Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Neuromagnetic Beta-Band Oscillations during Motor Imitation in Youth with Autism

Neuromagnetic Beta-Band Oscillations during Motor Imitation in Youth with Autism Hindawi Autism Research and Treatment Volume 2018, Article ID 9035793, 12 pages https://doi.org/10.1155/2018/9035793 Research Article Neuromagnetic Beta-Band Oscillations during Motor Imitation in Youth with Autism 1 1 1 2 3 I. Buard , E. Kronberg, S. Steinmetz, S. Hepburn, andD.C.Rojas Department of Neurology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Department of Psychiatry, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Department of Psychology, Colorado State University, Fort Collins, CO, USA Correspondence should be addressed to D. C. Rojas; don.rojas@colostate.edu Received 1 March 2018; Revised 22 June 2018; Accepted 12 July 2018; Published 25 July 2018 Academic Editor: Bennett L. Leventhal Copyright © 2018 I. Buard et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Children with ASD oen ft exhibit early difficulties with action imitation, possibly due to low-level sensory or motor impairments. Impaired cortical rhythms have been demonstrated in adults with ASD during motor imitation. While those oscillations reflect an age-dependent process, they have not been fully investigated in youth with ASD. We collected magnetoencephalography data to examine patterns of oscillatory activity in the mu (8-13 Hz) and beta frequency (15-30 Hz) range in 14 adolescents with and 14 adolescents without ASD during a fine motor imitation task. Typically developing adolescents exhibited adult-like patterns of motor signals, e.g., event-related beta and mu desynchronization (ERD) before and during the movement and a postmovement beta rebound(PMBR)aeft rthemovement.Incontrast,thosewithASDexhibitedstrongerbetaandmu-ERDandreducedPMBR.Behav- ioral performance was similar between groups despite differences in motor cortical oscillations. Finally, we observed age-related increases in PBMR and beta-ERD in the typically developing children, but this correlation was not present in the autism group. These results suggest reduced inhibitory drive in cortical rhythms in youth with autism during intact motor imitation. Furthermore, impairments in motor brain signals in autism may not be due to delayed brain development. In the context of the excitation- inhibition imbalance perspectives of autism, we oer ff new insights into altered organization of neurophysiological networks. 1. Introduction investigating not only the degree of impairment but also its underlying mechanism(s). Behavioral studies have investi- Autism Spectrum Disorder (ASD) is a complex disorder of gated potential links between degree of motor impairment brain development characterized, in varying degrees, by dif- and types of movements and/or movement contexts in autism ficulties in social interaction, verbal and nonverbal commu- (for reviews, see [11, 12]. However, there is a general lack nication, and repetitive behaviors [1]. As early as 20 months of of knowledge related to deficits among neural mechanisms age, children with autism exhibit a robust deficit in imitating responsible for orchestrating movements in ASD. the actions of other people [2, 3]. Diverse explanations for Voluntary movements are accompanied by changes in imitative difficulties in ASD have been proposed, including cortical rhythms that can be detected by electroencephalog- motor control [4] and sensory perception deficits [5]. Studies raphy (EEG) and magnetoencephalography (MEG). Distinct have found impairments in several aspects of motor function, oscillatory signals are associated with motor tasks but are differently modulated during movement imitation or obser- including coordination [6], gait [7], motor imitation [8], and movement preparation [9] in both adults and children with vation. autism. eTh term developmental dyspraxia has been used to First, movement-related changes in rhythmic activity in describe those deficits and has been proposed to be specific the mu-range (8-13 Hz) have been reported as early as infancy to autism [10]. While delayed or aberrant fine and gross [13]. Its pre- and perimovement suppression are known as motor movements in autism used to be popularly mistaken event-related desynchronization (ERD) during activation of for clumsiness, an increasing number of studies have been sensorimotor areas, followed by an increase after movement 2 Autism Research and Treatment onset, which has widely been reported as event-related improves from birth well into early adolescence [34, 35]. It synchronization (ERS) [14]. is unclear how maturational changes of the motor cortex may Second, rhythmic modulation in the ongoing beta (15- be aeff cted in autism spectrum disorders. Finally, while mu 30 Hz) rhythm follows a pattern similar to the mu rhythm and beta rhythms are generated around the same time relative [15] although ERS has been more specifically named post- to the movement but not from the same areas [36], their movement beta rebound (PMBR; [16]). It is known that many functional meanings are very distinct, which usually prevents experimental factors can aeff ct sensorimotor beta rhythms, from drawing conclusions based on results combined from including difficulty of the movement sequence, movement both oscillations. duration, and directional uncertainty (e.g., see [17, 18]). Beta In this study, we examined mu- and beta-band oscil- oscillations may also indicate the integrity of circuit-level lations in adolescents with ASD during a finger imitation and neurotransmitter function. eTh power of PMBR has task. eTh paradigm we chose involved simple finger-lifting rd previously been associated with inhibitory brain function. imitative movement performed from the3 -person perspec- For example, Gaetz et al. found that PMBR, but not beta-ERD, tive and from computer-generated human hand videos. We was correlated with the concentration of GABA measured predictedthatmotor-betarhythms wouldbeimpairedinthe from magnetic resonance spectroscopy in the sensorimotor autism group due to their motor and/or imitation problems. cortex [19]. Others have found that using direct pharmaco- Specicall fi y, we hypothesized that beta-ERD signal would be logical manipulation of GABA-A receptors, while not having higher (i.e., greater beta suppression) in ASD because of its direct eeff cts on ERD or PMBR, results in a general increase relevance to difficulty with movement preparation. Similarly, in spontaneous beta, which in turn predicts ERD and PMBR we expected a weaker beta-PMBR due to its association [20]. Oscillatory patterns in the beta-range of sensorimotor with cortical inhibitory processes, which are predicted to areas may therefore provide cortical signatures relevant to be impaired in autism. We expected to see lower mu- circuit dysfunction. suppressioninthe autism group, aspreviously showninthe Third, a high-gamma band ( ∼70-90 Hz) ERS is sometimes literature. observed at the onset of movement [21]. Abnormalities in mu and beta rhythms have been 2. Material and Methods described in ASD patients while performing motor imitation tasks, such as reduced mu-suppression during movement 2.1. Study Subjects. Participants were 28 right-handed ado- observation [22–24], although the relatively small sample lescents (Table 1). Subjects were matched for age and intel- sizes in these studies (less than 20 people per group) calls ligence quotient (FSIQ), using the 4-subtest version of the for caution in the generalization of those findings and Wechsler Abbreviated Scale of Intelligence (WASI; [37]). the need for replicative studies. These observations, among Handedness was assessed in all subjects using the Annett others, were interpreted as supportive of a “broken mirror” Handedness Questionnaire [38]. In the ASD group, ado- theory of autism involving mirror neuron circuitry [25]. lescents met DSM-IV criteria for ASD, as determined by Other studies, however, have revealed no group differences consensus of the Autism Diagnostic Observation Schedule in mu-band activity during action observation or imitation (ADOS, [39]), DSM-IV diagnosis and a parent report of [26, 27]. Reduction of PMBR during action observation has ASD symptoms using the Social Responsiveness Scale (SRS; been shown in adults with ASD compared to controls [28], [40]) and review of all available data by a clinical psy- although not in adults with Asperger syndrome [29]. To chologist (S.H.). Interobserver reliability of ADOS scores date, however, beta rhythms in children and adolescents with is assessed in 20% of cases, with ICCs ranging from .72 ASD have not been investigated during motor imitation. to .94. A second diagnostician independently completed a Oberman’s group has recently extended their work on mu- record review of 50% of cases concurred with ASD diag- suppression, reporting that it increases during childhood nosis for all cases reviewed. All subjects signed informed and adolescence and independently of an autism diagnosis consent and assent to participate in the study consistent with [30]. us, Th developmental delay of those motor-generated the guidelines of the Colorado Multiple Institution Review oscillations, rather than deviance from typical development, Board. does not support the “broken mirror” hypothesis in autism [25]. eTh strong developmental gradient in mu-suppression, as well as beta-ERD and PMBR, makes it important to 2.2. Stimuli and Experimental Design. The stimuli consisted distinguish studies involving children from those with adults. of a photorealistic animated right hand, presented in the Although transcranial magnetic stimulation studies suggest third-person perspective (Figure 1). eTh index or pinky that corticospinal motor pathways are fully developed in n fi gers from this hand were lifted brieyfl 3 s after the early adolescence [31, 32], there is other evidence suggesting beginning of the video (1 s duration for entire movement, further development of the motor cortex and its associated returning to rest) every 6 s. Subjects were asked to imitate cortical oscillations well into the adolescent period. PMBR, the finger movement with their right index or pinky finger for example, appears to strongly develop throughout adoles- as seen on the screen. Index and pinky imitation stimuli were cence. An MEG study observed limited PMBR in 4- to 6- presented in randomized order to the subject using E-prime year-olds and higher levels in adolescents aged 11 to 13, but 2.0(Psychology SoftwareTools,Inc.).Atotalof806-strials stillsignicfi antlylower compared with youngadults[33]. It were presented for each condition (160 total trials, for 16 is also known that ne fi motor control in healthy children minutes’ total experiment duration). Autism Research and Treatment 3 Table 1: Participants’ characteristics. ASD Controls 𝜒 / t value P value 7 Autistic Disorder N and DSM-IV diagnosis 5Asperger’s 14 2PDD-NOS Age 14.5±2. 8 13.8±2. 8 0.66 0.52 Male/female 13/1 11/3 n/a 0.16 Handedness∗ 0.8±0.3 0.8±0.2 0.46 0.65 IQ 106.5±19.2 110.3±15.8 0.57 0.58 SRS∗ 104.6±21.5 n/a n/a n/a SES∗ 48.2±10.5 49.2±9.2 0.23 0.82 ∗ handedness scores were obtained using the Annett handedness questionnaire [38]; SRS (Social Responsiveness Scale) is a brief informant-based measurement of autism traits [40]; SES (Socioeconomic Status) scores based on the Barratt modified measure of social status [41]. Figure 1: Right-hand third-person representation showing the hand at rest (left) and while performing an index lift movement (right). 2.3. MEG Data Acquisition, Preprocessing, and Coregistration a procedure adapted from the preprocessing of electromyo- with Structural MRI. MEG data were obtained in a magneti- graphy data for trigger definition [42]. eTh definition of cally shielded room (ETS-Lindgren, Cedar Park, TX, USA) movement onset was then defined as the point at which using a Magnes 3600 WH whole-head MEG device (4D the accelerometer signals exceeded 2.5 standard deviations Neuroimaging, San Diego, CA, USA), comprised of 248 first- of themeansignalwithaminimumdurationbetween order axial-gradiometer sensors (5 cm baseline) in a helmet- onsets of 5 s. MEG trials were den fi ed with an epoch shaped array. Five head position indicator coils attached to duration of 5500 ms, with 0 ms being the accelerometer- the subject’s scalp were used to determine the head position defined movement onset. Epochs were baseline corrected with respect to the sensor array. eTh locations of the coils with (-2500 to -1500 ms premovement onset) and those trials respect to three anatomical landmarks (nasion and preauric- contaminated by excessively large MEG amplitudes (±2,000 ular points, with the intersection of the tragus and daith of fT) were rejected from further analysis. A mean of 96 (±29) the ear defining the preauriculars) and 2 extra nonfiducial and 106 (±17) artifact-free epochs for the autism and control points as well as the scalp surface (approximately 500 points) groups, respectively, was used in further analyses. No group were determined with a 3D digitizer (Polhemus, Colchester, dieff rence was observed between groups for the remaining VT, USA). The MEG signals were acquired continuously in artifact-free trials, t(26)=0.33, p=0.74. Data from excessive a 0.1-200 Hz bandwidth and sampled at 678.17 Hz and 24-bit noise or movement artifacts were not included but small in- vertical resolution. scanner head movements have not been corrected. Single axis monolithic integrated circuits Leadless Chip Each participant’s MEG data were coregistered with Carrier (LCC) accelerometers (model ADXL103; Analog structural T1-weighted magnetic resonance imaging (MRI) Devices, Inc.) were attached to both index and pinky fin- data prior to source space analyses (see below MRI acquisi- gertips in order to precisely quantify movement onset. eTh tion procedures) using common landmarks from the MEG chips are wired to approximately 3.3 m of light weight, highly digitization procedure and MRI scan data. Structural MRI flexible, miniature cable (Cooner Wire NMVF 4/30-4046) data were aligned parallel to the anterior and posterior with local bypass capacitors (0.1 uf) and encapsulated in commissures and transformed into the Talairach coordinate heat-shrink. Accelerometer signals were high-pass filtered system[43]usingtheBrainElectricalSourceAnalysis(BESA) at20Hz,rectiefi d, andthenlow-passfilteredat10 Hz in MRI software (BESA MRI version 2.0). 4 Autism Research and Treatment 2.4. MEG Time-Frequency Transformation. MEG postpro- equivalent dipole over time and is a robust estimator for cessing was performed using BESA 5.3 (MEGIS Software source location, because it has only 3 degrees of freedom GmbH, Grafelfing, Germany). Artifact-free epochs (mean for the tfi ting procedure. To obtain the orientation of the per condition: 51 +/- 12) in the time-domain were trans- equivalent dipole, the regional source was rotated in such a formed to the time-frequency domain with a 2 Hz/25 ms way that one of the three dipoles represented the most dipole sampling in BESA using complex demodulation [44]. This orientation at the time of the motor-evoked peak. eTh n, complex demodulation consists of a multiplication of the normalized source power in the mu- and beta-range was time-domain signal with a complex periodic exponential computed from the corresponding source waveform in each function, with a frequency equal to the frequency under ROI (left and right motor regions) against noise estimated analysis, and a subsequent low-pass filter. This low-pass filter during baseline. is a ni fi te impulse response filter of Gaussian shape in the Mean power was extracted between -500 and 500 ms for time-domain, which is related to the envelope of the moving beta-ERD and 1000 ms for mu-ERD and between 750 and window in wavelet analysis. For our setting of a 2 Hz/25 2000 ms for PMBR. For statistical analyses, time-frequency ms time-frequency sampling, this filter has a width in the results were subjected to group statistical analysis in a 2 x 2 frequency domain of 5.7 Hz and in the time-domain of x 2 mixed design ANOVA statistical test (group by finger by 79 ms full width at half maximum [45]. Time-frequency hemisphere) with n fi ger and hemisphere treated as within- analyses were computed for each sensor individually, per subjects measures. Separate ANOVAs were computed for the trial, averaged over trials, and then normalized by dividing ERD and PMBR windows. the power of each time-frequency bin by the respective fre- quency’s mean baseline power. This normalization procedure 2.6. MR Data Acquisition. MR images and spectra were allowed task-related mu (8-13 Hz) and beta (15-30 Hz) power acquired using a 3.0T GE Signa HDx whole body, long uc fl tuations to be readily visualized in sensor space as the bore MR scanner (GE Healthcare, Waukesha, WI, USA) at following: a decrease (blue, Figure 2) was observed priorly the Brain Imaging Center, University of Colorado Denver. and extending to the movement and imaged using a -500 to Subjects were imaged in the supine position using a GE eight- 500ms(1000msformu-ERD)time window(withtime0 channel phased array head coil. To comply with age- and being the onset of the movement). Following the movement, population-related behaviors such as boredom and restless- increased beta-band power (red, Figure 2) was imaged using a ness, subjects watched a movie during the exams using MR- 750 to 2000 ms time window. Time-frequency bins of interest compatible goggles and headphones (Resonance Technology were chosen to focus on maximum ERD/PMBR responses in Inc., Northridge, CA, USA) during the procedure. A T1- the ipsilateral and contralateral sides of the MEG sensor array, weighted sequence was acquired for tissue segmentation as previously described [46, 47]. Baseline mu and beta power using a 3D inversion recovery fast, spoiled gradient echo (IR- were extracted within a -2500 to -1500 ms time window. SPGR) technique (matrix 256 x 256, FOV 22 cm, TR/TE/TI= 10/3/450 ms, NEX=1), resulting in 168 1.2 mm thick axial slices 2.5. MEG Source Reconstruction and Statistical Analyses. 2 with an in-plane resolution of .86 mm . Using BESA Research 5.3, cortical networks were imaged through an extension of the linearly constrained minimum 2.7. Finger Movement Accuracy. Behavioral analyses of cor- variance vector beamformer [48–50], which employs spatial rect responses were determined from the accelerometer data. filters in the frequency domain to calculate source power Correct trials were den fi ed as subject movements occurring for the entire brain volume. The single images are derived on the correctly indicated n fi ger within 3 s aer ft the move- from the cross-spectral densities of all combinations of ment onset from the video displayed on the screen (i.e., MEG sensors averaged over the time-frequency range of from 1 s to 4 s aer ft stimulus presentation). Subjects in both interest,andthesolutionoftheforwardproblem foreach groups failed to respond with either nge fi r on some trials, location on a grid specified by input voxel space (7 mm and these trials were excluded from the accuracy calculation. cubic voxels in Talairach space). Following convention, the We analyzed this factor separately as level of responsiveness, source power in these images was normalized per participant defined as the number of trials responded to as a percentage using a prestimulus noise period of equal duration and of the trials presented. Accuracy, response times, and level bandwidth [48]. In the dipole-fit model used in BESA, a set of responsiveness were extracted for each participant and of consecutive time points is considered in which dipoles averaged across trials. Separate 2 x 2 ANOVA designs (group areassumedtohavefixedpositionand xfi ed orvarying by finger, with finger as a repeated measure) were used to orientation. In the final analysis, we used a regional source assess each behavioral variable. with xfi ed location and orientation throughout the analysis window because this way the resulting source waveform represents the time course of activity in the region of interest. 3. Results The regional source was placed on one of the rfi st maxima of source activation, which typically was over in the motor 3.1. Behavioral Results. Participants with ASD performed area (identifiable by its “omega” or “epsilon” shape, as widely their movements around 3.23±.24 s aer ft the video movement reported [51]). A regional source is a set of three orthogonal onset, averaged across both fingers, while control children dipoles that represent the electrical activity of a small brain imitated the n fi ger-lifting movements aer ft 3.13 ±.12 s. No volume independent of changes of the net orientation of the significant main effect of group was observed, 𝐹(1,26) = Autism Research and Treatment 5 contralateral ipsilateral 30 30 50 50 40 40 25 25 30 30 20 20 20 20 10 10 CTL 15 15 −10 −10 −20 10 10 −20 −30 −30 −2000 −1000 0 1000 2000 −2000 −1000 0 1000 2000 Time (ms) Time (ms) 30 50 25 30 ASD −10 −10 10 −20 −20 −30 −30 −2000 −1000 0 1000 2000 −2000 −1000 0 1000 2000 Time (ms) Time (ms) Figure 2: Grand average of time-frequency spectra of CTL (control, top) and ASD (bottom) children. TFR plots are derived from beamforming source images localized in the contralateral and ipsilateral motor areas and time-locked to movement termination (accelerometer onset). Cortical oscillations in the mu- (8-13Hz) and beta- (15-30Hz) bands showing event-related desynchronization (ERD, blue) at movement onset and postmovement beta rebound (PMBR, yellow-red) time regions are characterized. Source power during the baseline period was subtracted from source power during movement or aer ft the movement intervals. Control subjects had a significant power decrease (ERD) in percent change from baseline in the beta frequency range beginning before movement and lasting throughout the duration of movement as well as a mu-ERD around the movement, which were followed by a strong beta power increase (PMBR) in percent change from baseline beginning shortly aer ft movement termination. On the contrary, ASD subjects had a significantly greater power decrease (ERD) compared to controls in the beta and mu frequencies range beginning before movement and lasting beyond the movement, which was followed by a weak or no beta power increase (PMBR) beginning shortly aer ft movement termination. .72, p> .05. The group by finger interaction term was also 3.2. Time-Frequency Results. As expected, we found relevant nonsignificant, 𝐹(1,26) = .11, p> .05. This suggests the appro- motor-associated beta and mu oscillations in both hemi- priateness of fixed time bins for motor-related oscillations in spheres, contralateral and ispilateral to the movement, during this specific study, since it is unlikely that group differences imitation of both fingers (Figures 2 and 3). in motor-related oscillations were due to delayed movements For beta-ERD, the main effect of group was signicfi ant, in the autism group. For accuracy, the main effect of group 𝐹(1,26) = 53.02, p < .001, indicating greater ERD in the was at the edge of statistical significance, 𝐹(1,26) =1.82, p = autism group relative to controls (i.e., greater suppression, .05 (Table 2). The group by n fi ger interaction term was non- see Figures 2 and 3, top). The hemisphere main effect was significant, 𝐹(1,26) = 1.53, p> .05. Finally, we looked at level also signica fi nt, 𝐹(1,26) =12.18, p < .001, indicating a of responsiveness and found that control children responded stronger beta-ERD in the left hemisphere, contralateral to on a significantly higher percentage of trials (73.39 +/- 17.04) the movement. The finger main effect was nonsignificant, than their aeff cted peers (66.48 +/- 14.91), 𝐹(1,26) =6.99, p< 𝐹(1,26) =.05, p> .05. The group by hemisphere eeff ct was .05. For responsiveness, no significant differences were noted nonsignificant, 𝐹(1,26) =2.14, p> .05. The group by n fi ger forthe maineeff ctoffinger,𝐹(1,26) =2.12, p> .05, or for the interaction was nonsignificant, 𝐹(1,26) =1.32, p> .05. The group by finger interaction term, 𝐹(1,26) =0.69, p> .05. hemisphere by finger effect was nonsignificant, 𝐹(1,26) = Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) 6 Autism Research and Treatment Table2:Behavioralresults. Finger ASD Controls Index Reaction time± SD 3.23± 0.24 3.13± 0.11 Accuracy± SD 94.10%± 6.95 98.80%± 1.84 Responsiveness± SD 60.70%± 14.40 72.3%± 13.48 Pinky Reaction time± SD 3.24%± 0.25 3.14%± 0.14 Accuracy± SD 94.00%± 7.15 95.90%± 6.51 Responsiveness± SD 65.90%± 18.16 74.80%± 15.19 SD, standard deviation. 1.77, p > .05. Finally, the group by finger by hemisphere interaction was also nonsignificant, 𝐹(1,26) =1.61, p> .05. Cohen’seeff ct sizeswere1.1 andhigherfor each beta-ERD group comparison, indicating that our significant results were likely meaningful. −10 For PMBR, the main effect of group was signicfi ant, 𝐹(1,26) =26.51, p< .001, indicating greater PMBR in the control group relative to the autism group (i.e., greater synchronization, see Figures 2 and 3, bottom). In contrary to −20 thebeta-ERD,thehemispheremaineeff ct wasnotsignicfi ant, 𝐹(1,26) =2.04, p> .05. The finger main effect was nonsignif- icant,𝐹(1,26) =.54, p> .05. The group by hemisphere effect was also nonsignificant, 𝐹(1,26) =.56, p> .05. The group by −30 index pinky index pinky finger interaction was nonsignificant, 𝐹(1,26) =.86, p> .05. The hemisphere by finger effect was nonsignificant, 𝐹(1,26) ASD =.02, p> .05. Finally, the group by finger by hemisphere Control interaction was also nonsignificant, 𝐹(1,26) =.62, p> .05. Cohen’s effect sizes were 0.8 and higher for each PMBR group comparison, indicating that our significant results were likely Contralateral Ipsilateral meaningful. For mu-ERD, the main effect of group was significant, 𝐹(1,26) =9.59, p < .05, indicating greater mu-ERD in the autism group relative to controls (i.e., greater mu- suppression, see Figures 2 and 4). On the contrary to beta- ERD, the hemisphere main eeff ct for mu-ERD was not 10 significant, 𝐹(1,26) =1.48, p> .05. eTh n fi ger main effect was also nonsignificant, 𝐹(1,26) = .88, p> .05. The group by hemisphere effect was nonsignificant, 𝐹(1,26) =.52, p > .05. The group by n fi ger interaction was marginally significant, 𝐹(1,26) = 3.28, p = 0.073, indicating that the greater mu-ERD in the ASD group was more prominent for an index movement. eTh hemisphere by n fi ger effect was nonsignificant, 𝐹(1,26) =.06, p > .05. Finally, the group −10 by finger by hemisphere interaction was also nonsignificant, index pinky index pinky 𝐹(1,26) =0.21, p > .05. Cohen’s effect sizes were 1.1 and higher for left hemisphere mu-ERD group comparison, but ASD Control as low as 0.2 in the right hemisphere. This suggests that mu-suppression results may only be meaningful in the Figure 3: Children with ASD exhibit greater beta-ERD than their contralateral hemisphere. control peers but beta-PMBR is absent. Mean (+/- sem) beta-band Correlations between age and beta-ERD, beta-PMBR, ERD (top) and -PMBR (bottom) to Index and Pinky Imitation from and mu-ERD were examined in each group and each hemi- virtual electrodes contralateral (left) and ipsilateral (right) to right sphere using a Pearson r correlation coefficient. In the control hand of participant. group, there was a significant negative correlation between beta-PMBR beta-ERD (% change from baseline) (% change from baseline) Autism Research and Treatment 7 r = -.6, p = 0.02 −10 −20 −10 −30 −40 −50 −20 r = .6, p = 0.02 index pinky index pinky ASD Control Contralateral Ipsilateral Figure 4: Children with ASD exhibit greater mu-ERD than their control peers. Mean (+/- sem) mu rhythm ERD to Index and Pinky −10 Imitation from virtual electrodes contralateral (left) and ipsilateral −20 (right) to right hand of participant. 810 12 14 16 18 age (years) Figure 5: Beta correlation results. Representative plots of the age and beta-ERD during imitation of the index finger, correlation results assessing the relationship between age of the r(12) = -.6; p< .05 (Figure 5, top). It should be noted that subjects and contralateral beta-ERD (top) and -PMBR (bottom) this was only the case in the left hemisphere, contralateral power during index imitation. eTh r values were -.6 for ERD and to themovement.In thesamegroup,beta-PMBRpower .6 for PMBR for control participants (closed black circles, red was significantly correlated with age during index imitation, trendline). No significance was found for ASD participants (open r(12) = .6, p < .05 (Figure 5, bottom), and during pinky black circles) for both ERD and PMBR. imitation, r(12)=.55, p< .05, in the left hemisphere only. No correlation with beta power was found in the right hemisphere (p> .10) in the control group. In the autism group, mu and beta power changes associated with the imitations no significant correlation was found between either beta- were signicfi antly different from those of controls. While ERDorbeta-PMBRandage forany imitated movement and both affected and nonaffected children were able to perform any hemisphere (p> .10). Lastly, no significant correlation was the simple action of lifting a finger, their cortical activity found between mu-ERD and age for any group, any imitated levels were strikingly different. In the motor cortex, induced movement, and any hemisphere (p> .10). powerrevealedanincreaseinmu-andbeta-ERDanda Baseline mu and beta power were calculated for the reductioninPMBRintheASDgroup compared to the source reconstructed waveforms and group differences were control group, during imitation of both n fi ger movements. examined using a 2-sample t-test. There was no significant Our results provide some physiological evidence of distinct difference between the control group and the autism group brain activity associated with imitation of hand movements for baseline power in the beta frequency band (p> .10). in children with autism. Below, we discuss the implications of these findings for understanding the pathological cortical activity in children with autism. 4. Discussion Surprisingly, we found greater mu-ERD in the group of In the current study, children without ASD exhibited a children with autism compared to their nonaeff cted peers. well-established pattern of oscillatory neural activity before Whether this greater ERD is restricted to the motor-related and aeft r movement onsets in brain areas associated with signals or rather linked to the mirror neurons remains to motor processing. Beta and mu-ERD were observed prior to be clariefi d. Following the “broken mirror” theory of autism movement onset and during movement execution, whereas a [25],actionobservation maycausethesamefiring eeff cts strong PMBR response emerged following movement termi- as action execution suggesting that self-other mapping leads nation. os Th e responses were observed though contralateral to imitation deficits in autism [52]. However, a recent meta- and ispilateral sensorimotor cortices. Children with autism analysis provides compelling evidence that neuroimaging also exhibited each of these neural responses, although the studies are far from providing clear support to this hypothesis mu-ERD (% change from baseline) beta-ERD beta-PMBR (% change from baseline) (% change from baseline) 8 Autism Research and Treatment [53] and neurophysiology studies on mu-suppression also increased errors could have been expected in the ASD group show similar inconsistencies, especially with regard to two due to a confusion of which finger is moving on the screen. potential mu subbands [54]. Of note, mu-suppression is We did report a weaker accuracy in the autism group, expected during movement observation, prior to movement but only during the imitation of an index movement. This execution but this was not clearly captured in our data. This partly confirms the relevance of our findings to imitation couldpossiblybeduetothemethodbeing used.While MEG problems in general in autism. Previous research has also studies like ours model mu source activity using an equivalent suggestedthatpeoplewithautismhavemoredicffi ultywhen dipole, i.e., assuming that a small number of dipoles recorded the imitated movement is meaningless or less goal-directed mu activity [55, 56], most EEG studies reporting strong [71]. It is important to note that the gestures imitated in and widespread mu-suppression assume that cortical areas the present study were not inherently meaningful from a contributing to mu activity are distributed throughout the communication perspective. Further investigation contrast- brain [57]. Mu data acquisition is therefore not reliably well ing anatomical and mirror motor imitation and exploring captured using MEG instead of EEG. Alternatively, it is meaningful manipulations might provide some explanation possible that our method prevents from collecting results on these aspects. from diverse sources. Beamforming uses a spatial filter Previous studies have shown that beta-ERD power during designedtobefully sensitivetoactivityfromthe targetvoxel, simple finger movements is correlated with age [33, 72]. while being as insensitive as possible to activity from other In accordance with those studies, we also obtained similar brain regions. Indeed, motor-related mu- and beta-ERD are correlations although only in the control group. In children generated around the same time but from distinct brain areas. with autism, where stronger ERD is observed, no correlation We observed signica fi ntly greater ERD in the beta-band was found with age. This makes sense from a developmental in children with ASD. Given the beta-ERD’s association standpoint; cortical rhythms are resulting from synchroniza- with movement preparation [14] and cognitive selection of tion of a massive number of neurons, which themselves are a proper motor response [58, 59], these results provide a fully mature. Higher maturation stages in the motor cortex possible physiological mechanism for the difficulty of indi- lead to higher beta power. But higher beta-ERD does not viduals with ASD to imitate movements [60]. Increased beta- mean that children with ASD have a more mature brain, or ERD has been characterized in some motor-related disorders, both beta-ERD and -PMBR powers would be higher. In other such as cerebral palsy [61], but a decrease has been reported words, the beta impairments observed in the ASD group in others [62, 63]. Greater beta-ERD in autism has been might not likely be due to delayed maturation of the motor reported in a previous EEG study while subjects observed cortex. This conrfi ms the dysfunctional integrative theory static hands [64], although passively watching hand actions of autism [73]. Alternatively, the aberrant beta-ERD might did not produce any significant differences in beta-band be linked to the reduction in beta-PMBR. In other words, activity. This difference in the static condition might suggest if the sensorimotor circuitry underlying beta oscillations that, to the extent that ERD reflects movement preparation, children with autism is failing to generate synchronized beta the participants with autism had greater dicul ffi ty imagining a oscillations at rest, then premovement beta-ERD would be static hand moving, but not while actually watching the hand limited by the low resting oscillatory power. However, the move. In the current study, there was no static condition and similar baseline beta power in both groups rules out this subjects did not passively watch the stimuli, so direct com- hypothesis in the context of our study. Similarly, while PMBR parisons with the previous EEG paper are dicffi ult. Because is thought to reflect an age-dependent inhibitory process latency of motor responses as well as neural activity in the [33, 72], our data showing increased PMBR are not explained sensorimotor cortex during motor preparation, especially by a greater age range in the ASD group. beta-band ERD, covaries with movement uncertainty [17, PMBR is proposed to be associated with motor deacti- 65], beta-ERD may be strongly associated with movement vation or inhibition of the motor cortex by somatosensory selection. What seems clear is that beta-ERD abnormalities afferents [74], or a “resetting” of the underlying cortical canbeobservedinautisminavarietyofconditionsrelatedto networks [75]. Transcranial magnetic stimulation studies movement observation or execution. have conrm fi ed that beta-PMBR corresponds to a period Alternatively, difficulties with body part orientation [66] of decreased corticospinal excitability [76], suggesting that or self-other mapping [67, 68] have been proposed to under- it may represent a state of cortical inhibition. It has also lie imitation problems in autism. It is therefore possible that been suggested that several beta rhythms exist, each with the third-person perspective of our imitation paradigm is a different functional significance [74, 77]. For instance, partly responsible for the increased beta-ERD in the ASD postmovement beta rebound (but not premovement beta- group. Behavioral imitation studies have shown that, in a ERD) has been shown to be related to a prolonged period third-person perspective, the movement that is imitated more of increased corticomuscular coherence following phasic easily is the mirror, or specular hand, versus the anatomical voluntary movements [78] that, in turn, may reflect the level hand [69].Theideathatvisuospatialinformation processing of attention to motor performance [79]. Although a couple deficits may be contributing to functional motor coordina- of autism studies report a reduction in PMBR during action tion deficits in autism has already been contemplated [70]. observation [28, 29] but not during performance of the In this context, while beta-ERD is strongly associated with action, we provide evidence that postmovement beta signals movement selection [17], a greater ERD could be due to are also affected during action imitation. Two explanations a greater difficulty to choose which finger to move, and rise from those opposite findings. First, while we did not Autism Research and Treatment 9 include an observation only condition, it is possible that extent changes in motor-beta rhythms are directly related to neuronal circuits activated in our task proceed independently observable changes in motor behavior. of the mirror neuron system. Second, our sample might include children with greater motor impairments than those 5. Limitations of the other studies. Current theories and experimental data strongly suggest Given the cross-sectional nature of the study and small that dysfunctional integrative mechanisms in ASD result sample size, we would like to warn about the highly prelim- from reduced neuronal synchronization [73]. eTh underlying inary nature of these findings. In addition, the low number cellular mechanisms seem to be an imbalance between excita- of females might limit the generalization of the results. tion and inhibition [80], which leads to hyperexcitability and By essence, autism isaspectrumsothecharacteristicsof unstable cortical networks, as abnormalities in GABAergic children with ASD and their life circumstances are mostly and glutamatergic transmitter systems has been characterized heterogeneous in nature. Addressing these issues may require in humans and animal models of autism [81, 82]. Recent stud- larger sample sizes and possibly interdisciplinary collabora- ies from our group and others have demonstrated reduced tion. GABA and increased glutamate in some regions of the brain in children with ASD [83–86] with possible evidence of reduced GABA in the motor cortex [84]. In typically 6. Conclusion developing participants, Gaetz and colleagues [19] reported a correlation between motor cortical GABA concentration and We have demonstrated that children and adolescents with PMBR power. Consequently, it is expected that changes in autism may have reduced inhibitory drive in cortical rhythms cortical oscillatory rhythms, especially a reduced PMBR, will as measured with MEG during motor imitation. Our results be foundinthe brainsofASDchildren. In this context,our support previous theories that inhibitory dysfunction could results are consistent with the interpretation of PMBR as a be one of the factors underlying abnormal behaviors in marker of inhibitory neuronal signaling and the excitation- autism. Further, changes in ERD suggest greater difficulty inhibition (E-I) imbalance theory of autism [80]. in movement planning in the autism group. Understand- Finally, it is critical to consider the mixed picture of ing these mechanisms may provide a potential target for behavioral resultsinthe currentstudy.Wefound that the future therapies to address motor-related symptoms, by both participants with ASD responded to the imitative stimuli less pharmacological and behavioral interventions. Whereas the oen ft than controls but that when the participants with ASD relevance of altered brain oscillations to motor imitation responded, their responses were as accurate as those in the problems in autism needs further claricfi ation, monitoring control group. Reduced responding could be interpreted as pathological beta-bands features with MEG might hold evidence of confusion over the imitative action requested but promise as a biomarker for motor impairments in ASD. equally could be considered evidence of greater lapsing of On this last point, although a large number of individuals attention during the task. In the current study, we cannot with ASD have motor difficulties, they are not universally discern between these possibilities. Since we were focused observed [87]. Due to this heterogeneity, specicfi ation of on response-locked beta-band ERD and PMBR, we could motor impairments in autism may be useful for the identi- not analyze trials on which the subjects did not respond fication of clinically relevant subgroups in ASD. Moreover, to the stimuli, limiting our understanding of whether such a better understanding of the neurobiology of motor and/or trials were associated with additional differences in beta-band imitation impairments is vital for identifying treatments to activity. improve outcomes related to motor deficiencies. Whilebeta-ERDand-PBMRaregeneratedfromthesame regions, it is not clear whether they result from similar events Abbreviations at theneuronalornetworklevel. Ourcohortofchildrenwith ASD did not exhibit significant motor defects. We interpret ASD: Autism spectrum disorders the aberrant pattern of beta rhythms observed in our ASD MEG: Magnetoencephalography group, especially the increased ERD, as most likely associated ERD: Event-related desynchronization with the dicffi ulty of cognitive processes involved in selecting PMBR: Postmovement beta rebound the motor response rather than with a motor decfi it itself. EEG: Electroencephalography In contrast, the reduced PMBR may be related to reduced GABA: Gamma-aminobutyric acid inhibition in the motor cortices. Indeed, PMBR is absent DSM: Diagnostic and statistical manual of in young children [33], which suggests that the reduced mental disorders influence of inhibition in the motor cortices may represent an ADOS: Autism diagnostic observation schedule optimal physiological environment to facilitate motor learn- MR(I): Magnetic resonance (imaging). ing or to recover from motor delay. It could therefore also be interpreted to be a compensatory consequence of the ERD changes in the autism group. Another possible interpretation Data Availability ofreducedPMBRintheASDgroup isthat ofdevelopmental delay. The mixed pattern of aberrant beta oscillations poses Deidentiefi d data is available on our laboratory server in.mat an interesting question for further exploration—i.e., to what and.xlsx file format readily usable by any requester. 10 Autism Research and Treatment Additional Points [10] L. K. MacNeil and S. H. Mostofsky, “Specificity of dyspraxia in children with autism,” Neuropsychology,vol.26,no.2,pp.165– Highlights. (i) Motor oscillations are impaired in children 171, 2012. with ASD during motor imitation. (ii) Increased beta-ERD [11] R. Downey and M. J. K. Rapport, “Motor activity in children may be related to difficulties with movement selection with autism: A review of current literature,” Pediatric Physical and preparation. (iii) Weaker or absent beta-ERS suggests Therapy ,vol.24, no.1,pp. 2–20,2012. reduced inhibitory drive. [12] K.J. Steinman,S.H.Mostofsky,andM. B. Denckla, “Toward a narrower, more pragmatic view of developmental dyspraxia,” Journal of Child Neurology,vol.25,no.1,pp. 71–81,2010. Conflicts of Interest [13] M. Berchicci, T. Zhang, L. Romero et al., “Development of None of the authors have potential conflicts of interest to be Mu Rhythm in infants and preschool children,” Developmental Neuroscience, vol. 33, no. 2, pp. 130–143, 2011. disclosed. [14] G. Pfurtscheller, M. Pregenzer, and C. Neuper, “Visualization of sensorimotor areas involved in preparation for hand movement Acknowledgments basedonclassificationof𝜇 and central𝛽 rhythms in single EEG trials in man,” Neuroscience Letters,vol.181,no.1-2, pp.43–46, The authors’ work was supported by Autism Speaks (Post- doctoral Fellowship #7592, I. Buard), the Colorado Clinical [15] G. Pfurtscheller, “Central beta rhythm during sensorimotor andTranslationalScienceInstitute(Co-PilotAward,I.Buard) activities in man,” ElectroencephalographyandClinicalNeuro- and the National Institute of Health (R01 MH082820, D. C. physiology,vol.51,no.3,pp. 253–264, 1981. Rojas). eTh authors would like to thank Himaja Gaddipati [16] G. Pfurtscheller and F. H. L. da Silva, “Event-related EEG/MEG for technical assistance with some of the MEG recordings synchronization and desynchronization: basic principles,” Clin- andIan Southwellfor therealization of theanimatedhands ical Neurophysiology,vol.110,no.11, pp.1842–1857,1999. (http://www.southwellart.com). [17] C. Tzagarakis, N. F. Ince, A. C. Leuthold, and G. Pellizzer, “Beta-band activity during motor planning reflects response References uncertainty,” The Journal of Neuroscience , vol. 30, no. 34, pp. 11270–11277, 2010. [1] American Psychiatric Association (APA), Diagnostic and statis- [18] E. Heinrichs-Graham and T. W. Wilson, “Is an absolute level ticalmanualofmentaldisorders, American Psychiatric Publish- of cortical beta suppression required for proper movement? ing, Arlington, Va, USA, 5th edition, 2013. Magnetoencephalographic evidence from healthy aging,” Neu- [2] T. Charman, J. Swettenham, S. Baron-Cohen, A. Cox, G. Baird, roImage,vol.134,pp.514–521,2016. and A. Drew, “Infants with autism: An investigation of empathy, [19] W. Gaetz, J. C. Edgar, D. J. Wang, and T. P. L. Roberts, pretend play, joint attention, and imitation.,” Developmental “Relating MEG measured motor cortical oscillations to resting Psychology,vol.33,no.5,pp.781–789, 1997. 𝛾-a minobutyric acid (GABA) concentration,” NeuroImage,vol. [3] W.L.Stone,O.Y.Ousley,andC.D.Littleford, “Motorimitation 55,no.2, pp.616–621,2011. in young children with autism: what’s the object?” Journal of [20] S. D. Hall, I. M. Stanford, N. Yamawaki et al., “eTh role of Abnormal Child Psychology,vol.25,no.6,pp. 475–485, 1997. GABAergic modulation in motor function related neuronal [4] M.Vanvuchelen,H.Roeyers,and W. De Weerdt,“Nature of network activity,” NeuroImage, vol. 56, no. 3, pp. 1506–1510, 2011. motor imitation problems in school-aged boys with autism: a [21] T. Ball, E. Demandt, I. Mutschler et al., “Movement related activ- motororacognitiveproblem?” Autism,vol.11,no.3,pp. 225– ityinthehighgammarangeofthehumanEEG,” NeuroImage, 240, 2007. vol.41,no.2, pp.302–310,2008. [5] B.Ingersoll,L.Schreibman, andQ.H.Tran,“Eeff ctof Sensory [22] J. Martineau, S. Cochin, R. Magne, and C. Barthelemy, Feedback on Immediate Object Imitation in Children with “Impaired cortical activation in autistic children: Is the mirror Autism,” Journal of Autism and Developmental Disorders,vol. neuron system involved?” International Journal of Psychophysi- 33,no.6, pp.673–683,2003. ology,vol.68, no.1,pp.35–40,2008. [6] K.A.Fournier, C. J. Hass,S.K.Naik,N.Lodha,andJ.H. [23] L. M. Oberman, E. M. Hubbard, J. P. McCleery, E. L. Altschuler, Cauraugh, “Motor coordination in autism spectrum disorders: V. S. Ramachandran, and J. A. Pineda, “EEG evidence for mirror a synthesis and meta-analysis,” Journal of Autism and Develop- neuron dysfunction in autism spectrum disorders,” Cognitive mental Disorders, vol. 40, no. 10, pp. 1227–1240, 2010. Brain Research, vol. 24, no. 2, pp. 190–198, 2005. [7] J.A.Vilensky, A.R.Damasio, andR.G.Maurer,“Gaitdistur- [24] R. Bernier, G. Dawson, S. Webb, and M. Murias, “EEG mu bances in patients with autistic behavior. A preliminary study,” rhythm and imitation impairments in individuals with autism JAMA Neurology,vol.38,no.10,pp.646–649,1981. spectrum disorder,” Brain and Cognition,vol.64, no.3,pp. 228– [8] J. H. G. Williams, A. Whiten, and T. Singh, “A systematic review 237, 2007. of action imitation in autistic spectrum disorder,” Journal of Autism and Developmental Disorders,vol.34,no.3,pp.285–299, [25] L. M. Oberman and V. S. Ramachandran, “Preliminary evidence for deficits in multisensory integration in autism spectrum disorders: eTh mirror neuron hypothesis,” Social Neuroscience, [9] N. J. Rinehart, J. L. Bradshaw, A. V. Brereton, and B. J. vol. 3, no. 3-4, pp. 348–355, 2008. Tonge, “Movement Preparation in High-Functioning Autism and Asperger Disorder: A Serial Choice Reaction Time Task [26] R. Raymaekers, J. R. Wiersema, and H. Roeyers, “EEG study Involving Motor Reprogramming,” Journal of Autism and of the mirror neuron system in children with high functioning Developmental Disorders,vol.31,no.1,pp. 79–88, 2001. autism,” Brain Research, vol. 1304, pp. 113–121, 2009. Autism Research and Treatment 11 [27] L. Ruysschaert, P. Warreyn, J. R. Wiersema, A. Oostra, and H. [44] N. Papp and P. Ktonas, “Critical evaluation of complex demodu- Roeyers, “Exploring the Role of Neural Mirroring in Children lation techniques for the quantification of bioelectrical activity,” with Autism Spectrum Disorder,” Autism Research,vol.7,no.2, Biomed Sci Instrum,vol.13, pp.135–145,1977. pp.197–206,2014. [45] K.Hoechstetter,H.Bornfleth,D.Weckesser,N.Ille,P. Berg,and [28] E. Honaga, R. Ishii, R. Kurimoto et al., “Post-movement beta M. Scherg, “BESA source coherence: A new method to study rebound abnormality as indicator of mirror neuron system cortical oscillatory coupling,” Brain Topography,vol.16, no.4, dysfunctioninautisticspectrumdisorder:An MEGstudy,” pp. 233–238, 2004. Neuroscience Letters,vol.478,no. 3, pp.141–145,2010. [46] T. W. Wilson, E. Slason, R. Asherin et al., “Abnormal gamma [29] S. Avikainen, T. Kulomaki ¨ , and R. Hari, “Normal movement and beta MEG activity during finger movements in early-onset psychosis,” Developmental Neuropsychology,vol.36, no.5,pp. reading in Asperger subjects,” NeuroReport,vol.10,no.17,pp. 3467–3470, 1999. 596–613, 2011. [30] L. M. Oberman, J. P. McCleery, E. M. Hubbard et al., “Develop- [47] Y. Nishimura, Y. Ikeda, A. Suematsu, and S. Higuchi, “Eeff ct of mental changes in mu suppression to observed and executed visual orientation on mu suppression in children: a comparative actions in autism spectrum disorders,” Social Cognitive and EEG study with adults,” JournalofPhysiological Anthropology, vol. 37, no. 1, 2018. Affective Neuroscience ,vol.8,no. 3,ArticleIDnsr097,pp.300– 304, 2013. [48] B. D. Van Veen, W. Van Drongelen, M. Yuchtman, and A. Suzuki, “Localization of brain electrical activity via linearly con- [31] T. H. H. G. Koh and J. A. Eyre, “Maturation of corticospinal tracts assessed by electromagnetic stimulation of the motor strained minimum variance spatial filtering,” IEEE Transactions on Biomedical Engineering, vol. 44, no. 9, pp. 867–880, 1997. cortex,” Archives of Disease in Childhood,vol.63,no.11,pp.1347– 1352, 1988. [49] A. Hillebrand and G. R. Barnes, “Beamformer Analysis of MEG Data,” International Review of Neurobiology,vol.68, pp.149–171, [32] A. Nezu, S. Kimura, S. Uehara, T. Kobayashi, M. Tanaka, and K. Saito, “Magnetic stimulation of motor cortex in children: Maturity of corticospinal pathway and problem of clinical [50] M. Liljestrom, ¨ J. Kujala, O. Jensen, and R. Salmelin, “Neuromag- application,” Brain & Development,vol.19,no.3,pp.176–180, netic localization of rhythmic activity in the human brain: A comparison of three methods,” NeuroImage,vol.25, no.3,pp. 734–745, 2005. [33] W. Gaetz, M. MacDonald, D. Cheyne, and O. C. Snead, “Neu- romagnetic imaging of movement-related cortical oscillations [51] O. F. Kaneko, N. J. Fischbein, J. Rosenberg, M. Wintermark, in children and adults: Age predicts post-movement beta and M. M. Zeineh, “eTh “white gray sign” identifies the central rebound,” NeuroImage,vol.51, no.2,pp.792–807, 2010. sulcus on 3T high-resolution T1-weighted images,” American Journal of Neuroradiology, vol. 38, no. 2, pp. 276–280, 2017. [34] M. B. Denckla, “Development of Speed in Repetitive and Suc- cessive Finger-movements in Normal Children,” Developmental [52] M. Iacoboni and M. Dapretto, “eTh mirror neuron system Medicine & Child Neurology, vol. 15, no. 5, pp. 635–645, 1973. and the consequences of its dysfunction,” Nature Reviews Neuroscience,vol.7,no. 12,pp. 942–951, 2006. [35] P. H.Wol,C ff .E.Gunnoe,andC. Cohen,“ASSOCIATED MOVEMENTS AS A MEASURE OF DEVELOPMENTAL [53] A. F. D. C. Hamilton, “Reflecting on the mirror neuron system in AGE,” DevelopmentalMedicine&ChildNeurology,vol.25, no. autism: A systematic review of current theories,” Developmental 4, pp. 417–429, 1983. Cognitive Neuroscience,vol.3,no. 1,pp.91–105, 2013. [36] D. J. McFarland, L. A. Miner, T. M. Vaughan, and J. R. Wolpaw, [54] G. Dumas, R. Soussignan, L. Hugueville, J. Martinerie, and J. “Mu and beta rhythm topographies during motor imagery and Nadel, “Revisiting mu suppression in autism spectrum disor- actual movements,” Brain Topography,vol.12, no.3,pp.177–186, der,” Brain Research,vol.1585, pp.108–119,2014. [55] R. Salmelin, M. Ham ¨ al ¨ ainen, ¨ M. Kajola, and R. Hari, “Func- [37] D. Wechsler, Wechsler Abbreviated Scale of Intelligence,The tional segregation of movement-related rhythmic activity in the Psychological Corporation: Harcourt Brace & Company, New human brain,” NeuroImage,vol.2,no.4, pp.237–243,1995. York, NY, USA, 1999. [56] R. Hari, R. Salmelin, J. P. Mak ¨ ela, ¨ S. Salenius, and M. Helle, [38] M. Annett, LefT, Right, Hand, and Brain: eTh Right Shift eTh ory , “Magnetoencephalographic cortical rhythms,” International Hove, UK, 1985. Journal of Psychophysiology,vol.26,no.1-3,pp. 51–62,1997. [39] K. Gotham, S. Risi, A. Pickles, and C. Lord, “eTh Autism Diag- [57] S. G. Thorpe, E. N. Cannon, and N. A. Fox, “Spectral and source nostic Observation Schedule: Revised Algorithms for Improved structural development of mu and alpha rhythms from infancy Diagnostic Validity,” Journal of Autism and Developmental through adulthood,” Clinical Neurophysiology,vol.127,no. 1, pp. Disorders,vol.37, no.4,pp.613–627, 2007. 254–269, 2016. [40] J. N. Constantino, “Validation of a brief quantitative measure [58] J. Kaiser, N. Birbaumer, and W. Lutzenberger, “Event-related of autistic traits: comparison of the social responsiveness scale beta desynchronization indicates timing of response selection with the autism diagnostic interview-revised,” Journal of Autism in a delayed-response paradigm in humans,” Neuroscience and Developmental Disorders,vol.33, no.4,pp.427–433, 2003. Letters,vol.312,no. 3,pp.149–152,2001. [41] W. Barratt, “eTh Barratt Simplified Measure of Social Status,” [59] L. M. F. Doyle, K. Yarrow, and P. Brown, “Lateralization of http://wbarratt.indstate.edu/socialclass/Barratt 2006. event-related beta desynchronization in the EEG during pre- cued reaction time tasks,” Clinical Neurophysiology,vol.116,no. [42] G. Barrett, H. Shibasaki, and R. Neshige, “A computer-assisted 8, pp. 1879–1888, 2005. method for averaging movement-related cortical potentials with respect to EMG onset,” Electroencephalography and Clin- [60] E. Gowen, “Imitation in autism: Why action kinematics matter,” ical Neurophysiology,vol.60,no.3,pp. 276–281, 1985. Frontiers in Integrative Neuroscience,2012. [43] J. T. P. Talairach, Co-Planar Stereotaxic Atlas of the Human [61] M. J. Kurz, K. M. Becker, E. Heinrichs-Graham, and T. W. Brain, Thieme, New York, NY, USA, 1988. Wilson, “Neurophysiological abnormalities in the sensorimotor 12 Autism Research and Treatment cortices during the motor planning and movement execution [78] B. Feige, A. Aertsen, and R. Kristeva-Feige, “Dynamic syn- stages of children with cerebral palsy,” Developmental Medicine chronization between multiple cortical motor areas and muscle &Child Neurology, vol. 56, no. 11, pp. 1072–1077, 2014. activity in phasic voluntary movements,” Journal of Neurophys- iology,vol.84, no.5,pp. 2622–2629, 2000. [62] E. Heinrichs-Graham, T. W. Wilson, P. M. Santamaria et al., “Neuromagnetic evidence of abnormal movement-related beta [79] R.Kristeva-Feige,C.Fritsch,J.Timmer,andC.-H.Luc ¨ king, desynchronization in Parkinson’s disease,” Cerebral Cortex,vol. “Eeff cts of attention and precision of exerted force on beta 24,no. 10,pp. 2669–2678, 2014. range EEG-EMG synchronization during a maintained motor contraction task,” Clinical Neurophysiology, vol. 113, no. 1, pp. [63] E. C. Dias, S. Bickel, M. L. Epstein, P. Sehatpour, and D. C. 124–131, 2002. Javitt, “Abnormal task modulation of oscillatory neural activity [80] J. L. R. Rubenstein and M. M. Merzenich, “Model of autism: in schizophrenia,” Frontiers in Psychology,vol.4,2013. increased ratio of excitation/inhibition in key neural systems,” [64] I. Puzzo, N. R. Cooper, P. Vetter, and R. Russo, “EEG activation Genes, Brain and Behavior,vol.2,no.5, pp.255–267,2003. differences in the pre-motor cortex and supplementary motor area between normal individuals with high and low traits of [81] G. C. Carlson, “Glutamate receptor dysfunction and drug autism,” Brain Research,vol.1342,pp.104–110,2010. targets across models of autism spectrum disorders,” Pharma- cology Biochemistry & Behavior,vol.100,no.4,pp.850–854, [65] A. K. Churchland, R. Kiani, and M. N. Shadlen, “Decision- making with multiple alternatives,” Nature Neuroscience, vol. 11, no. 6, pp. 693–702, 2008. [82] S.H. Fatemi,T.J.Reutiman,T. D.Folsom,and P. D.Thuras, “GABA receptor downregulation in brains of subjects with [66] H. Stieglitz Ham, M. Corley, G. Rajendran, J. Carletta, and 𝐴 autism,” Journal of Autism and Developmental Disorders,vol.39, S. Swanson, “Brief report: Imitation of meaningless gestures no. 2, pp. 223–230, 2009. in individuals with Asperger syndrome and high-functioning autism,” Journal of Autism and Developmental Disorders,vol.38, [83] M.S. Brown,D.Singel,S.Hepburn,andD.C. Rojas, “Increased no. 3, pp. 569–573, 2008. glutamate concentration in the auditory cortex of persons with autism and first-degree relatives: A 1H-MRS study,” Autism [67] H. J. Stewart, R. D. Mcintosh, and J. H. G. Williams, “A spe- Research,vol.6,no. 1,pp.1–10,2013. cific deficit of imitation in autism spectrum disorder,” Autism Research,vol.6,no.6, pp.522–530,2013. [84] W. Gaetz, L. Bloy, D. J. Wang et al., “GABA estimation in the brains of children on the autism spectrum: Measurement [68] S. J. Rogers and B. F. Pennington, “A theoretical approach to the precision and regional cortical variation,” NeuroImage,vol.86, deficits in infantile autism,” Development and Psychopathology, pp.1–9,2014. vol. 3, no. 2, pp. 137–162, 1991. [85] D. C. Rojas, D. Singel, S. Steinmetz, S. Hepburn, and M. S. [69] W. N. Schofield, “Hand movements which cross the body mid- Brown, “Decreased left perisylvian GABA concentration in line: findings relating age differences to handedness,” Perceptual children with autism and unaeff cted siblings,” NeuroImage,vol. and Motor Skills,vol.42,no.2,pp.643–646,1976. 86, pp. 28–34, 2014. [70] N.M.G.Salowitz,P.Eccarius,J.Karstetal.,“Briefreport:Visuo- [86] N. A.J. Puts,E.L.Wodka,A.D.Harrisetal.,“ReducedGABA spatial guidance of movement during gesture imitation and and altered somatosensory function in children with autism mirror drawing in children with autism spectrum disorders,” spectrum disorder,” Autism Research,vol.10, no.4,pp. 608–619, Journal of Autism and Developmental Disorders,vol.43,no.4, pp.985–995,2013. [87] G. Dawson, S. Webb, G. D. Schellenberg et al., “Defining the [71] K. S. Wild, E. Poliako,ff A. Jerrison, and E. Gowen, “Goal- broader phenotype of autism: Genetic, brain, and behavioral directed and goal-less imitation in autism spectrum disorder,” perspectives,” Development and Psychopathology,vol.14, no.3, JournalofAutismandDevelopmentalDisorders,vol.42, no.8, pp. 581–611, 2002. pp.1739–1749,2012. [72] T. W. Wilson, E. Slason, R. Asherin et al., “An extended motor network generates beta and gamma oscillatory perturbations during development,” Brain and Cognition,vol.73,no.2,pp.75– 84, 2010. [73] P. J. Uhlhaas and W. Singer, “Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophys- iology,” Neuron,vol.52,no.1,pp.155–168, 2006. [74] F. Cassim, C. Monaca, W. Szurhaj et al., “Does post-movement beta synchronization reflect an idling motor cortex?” NeuroRe- port,vol.12, no.17,pp.3859–3863,2001. [75] G. Pfurtscheller, C. Neuper, C. Brunner, and F. Lopes Da Silva, “Beta rebound aer ft different types of motor imagery in man,” Neuroscience Letters,vol.378,no.3, pp.156–159,2005. [76] R. Chen, Z. Yaseen, L. G. Cohen, and M. Hallett, “Time course of corticospinal excitability in reaction time and self- paced movements,” Annals of Neurology,vol.44,no.3,pp.317–325, [77] B. Feige, R. Kristeva-Feige, S. Rossi, V. Pizzella, and P.-M. Rossini, “Neuromagnetic study of movement-related changes in rhythmic brain activity,” Brain Research,vol.734,no.1-2, pp. 252–260, 1996. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 International Journal of Journal of Immunology Research Endocrinology Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com BioMed PPAR Research Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Neurology Research and Treatment Cellular Longevity Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Autism Research and Treatment Hindawi Publishing Corporation

Neuromagnetic Beta-Band Oscillations during Motor Imitation in Youth with Autism

Loading next page...
 
/lp/hindawi-publishing-corporation/neuromagnetic-beta-band-oscillations-during-motor-imitation-in-youth-hHsD2pwiLD
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2018 I. Buard et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2090-1925
eISSN
2090-1933
DOI
10.1155/2018/9035793
Publisher site
See Article on Publisher Site

Abstract

Hindawi Autism Research and Treatment Volume 2018, Article ID 9035793, 12 pages https://doi.org/10.1155/2018/9035793 Research Article Neuromagnetic Beta-Band Oscillations during Motor Imitation in Youth with Autism 1 1 1 2 3 I. Buard , E. Kronberg, S. Steinmetz, S. Hepburn, andD.C.Rojas Department of Neurology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Department of Psychiatry, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Department of Psychology, Colorado State University, Fort Collins, CO, USA Correspondence should be addressed to D. C. Rojas; don.rojas@colostate.edu Received 1 March 2018; Revised 22 June 2018; Accepted 12 July 2018; Published 25 July 2018 Academic Editor: Bennett L. Leventhal Copyright © 2018 I. Buard et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Children with ASD oen ft exhibit early difficulties with action imitation, possibly due to low-level sensory or motor impairments. Impaired cortical rhythms have been demonstrated in adults with ASD during motor imitation. While those oscillations reflect an age-dependent process, they have not been fully investigated in youth with ASD. We collected magnetoencephalography data to examine patterns of oscillatory activity in the mu (8-13 Hz) and beta frequency (15-30 Hz) range in 14 adolescents with and 14 adolescents without ASD during a fine motor imitation task. Typically developing adolescents exhibited adult-like patterns of motor signals, e.g., event-related beta and mu desynchronization (ERD) before and during the movement and a postmovement beta rebound(PMBR)aeft rthemovement.Incontrast,thosewithASDexhibitedstrongerbetaandmu-ERDandreducedPMBR.Behav- ioral performance was similar between groups despite differences in motor cortical oscillations. Finally, we observed age-related increases in PBMR and beta-ERD in the typically developing children, but this correlation was not present in the autism group. These results suggest reduced inhibitory drive in cortical rhythms in youth with autism during intact motor imitation. Furthermore, impairments in motor brain signals in autism may not be due to delayed brain development. In the context of the excitation- inhibition imbalance perspectives of autism, we oer ff new insights into altered organization of neurophysiological networks. 1. Introduction investigating not only the degree of impairment but also its underlying mechanism(s). Behavioral studies have investi- Autism Spectrum Disorder (ASD) is a complex disorder of gated potential links between degree of motor impairment brain development characterized, in varying degrees, by dif- and types of movements and/or movement contexts in autism ficulties in social interaction, verbal and nonverbal commu- (for reviews, see [11, 12]. However, there is a general lack nication, and repetitive behaviors [1]. As early as 20 months of of knowledge related to deficits among neural mechanisms age, children with autism exhibit a robust deficit in imitating responsible for orchestrating movements in ASD. the actions of other people [2, 3]. Diverse explanations for Voluntary movements are accompanied by changes in imitative difficulties in ASD have been proposed, including cortical rhythms that can be detected by electroencephalog- motor control [4] and sensory perception deficits [5]. Studies raphy (EEG) and magnetoencephalography (MEG). Distinct have found impairments in several aspects of motor function, oscillatory signals are associated with motor tasks but are differently modulated during movement imitation or obser- including coordination [6], gait [7], motor imitation [8], and movement preparation [9] in both adults and children with vation. autism. eTh term developmental dyspraxia has been used to First, movement-related changes in rhythmic activity in describe those deficits and has been proposed to be specific the mu-range (8-13 Hz) have been reported as early as infancy to autism [10]. While delayed or aberrant fine and gross [13]. Its pre- and perimovement suppression are known as motor movements in autism used to be popularly mistaken event-related desynchronization (ERD) during activation of for clumsiness, an increasing number of studies have been sensorimotor areas, followed by an increase after movement 2 Autism Research and Treatment onset, which has widely been reported as event-related improves from birth well into early adolescence [34, 35]. It synchronization (ERS) [14]. is unclear how maturational changes of the motor cortex may Second, rhythmic modulation in the ongoing beta (15- be aeff cted in autism spectrum disorders. Finally, while mu 30 Hz) rhythm follows a pattern similar to the mu rhythm and beta rhythms are generated around the same time relative [15] although ERS has been more specifically named post- to the movement but not from the same areas [36], their movement beta rebound (PMBR; [16]). It is known that many functional meanings are very distinct, which usually prevents experimental factors can aeff ct sensorimotor beta rhythms, from drawing conclusions based on results combined from including difficulty of the movement sequence, movement both oscillations. duration, and directional uncertainty (e.g., see [17, 18]). Beta In this study, we examined mu- and beta-band oscil- oscillations may also indicate the integrity of circuit-level lations in adolescents with ASD during a finger imitation and neurotransmitter function. eTh power of PMBR has task. eTh paradigm we chose involved simple finger-lifting rd previously been associated with inhibitory brain function. imitative movement performed from the3 -person perspec- For example, Gaetz et al. found that PMBR, but not beta-ERD, tive and from computer-generated human hand videos. We was correlated with the concentration of GABA measured predictedthatmotor-betarhythms wouldbeimpairedinthe from magnetic resonance spectroscopy in the sensorimotor autism group due to their motor and/or imitation problems. cortex [19]. Others have found that using direct pharmaco- Specicall fi y, we hypothesized that beta-ERD signal would be logical manipulation of GABA-A receptors, while not having higher (i.e., greater beta suppression) in ASD because of its direct eeff cts on ERD or PMBR, results in a general increase relevance to difficulty with movement preparation. Similarly, in spontaneous beta, which in turn predicts ERD and PMBR we expected a weaker beta-PMBR due to its association [20]. Oscillatory patterns in the beta-range of sensorimotor with cortical inhibitory processes, which are predicted to areas may therefore provide cortical signatures relevant to be impaired in autism. We expected to see lower mu- circuit dysfunction. suppressioninthe autism group, aspreviously showninthe Third, a high-gamma band ( ∼70-90 Hz) ERS is sometimes literature. observed at the onset of movement [21]. Abnormalities in mu and beta rhythms have been 2. Material and Methods described in ASD patients while performing motor imitation tasks, such as reduced mu-suppression during movement 2.1. Study Subjects. Participants were 28 right-handed ado- observation [22–24], although the relatively small sample lescents (Table 1). Subjects were matched for age and intel- sizes in these studies (less than 20 people per group) calls ligence quotient (FSIQ), using the 4-subtest version of the for caution in the generalization of those findings and Wechsler Abbreviated Scale of Intelligence (WASI; [37]). the need for replicative studies. These observations, among Handedness was assessed in all subjects using the Annett others, were interpreted as supportive of a “broken mirror” Handedness Questionnaire [38]. In the ASD group, ado- theory of autism involving mirror neuron circuitry [25]. lescents met DSM-IV criteria for ASD, as determined by Other studies, however, have revealed no group differences consensus of the Autism Diagnostic Observation Schedule in mu-band activity during action observation or imitation (ADOS, [39]), DSM-IV diagnosis and a parent report of [26, 27]. Reduction of PMBR during action observation has ASD symptoms using the Social Responsiveness Scale (SRS; been shown in adults with ASD compared to controls [28], [40]) and review of all available data by a clinical psy- although not in adults with Asperger syndrome [29]. To chologist (S.H.). Interobserver reliability of ADOS scores date, however, beta rhythms in children and adolescents with is assessed in 20% of cases, with ICCs ranging from .72 ASD have not been investigated during motor imitation. to .94. A second diagnostician independently completed a Oberman’s group has recently extended their work on mu- record review of 50% of cases concurred with ASD diag- suppression, reporting that it increases during childhood nosis for all cases reviewed. All subjects signed informed and adolescence and independently of an autism diagnosis consent and assent to participate in the study consistent with [30]. us, Th developmental delay of those motor-generated the guidelines of the Colorado Multiple Institution Review oscillations, rather than deviance from typical development, Board. does not support the “broken mirror” hypothesis in autism [25]. eTh strong developmental gradient in mu-suppression, as well as beta-ERD and PMBR, makes it important to 2.2. Stimuli and Experimental Design. The stimuli consisted distinguish studies involving children from those with adults. of a photorealistic animated right hand, presented in the Although transcranial magnetic stimulation studies suggest third-person perspective (Figure 1). eTh index or pinky that corticospinal motor pathways are fully developed in n fi gers from this hand were lifted brieyfl 3 s after the early adolescence [31, 32], there is other evidence suggesting beginning of the video (1 s duration for entire movement, further development of the motor cortex and its associated returning to rest) every 6 s. Subjects were asked to imitate cortical oscillations well into the adolescent period. PMBR, the finger movement with their right index or pinky finger for example, appears to strongly develop throughout adoles- as seen on the screen. Index and pinky imitation stimuli were cence. An MEG study observed limited PMBR in 4- to 6- presented in randomized order to the subject using E-prime year-olds and higher levels in adolescents aged 11 to 13, but 2.0(Psychology SoftwareTools,Inc.).Atotalof806-strials stillsignicfi antlylower compared with youngadults[33]. It were presented for each condition (160 total trials, for 16 is also known that ne fi motor control in healthy children minutes’ total experiment duration). Autism Research and Treatment 3 Table 1: Participants’ characteristics. ASD Controls 𝜒 / t value P value 7 Autistic Disorder N and DSM-IV diagnosis 5Asperger’s 14 2PDD-NOS Age 14.5±2. 8 13.8±2. 8 0.66 0.52 Male/female 13/1 11/3 n/a 0.16 Handedness∗ 0.8±0.3 0.8±0.2 0.46 0.65 IQ 106.5±19.2 110.3±15.8 0.57 0.58 SRS∗ 104.6±21.5 n/a n/a n/a SES∗ 48.2±10.5 49.2±9.2 0.23 0.82 ∗ handedness scores were obtained using the Annett handedness questionnaire [38]; SRS (Social Responsiveness Scale) is a brief informant-based measurement of autism traits [40]; SES (Socioeconomic Status) scores based on the Barratt modified measure of social status [41]. Figure 1: Right-hand third-person representation showing the hand at rest (left) and while performing an index lift movement (right). 2.3. MEG Data Acquisition, Preprocessing, and Coregistration a procedure adapted from the preprocessing of electromyo- with Structural MRI. MEG data were obtained in a magneti- graphy data for trigger definition [42]. eTh definition of cally shielded room (ETS-Lindgren, Cedar Park, TX, USA) movement onset was then defined as the point at which using a Magnes 3600 WH whole-head MEG device (4D the accelerometer signals exceeded 2.5 standard deviations Neuroimaging, San Diego, CA, USA), comprised of 248 first- of themeansignalwithaminimumdurationbetween order axial-gradiometer sensors (5 cm baseline) in a helmet- onsets of 5 s. MEG trials were den fi ed with an epoch shaped array. Five head position indicator coils attached to duration of 5500 ms, with 0 ms being the accelerometer- the subject’s scalp were used to determine the head position defined movement onset. Epochs were baseline corrected with respect to the sensor array. eTh locations of the coils with (-2500 to -1500 ms premovement onset) and those trials respect to three anatomical landmarks (nasion and preauric- contaminated by excessively large MEG amplitudes (±2,000 ular points, with the intersection of the tragus and daith of fT) were rejected from further analysis. A mean of 96 (±29) the ear defining the preauriculars) and 2 extra nonfiducial and 106 (±17) artifact-free epochs for the autism and control points as well as the scalp surface (approximately 500 points) groups, respectively, was used in further analyses. No group were determined with a 3D digitizer (Polhemus, Colchester, dieff rence was observed between groups for the remaining VT, USA). The MEG signals were acquired continuously in artifact-free trials, t(26)=0.33, p=0.74. Data from excessive a 0.1-200 Hz bandwidth and sampled at 678.17 Hz and 24-bit noise or movement artifacts were not included but small in- vertical resolution. scanner head movements have not been corrected. Single axis monolithic integrated circuits Leadless Chip Each participant’s MEG data were coregistered with Carrier (LCC) accelerometers (model ADXL103; Analog structural T1-weighted magnetic resonance imaging (MRI) Devices, Inc.) were attached to both index and pinky fin- data prior to source space analyses (see below MRI acquisi- gertips in order to precisely quantify movement onset. eTh tion procedures) using common landmarks from the MEG chips are wired to approximately 3.3 m of light weight, highly digitization procedure and MRI scan data. Structural MRI flexible, miniature cable (Cooner Wire NMVF 4/30-4046) data were aligned parallel to the anterior and posterior with local bypass capacitors (0.1 uf) and encapsulated in commissures and transformed into the Talairach coordinate heat-shrink. Accelerometer signals were high-pass filtered system[43]usingtheBrainElectricalSourceAnalysis(BESA) at20Hz,rectiefi d, andthenlow-passfilteredat10 Hz in MRI software (BESA MRI version 2.0). 4 Autism Research and Treatment 2.4. MEG Time-Frequency Transformation. MEG postpro- equivalent dipole over time and is a robust estimator for cessing was performed using BESA 5.3 (MEGIS Software source location, because it has only 3 degrees of freedom GmbH, Grafelfing, Germany). Artifact-free epochs (mean for the tfi ting procedure. To obtain the orientation of the per condition: 51 +/- 12) in the time-domain were trans- equivalent dipole, the regional source was rotated in such a formed to the time-frequency domain with a 2 Hz/25 ms way that one of the three dipoles represented the most dipole sampling in BESA using complex demodulation [44]. This orientation at the time of the motor-evoked peak. eTh n, complex demodulation consists of a multiplication of the normalized source power in the mu- and beta-range was time-domain signal with a complex periodic exponential computed from the corresponding source waveform in each function, with a frequency equal to the frequency under ROI (left and right motor regions) against noise estimated analysis, and a subsequent low-pass filter. This low-pass filter during baseline. is a ni fi te impulse response filter of Gaussian shape in the Mean power was extracted between -500 and 500 ms for time-domain, which is related to the envelope of the moving beta-ERD and 1000 ms for mu-ERD and between 750 and window in wavelet analysis. For our setting of a 2 Hz/25 2000 ms for PMBR. For statistical analyses, time-frequency ms time-frequency sampling, this filter has a width in the results were subjected to group statistical analysis in a 2 x 2 frequency domain of 5.7 Hz and in the time-domain of x 2 mixed design ANOVA statistical test (group by finger by 79 ms full width at half maximum [45]. Time-frequency hemisphere) with n fi ger and hemisphere treated as within- analyses were computed for each sensor individually, per subjects measures. Separate ANOVAs were computed for the trial, averaged over trials, and then normalized by dividing ERD and PMBR windows. the power of each time-frequency bin by the respective fre- quency’s mean baseline power. This normalization procedure 2.6. MR Data Acquisition. MR images and spectra were allowed task-related mu (8-13 Hz) and beta (15-30 Hz) power acquired using a 3.0T GE Signa HDx whole body, long uc fl tuations to be readily visualized in sensor space as the bore MR scanner (GE Healthcare, Waukesha, WI, USA) at following: a decrease (blue, Figure 2) was observed priorly the Brain Imaging Center, University of Colorado Denver. and extending to the movement and imaged using a -500 to Subjects were imaged in the supine position using a GE eight- 500ms(1000msformu-ERD)time window(withtime0 channel phased array head coil. To comply with age- and being the onset of the movement). Following the movement, population-related behaviors such as boredom and restless- increased beta-band power (red, Figure 2) was imaged using a ness, subjects watched a movie during the exams using MR- 750 to 2000 ms time window. Time-frequency bins of interest compatible goggles and headphones (Resonance Technology were chosen to focus on maximum ERD/PMBR responses in Inc., Northridge, CA, USA) during the procedure. A T1- the ipsilateral and contralateral sides of the MEG sensor array, weighted sequence was acquired for tissue segmentation as previously described [46, 47]. Baseline mu and beta power using a 3D inversion recovery fast, spoiled gradient echo (IR- were extracted within a -2500 to -1500 ms time window. SPGR) technique (matrix 256 x 256, FOV 22 cm, TR/TE/TI= 10/3/450 ms, NEX=1), resulting in 168 1.2 mm thick axial slices 2.5. MEG Source Reconstruction and Statistical Analyses. 2 with an in-plane resolution of .86 mm . Using BESA Research 5.3, cortical networks were imaged through an extension of the linearly constrained minimum 2.7. Finger Movement Accuracy. Behavioral analyses of cor- variance vector beamformer [48–50], which employs spatial rect responses were determined from the accelerometer data. filters in the frequency domain to calculate source power Correct trials were den fi ed as subject movements occurring for the entire brain volume. The single images are derived on the correctly indicated n fi ger within 3 s aer ft the move- from the cross-spectral densities of all combinations of ment onset from the video displayed on the screen (i.e., MEG sensors averaged over the time-frequency range of from 1 s to 4 s aer ft stimulus presentation). Subjects in both interest,andthesolutionoftheforwardproblem foreach groups failed to respond with either nge fi r on some trials, location on a grid specified by input voxel space (7 mm and these trials were excluded from the accuracy calculation. cubic voxels in Talairach space). Following convention, the We analyzed this factor separately as level of responsiveness, source power in these images was normalized per participant defined as the number of trials responded to as a percentage using a prestimulus noise period of equal duration and of the trials presented. Accuracy, response times, and level bandwidth [48]. In the dipole-fit model used in BESA, a set of responsiveness were extracted for each participant and of consecutive time points is considered in which dipoles averaged across trials. Separate 2 x 2 ANOVA designs (group areassumedtohavefixedpositionand xfi ed orvarying by finger, with finger as a repeated measure) were used to orientation. In the final analysis, we used a regional source assess each behavioral variable. with xfi ed location and orientation throughout the analysis window because this way the resulting source waveform represents the time course of activity in the region of interest. 3. Results The regional source was placed on one of the rfi st maxima of source activation, which typically was over in the motor 3.1. Behavioral Results. Participants with ASD performed area (identifiable by its “omega” or “epsilon” shape, as widely their movements around 3.23±.24 s aer ft the video movement reported [51]). A regional source is a set of three orthogonal onset, averaged across both fingers, while control children dipoles that represent the electrical activity of a small brain imitated the n fi ger-lifting movements aer ft 3.13 ±.12 s. No volume independent of changes of the net orientation of the significant main effect of group was observed, 𝐹(1,26) = Autism Research and Treatment 5 contralateral ipsilateral 30 30 50 50 40 40 25 25 30 30 20 20 20 20 10 10 CTL 15 15 −10 −10 −20 10 10 −20 −30 −30 −2000 −1000 0 1000 2000 −2000 −1000 0 1000 2000 Time (ms) Time (ms) 30 50 25 30 ASD −10 −10 10 −20 −20 −30 −30 −2000 −1000 0 1000 2000 −2000 −1000 0 1000 2000 Time (ms) Time (ms) Figure 2: Grand average of time-frequency spectra of CTL (control, top) and ASD (bottom) children. TFR plots are derived from beamforming source images localized in the contralateral and ipsilateral motor areas and time-locked to movement termination (accelerometer onset). Cortical oscillations in the mu- (8-13Hz) and beta- (15-30Hz) bands showing event-related desynchronization (ERD, blue) at movement onset and postmovement beta rebound (PMBR, yellow-red) time regions are characterized. Source power during the baseline period was subtracted from source power during movement or aer ft the movement intervals. Control subjects had a significant power decrease (ERD) in percent change from baseline in the beta frequency range beginning before movement and lasting throughout the duration of movement as well as a mu-ERD around the movement, which were followed by a strong beta power increase (PMBR) in percent change from baseline beginning shortly aer ft movement termination. On the contrary, ASD subjects had a significantly greater power decrease (ERD) compared to controls in the beta and mu frequencies range beginning before movement and lasting beyond the movement, which was followed by a weak or no beta power increase (PMBR) beginning shortly aer ft movement termination. .72, p> .05. The group by finger interaction term was also 3.2. Time-Frequency Results. As expected, we found relevant nonsignificant, 𝐹(1,26) = .11, p> .05. This suggests the appro- motor-associated beta and mu oscillations in both hemi- priateness of fixed time bins for motor-related oscillations in spheres, contralateral and ispilateral to the movement, during this specific study, since it is unlikely that group differences imitation of both fingers (Figures 2 and 3). in motor-related oscillations were due to delayed movements For beta-ERD, the main effect of group was signicfi ant, in the autism group. For accuracy, the main effect of group 𝐹(1,26) = 53.02, p < .001, indicating greater ERD in the was at the edge of statistical significance, 𝐹(1,26) =1.82, p = autism group relative to controls (i.e., greater suppression, .05 (Table 2). The group by n fi ger interaction term was non- see Figures 2 and 3, top). The hemisphere main effect was significant, 𝐹(1,26) = 1.53, p> .05. Finally, we looked at level also signica fi nt, 𝐹(1,26) =12.18, p < .001, indicating a of responsiveness and found that control children responded stronger beta-ERD in the left hemisphere, contralateral to on a significantly higher percentage of trials (73.39 +/- 17.04) the movement. The finger main effect was nonsignificant, than their aeff cted peers (66.48 +/- 14.91), 𝐹(1,26) =6.99, p< 𝐹(1,26) =.05, p> .05. The group by hemisphere eeff ct was .05. For responsiveness, no significant differences were noted nonsignificant, 𝐹(1,26) =2.14, p> .05. The group by n fi ger forthe maineeff ctoffinger,𝐹(1,26) =2.12, p> .05, or for the interaction was nonsignificant, 𝐹(1,26) =1.32, p> .05. The group by finger interaction term, 𝐹(1,26) =0.69, p> .05. hemisphere by finger effect was nonsignificant, 𝐹(1,26) = Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) 6 Autism Research and Treatment Table2:Behavioralresults. Finger ASD Controls Index Reaction time± SD 3.23± 0.24 3.13± 0.11 Accuracy± SD 94.10%± 6.95 98.80%± 1.84 Responsiveness± SD 60.70%± 14.40 72.3%± 13.48 Pinky Reaction time± SD 3.24%± 0.25 3.14%± 0.14 Accuracy± SD 94.00%± 7.15 95.90%± 6.51 Responsiveness± SD 65.90%± 18.16 74.80%± 15.19 SD, standard deviation. 1.77, p > .05. Finally, the group by finger by hemisphere interaction was also nonsignificant, 𝐹(1,26) =1.61, p> .05. Cohen’seeff ct sizeswere1.1 andhigherfor each beta-ERD group comparison, indicating that our significant results were likely meaningful. −10 For PMBR, the main effect of group was signicfi ant, 𝐹(1,26) =26.51, p< .001, indicating greater PMBR in the control group relative to the autism group (i.e., greater synchronization, see Figures 2 and 3, bottom). In contrary to −20 thebeta-ERD,thehemispheremaineeff ct wasnotsignicfi ant, 𝐹(1,26) =2.04, p> .05. The finger main effect was nonsignif- icant,𝐹(1,26) =.54, p> .05. The group by hemisphere effect was also nonsignificant, 𝐹(1,26) =.56, p> .05. The group by −30 index pinky index pinky finger interaction was nonsignificant, 𝐹(1,26) =.86, p> .05. The hemisphere by finger effect was nonsignificant, 𝐹(1,26) ASD =.02, p> .05. Finally, the group by finger by hemisphere Control interaction was also nonsignificant, 𝐹(1,26) =.62, p> .05. Cohen’s effect sizes were 0.8 and higher for each PMBR group comparison, indicating that our significant results were likely Contralateral Ipsilateral meaningful. For mu-ERD, the main effect of group was significant, 𝐹(1,26) =9.59, p < .05, indicating greater mu-ERD in the autism group relative to controls (i.e., greater mu- suppression, see Figures 2 and 4). On the contrary to beta- ERD, the hemisphere main eeff ct for mu-ERD was not 10 significant, 𝐹(1,26) =1.48, p> .05. eTh n fi ger main effect was also nonsignificant, 𝐹(1,26) = .88, p> .05. The group by hemisphere effect was nonsignificant, 𝐹(1,26) =.52, p > .05. The group by n fi ger interaction was marginally significant, 𝐹(1,26) = 3.28, p = 0.073, indicating that the greater mu-ERD in the ASD group was more prominent for an index movement. eTh hemisphere by n fi ger effect was nonsignificant, 𝐹(1,26) =.06, p > .05. Finally, the group −10 by finger by hemisphere interaction was also nonsignificant, index pinky index pinky 𝐹(1,26) =0.21, p > .05. Cohen’s effect sizes were 1.1 and higher for left hemisphere mu-ERD group comparison, but ASD Control as low as 0.2 in the right hemisphere. This suggests that mu-suppression results may only be meaningful in the Figure 3: Children with ASD exhibit greater beta-ERD than their contralateral hemisphere. control peers but beta-PMBR is absent. Mean (+/- sem) beta-band Correlations between age and beta-ERD, beta-PMBR, ERD (top) and -PMBR (bottom) to Index and Pinky Imitation from and mu-ERD were examined in each group and each hemi- virtual electrodes contralateral (left) and ipsilateral (right) to right sphere using a Pearson r correlation coefficient. In the control hand of participant. group, there was a significant negative correlation between beta-PMBR beta-ERD (% change from baseline) (% change from baseline) Autism Research and Treatment 7 r = -.6, p = 0.02 −10 −20 −10 −30 −40 −50 −20 r = .6, p = 0.02 index pinky index pinky ASD Control Contralateral Ipsilateral Figure 4: Children with ASD exhibit greater mu-ERD than their control peers. Mean (+/- sem) mu rhythm ERD to Index and Pinky −10 Imitation from virtual electrodes contralateral (left) and ipsilateral −20 (right) to right hand of participant. 810 12 14 16 18 age (years) Figure 5: Beta correlation results. Representative plots of the age and beta-ERD during imitation of the index finger, correlation results assessing the relationship between age of the r(12) = -.6; p< .05 (Figure 5, top). It should be noted that subjects and contralateral beta-ERD (top) and -PMBR (bottom) this was only the case in the left hemisphere, contralateral power during index imitation. eTh r values were -.6 for ERD and to themovement.In thesamegroup,beta-PMBRpower .6 for PMBR for control participants (closed black circles, red was significantly correlated with age during index imitation, trendline). No significance was found for ASD participants (open r(12) = .6, p < .05 (Figure 5, bottom), and during pinky black circles) for both ERD and PMBR. imitation, r(12)=.55, p< .05, in the left hemisphere only. No correlation with beta power was found in the right hemisphere (p> .10) in the control group. In the autism group, mu and beta power changes associated with the imitations no significant correlation was found between either beta- were signicfi antly different from those of controls. While ERDorbeta-PMBRandage forany imitated movement and both affected and nonaffected children were able to perform any hemisphere (p> .10). Lastly, no significant correlation was the simple action of lifting a finger, their cortical activity found between mu-ERD and age for any group, any imitated levels were strikingly different. In the motor cortex, induced movement, and any hemisphere (p> .10). powerrevealedanincreaseinmu-andbeta-ERDanda Baseline mu and beta power were calculated for the reductioninPMBRintheASDgroup compared to the source reconstructed waveforms and group differences were control group, during imitation of both n fi ger movements. examined using a 2-sample t-test. There was no significant Our results provide some physiological evidence of distinct difference between the control group and the autism group brain activity associated with imitation of hand movements for baseline power in the beta frequency band (p> .10). in children with autism. Below, we discuss the implications of these findings for understanding the pathological cortical activity in children with autism. 4. Discussion Surprisingly, we found greater mu-ERD in the group of In the current study, children without ASD exhibited a children with autism compared to their nonaeff cted peers. well-established pattern of oscillatory neural activity before Whether this greater ERD is restricted to the motor-related and aeft r movement onsets in brain areas associated with signals or rather linked to the mirror neurons remains to motor processing. Beta and mu-ERD were observed prior to be clariefi d. Following the “broken mirror” theory of autism movement onset and during movement execution, whereas a [25],actionobservation maycausethesamefiring eeff cts strong PMBR response emerged following movement termi- as action execution suggesting that self-other mapping leads nation. os Th e responses were observed though contralateral to imitation deficits in autism [52]. However, a recent meta- and ispilateral sensorimotor cortices. Children with autism analysis provides compelling evidence that neuroimaging also exhibited each of these neural responses, although the studies are far from providing clear support to this hypothesis mu-ERD (% change from baseline) beta-ERD beta-PMBR (% change from baseline) (% change from baseline) 8 Autism Research and Treatment [53] and neurophysiology studies on mu-suppression also increased errors could have been expected in the ASD group show similar inconsistencies, especially with regard to two due to a confusion of which finger is moving on the screen. potential mu subbands [54]. Of note, mu-suppression is We did report a weaker accuracy in the autism group, expected during movement observation, prior to movement but only during the imitation of an index movement. This execution but this was not clearly captured in our data. This partly confirms the relevance of our findings to imitation couldpossiblybeduetothemethodbeing used.While MEG problems in general in autism. Previous research has also studies like ours model mu source activity using an equivalent suggestedthatpeoplewithautismhavemoredicffi ultywhen dipole, i.e., assuming that a small number of dipoles recorded the imitated movement is meaningless or less goal-directed mu activity [55, 56], most EEG studies reporting strong [71]. It is important to note that the gestures imitated in and widespread mu-suppression assume that cortical areas the present study were not inherently meaningful from a contributing to mu activity are distributed throughout the communication perspective. Further investigation contrast- brain [57]. Mu data acquisition is therefore not reliably well ing anatomical and mirror motor imitation and exploring captured using MEG instead of EEG. Alternatively, it is meaningful manipulations might provide some explanation possible that our method prevents from collecting results on these aspects. from diverse sources. Beamforming uses a spatial filter Previous studies have shown that beta-ERD power during designedtobefully sensitivetoactivityfromthe targetvoxel, simple finger movements is correlated with age [33, 72]. while being as insensitive as possible to activity from other In accordance with those studies, we also obtained similar brain regions. Indeed, motor-related mu- and beta-ERD are correlations although only in the control group. In children generated around the same time but from distinct brain areas. with autism, where stronger ERD is observed, no correlation We observed signica fi ntly greater ERD in the beta-band was found with age. This makes sense from a developmental in children with ASD. Given the beta-ERD’s association standpoint; cortical rhythms are resulting from synchroniza- with movement preparation [14] and cognitive selection of tion of a massive number of neurons, which themselves are a proper motor response [58, 59], these results provide a fully mature. Higher maturation stages in the motor cortex possible physiological mechanism for the difficulty of indi- lead to higher beta power. But higher beta-ERD does not viduals with ASD to imitate movements [60]. Increased beta- mean that children with ASD have a more mature brain, or ERD has been characterized in some motor-related disorders, both beta-ERD and -PMBR powers would be higher. In other such as cerebral palsy [61], but a decrease has been reported words, the beta impairments observed in the ASD group in others [62, 63]. Greater beta-ERD in autism has been might not likely be due to delayed maturation of the motor reported in a previous EEG study while subjects observed cortex. This conrfi ms the dysfunctional integrative theory static hands [64], although passively watching hand actions of autism [73]. Alternatively, the aberrant beta-ERD might did not produce any significant differences in beta-band be linked to the reduction in beta-PMBR. In other words, activity. This difference in the static condition might suggest if the sensorimotor circuitry underlying beta oscillations that, to the extent that ERD reflects movement preparation, children with autism is failing to generate synchronized beta the participants with autism had greater dicul ffi ty imagining a oscillations at rest, then premovement beta-ERD would be static hand moving, but not while actually watching the hand limited by the low resting oscillatory power. However, the move. In the current study, there was no static condition and similar baseline beta power in both groups rules out this subjects did not passively watch the stimuli, so direct com- hypothesis in the context of our study. Similarly, while PMBR parisons with the previous EEG paper are dicffi ult. Because is thought to reflect an age-dependent inhibitory process latency of motor responses as well as neural activity in the [33, 72], our data showing increased PMBR are not explained sensorimotor cortex during motor preparation, especially by a greater age range in the ASD group. beta-band ERD, covaries with movement uncertainty [17, PMBR is proposed to be associated with motor deacti- 65], beta-ERD may be strongly associated with movement vation or inhibition of the motor cortex by somatosensory selection. What seems clear is that beta-ERD abnormalities afferents [74], or a “resetting” of the underlying cortical canbeobservedinautisminavarietyofconditionsrelatedto networks [75]. Transcranial magnetic stimulation studies movement observation or execution. have conrm fi ed that beta-PMBR corresponds to a period Alternatively, difficulties with body part orientation [66] of decreased corticospinal excitability [76], suggesting that or self-other mapping [67, 68] have been proposed to under- it may represent a state of cortical inhibition. It has also lie imitation problems in autism. It is therefore possible that been suggested that several beta rhythms exist, each with the third-person perspective of our imitation paradigm is a different functional significance [74, 77]. For instance, partly responsible for the increased beta-ERD in the ASD postmovement beta rebound (but not premovement beta- group. Behavioral imitation studies have shown that, in a ERD) has been shown to be related to a prolonged period third-person perspective, the movement that is imitated more of increased corticomuscular coherence following phasic easily is the mirror, or specular hand, versus the anatomical voluntary movements [78] that, in turn, may reflect the level hand [69].Theideathatvisuospatialinformation processing of attention to motor performance [79]. Although a couple deficits may be contributing to functional motor coordina- of autism studies report a reduction in PMBR during action tion deficits in autism has already been contemplated [70]. observation [28, 29] but not during performance of the In this context, while beta-ERD is strongly associated with action, we provide evidence that postmovement beta signals movement selection [17], a greater ERD could be due to are also affected during action imitation. Two explanations a greater difficulty to choose which finger to move, and rise from those opposite findings. First, while we did not Autism Research and Treatment 9 include an observation only condition, it is possible that extent changes in motor-beta rhythms are directly related to neuronal circuits activated in our task proceed independently observable changes in motor behavior. of the mirror neuron system. Second, our sample might include children with greater motor impairments than those 5. Limitations of the other studies. Current theories and experimental data strongly suggest Given the cross-sectional nature of the study and small that dysfunctional integrative mechanisms in ASD result sample size, we would like to warn about the highly prelim- from reduced neuronal synchronization [73]. eTh underlying inary nature of these findings. In addition, the low number cellular mechanisms seem to be an imbalance between excita- of females might limit the generalization of the results. tion and inhibition [80], which leads to hyperexcitability and By essence, autism isaspectrumsothecharacteristicsof unstable cortical networks, as abnormalities in GABAergic children with ASD and their life circumstances are mostly and glutamatergic transmitter systems has been characterized heterogeneous in nature. Addressing these issues may require in humans and animal models of autism [81, 82]. Recent stud- larger sample sizes and possibly interdisciplinary collabora- ies from our group and others have demonstrated reduced tion. GABA and increased glutamate in some regions of the brain in children with ASD [83–86] with possible evidence of reduced GABA in the motor cortex [84]. In typically 6. Conclusion developing participants, Gaetz and colleagues [19] reported a correlation between motor cortical GABA concentration and We have demonstrated that children and adolescents with PMBR power. Consequently, it is expected that changes in autism may have reduced inhibitory drive in cortical rhythms cortical oscillatory rhythms, especially a reduced PMBR, will as measured with MEG during motor imitation. Our results be foundinthe brainsofASDchildren. In this context,our support previous theories that inhibitory dysfunction could results are consistent with the interpretation of PMBR as a be one of the factors underlying abnormal behaviors in marker of inhibitory neuronal signaling and the excitation- autism. Further, changes in ERD suggest greater difficulty inhibition (E-I) imbalance theory of autism [80]. in movement planning in the autism group. Understand- Finally, it is critical to consider the mixed picture of ing these mechanisms may provide a potential target for behavioral resultsinthe currentstudy.Wefound that the future therapies to address motor-related symptoms, by both participants with ASD responded to the imitative stimuli less pharmacological and behavioral interventions. Whereas the oen ft than controls but that when the participants with ASD relevance of altered brain oscillations to motor imitation responded, their responses were as accurate as those in the problems in autism needs further claricfi ation, monitoring control group. Reduced responding could be interpreted as pathological beta-bands features with MEG might hold evidence of confusion over the imitative action requested but promise as a biomarker for motor impairments in ASD. equally could be considered evidence of greater lapsing of On this last point, although a large number of individuals attention during the task. In the current study, we cannot with ASD have motor difficulties, they are not universally discern between these possibilities. Since we were focused observed [87]. Due to this heterogeneity, specicfi ation of on response-locked beta-band ERD and PMBR, we could motor impairments in autism may be useful for the identi- not analyze trials on which the subjects did not respond fication of clinically relevant subgroups in ASD. Moreover, to the stimuli, limiting our understanding of whether such a better understanding of the neurobiology of motor and/or trials were associated with additional differences in beta-band imitation impairments is vital for identifying treatments to activity. improve outcomes related to motor deficiencies. Whilebeta-ERDand-PBMRaregeneratedfromthesame regions, it is not clear whether they result from similar events Abbreviations at theneuronalornetworklevel. Ourcohortofchildrenwith ASD did not exhibit significant motor defects. We interpret ASD: Autism spectrum disorders the aberrant pattern of beta rhythms observed in our ASD MEG: Magnetoencephalography group, especially the increased ERD, as most likely associated ERD: Event-related desynchronization with the dicffi ulty of cognitive processes involved in selecting PMBR: Postmovement beta rebound the motor response rather than with a motor decfi it itself. EEG: Electroencephalography In contrast, the reduced PMBR may be related to reduced GABA: Gamma-aminobutyric acid inhibition in the motor cortices. Indeed, PMBR is absent DSM: Diagnostic and statistical manual of in young children [33], which suggests that the reduced mental disorders influence of inhibition in the motor cortices may represent an ADOS: Autism diagnostic observation schedule optimal physiological environment to facilitate motor learn- MR(I): Magnetic resonance (imaging). ing or to recover from motor delay. It could therefore also be interpreted to be a compensatory consequence of the ERD changes in the autism group. Another possible interpretation Data Availability ofreducedPMBRintheASDgroup isthat ofdevelopmental delay. The mixed pattern of aberrant beta oscillations poses Deidentiefi d data is available on our laboratory server in.mat an interesting question for further exploration—i.e., to what and.xlsx file format readily usable by any requester. 10 Autism Research and Treatment Additional Points [10] L. K. MacNeil and S. H. Mostofsky, “Specificity of dyspraxia in children with autism,” Neuropsychology,vol.26,no.2,pp.165– Highlights. (i) Motor oscillations are impaired in children 171, 2012. with ASD during motor imitation. (ii) Increased beta-ERD [11] R. Downey and M. J. K. Rapport, “Motor activity in children may be related to difficulties with movement selection with autism: A review of current literature,” Pediatric Physical and preparation. (iii) Weaker or absent beta-ERS suggests Therapy ,vol.24, no.1,pp. 2–20,2012. reduced inhibitory drive. [12] K.J. Steinman,S.H.Mostofsky,andM. B. Denckla, “Toward a narrower, more pragmatic view of developmental dyspraxia,” Journal of Child Neurology,vol.25,no.1,pp. 71–81,2010. Conflicts of Interest [13] M. Berchicci, T. Zhang, L. Romero et al., “Development of None of the authors have potential conflicts of interest to be Mu Rhythm in infants and preschool children,” Developmental Neuroscience, vol. 33, no. 2, pp. 130–143, 2011. disclosed. [14] G. Pfurtscheller, M. Pregenzer, and C. Neuper, “Visualization of sensorimotor areas involved in preparation for hand movement Acknowledgments basedonclassificationof𝜇 and central𝛽 rhythms in single EEG trials in man,” Neuroscience Letters,vol.181,no.1-2, pp.43–46, The authors’ work was supported by Autism Speaks (Post- doctoral Fellowship #7592, I. Buard), the Colorado Clinical [15] G. Pfurtscheller, “Central beta rhythm during sensorimotor andTranslationalScienceInstitute(Co-PilotAward,I.Buard) activities in man,” ElectroencephalographyandClinicalNeuro- and the National Institute of Health (R01 MH082820, D. C. physiology,vol.51,no.3,pp. 253–264, 1981. Rojas). eTh authors would like to thank Himaja Gaddipati [16] G. Pfurtscheller and F. H. L. da Silva, “Event-related EEG/MEG for technical assistance with some of the MEG recordings synchronization and desynchronization: basic principles,” Clin- andIan Southwellfor therealization of theanimatedhands ical Neurophysiology,vol.110,no.11, pp.1842–1857,1999. (http://www.southwellart.com). [17] C. Tzagarakis, N. F. Ince, A. C. Leuthold, and G. Pellizzer, “Beta-band activity during motor planning reflects response References uncertainty,” The Journal of Neuroscience , vol. 30, no. 34, pp. 11270–11277, 2010. [1] American Psychiatric Association (APA), Diagnostic and statis- [18] E. Heinrichs-Graham and T. W. Wilson, “Is an absolute level ticalmanualofmentaldisorders, American Psychiatric Publish- of cortical beta suppression required for proper movement? ing, Arlington, Va, USA, 5th edition, 2013. Magnetoencephalographic evidence from healthy aging,” Neu- [2] T. Charman, J. Swettenham, S. Baron-Cohen, A. Cox, G. Baird, roImage,vol.134,pp.514–521,2016. and A. Drew, “Infants with autism: An investigation of empathy, [19] W. Gaetz, J. C. Edgar, D. J. Wang, and T. P. L. Roberts, pretend play, joint attention, and imitation.,” Developmental “Relating MEG measured motor cortical oscillations to resting Psychology,vol.33,no.5,pp.781–789, 1997. 𝛾-a minobutyric acid (GABA) concentration,” NeuroImage,vol. [3] W.L.Stone,O.Y.Ousley,andC.D.Littleford, “Motorimitation 55,no.2, pp.616–621,2011. in young children with autism: what’s the object?” Journal of [20] S. D. Hall, I. M. Stanford, N. Yamawaki et al., “eTh role of Abnormal Child Psychology,vol.25,no.6,pp. 475–485, 1997. GABAergic modulation in motor function related neuronal [4] M.Vanvuchelen,H.Roeyers,and W. De Weerdt,“Nature of network activity,” NeuroImage, vol. 56, no. 3, pp. 1506–1510, 2011. motor imitation problems in school-aged boys with autism: a [21] T. Ball, E. Demandt, I. Mutschler et al., “Movement related activ- motororacognitiveproblem?” Autism,vol.11,no.3,pp. 225– ityinthehighgammarangeofthehumanEEG,” NeuroImage, 240, 2007. vol.41,no.2, pp.302–310,2008. [5] B.Ingersoll,L.Schreibman, andQ.H.Tran,“Eeff ctof Sensory [22] J. Martineau, S. Cochin, R. Magne, and C. Barthelemy, Feedback on Immediate Object Imitation in Children with “Impaired cortical activation in autistic children: Is the mirror Autism,” Journal of Autism and Developmental Disorders,vol. neuron system involved?” International Journal of Psychophysi- 33,no.6, pp.673–683,2003. ology,vol.68, no.1,pp.35–40,2008. [6] K.A.Fournier, C. J. Hass,S.K.Naik,N.Lodha,andJ.H. [23] L. M. Oberman, E. M. Hubbard, J. P. McCleery, E. L. Altschuler, Cauraugh, “Motor coordination in autism spectrum disorders: V. S. Ramachandran, and J. A. Pineda, “EEG evidence for mirror a synthesis and meta-analysis,” Journal of Autism and Develop- neuron dysfunction in autism spectrum disorders,” Cognitive mental Disorders, vol. 40, no. 10, pp. 1227–1240, 2010. Brain Research, vol. 24, no. 2, pp. 190–198, 2005. [7] J.A.Vilensky, A.R.Damasio, andR.G.Maurer,“Gaitdistur- [24] R. Bernier, G. Dawson, S. Webb, and M. Murias, “EEG mu bances in patients with autistic behavior. A preliminary study,” rhythm and imitation impairments in individuals with autism JAMA Neurology,vol.38,no.10,pp.646–649,1981. spectrum disorder,” Brain and Cognition,vol.64, no.3,pp. 228– [8] J. H. G. Williams, A. Whiten, and T. Singh, “A systematic review 237, 2007. of action imitation in autistic spectrum disorder,” Journal of Autism and Developmental Disorders,vol.34,no.3,pp.285–299, [25] L. M. Oberman and V. S. Ramachandran, “Preliminary evidence for deficits in multisensory integration in autism spectrum disorders: eTh mirror neuron hypothesis,” Social Neuroscience, [9] N. J. Rinehart, J. L. Bradshaw, A. V. Brereton, and B. J. vol. 3, no. 3-4, pp. 348–355, 2008. Tonge, “Movement Preparation in High-Functioning Autism and Asperger Disorder: A Serial Choice Reaction Time Task [26] R. Raymaekers, J. R. Wiersema, and H. Roeyers, “EEG study Involving Motor Reprogramming,” Journal of Autism and of the mirror neuron system in children with high functioning Developmental Disorders,vol.31,no.1,pp. 79–88, 2001. autism,” Brain Research, vol. 1304, pp. 113–121, 2009. Autism Research and Treatment 11 [27] L. Ruysschaert, P. Warreyn, J. R. Wiersema, A. Oostra, and H. [44] N. Papp and P. Ktonas, “Critical evaluation of complex demodu- Roeyers, “Exploring the Role of Neural Mirroring in Children lation techniques for the quantification of bioelectrical activity,” with Autism Spectrum Disorder,” Autism Research,vol.7,no.2, Biomed Sci Instrum,vol.13, pp.135–145,1977. pp.197–206,2014. [45] K.Hoechstetter,H.Bornfleth,D.Weckesser,N.Ille,P. Berg,and [28] E. Honaga, R. Ishii, R. Kurimoto et al., “Post-movement beta M. Scherg, “BESA source coherence: A new method to study rebound abnormality as indicator of mirror neuron system cortical oscillatory coupling,” Brain Topography,vol.16, no.4, dysfunctioninautisticspectrumdisorder:An MEGstudy,” pp. 233–238, 2004. Neuroscience Letters,vol.478,no. 3, pp.141–145,2010. [46] T. W. Wilson, E. Slason, R. Asherin et al., “Abnormal gamma [29] S. Avikainen, T. Kulomaki ¨ , and R. Hari, “Normal movement and beta MEG activity during finger movements in early-onset psychosis,” Developmental Neuropsychology,vol.36, no.5,pp. reading in Asperger subjects,” NeuroReport,vol.10,no.17,pp. 3467–3470, 1999. 596–613, 2011. [30] L. M. Oberman, J. P. McCleery, E. M. Hubbard et al., “Develop- [47] Y. Nishimura, Y. Ikeda, A. Suematsu, and S. Higuchi, “Eeff ct of mental changes in mu suppression to observed and executed visual orientation on mu suppression in children: a comparative actions in autism spectrum disorders,” Social Cognitive and EEG study with adults,” JournalofPhysiological Anthropology, vol. 37, no. 1, 2018. Affective Neuroscience ,vol.8,no. 3,ArticleIDnsr097,pp.300– 304, 2013. [48] B. D. Van Veen, W. Van Drongelen, M. Yuchtman, and A. Suzuki, “Localization of brain electrical activity via linearly con- [31] T. H. H. G. Koh and J. A. Eyre, “Maturation of corticospinal tracts assessed by electromagnetic stimulation of the motor strained minimum variance spatial filtering,” IEEE Transactions on Biomedical Engineering, vol. 44, no. 9, pp. 867–880, 1997. cortex,” Archives of Disease in Childhood,vol.63,no.11,pp.1347– 1352, 1988. [49] A. Hillebrand and G. R. Barnes, “Beamformer Analysis of MEG Data,” International Review of Neurobiology,vol.68, pp.149–171, [32] A. Nezu, S. Kimura, S. Uehara, T. Kobayashi, M. Tanaka, and K. Saito, “Magnetic stimulation of motor cortex in children: Maturity of corticospinal pathway and problem of clinical [50] M. Liljestrom, ¨ J. Kujala, O. Jensen, and R. Salmelin, “Neuromag- application,” Brain & Development,vol.19,no.3,pp.176–180, netic localization of rhythmic activity in the human brain: A comparison of three methods,” NeuroImage,vol.25, no.3,pp. 734–745, 2005. [33] W. Gaetz, M. MacDonald, D. Cheyne, and O. C. Snead, “Neu- romagnetic imaging of movement-related cortical oscillations [51] O. F. Kaneko, N. J. Fischbein, J. Rosenberg, M. Wintermark, in children and adults: Age predicts post-movement beta and M. M. Zeineh, “eTh “white gray sign” identifies the central rebound,” NeuroImage,vol.51, no.2,pp.792–807, 2010. sulcus on 3T high-resolution T1-weighted images,” American Journal of Neuroradiology, vol. 38, no. 2, pp. 276–280, 2017. [34] M. B. Denckla, “Development of Speed in Repetitive and Suc- cessive Finger-movements in Normal Children,” Developmental [52] M. Iacoboni and M. Dapretto, “eTh mirror neuron system Medicine & Child Neurology, vol. 15, no. 5, pp. 635–645, 1973. and the consequences of its dysfunction,” Nature Reviews Neuroscience,vol.7,no. 12,pp. 942–951, 2006. [35] P. H.Wol,C ff .E.Gunnoe,andC. Cohen,“ASSOCIATED MOVEMENTS AS A MEASURE OF DEVELOPMENTAL [53] A. F. D. C. Hamilton, “Reflecting on the mirror neuron system in AGE,” DevelopmentalMedicine&ChildNeurology,vol.25, no. autism: A systematic review of current theories,” Developmental 4, pp. 417–429, 1983. Cognitive Neuroscience,vol.3,no. 1,pp.91–105, 2013. [36] D. J. McFarland, L. A. Miner, T. M. Vaughan, and J. R. Wolpaw, [54] G. Dumas, R. Soussignan, L. Hugueville, J. Martinerie, and J. “Mu and beta rhythm topographies during motor imagery and Nadel, “Revisiting mu suppression in autism spectrum disor- actual movements,” Brain Topography,vol.12, no.3,pp.177–186, der,” Brain Research,vol.1585, pp.108–119,2014. [55] R. Salmelin, M. Ham ¨ al ¨ ainen, ¨ M. Kajola, and R. Hari, “Func- [37] D. Wechsler, Wechsler Abbreviated Scale of Intelligence,The tional segregation of movement-related rhythmic activity in the Psychological Corporation: Harcourt Brace & Company, New human brain,” NeuroImage,vol.2,no.4, pp.237–243,1995. York, NY, USA, 1999. [56] R. Hari, R. Salmelin, J. P. Mak ¨ ela, ¨ S. Salenius, and M. Helle, [38] M. Annett, LefT, Right, Hand, and Brain: eTh Right Shift eTh ory , “Magnetoencephalographic cortical rhythms,” International Hove, UK, 1985. Journal of Psychophysiology,vol.26,no.1-3,pp. 51–62,1997. [39] K. Gotham, S. Risi, A. Pickles, and C. Lord, “eTh Autism Diag- [57] S. G. Thorpe, E. N. Cannon, and N. A. Fox, “Spectral and source nostic Observation Schedule: Revised Algorithms for Improved structural development of mu and alpha rhythms from infancy Diagnostic Validity,” Journal of Autism and Developmental through adulthood,” Clinical Neurophysiology,vol.127,no. 1, pp. Disorders,vol.37, no.4,pp.613–627, 2007. 254–269, 2016. [40] J. N. Constantino, “Validation of a brief quantitative measure [58] J. Kaiser, N. Birbaumer, and W. Lutzenberger, “Event-related of autistic traits: comparison of the social responsiveness scale beta desynchronization indicates timing of response selection with the autism diagnostic interview-revised,” Journal of Autism in a delayed-response paradigm in humans,” Neuroscience and Developmental Disorders,vol.33, no.4,pp.427–433, 2003. Letters,vol.312,no. 3,pp.149–152,2001. [41] W. Barratt, “eTh Barratt Simplified Measure of Social Status,” [59] L. M. F. Doyle, K. Yarrow, and P. Brown, “Lateralization of http://wbarratt.indstate.edu/socialclass/Barratt 2006. event-related beta desynchronization in the EEG during pre- cued reaction time tasks,” Clinical Neurophysiology,vol.116,no. [42] G. Barrett, H. Shibasaki, and R. Neshige, “A computer-assisted 8, pp. 1879–1888, 2005. method for averaging movement-related cortical potentials with respect to EMG onset,” Electroencephalography and Clin- [60] E. Gowen, “Imitation in autism: Why action kinematics matter,” ical Neurophysiology,vol.60,no.3,pp. 276–281, 1985. Frontiers in Integrative Neuroscience,2012. [43] J. T. P. Talairach, Co-Planar Stereotaxic Atlas of the Human [61] M. J. Kurz, K. M. Becker, E. Heinrichs-Graham, and T. W. Brain, Thieme, New York, NY, USA, 1988. Wilson, “Neurophysiological abnormalities in the sensorimotor 12 Autism Research and Treatment cortices during the motor planning and movement execution [78] B. Feige, A. Aertsen, and R. Kristeva-Feige, “Dynamic syn- stages of children with cerebral palsy,” Developmental Medicine chronization between multiple cortical motor areas and muscle &Child Neurology, vol. 56, no. 11, pp. 1072–1077, 2014. activity in phasic voluntary movements,” Journal of Neurophys- iology,vol.84, no.5,pp. 2622–2629, 2000. [62] E. Heinrichs-Graham, T. W. Wilson, P. M. Santamaria et al., “Neuromagnetic evidence of abnormal movement-related beta [79] R.Kristeva-Feige,C.Fritsch,J.Timmer,andC.-H.Luc ¨ king, desynchronization in Parkinson’s disease,” Cerebral Cortex,vol. “Eeff cts of attention and precision of exerted force on beta 24,no. 10,pp. 2669–2678, 2014. range EEG-EMG synchronization during a maintained motor contraction task,” Clinical Neurophysiology, vol. 113, no. 1, pp. [63] E. C. Dias, S. Bickel, M. L. Epstein, P. Sehatpour, and D. C. 124–131, 2002. Javitt, “Abnormal task modulation of oscillatory neural activity [80] J. L. R. Rubenstein and M. M. Merzenich, “Model of autism: in schizophrenia,” Frontiers in Psychology,vol.4,2013. increased ratio of excitation/inhibition in key neural systems,” [64] I. Puzzo, N. R. Cooper, P. Vetter, and R. Russo, “EEG activation Genes, Brain and Behavior,vol.2,no.5, pp.255–267,2003. differences in the pre-motor cortex and supplementary motor area between normal individuals with high and low traits of [81] G. C. Carlson, “Glutamate receptor dysfunction and drug autism,” Brain Research,vol.1342,pp.104–110,2010. targets across models of autism spectrum disorders,” Pharma- cology Biochemistry & Behavior,vol.100,no.4,pp.850–854, [65] A. K. Churchland, R. Kiani, and M. N. Shadlen, “Decision- making with multiple alternatives,” Nature Neuroscience, vol. 11, no. 6, pp. 693–702, 2008. [82] S.H. Fatemi,T.J.Reutiman,T. D.Folsom,and P. D.Thuras, “GABA receptor downregulation in brains of subjects with [66] H. Stieglitz Ham, M. Corley, G. Rajendran, J. Carletta, and 𝐴 autism,” Journal of Autism and Developmental Disorders,vol.39, S. Swanson, “Brief report: Imitation of meaningless gestures no. 2, pp. 223–230, 2009. in individuals with Asperger syndrome and high-functioning autism,” Journal of Autism and Developmental Disorders,vol.38, [83] M.S. Brown,D.Singel,S.Hepburn,andD.C. Rojas, “Increased no. 3, pp. 569–573, 2008. glutamate concentration in the auditory cortex of persons with autism and first-degree relatives: A 1H-MRS study,” Autism [67] H. J. Stewart, R. D. Mcintosh, and J. H. G. Williams, “A spe- Research,vol.6,no. 1,pp.1–10,2013. cific deficit of imitation in autism spectrum disorder,” Autism Research,vol.6,no.6, pp.522–530,2013. [84] W. Gaetz, L. Bloy, D. J. Wang et al., “GABA estimation in the brains of children on the autism spectrum: Measurement [68] S. J. Rogers and B. F. Pennington, “A theoretical approach to the precision and regional cortical variation,” NeuroImage,vol.86, deficits in infantile autism,” Development and Psychopathology, pp.1–9,2014. vol. 3, no. 2, pp. 137–162, 1991. [85] D. C. Rojas, D. Singel, S. Steinmetz, S. Hepburn, and M. S. [69] W. N. Schofield, “Hand movements which cross the body mid- Brown, “Decreased left perisylvian GABA concentration in line: findings relating age differences to handedness,” Perceptual children with autism and unaeff cted siblings,” NeuroImage,vol. and Motor Skills,vol.42,no.2,pp.643–646,1976. 86, pp. 28–34, 2014. [70] N.M.G.Salowitz,P.Eccarius,J.Karstetal.,“Briefreport:Visuo- [86] N. A.J. Puts,E.L.Wodka,A.D.Harrisetal.,“ReducedGABA spatial guidance of movement during gesture imitation and and altered somatosensory function in children with autism mirror drawing in children with autism spectrum disorders,” spectrum disorder,” Autism Research,vol.10, no.4,pp. 608–619, Journal of Autism and Developmental Disorders,vol.43,no.4, pp.985–995,2013. [87] G. Dawson, S. Webb, G. D. Schellenberg et al., “Defining the [71] K. S. Wild, E. Poliako,ff A. Jerrison, and E. Gowen, “Goal- broader phenotype of autism: Genetic, brain, and behavioral directed and goal-less imitation in autism spectrum disorder,” perspectives,” Development and Psychopathology,vol.14, no.3, JournalofAutismandDevelopmentalDisorders,vol.42, no.8, pp. 581–611, 2002. pp.1739–1749,2012. [72] T. W. Wilson, E. Slason, R. Asherin et al., “An extended motor network generates beta and gamma oscillatory perturbations during development,” Brain and Cognition,vol.73,no.2,pp.75– 84, 2010. [73] P. J. Uhlhaas and W. Singer, “Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophys- iology,” Neuron,vol.52,no.1,pp.155–168, 2006. [74] F. Cassim, C. Monaca, W. Szurhaj et al., “Does post-movement beta synchronization reflect an idling motor cortex?” NeuroRe- port,vol.12, no.17,pp.3859–3863,2001. [75] G. Pfurtscheller, C. Neuper, C. Brunner, and F. Lopes Da Silva, “Beta rebound aer ft different types of motor imagery in man,” Neuroscience Letters,vol.378,no.3, pp.156–159,2005. [76] R. Chen, Z. Yaseen, L. G. Cohen, and M. Hallett, “Time course of corticospinal excitability in reaction time and self- paced movements,” Annals of Neurology,vol.44,no.3,pp.317–325, [77] B. Feige, R. Kristeva-Feige, S. Rossi, V. Pizzella, and P.-M. Rossini, “Neuromagnetic study of movement-related changes in rhythmic brain activity,” Brain Research,vol.734,no.1-2, pp. 252–260, 1996. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 International Journal of Journal of Immunology Research Endocrinology Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com BioMed PPAR Research Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Neurology Research and Treatment Cellular Longevity Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal

Autism Research and TreatmentHindawi Publishing Corporation

Published: Jul 25, 2018

References