2012). Thaut et al. (1998a, b) postulated that rhythmic en-trainment can be mediated by adaptive neural timing networks that arephysiologically equipped to lock motor responses to rhythmic sensorypatterns and create temporal response stability over time.

Studies examining entrainment have often used finger tapping as

https://doi.org/10.1016/j.neuropsychologia.2018.05.017Received 30 August 2017; Received in revised form 18 May 2018; Accepted 20 May 2018

⁎ Corresponding author. Address: 716 N Broadway, Room 306, Baltimore, MD 21205, USA.E-mail addresses: crasta@kennedykrieger.org (J.E. Crasta), michael.thaut@utoronto.ca (M.H. Thaut), chuck.anderson@colostate.edu (C.W. Anderson),

patricia.davies@colostate.edu (P.L. Davies), william.gavin@colostate.edu (W.J. Gavin).

Neuropsychologia 117 (2018) 102–112

Available online 22 May 20180028-3932/ © 2018 Elsevier Ltd. All rights reserved.

T

the experimental movement (Repp, 2005). Evenly-spaced rhythmicauditory stimuli are often used in finger tapping paradigms, which areconsidered fairly easy to perform by most neurotypical individuals.Several researchers have shown in tasks that require tapping to audi-tory or visual stimuli occurring in isochronous sequence motor re-sponses are effectively entrained to the onset of sensory stimuli (Repp,2005; Thaut et al., 1998a, b). Furthermore, Thaut et al. (1998a, b)demonstrated that a steady synchronization state is typically attainedwithin two to three stimulus repetitions and can be reliably maintainedover long timing sequences. Synchronized tapping to a beat has beenshown to be accurately maintained at intervals of 200–1800ms, with amaximum accuracy at about 600ms (Repp, 2005). Within relativelyshort (200–1800ms) inter-trial intervals (ITI), the responses turn intoanticipations with the tendency of the motor taps to precede the stimuli(Fraisse, 1982). This phenomenon of negative mean asynchrony hasbeen shown to be ubiquitous to finger tapping paradigms (Repp, 2005).These studies suggest that entrainment of motor responses occurs due tothe prediction of future events and may occur unintentionally (Schmidtand O’Brien, 1997). Thus, the finger taps in a synchronization paradigmdiffer from simple reaction time tasks because the former occurs as atimed response that is delayed to coincide with the next auditory sti-muli (Repp, 2005). While researchers have begun to elucidate the be-havioral mechanisms underlying sensorimotor synchronization (Repp,2005; Repp and Su, 2013), the effect of entrainment at the neural levelstill remains unclear.

Neural oscillation signals obtained through electroencephalography(EEG) consistently reflect specific cognitive and sensory processes (Thutet al., 2012). Frequencies in the delta (1–3 Hz), theta (4–7 Hz) andalpha (8–12 Hz) bands are shown to be part of a hierarchical network ofoscillators involved in predictive sensory processing (Arnal et al.,2014). Delta oscillations are associated with decision making and at-tentional processes (Güntekin and Başar, 2016), while theta band ac-tivity is associated with cognitive control across brain networks(Cavanagh and Frank, 2014). Frequencies in the alpha range are asso-ciated with entrainment using visual stimuli (Keitel et al., 2014; Spaaket al., 2014). Beta activity (13–30 Hz) is associated with anticipationand predictive timing (Arnal et al., 2011), and gamma band activity(30–80 Hz) is associated with top-down attentional control (Debeneret al., 2003; Tiitinen et al., 1993). When recorded EEG signals aresegmented and time-locked to a specific re-occurring event such as anauditory tone, the event-driven data can be averaged. The process ofaveraging data produces information about what is phase-locked acrosstrials and represents evoked responses. Additionally, across segments,the degree of phase-locking for frequency-specific EEG oscillations canbe measured (Lachaux et al., 1999; Roach and Mathalon, 2008). Whendata are summed across trials, the activity represents evoked and in-duced responses. To obtain a measure of induced activity, the evokedactivity is subtracted from the summed activity. Participants listeningto a sequence of isochronous beats appear to have distinct cortical os-cillations, an evoked gamma band which occurs in response to theauditory stimuli, and an induced gamma band which occurs in antici-pation of the onset of the auditory stimuli (Snyder and Large, 2005).

Specialized connections between the auditory and motor systemsunderlie predictive timing processing during rhythmic finger tapping(Thaut et al., 2009). Multiple brain regions such as the cerebellum(Molinari et al., 2007; Grahn et al., 2011), inferior colliculus (Tierneyand Kraus, 2013) and basal ganglia are involved during auditory-motorentrainment (Thaut et al., 2009). At the basic level, neurons in theprimary auditory cortex have a dual representation of the repetitionrate of a stimulus, by discharge rate, and by synchronized firing pat-terns of a group of neurons (Bendor and Wang, 2007; Eggermont,2001). Serrien (2008) found higher tapping variability and greaterfunctional connectivity in the beta band during unpaced versus syn-chronized tapping. Thus, when external pacing stimuli were absent,there was a higher demand for motor timing in areas such as the sup-plementary motor area (Serrien, 2008). Others have found similar

results with significant activity in the prefrontal-parietal-temporalnetwork during unpaced tapping (without external pacing stimuli)alone but not during synchronized tapping (Jantzen et al., 2007). Ac-tivity in these areas is linked with working memory load, which impliesan increased demand on working memory to maintain temporal paceduring tapping in the absence of the rhythmic input (Jantzen et al.,2007). Collectively, these studies indicate that synchronized tappingmay require less neural activity compared to unpaced tapping.

The induced gamma band activity can be observed even duringoccasional tone omissions, which suggests that auditory-motor en-trainment is linked to temporal expectancy (Snyder and Large, 2005).Studies using EEG and magnetoencephalography (MEG) measures showthat the brain synchronizes to an external rhythm even in the absence ofmotor response (Lakatos et al., 2008). These oscillatory neural beha-viors represent endogenous entrainment to the beat of the auditoryrhythm (Large and Snyder, 2009). Electrophysiological recordings fromthe monkey auditory cortex have indicated that entrained oscillatoryactivity to a train of pure tone auditory stimuli persisted even after thestimulation ceased (Lakatos et al., 2013). This finding suggests thatexposure to rhythmic contexts could influence subsequent perception.Additionally, researchers propose that the rhythmicity in the motorcortical oscillations (oscillatory rhythms in the delta, theta, mu [~10Hz activity in the sensorimotor areas reflecting sensorimotor pro-cessing in the frontoparietal networks; Pineda, 2005], and beta bands)modulates the neural oscillations of the auditory cortex, which is be-lieved to in turn influence auditory perceptual processes (Schroederet al., 2008, 2010). This behavioral change associated with the repeatedprocessing of a stimulus also referred to as “priming effects”, is ob-served in numerous regions of the human brain (Henson, 2003). Adecreased hemodynamic response or a reduced response is seen usingfunctional imaging for primed versus unprimed auditory stimuli withreference to memory tasks (Bergerbest et al., 2004; Henson, 2003;Schacter and Buckner, 1998). This reduced hemodynamic response isassumed to reflect faster or more efficient processing of the primedstimulus (Henson, 2003). However, the effect of priming on neuraloscillatory behavior during auditory-motor entrainment has not beenexamined.

Interventions using rhythmic auditory stimulation have been shownto benefit motor performance in various clinical populations with motorimpairments such as Parkinson's disease (Thaut et al., 1996), traumaticbrain injury (Thaut and Abiru, 2010), and stroke (Thaut and McIntosh,2014). Thus, the addition of rhythmic auditory stimuli results in moreefficient motor processing at the behavioral level. However, little isknown about the neural mechanisms by which auditory entrainmentenhances motor performance. Researchers have yet to determine theneural oscillations that enable people to achieve and maintain the stateof entrainment. The oscillations through which the brain spontaneouslycreates internal beat representations from rhythmic auditory stimuliremain unknown. Understanding these mechanisms may shed light onthe theoretical underpinnings of rhythmic auditory-motor entrainment.Understanding the neural oscillations related to auditory-motor en-trainment can provide useful biomarkers and guide the rehabilitation ofindividuals with motor impairments. Thus, there exists a critical needfor systematic research aimed at exploring this complex neural inter-action to understand the neural substrates of entrainment and move-ment. Recent advances in EEG analyses using time-frequency measureshave suggested that neural oscillations and their synchronizationconvey crucial information about mechanisms for inter-neuronal com-munication and integration of information across brain regions.

The purpose of this study is to provide evidence of neural entrain-ment of auditory and motor information using measures of EEG oscil-lations. Additionally, to better understand the neural oscillations thatrepresent the synchrony of auditory and motor systems, we first ex-amine the neural oscillatory activity of a motor-only condition and anauditory-only condition. Then we compare each of the two controlconditions to the respective within-group combined auditory-motor

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

103

synchronization condition (i.e., combined condition). This study's re-search questions include:

1. How do the neural oscillations for an auditory-only condition differfrom a motor-only condition and are they qualitatively and quan-titatively different?

2. What neural oscillations represent the synchrony of auditory andmotor systems when both systems are active together during en-trainment?

3. Does prior exposure to an auditory-only entrainment condition(auditory priming) or a motor-only condition (motor priming) resultin differences in neural synchronization in an auditory-motor en-trainment condition?

We hypothesized that participants who are engaged in tapping torhythmic auditory stimuli will have different synchronized brain ac-tivity in the alpha, beta, and gamma band oscillations compared towhen participants only hear rhythmic auditory stimuli or when onlyengaged in rhythmic tapping with no auditory stimuli. Additionally,during the combined auditory-motor synchronization condition (alsoreferred to as the combined condition), participants who hear rhythmicauditory stimuli (auditory-only condition) prior to tapping to auditorystimuli will have brain activity that is more synchronized and of lesspower in alpha, beta, and gamma oscillations than participants in themotor-only group (motor-only condition) who do not receive this initialauditory experience.

2. Methods

2.1. Participants

Forty neurotypical young adults, ages 18–39 years (M = 23.78years, SD = 4.95; 18 males, 23 females) participated in this study.Participants were recruited through word-of-mouth from the localuniversity. All participants were screened using a self-report screeningquestionnaire developed in our lab to ensure that they were free ofneurological injuries, disabilities, and family histories of psychologicaldisorders. All participants met these inclusion criteria and no partici-pants were excluded from the analysis based on these criteria.Participants were randomly assigned to one of two groups each withdifferent control conditions: an auditory-first group (n= 20; M = 24.7years, SD = 4.5; 8 males) or a motor-first group (n= 20; M = 22.5years, SD = 4.9; 10 males). The groups were not significantly differenton age, t (38) = 1.5, p= .14. Twenty-two of the 40 participants reportedhaving played a musical instrument. Of the 22 participants who re-ported having played an instrument, 12 belonged to the auditory-firstgroup, while 10 belonged to the motor-first group. Participants reportedtheir level of musical expertise on a Likert scale ranging from 0 (noexpertise) to 5 (expert). The groups were not significantly different onmusical expertise, t (38) = 1.7, p= .10, with both groups having par-ticipants along the continuum from no expertise to expert level.

2.2. Data collection

A cross-sectional quasi-experimental quantitative study design withconvenience sampling procedures was employed to compare twogroups. Participants in the auditory-first group completed an auditory-only control condition where they listened to 400 trials of auditorystimuli presented every 800ms. Participants assigned to the motor-firstgroup completed a motor-only control condition and were asked to self-pace their tapping to the desired frequency of 800ms which wasmodeled by the researcher for 5 beats (Krause et al., 2010). After theinitial first 5 beats, participants were not given any other feedbackrelated to their timing. All participants were able to maintain consistenttiming following the removal of the cue, which is consistent with pre-vious finger tapping studies (Thaut et al., 1998a, b). Following a short

break after their respective control condition, participants in bothgroups completed an auditory-motor combined condition which re-quired the participant to tap along with auditory stimuli every 800ms.The stimulus intensity of all auditory tones was set to 70 dB soundpressure level (SPL) for this study. The tones were binaurally presentedthrough ER-3A inserted earphones (Etymotic Research) using E-Primesoftware. The tones were generated using a 1 kHz sinusoidal wave. Eachtone had a 50ms duration with a 10ms rise/fall time. Participants wereinstructed to tap the right key on the touchpad using their right handfor the motor response. The participants were seated in a relaxed po-sition in a high-back chair in a sound-attenuated and electricallyshielded chamber. Prior to the EEG recording, they were provided witha short training to reduce artifacts caused by eye blinks and muscleactivity. Ethical approval was obtained from the Institutional ReviewBoard of Colorado State University and all participants providedwritten consent before the study session.

2.3. EEG data recording

All EEG data were collected using the BioSemi ActiveTwo EEGAcquisition System (BioSemi, Wg-Plein 129, 1054 SC Amsterdam,Netherlands). This system includes 64 pin-type Ag/AgCl sintered ac-tive-electrodes inserted into an Active Two Lycra head cap based on theAmerican Electroencephalographic Society nomenclature guidelines.The system also has 8 additional flat electrodes. Four flat electrodesmeasured electrooculograms (EOGs); two electrodes were placed on theleft supraorbital and infraorbital region to record vertical eye move-ments, and two were placed lateral to the external canthus of each eyeto measure horizontal eye movements. Two flat electrodes were placedon the left and right earlobes and used as the offline reference.Additionally, two flat electrodes were placed on the left and rightmastoids. A Common Mode Sense (CMS) active electrode and a DrivenRight Leg (DRL) passive electrode served as the online reference andground (http://www.biosemi.com/faq/cms&drl.htm). Data were sam-pled at a rate of 1024 Hz with a bandwidth of 0–268 Hz.

2.4. Electrophysiological data reduction

Brain Vision Analyzer software (Brain Products GmbH, München,Germany) was used to conduct all offline EEG analyses. First, the fourEOG channels were converted to a vertical and a horizontal bipolarEOG. Data were filtered offline from the continuous EEG with a band-pass setting of .1–80 Hz. Following this, data were segmented time-locked to the onset of the target event, either the auditory stimulus orthe motor response with a duration of 1800ms pre-stimulus onset to1800ms post-stimulus onset. Baseline correction relative to a baselineof − 200–0ms was performed. Next, an eye regression technique de-signed to remove eye movement from trials was performed (Segalowitz,1996). Baseline correction was performed again relative to a baseline of− 200–0ms for the non-rejected segments. Next, segments with voltagedeviations greater than± 100 microvolts (μV) on any of the EEGchannels or the bipolar EOG channels were eliminated. No more than5% of segments out of 400 were lost for any individual for any condi-tion. Time-frequency analyses were done using software written inMATLAB according to equations outlined by Roach and Mathalon(2008).

2.5. Time-frequency analysis

Time-frequency analysis involves spectral decomposition of event-locked EEG oscillations into magnitude and phase information for eachfrequency present in the processed EEG segment. A continuous wavelettransform based on the Morlet wavelet was applied using a MATLABprogram designed to convert waveforms into sine waves of variousphase angles and frequencies. The continuous wavelet transform wasused as it provides a time-frequency representation of EEG with a

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

104

varying time resolution that increases with frequency (Addison, 2016).The frequency characteristics measured included (1) total power, (2)evoked power, and (3) phase-locking factor. Power represents themagnitude of neuro-electric oscillations at specific frequencies (Roachand Mathalon, 2008). Total power represents the magnitude of oscil-lations irrespective of the phase angles. Evoked power represents thesignal intensity of the event-related potential (ERP) and is obtained byaveraging the time-frequency data of each segment locked to the onsetof the eliciting event. Phase-locking factor is a measure of the degree ofconsistency of the phase angle of the oscillations at each frequencyelicited by the event onset across trials (Roach and Mathalon, 2008).For each individual, the time-frequency plots obtained for each segmentwere averaged across each condition. Measures of the various regions ofinterest (ROI) were obtained from these plots. Grand averaged time-frequency plots were then obtained for each experimental condition byaveraging across individuals as shown in Figs. 2, 3, and 4. In thesefigures, the Y-axis represents the frequency ranges from 0–80 Hz andthe X-axis represents time in milliseconds. For all conditions except themotor-only condition, time 0 represents the onset of the auditory sti-mulus. For the motor-only condition, time 0 represents the tap. Sincethe auditory stimulus was presented every 800ms, three events are seenin the figures below. Due to distortion of the epoch edges caused by thesliding filter, these plots were trimmed from±1800ms to± 1200msto eliminate the distortions. The intensity scale depicts greater activityin warm colors (red-yellow).

For descriptive purposes only, paired comparisons of data used togenerate each grand averaged time-frequency plot were calculatedusing either independent-samples t-tests (for between-group compar-isons) or paired-samples t-tests (for within group comparisons). Thealpha value was set at .05 for each t-test conducted. However, to pro-vide partial control for Type I error inflation, at least eight consecutivesignificant comparisons around the target value were required before aspecific value was portrayed on the graph. This value was chosen as itprovided the best visual representation of the differences between theconditions that were to be evaluated statistically using ROI measures.Boxes C, F, G, and H in the figures represent the t-maps. The site C3 waschosen due to its location over the primary motor cortex of the lefthemisphere as well as its proximity to the primary auditory cortex.Additionally, since all participants used their right hand to make themotor response, the site C3 represents motor activity during the fingertap. Broadband topographical maps representing frequency ranges from0 to 80 Hz for both groups were used to confirm that the site C3 wasappropriate for use (see Fig. 1).

2.6. Behavioral data analysis

Inter-response intervals (IRI) were obtained during the motor-onlycondition to ensure that participants were performing the task cor-rectly. IRIs were also obtained for the combined conditions for bothgroups. Synchronization error (SE) – the time lag between the motorresponse and the auditory tones, was obtained by averaging responsesacross all individual trials within each participant, and then averagedacross all participants to obtain group mean data (Thaut et al., 1998a,b).

2.7. Statistical analysis

Synchronization error for the combined condition for the auditory-first and motor-first groups were compared using independent samples t-tests. Analysis of the time-frequency data was based on our literaturereview to identify principle ROIs. Five ROIs were identified and thedifferences across conditions were statistically compared using Analysisof Variance (ANOVA) procedures. For evoked power, two ROI's wereidentified, 1) 20–45 Hz from 10 to 60ms which has been associatedwith motor activation, and 2) 4–20 Hz from 60 to 250ms which hasbeen associated with sensorimotor processing in the frontoparietal

networks. For phase locking, the same two ROI's were identified: 1)20–45 Hz from 10 to 60ms, and 2) 4–20 Hz from 60 to 250ms. One ROIwas identified for total power in the low beta band region, 13–16 Hz,from 50 to 150ms which has been associated with anticipation andpredictive timing. To answer the first two research questions, data fromeach ROI was analyzed using a two-factor ANOVA design, one between,and one within factor. The two levels of the between factor, “Group”,were the auditory-first and the motor-first groups. The levels of the withinfactor, Condition, were the two task conditions: auditory-only or motor-only task and the auditory-motor combined task.

To answer the third research question, comparisons of the differ-ences in the oscillatory activity between the auditory-first group and themotor-first group during the combined auditory-motor task were eval-uated for each of each of the 5 ROIs using univariate ANOVA procedurewith a single covariate. In each ANOVA the covariate was the corre-sponding ROI measure from the first task which served as a control forthe individual differences.

To maintain a family-wise alpha level of .05, the testwise alphavalue was set at .005 using the Bonferroni correction (.05/10) for theten ANOVAs. After conducting the two-factor ANOVAs, a posterioriTukey's honestly significant difference (HSD) post-hoc t-tests werecalculated to examine differences between cell means when significantinteraction effects were found (Kirk, 1968, pp. 265–269). All statisticalanalyses were performed using the Statistical Package for Social Sci-ences (SPSS), version 24.0.

3. Results

3.1. Motor performance

Participants were asked to self-pace their motor responses to around800ms for 400 trials during the motor-only condition. The mean of theIRI (SD) for the tapping for the motor-only condition was 819.22(222.13) ms. The synchronization error for the combined condition forthe motor-first group had a negative mean asynchrony of − 52.53(136.23), and the IRI for the tapping had a mean (SD) of 790.68(37.97). Similarly, the synchronization error for the combined condi-tion, for the auditory-first group demonstrated the well-documentednegative mean asynchrony of the tapping slightly prior to the onset ofthe beat, with a mean (SD) of − 55.86 (84.10), and the IRI for thetapping had a mean (SD) of 779.24 (24.91). There were no significantgroup differences on the synchronization error (t (38) = .093, p= .93)or the IRI (t (38) = 1.098, p= .28).

3.2. Neural oscillations involved in processing auditory-only and motor-onlyevents

To answer our first research question, “How do the neural oscilla-tions for an auditory-only condition differ from a motor-only conditionand are they qualitatively and quantitatively different?” we examinedthe evoked power and phase synchrony observed for each condition andthen report the statistical outcomes for differences in ROIs across theconditions.

3.2.1. Evoked powerVisual inspection of the time-frequency plot depicting evoked power

indicates differences in brain activity for the auditory-only condition(Fig. 2-A) compared to the motor-only condition (Fig. 2-B). Fig. 2-Adepicts the evoked power of the participants’ brain activity to thepresentation of the rhythmic tones and reveals evoked power primarilyin the theta, alpha, and beta bands (4 – 30 Hz) from tone onset to ap-proximately 500ms with high gamma band activity (45–80 Hz) evidentthroughout the trials. Fig. 2-B, shows that the evoked power of theparticipants’ self-paced motor responses is obtained primarily in thehigh beta and low gamma bands (20–45 Hz) from 10ms to approxi-mately 60ms and then in theta, alpha and low beta bands (4–20 Hz)

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

105

from 60 to 250ms when time-locked to the motor response.A comparison of these two control conditions is shown in Fig. 2-C.

For the ROI related to motor activation, this figure reveals lower evokedpower for high beta and low gamma band oscillations (i.e., 20–45 Hz)between 10 and 60ms during auditory-only condition compared to themotor-only condition with the differences depicted in blue. The ANOVAevaluating the averaged evoked power for this ROI revealed a sig-nificant Group-by-Condition interaction; F (1, 36) = 15.34, p < .0005,ŋ2p = .30. The posteriori test indicated significantly lower activity in thisregion for the auditory-only condition (M = .12 μV2, SD = .06) com-pared to the motor-only condition (M = .22 μV2, SD = .08); q (1, 19)

=− 7.42, p < .01. Fig. 2-C also reveals that for the ROI associatedwith sensorimotor processing, greater evoked power was found in thedelta, alpha and low beta bands (i.e., 4–20 Hz) between 60 and 250msin the auditory-only condition compared to the motor-only condition(depicted as yellow-red). There was also a significant Group-by-Con-dition interaction effect for this ROI, F (1, 36) = 54.53, p < .0005, ŋ2p= .60. The posteriori test revealed significantly greater activity in thisregion for the auditory-only condition (M = .53 μV2, SD = .21) com-pared to the motor-only condition (M = .31 μV2, SD = .17); q (1, 19)

= 5.80, p < .01.

3.2.2. Phase locking factorVisual inspection of the time-frequency plot depicting phase locking

values also indicates differences in brain activity for the auditory-onlycondition (Fig. 3-A) compared to the motor-only condition (Fig. 3-B).Fig. 3-A depicts the degree of phase locking across trials to the pre-sentation of the periodic tones and shows synchrony of oscillations inthe high beta and low gamma bands (20–45 Hz) shortly after tone onsetto about 150ms. Even greater synchrony is shown for the theta, alpha

and low beta bands (4–20 Hz) from about 60–350ms. Low synchronyvalues are also visible in the high gamma band (45–80 Hz). Fig. 3-Bshows that the synchrony of brain activity of the participants acrosstrials during the self-paced motor responses was brief and of low levelsin the high beta and low gamma bands (20–45 Hz) shortly after the tapto around 60ms. Low levels of synchrony were also present in the theta,alpha, and low beta bands (4–20 Hz) from 0 to 250ms.

A comparison of phase locking values obtained for these two controlconditions is shown in Fig. 3-C. For the ROI related to motor activation,this figure reveals lower phase locking values in the high beta and lowgamma bands (i.e., 20–45 Hz) during the auditory-only conditioncompared to the motor-only condition between 10 and 60ms (depictedin blue). The ANOVA evaluating the averaged phase locking values forthis ROI revealed a significant Group-by-Condition interaction; F (1, 36)

= 19.07, p < .0005, ŋ2p = .35. The posteriori test indicated sig-nificantly lower phase values in this region for the auditory-only con-dition (M= .06 μV2, SD = .017) compared to the motor-only condition(M = .11 μV2, SD = .040); q (1, 19) =− 9.26, p < .01. Fig. 3-C alsoreveals that for the region associated with sensorimotor processing,greater phase locking values were found in the delta, alpha and lowbeta bands (i.e., 4− 20 Hz) between 60 and 250ms in the auditory-only condition compared to the motor-only condition (depicted asyellow-red). There was a significant Group-by-Condition interactioneffect for this ROI, F (1, 36) = 53.68, p < .0005, ŋ2p = .60. The posterioritest revealed significantly greater phase synchrony in this region for theauditory-only condition (M = .17 μV2, SD = .06) compared to themotor-only condition (M = .10 μV2, SD = .04); q (1, 19) = 7.08,p < .01.

Fig. 1. Broadband topographical maps representing a frequency range from 0 to 80 Hz for the different experimental conditions. Figure inlet shows Site C3.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

106

3.3. Neural oscillations during the auditory-motor combined condition

To answer our second research question, “What neural oscillationsrepresent the synchrony of auditory and motor systems when bothsystems are active together during entrainment?” we examined theevoked power, phase synchrony, and total power of the auditory-motorcombined condition and made within-subject statistical comparisons tothe measures obtained from the motor-only or auditory-only conditions,respectively.

3.3.1. Evoked powerFigs. 2-D and 2-E, representing the combined condition for the au-

ditory-first group and the motor-first group respectively, show evokedpower primarily in the theta, alpha and low beta bands (4–20 Hz) be-tween 60 and 250ms with the greatest power in the theta band(4–7 Hz) around 200ms. A comparison of the evoked power for thecombined condition (Fig. 2-E) to that of the motor-only condition of themotor-first group (Fig. 2-B) is shown in Fig. 2-H.

To statistically compare the evoked power for the combined con-dition (Fig. 2-E) to that of the motor-only condition of the motor-first

group (Fig. 2-B) follow-up posteriori t-tests were conducted for each ofthe two ROIs for evoked power based on the significant Group-by-Condition interactions reported above. In the ROI related to motoractivation, when the rhythmic tone is added to paced tapping, theevoked power in the high beta and low gamma bands (20 – 45 Hz)between 10 and 60ms is significantly less in the combined condition (M= .10 μV2, SD= .034) compared to the motor-only condition (M= .22μV2, SD = .085); q (1, 19) = 8.77, p < .01. But as expected, in the ROIassociated with sensorimotor processing, the evoked power in the delta,alpha and low beta bands (i.e., 4– 20 Hz) between 60 and 250ms issignificantly greater in the combined condition (M = .45 μV2, SD =.112) compared to the motor-only condition (M= .31 μV2, SD= .17); q(1, 19) =−3.84, p < .05.

A similar pattern of oscillations for auditory and motor activity wasexpected to be seen in Fig. 2-G which depicts the differences betweenthe auditory-only condition and the combined condition for the audi-tory-first group. However, despite the additional requirement of motoractivation for tapping along with the tone presentations, follow-upposteriori t-test revealed that the evoked power in the ROI related tomotor activation (i.e., 20 – 45 Hz between 10 and 60ms) is not

Fig. 2. Evoked power and tmaps for the four conditions at site C3. The X-axis represents time in milliseconds. The Y-axis represents frequency ranges from 0 to 80 Hz.For all conditions except the motor-only condition (Plot B), time 0 represents the auditory stimulus. For the motor-only condition, time 0 represents the finger tap.The color bar for plot A, B, D & E represents the amplitude of the oscillations depicting greater activity in warm colors. The color bar for plot C, F, G 7 H represents theT values. A. Auditory-only condition. B. Motor-only condition. C. T map comparing auditory-only and motor-only. D. Combined condition for the auditory-first group.E. Combined condition for the motor-first group. F. T map comparing combined condition for the auditory-first group with the motor-first group. G. T maps comparingauditory-only condition to combined condition for the auditory-first group. H. T map comparing motor-only condition to combined condition for the motor-first group.The boxes represent two ROIs, namely 1) 20–45 Hz from 10 to 60ms, and 2) 4–20 Hz from 60 to 250ms.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

107

significantly different in the combined condition (M = .10 μV2, SD =.045) when compared to the auditory-only condition (M = .12 μV2, SD= .064); q (1, 19) = 1.41. However, the follow-up posteriori t-test for theROI associated with sensorimotor processing (i.e., 4–20 Hz between 60and 250ms) revealed that the evoked power is significantly less incombined condition (M = .37 μV2, SD = .161) compared to the au-ditory-only condition (M = .53 μV2, SD = .208); q (1, 19) = 4.19,p < .01.

3.3.2. Phase Locking FactorFig. 3-H compares the phase locking values for the auditory-motor

combined condition (Fig. 3-E) to the motor-only condition (Fig. 3-B) forthe motor-first group. In keeping with the findings for evoked power,when the rhythmic tone is added to paced tapping, the follow-up pos-teriori t-test revealed that the phase synchronization for the ROI relatedto motor activation (20 – 45 Hz between 10 and 60ms) is significantlyless for the combined condition (M = .05, SD = .012) compared to thephase synchronization of the motor-only condition (M = .11, SD =.040); q (1, 19) = 10.13, p < .01. The follow-up posteriori t-test alsorevealed that mean phase locking value in the ROI associated withsensorimotor processing (i.e., 4–20 Hz between 60 and 250ms) is

significantly greater in the combined condition (M = .15, SD = .029)compared to the motor-only condition (M = .10, SD = .036); q (1, 19)

=−5.09, p < .01. Thus, the addition of the rhythmic tones to fingertapping leads to greater synchrony of the oscillations than associatedwith self-paced rhythmic tapping in this region as would be expected.

Fig. 3-G shows the comparison of the combined condition (Fig. 3-D)to the auditory-only condition (Fig. 3-A) for the auditory-first group. Thefollow-up posteriori t-tests revealed that despite the additional require-ment of motor activation for tapping along with the tone presentations,the phase synchronization in ROI related to motor activation (i.e., 20 –45 Hz between 10 and 60ms) was not significantly different in thecombined condition (M = .05, SD = .015) when compared to the au-ditory-only condition (M = .06, SD = .017); q (1, 19) = 1.45. But in theROI associated with sensorimotor processing (i.e., 4− 20 Hz between60 and 250ms), the follow-up posteriori t-test revealed that the phasesynchronization is significantly less in combined condition (M = .13,SD = .048) compared to the auditory-only condition (M = .17, SD =.058); q (1, 19) = 3.69, p < .05.

3.3.3. Total powerFigs. 4-D and 4-E, representing the combined condition for the

Fig. 3. Phase locking and t maps for the four conditions at site C3. The X-axis depicts time in milliseconds. The Y-axis depicts the frequency ranges from 0 to 80 Hz.For all conditions except the motor-only condition (Plot B), time 0 represents the auditory stimulus. For the motor-only condition, time 0 represents the finger tap.The color bar for plot A, B, D & E represents the amplitude of the oscillations depicting greater activity in warm colors. The color bar for plot C, F, G 7 H represents theT values. A. Auditory-only condition. B. Motor-only condition. C. T map comparing auditory-only and motor-only. D. Combined condition for the auditory-first group.E. Combined condition for the motor-first group. F. T map comparing combined condition for the auditory-first group with the motor-first group. G. T maps comparingauditory-only condition to combined condition for the auditory-first group. H. T map comparing motor-only condition to combined condition for the motor-first group.The boxes represent two ROIs, namely 1) 20–45 Hz from 10 to 60ms, and 2) 4–20 Hz from 60 to 250ms.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

108

auditory-first group and the motor-first group respectively, show that totalpower is concentrated in the lower oscillation bands below 30 Hzthroughout the inter-trial interval with greatest power in the low betaband (13–16 Hz) between 50 and 150ms. For this ROI associated withanticipation and predictive timing, the ANOVA evaluating the averagedtotal power also revealed a significant Group-by-Condition interaction;F(1, 36) = 34.88, p < .0005, ŋ2p = .49.

Comparisons of the total power for the combined condition (Fig. 4-E) to that of the motor-only condition of the motor-first group (Fig. 4-B)is shown in Fig. 4-H. The posteriori test revealed that the brain activityin this ROI is slightly greater but not significantly so for the combinedcondition (M = 2.30 μV2, SD = .62) compared to the motor-onlycondition (M = 2.23 μV2, SD = .54); q (1, 19) =−.59, p > .05.Comparisons of the total power for the combined condition (Fig. 4-D) tothat of the auditory-only condition of the auditory-first group (Fig. 4-A)is shown in Fig. 4-G. The posteriori test revealed significantly less ac-tivity in this ROI for the combined condition (M = 2.08 μV2, SD = .40)compared to the auditory-only condition of the auditory-first group (M= 2.56 μV2, SD = .63); q (1, 19) = 3.79, p < .05.

3.4. Prior exposure leads to differences in neural synchronization duringauditory-motor entrainment

To answer the third research question, “Does prior exposure to anauditory-only entrainment condition (auditory priming) versus amotor-only condition (motor priming) result in differences in neuralsynchronization in an auditory-motor entrainment condition?” statis-tical comparisons of the combined condition of the motor-first group tothe combined condition the auditory-first group were made using uni-variate ANOVAs with the measure from the first task serving as thecovariate. Each of the ROIs from the evoked power, phase synchronyand total power from the time-frequency analyses were evaluated.

For the two ROIs associated with motor activation (i.e., 20–45 Hzbetween 10 and 60ms), statistically significant differences were notfound between the combined condition of the auditory-first group andthe combined condition of the motor-first group. Specifically, compar-isons of evoked power revealed that the combined condition of theauditory-first group (M = .10 μV2, SD = .045) was not significantlydifferent from that of the motor-first group (M = .10 μV2, SD = .034);i.e., the Group effect was not significant, F (1, 35) = .19, p= .67, ŋ2p

Fig. 4. Total power and tmaps for the four conditions at site C3. The X-axis depicts time in milliseconds. The Y-axis depicts the frequency ranges from 0 to 80 Hz. Forall conditions except the motor-only condition (Plot B), time 0 represents the auditory stimulus. For the motor-only condition, time 0 represents the finger tap. Thecolor bar for plot A, B, D & E represents the amplitude of the oscillations depicting greater activity in warm colors. The color bar for plot C, F, G 7 H represents the Tvalues. A. Auditory-only condition. B. Motor-only condition. C. T map comparing auditory-only and motor-only. D. Combined condition for the auditory-first group. E.Combined condition for the motor-first group. F. T map comparing combined condition for the auditory-first group with the motor-first group. G. T maps comparingauditory-only condition to combined condition for the auditory-first group. H. T map comparing motor-only condition to combined condition for the motor-first group.The boxes represent an ROI at 13–16 Hz from 50 to 150ms.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

109

= .005. And the comparison of phase synchrony revealed that thecombined condition of the auditory-first group (M = .05 μV2, SD =.015) was not significantly different from that of the motor-first group (M= .05 μV2, SD= .012); i.e., the Group effect was not significant, F (1, 35)

= .24, p= .63, ŋ2p = .007.In contrast, for the two ROIs associated with sensorimotor proces-

sing (i.e., 4− 20 Hz between 60 and 250ms), the mean ROI values forthe combined condition of the auditory-first group were significantlylower than for the combined condition of the motor-first group.Comparisons of evoked power revealed that the combined condition ofthe auditory-first group (M = .37 μV2, SD = .16) was significantly lowerthan that of the motor-first group (M = .45 μV2, SD = .11); i.e., a sig-nificant Group effect was found, F (1, 35) = 32.37, p < .0005, ŋ2p = .48.The comparison of phase synchrony also revealed that the combinedcondition of the auditory-first group (M = .13, SD = .048) was sig-nificantly less than that of the motor-first group (M = .15, SD = .029);i.e., a significant Group effect was found, F (1, 35) = 20.19, p < .0005,ŋ2p = .37.

In addition, the comparison of total power in the low beta bandregion for the one ROI associated with anticipation and predictivetiming (i.e., 13 – 16 Hz, from 50 to 150ms) revealed that the combinedcondition of the auditory-first group (M = 2.08 μV2, SD = .40) wassignificantly lower than the combined condition of the motor-first group(M = 2.30 μV2, SD = .62); i.e., a significant Group effect was found, F(1, 35) = 29.96, p < .0005, ŋ2p = .46.

4. Discussion

The purpose of this study was to examine the neural oscillationsunderlying auditory-motor entrainment using EEG. Specifically, weexamined the oscillations for auditory-only and motor-only conditionsand compared those results to the oscillations in a combined auditory-motor synchronization condition. Finally, we examined priming effectsof auditory-only and motor-only conditions on combined auditory-motor synchronization. To empirically examine the influence of audi-tory stimuli on motor performance, two groups of participants wereengaged in a condition requiring tapping to rhythmic auditory beats,but the groups differed in regard to the control condition. One groupparticipated in the auditory-only control condition, which involvedlistening to 400 trials of auditory tones presented every 800ms, and asecond group participated in a motor-only control condition, whichinvolved finger tapping every 800ms without any external stimuli.

The first question of this study involved comparing the neural os-cillations for an auditory-only and a motor-only condition. Time-fre-quency analysis of the auditory-only condition revealed activity in thetheta, alpha, and low beta bands to the presentation of the auditorytone (See the A plots in all time-frequency figures). Additionally, duringthe auditory-only condition, continuous activity around 10 Hz oscilla-tions is seen in the plot of total power in Fig. 4-A indicating the pre-sence of an entrained alpha rhythm. Some of the observed power isevoked shortly after the onset of the auditory stimuli (See Figs. 2-A and3-A) but this alpha oscillation is also present in the induced power (plotnot included in the article). Several researchers examining visual sti-muli have found similar results of alpha entrainment (De Graaf et al.,2013; Spaak et al., 2014). Time-frequency plots for the motor-onlycondition showed oscillatory activity in theta, alpha, and beta bandscorresponding to the motor response (See the B plots in all time-fre-quency figures). The evoked power in beta and gamma bands for themotor-only condition was found to be greater than the auditory-onlycondition and the evoked delta, alpha, and low beta activity was foundto be reduced compared to the auditory-only condition. These findingsare consistent with research examining the role of the beta band (Romeiet al., 2016). Activity in the beta band has been associated with thecortical control of motor output (Engel and Fries, 2010). Thus, an-swering the first question, the neural oscillations for an auditory-onlycondition compared to a motor-only condition are qualitatively and

quantitatively different.The second question related to determining which neural oscilla-

tions represent the synchrony of auditory and motor systems when bothsystems are active together during entrainment. This was examinedusing within-subject statistical comparisons comparing the auditory-motor combined condition of each group to its control condition,namely the auditory-only or the motor-only condition. For the motor-first group, the addition of the auditory stimuli to the motor responseresulted in an increase in phase synchrony and evoked power in thetheta and alpha and low beta (4 – 20 Hz) frequencies in the combinedcondition compared to the motor-only condition. This result was ex-pected based on the frequencies observed in the auditory-only condi-tion. In contrast, the auditory-first group displayed less evoked powerand phase synchrony in the delta, alpha and low beta (4 – 20 Hz) in thecombined condition compared to the auditory-only condition. Thus,these results demonstrate the neural oscillations in the delta, alpha, andlow beta 60 – 250ms post-stimulus are characteristic of entrainment.

Our findings regarding which neural oscillations reflect the syn-chrony of auditory and motor systems are similar to other reports in theliterature related to beta and gamma oscillations. Several researchershave hypothesized that beta oscillations are associated with timing andpredictive sensory processing (Arnal and Giraud, 2012; Leventhal et al.,2012). A study examining passive listening to a regular musical beatwith the occasional omission of single tones showed that beta activity(~ 20 Hz) decreased after each tone, followed by an increase, resultingin a periodic modulation synchronized to the auditory stimulus (Fujiokaet al., 2009). The beta decrease was absent after omissions. Ad-ditionally, gamma activity (~ 40 Hz) showed a peak after tone andomission. The decrease in beta and increase in gamma activity fol-lowing the presentation of a tone have led researchers to believe thatthese neural frequencies have different roles, with gamma involved inauditory stimulus encoding and beta involved in auditory-motor in-teractions (Fujioka et al., 2009).

In our study, delta, alpha and beta oscillations (4–20 Hz) showed asignificant increase in evoked power and synchrony when auditorystimuli were added to motor tapping, as seen in the combined conditionfor the motor-first group (See E plots in Figs. 2 and 3). However, for theauditory-first group, there was a significant decrease in the delta, alphaand low beta power and synchrony when adding tapping to the audi-tory stimuli (See D plots in Figs. 2 and 3). These findings suggest thatauditory stimuli drive delta, alpha, and low beta oscillations and arelikely to be instrumental in the auditory-motor interactions as seen inthe D plots of Figs. 2 and 3. Unique to this study based on the studydesign, our results directly demonstrate increased gamma (45–80 Hz)activity and synchrony following the presentation of an auditory sti-mulus with the highest level of gamma activity and synchrony seen inauditory-only (See A plots in Figs. 2–4) compared to all other condi-tions. Visual inspection of the time-frequency plots of the combinedconditions (See the D and E plots in Figs. 2–4) indicate that the auditory-first group had greater gamma activity and synchrony compared to themotor-first group. In addition, a significant increase in gamma activityand synchrony occurred in the motor-first group for the combined con-dition compared to the motor-only condition. Collectively, the results ofthis study demonstrate that the auditory stimuli produce neural oscil-lations that drive the auditory-motor entrainment.

Related to the third question about prior exposure influencing au-ditory-motor entrainment, time-frequency analysis of the combinedcondition for the two groups indicated a robust priming effect.Specifically, the group that had participated in the auditory-only con-dition first, had less neural synchronization of delta, alpha and low betafrequencies (4 – 20 Hz), and reduced evoked power in these same fre-quencies and reduced total power in low beta (13 – 16 Hz) bandscompared to the group that had participated in the motor-only condi-tion. Thus, the auditory-only condition in the auditory-first group ap-pears to have primed the neural system such that less power (i.e., neuralactivity) was required in the combined condition compared to the

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

110

motor-first group that did not hear the auditory beats prior to the com-bined condition.

As shown in our study using time-frequency analysis of the evokedand total power, a reduced neural response occurred during the com-bined condition for the auditory-first group compared to the auditory-only condition. Studies using functional magnetic resonance imaging(fMRI) in humans and single cell recording in non-human primatesprovide some insight into the underlying mechanisms in the reductionof brain activity following priming. Using fMRI techniques in humans,Bergerbest and colleagues (2004) found that repeating an auditorystimulus (i.e., repetition priming or suppression) resulted in reducedactivation of the auditory cortex and brain regions other than the au-ditory cortex suggesting a network underlying the priming effect.Researchers have also found decreased hemodynamic responses usingfunctional imaging for primed versus unprimed auditory stimuli withreference to memory conditions (Bergerbest et al., 2004; Henson, 2003;Schacter and Buckner, 1998). This reduced hemodynamic response isassumed to reflect faster or more efficient processing of the primedstimulus (Henson, 2003). In non-human primate studies, it has beenshown that when a stimulus is repeated, neurons show a decreasedfiring rate even after a single exposure to the stimulus (for reviews seeBrown and Xiang, 1998; Desimone, 1996). This repetition suppression,an intrinsic property of cortical neurons, has been hypothesized tooccur due to the sharpening of neuronal connections, such that ex-perience leads to both a smaller representation and a more selectiverepresentation in the sensory cortex (Desimone, 1996; Wiggs andMartin, 1998). This sharpened representation would, in turn, lead to abetter neural synchronization of primed stimuli.

While previous work using functional imaging or behavioral mea-sures have shown evidence of priming, our study has two novel find-ings. Firstly, this study demonstrated that the auditory system drivesthe motor system through the process of entrainment. Specifically, dueto the temporal resolution of EEG and the time-frequency analysis, ourresults demonstrated sharpening of neural signals in oscillations asso-ciated with sensorimotor processing (i.e., 4–20 Hz) during the com-bined auditory-motor condition as compared to the motor-only condi-tion for the motor-first group. Thus, we have shown more synchronizedbrain activity during auditory-motor entrainment (external clock)compared to unpaced tapping (internal clock). Secondly, the results ofthis study have indicated that auditory priming requires less neuralactivity during an auditory-motor combined task compared to motorpriming.

5. Conclusion

The results of this study demonstrate that neural oscillations arequalitatively and quantitatively different for auditory-only conditionscompared to motor-only conditions. Furthermore, this study suggeststhat auditory stimuli produce neural oscillations that drive auditory-motor entrainment. Finally, even brief periods of rhythmic priming ofthe auditory system facilitates the neural efficiency of the motor systemduring the process of entrainment. This is the first study to elucidateneural entrainment mechanisms underlying the clinical applications ofrhythmic auditory stimulation to mobility training in motor re-habilitation. The understanding and validity of a clinical applicationwithin medicine and rehabilitation are greatly enhanced when under-lying neural mechanisms are known and can be demonstrated. In thecase of this research, the unique findings will be of great importance forclinical applications. The data regarding auditory priming stronglysuggest that before a patient is asked to move to the rhythmic auditorystimulus a ‘brain priming’ period of listening may be beneficial inpreparing for movement readiness and thus result in better motor ex-ecution by the patient, and hence better rehabilitation results.

Acknowledgments

We thank Donald Rojas, Ph.D., and Peter Teale, M.S.E.E for theMATLAB code used to conduct the time-frequency analyses.

Funding

This study was funded in part by the Academy of Neurologic MusicTherapy awarded to MHT and by an NSF grant (IIS-1065513) awardedto CWA, PLD, and WJG.

Data statement

The raw EEG data and e-prime code will be made available to in-terested researchers. To access the data the researchers should directlycontact the lab PI – patricia.davies@colostate.edu.

Conflicts of interest

None.

References

Addison, P.S., 2016. The Illustrated Wavelet Transform Handbook: Introductory Theoryand Applications in Science, Engineering, Medicine and Finance, second ed. CRCPress, Boca Raton, FL.

Arnal, L.H., Giraud, A.L., 2012. Cortical oscillations and sensory predictions. TrendsCogn. Sci. 16 (7), 390–398. http://dx.doi.org/10.1016/j.tics.2012.05.003.

Arnal, L.H., Doelling, K.B., Poeppel, D., 2014. Delta–beta coupled oscillations underlietemporal prediction accuracy. Cereb. Cortex. http://dx.doi.org/10.1093/cercor/bhu103. (bhu103).

Arnal, L.H., Wyart, V., Giraud, A.L., 2011. Transitions in neural oscillations reflect pre-diction errors generated in audiovisual speech. Nat. Neurosci. 14 (6), 797–801.

Bendor, D., Wang, X., 2007. Differential neural coding of acoustic flutter within primateauditory cortex. Nat. Neurosci. 10 (6), 763–771. http://dx.doi.org/10.1038/nn.2810.

Bergerbest, D., Ghahremani, D.G., Gabrieli, J.D., 2004. Neural correlates of auditory re-petition priming: reduced fMRI activation in the auditory cortex. J. Cogn. Neurosci.16 (6), 966–977.

Brown, M.W., Xiang, J.Z., 1998. Recognition memory: neuronal substrates of the judge-ment of prior occurrence. Prog. Neurobiol. 55 (2), 149–189. http://dx.doi.org/10.1016/S0301-0082(98)00002-1.

Cavanagh, J.F., Frank, M.J., 2014. Frontal theta as a mechanism for cognitive control.Trends Cogn. Sci. 18 (8), 414–421. http://dx.doi.org/10.1016/j.tics.2014.04.012.

De Graaf, T.A., Gross, J., Paterson, G., Rusch, T., Sack, A.T., Thut, G., 2013. Alpha-bandrhythms in visual task performance: phase-locking by rhythmic sensory stimulation.PLoS One 8 (3), e60035. http://dx.doi.org/10.1371/journal.pone.0060035.

Debener, S., Herrmann, C.S., Kranczioch, C., Gembris, D., Engel, A.K., 2003. Top-downattentional processing enhances auditory evoked gamma band activity. NeuroReport14 (5), 683–686. http://dx.doi.org/10.1097/01.wnr.0000064987.96259.5c.

Desimone, R., 1996. Neural mechanisms for visual memory and their role in attention.Proc. Natl. Acad. Sci. USA 93 (24), 13494–13499.

Eggermont, J.J., 2001. Between sound and perception: reviewing the search for a neuralcode. Hear. Res. 157 (1), 1–42. http://dx.doi.org/10.1016/S0378-5955(01)00259-3.

Engel, A.K., Fries, P., 2010. Beta-band oscillations—signaling the status quo? Curr. Opin.Neurobiol. 20 (2), 156–165. http://dx.doi.org/10.1016/j.conb.2010.02.015.

Fraisse, P., 1982. Rhythm and tempo. In: Deutsch, D. (Ed.), The Psychology of Music.Academic Press, Orlando, FL, pp. 149–180.

Fujioka, T., Trainor, L.J., Large, E.W., Ross, B., 2009. Beta and gamma rhythms in humanauditory cortex during musical beat processing. Ann. N. Y. Acad. Sci. 1169 (1),89–92. http://dx.doi.org/10.1111/j.1749-6632.2009.04779.x.

Giraud, A.L., Poeppel, D., 2012. Cortical oscillations and speech processing: emergingcomputational principles and operations. Nat. Neurosci. 15 (4), 511–517. http://dx.doi.org/10.1038/nn.3063.

Grahn, J.A., Henry, M.J., McAuley, J.D., 2011. FMRI investigation of cross-modal inter-actions in beat perception: audition primes vision, but not vice versa. NeuroImage 54(2), 1231–1243. http://dx.doi.org/10.1016/j.neuroimage.2010.09.033.

Güntekin, B., Başar, E., 2016. Review of evoked and event-related delta responses in thehuman brain. Int. J. Psychophysiol. 103, 43–52. http://dx.doi.org/10.1016/j.ijpsycho.2015.02.001.

Henson, R.N.A., 2003. Neuroimaging studies of priming. Prog. Neurobiol. 70 (1), 53–81.http://dx.doi.org/10.1016/S0301-0082(03)00086-8.

Jantzen, K.J., Oullier, O., Marshall, M.L., Steinberg, F.L., Kelso, J.A.S., 2007. A parametricfMRI investigation of context effects in sensorimotor timing and coordination.Neuropsychologia 45 (4), 673–684. http://dx.doi.org/10.1016/j.neuropsychologia.2006.07.020.

Keitel, C., Quigley, C., Ruhnau, P., 2014. Stimulus-driven brain oscillations in the alpharange: entrainment of intrinsic rhythms or frequency-following response? J.Neurosci. 34 (31), 10137–10140. http://dx.doi.org/10.1523/JNEUROSCI.1904-14.2014.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

111

Kirk, R.E., 1968. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole Publications, Monterey, CA.

Krause, V., Pollok, B., Schnitzler, A., 2010. Perception in action: the impact of sensoryinformation on sensorimotor synchronization in musicians and non-musicians. ActaPsychol. 133 (1), 28–37. http://dx.doi.org/10.1016/j.actpsy.2009.08.003.

Lachaux, J.P., Rodriguez, E., Martinerie, J., Varela, F.J., 1999. Measuring phase syn-chrony in brain signals. Human. Brain Mapp. 8 (4), 194–208. http://dx.doi.org/10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C.

Lakatos, P., Karmos, G., Mehta, A.D., Ulbert, I., Schroeder, C.E., 2008. Entrainment ofneuronal oscillations as a mechanism of attentional selection. Science 320 (5872),110–113. http://dx.doi.org/10.1126/science.1154735.

Lakatos, P., Musacchia, G., O’Connel, M.N., Falchier, A.Y., Javitt, D.C., Schroeder, C.E.,2013. The spectrotemporal filter mechanism of auditory selective attention. Neuron77 (4), 750–761. http://dx.doi.org/10.1016/j.neuron.2012.11.034.

Large, E.W., Snyder, J.S., 2009. Pulse and meter as neural resonance. Ann. N.Y. Acad. Sci.1169 (1), 46–57. http://dx.doi.org/10.1111/j.1749-6632.2009.04550.x.

Leventhal, D.K., Gage, G.J., Schmidt, R., Pettibone, J.R., Case, A.C., Berke, J.D., 2012.Basal ganglia beta oscillations accompany cue utilization. Neuron 73 (3), 523–536.http://dx.doi.org/10.1016/j.neuron.2011.11.032.

Molinari, M., Leggio, M.G., Thaut, M.H., 2007. The cerebellum and neural networks forrhythmic sensorimotor synchronization in the human brain. Cerebellum 6 (1), 18–23.http://dx.doi.org/10.1080/14734220601142886.

Molinari, M., Leggio, M.G., Martin, M., Cerasa, A., Thaut, M., 2003. Neurobiology ofrhythmic motor entrainment. Ann. N. Y. Acad. Sci. 999 (1), 313–321. http://dx.doi.org/10.1196/annals.1284.042.

Peelle, J.E., Davis, M.H., 2012. Neural oscillations carry speech rhythm through tocomprehension. Front. Psychol. 3, 320. http://dx.doi.org/10.3389/fpsyg.2012.00320.

Pineda, J.A., 2005. The functional significance of mu rhythms: translating “seeing” and“hearing” into “doing”. Brain Res. Rev. 50 (1), 57–68. http://dx.doi.org/10.1016/j.brainresrev.2005.04.005.

Pressing, J., 1999. The referential dynamics of cognition and action. Psychol. Rev. 106(4), 714–747. http://dx.doi.org/10.1037/0033-295X.106.4.714.

Repp, B.H., 2005. Sensorimotor synchronization: a review of the tapping literature.Psychon. Bull. Rev. 12 (6), 969–992. http://dx.doi.org/10.3758/BF03206433.

Repp, B.H., Su, Y.H., 2013. Sensorimotor synchronization: a review of recent research(2006–2012). Psychon. Bull. Rev. 20 (3), 403–452. http://dx.doi.org/10.3758/s13423-012-0371-2.

Roach, B.J., Mathalon, D.H., 2008. Event-related EEG time-frequency analysis: an over-view of measures and an analysis of early gamma band phase locking in schizo-phrenia. Schizophr. Bull. 34 (5), 907–926. http://dx.doi.org/10.1093/schbul/sbn093.

Romei, V., Bauer, M., Brooks, J.L., Economides, M., Penny, W., Thut, G., Bestmann, S.,2016. Causal evidence that intrinsic beta-frequency is relevant for enhanced signalpropagation in the motor system as shown through rhythmic TMS. NeuroImage 126,120–130. http://dx.doi.org/10.1016/j.neuroimage.2015.11.020.

Schacter, D.L., Buckner, R.L., 1998. Priming and the brain. Neuron 20 (2), 185–195.http://dx.doi.org/10.1016/S0896-6273(00)80448-1.

Schmidt, R.C., O'Brien, B., 1997. Evaluating the dynamics of unintended interpersonalcoordination. Ecol. Psychol. 9 (3), 189–206. http://dx.doi.org/10.1207/

s15326969eco0903_2.Schroeder, C.E., Lakatos, P., Kajikawa, Y., Partan, S., Puce, A., 2008. Neuronal oscillations

and visual amplification of speech. Trends Cogn. Sci. 12 (3), 106–113. http://dx.doi.org/10.1016/j.tics.2008.01.002.

Schroeder, C.E., Wilson, D.A., Radman, T., Scharfman, H., Lakatos, P., 2010. Dynamics ofactive sensing and perceptual selection. Curr. Opin. Neurobiol. 20 (2), 172–176.http://dx.doi.org/10.1016/j.conb.2010.02.010.

Segalowitz, S.J., 1996. EYEREG.EXE program for epoch-based eye-channel correction ofERPs. Brock University, St. Catharines, Canada.

Serrien, D.J., 2008. The neural dynamics of timed motor tasks: evidence from a syn-chronization–continuation paradigm. Eur. J. Neurosci. 27 (6), 1553–1560. http://dx.doi.org/10.1111/j.1460-9568.2008.06110.x.

Snyder, J.S., Large, E.W., 2005. Gamma-band activity reflects the metric structure ofrhythmic tone sequences. Cogn. Brain Res. 24 (1), 117–126. http://dx.doi.org/10.1016/j.cogbrainres.2004.12.014.

Spaak, E., de Lange, F.P., Jensen, O., 2014. Local entrainment of alpha oscillations byvisual stimuli causes cyclic modulation of perception. J. Neurosci. 34 (10),3536–3544. http://dx.doi.org/10.1523/jneurosci.4385-13.2014.

Thaut, M.H., Abiru, M., 2010. Rhythmic auditory stimulation in rehabilitation of move-ment disorders: a review of current research. Music Percept.: Interdiscip. J. 27 (4),263–269. http://dx.doi.org/10.1525/mp.2010.27.4.263.

Thaut, M.H., McIntosh, G.C., 2014. Neurologic music therapy in stroke rehabilitation.Curr. Phys. Med. Rehabil. Rep. 2 (2), 106–113. http://dx.doi.org/10.1007/s40141-014-0049-y.

Thaut, M.H., McIntosh, G.C., Rice, R.R., Miller, R.A., Rathbun, J., Brault, J.M., 1996.Rhythmic auditory stimulation in gait training for Parkinson's disease patients. Mov.Disord. 11 (2), 193–200. http://dx.doi.org/10.1002/mds.870110213.

Thaut, M.H., Miller, R.A., Schauer, L.M., 1998a. Multiple synchronization strategies inrhythmic sensorimotor tasks: phase vs period correction. Biol. Cybern. 79 (3),241–250. http://dx.doi.org/10.1007/s004220050474.

Thaut, M.H., Stephan, K.M., Wunderlich, G., Schicks, W., Tellmann, L., Herzog, H.,Hömberg, V., 2009. Distinct cortico-cerebellar activations in rhythmic auditorymotor synchronization. Cortex 45 (1), 44–53. http://dx.doi.org/10.1016/j.cortex.2007.09.009.

Thaut, M.H., Tian, B., Azimi-Sadjadi, M.R., 1998b. Rhythmic finger tapping to cosine-wave modulated metronome sequences: evidence of subliminal entrainment. Human.Mov. Sci. 17 (6), 839–863. http://dx.doi.org/10.1016/S0167-9457(98)00031-1.

Thut, G., Miniussi, C., Gross, J., 2012. The functional importance of rhythmic activity inthe brain. Curr. Biol. 22 (16), R658–R663. http://dx.doi.org/10.1016/j.cub.2012.06.061.

Tierney, A., Kraus, N., 2013. The ability to move to a beat is linked to the consistency ofneural responses to sound. J. Neurosci. 33 (38), 14981–14988. http://dx.doi.org/10.1523/JNEUROSCI.0612-13.2013.

Tiitinen, H.T., Sinkkonen, J., Reinikainen, K., Alho, K., Lavikainen, J., Näätänen, R.,1993. Selective attention enhances the auditory 40-Hz transient response in humans.Nature 364 (6432), 59–60. http://dx.doi.org/10.1038/364059a0.

Wiggs, C.L., Martin, A., 1998. Properties and mechanisms of perceptual priming. Curr.Opin. Neurobiol. 8 (2), 227–233. http://dx.doi.org/10.1016/S0959-4388(98)80144-X.

J.E. Crasta et al. Neuropsychologia 117 (2018) 102–112

112