Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 1.
Published in final edited form as: Clin Neurophysiol. 2020 Oct 16;132(1):180–190. doi: 10.1016/j.clinph.2020.09.020

Expanding our understanding of sensory gating in children with autism spectrum disorders

Jewel E Crasta a,1, William J Gavin b,c, Patricia L Davies a,c
PMCID: PMC7856194  NIHMSID: NIHMS1655047  PMID: 33310588

Abstract

Objective:

This study examined sensory gating in children with autism spectrum disorders (ASD). Gating is usually examined at the P50 component and rarely at mid- and late-latency components.

Methods:

Electroencephalography data were recorded during a paired-click paradigm, from 18 children with ASD (5–12 years), and 18 typically-developing (TD) children. Gating was assessed at the P50, N1, P2, and N2 event-related potential components. Parents of all participants completed the Short Sensory Profile (SSP).

Results:

TD children showed gating at all components while children with ASD showed gating only at P2 and N2. Compared to TD children, the ASD group showed significantly reduced gating at P50, N1, and P2. No group differences were found at N2, suggesting typical N2 gating in the ASD group. Time-frequency analyses showed reduced orientation and neural synchronization of auditory stimuli. P50 and N1 gating significantly correlated with the SSP.

Conclusion:

Although children with ASD have impaired early orientation and filtering of auditory stimuli, they exhibited gating at P2 and N2 components suggesting use of different gating mechanisms compared to TD children. Sensory deficits in ASD may relate to gating.

Significance:

The data provide novel evidence for impaired neural orientation, filtering, and synchronization in children with ASD.

Keywords: Sensory gating, autism spectrum disorders, electroencephalography (EEG), auditory filtering, time-frequency analysis

1. Introduction

Sensory processing deficits, which include hyper- or hyporeactivity to sensory input are part of the diagnostic criteria for children with autism spectrum disorders (ASD) (American Psychiatric Association, 2013). Sensory processing refers to both, the neurological mechanism which includes the reception, modulation, integration, and organization of incoming sensory stimuli and the behavioral response to sensory information (Miller and Lane, 2000). Sensory processing difficulties have been associated with social, emotional, and behavioral responsiveness in children with ASD (Baker et al., 2008). Specifically, sensory processing dysfunction can lead to aberrant behaviors in an attempt to make sense of and regulate environmental stimulation (Iarocci and McDonald, 2006). Researchers have typically examined the behavioral responses of sensory processing in ASD using parent-report behavioral measures such as the Sensory Profile (Dunn, 1999), which examines the child’s response to a variety of sensory information in everyday contexts; however, there is limited research examining the neurological mechanisms of sensory processing.

Among the sensory domains affected, deficits in auditory processing are one of the most commonly reported impairments in ASD (Tomchek and Dunn, 2007). Deficits in auditory processing have been associated with impairments in neural inhibition and filtering of sensory input (Orekhova et al., 2008). However, there is a paucity of research identifying the specific neuropathology underlying sensory processing dysfunction, which warrants further study. Neural aspects of sensory processing can be explored using functional neurophysiological methods like magnetic resonance imaging (MRI; spatial information) electroencephalography (EEG) and event-related potentials (ERPs; temporal information), which examine real-time brain activation while the brain processes sensory stimuli (Davies and Gavin, 2007). The sensory gating EEG paradigm has been used to examine the neural mechanisms underlying auditory information processing.

Sensory gating is a neurological process that filters out irrelevant stimuli, thus preventing sensory overload of higher brain functions. Deficits in sensory gating have been associated with difficulties in organizing sensory information and with the development of severe behavioral aberrations in disorders like schizophrenia (Freedman et al., 1991). For the auditory sensory gating EEG paradigm, the participant listens to repeated presentations of a pair of identical click sounds, presented within a short period of time from each other. In the resulting ERP, the P50 component presents as a positive deflection that occurs around 40 – 80 milliseconds (ms) after stimulus presentation. The P50 develops as a neural orienting response to the first click while simultaneously activating inhibitory pathways. When the second click is presented immediately thereafter, the active inhibitory pathways suppress the P50 potential in response to the second click stimulus (Freedman et al., 1987). The reduction in amplitude of the P50 to the second click (a.k.a., test click) compared to the first click (a.k.a., conditioning click) represents gating which is sometimes described as suppression. Traditionally, sensory gating has been measured as a ratio, the click 2 (Test) divided by the click 1 (Conditioning), labeled as T/C ratio. Lower ratios of click 2 to click 1 indicate stronger attenuation of irrelevant or redundant information, and thus better gating ability. More recently, gating has also been assessed using difference scores instead of ratios, which is considered more reliable (Dalecki et al., 2011), wherein the amplitude of click 2 is subtracted from the amplitude of click 1 (C – T).

The P50 auditory evoked potential component has been widely used to evaluate sensory gating in schizophrenia (Clementz et al., 1998; Freedman et al., 1996). Due to the pervasive and quantifiable nature of gating, impaired P50 suppression is considered an endophenotype for schizophrenia (Calkins et al., 2007). Researchers have proposed that gating deficits observed in schizophrenia are due to two mechanisms; reduced responsiveness leading to smaller averaged amplitudes of click 1 and reduced gating of click 2 (Jansen et al., 2010). The reduced responsiveness to click 1 in schizophrenia suggests deficits in orientation to a novel stimulus and a lack of basic attentiveness (Jansen et al., 2010).

More recently, sensory gating has been examined in children with various disorders associated with sensory processing difficulties such as with sensory processing disorders (SPD) (Davies et al., 2009) and ASD (Kemner et al., 2002). Children with SPD (without a co-diagnosis of ASD) demonstrated less robust sensory gating than an age-matched group of neurotypical children (Davies et al., 2009). Some researchers have shown intact P50 suppression in adult males with ASD (Magnée et al., 2009). A study by Kemner et al. (2002) showed no differences with regard to P50 amplitudes and P50 suppression between children with ASD and neurotypical peers. Orekhova et al. (2008) report comparable findings of similar gating in children with ASD compared to neurotypical peers, ages 3 – 8 years; though, sensory gating was significantly reduced in children with intellectual impairment (IQ < 72) and young children with and without ASD. Madsen et al. (2015) also found similar P50 gating in children ages 8 – 12 years when compared to matched controls. However, they found attenuated P50 amplitude to click 1 in the Asperger subcategory of their sample. These results of sensory gating in children with various disorders associated with sensory processing are mixed and no clear description of sensory processing function or dysfunction has emerged warranting further investigation.

In addition to the P50 component, researchers have examined mid- and late-latency N1, P2, and N2 components in the sensory gating paradigm, which have been shown to have better test-retest reliability than P50 measures (Rentzsch et al., 2008). These later components have been associated with higher-order cognitive processes (Kisley and Cornwell, 2006). While the P50 is thought to represent pre-attentive processing, the N1 and P2 are thought to reflect early attention, and the N2 is thought to reflect later attentive processes (Boutros et al., 2004). The N1 component presents as a negative deflection occurring around 100 ms post-stimulus onset. The N1 amplitude to auditory stimuli has been found to be normal, decreased, or even increased in people with ASD compared to controls (Bomba and Pang, 2004). The P2 component is a positive deflection following the N1, and the N2 component is a negative deflection around 200 ms post-stimulus onset. Some researchers have shown that children with ASD have significant gating deficits at the N1 component (Orekhova et al., 2008), while others have found typical N1 gating in children with ASD (Madsen et al., 2015). Additionally, deficits in gating at the N1 and P2 components have been shown in individuals with schizophrenia (Boutros et al., 2004). However, to our knowledge, gating at the late-latency P2 and N2 components have not been examined in children with ASD.

In addition to ERPs, the brain’s oscillatory patterns in response to sensory input provide valuable information about the different stages of information processing. Analysis of both ERPs and time-frequency characteristics may provide comprehensive information regarding neural auditory information processing. Time-frequency information can provide information about both power and synchrony. Researchers examining sensory gating in neurotypical adults using time-frequency analysis have shown that the auditory stimulus evokes early gamma activity, followed by beta activity measured by evoked power. Examining synchrony, the early phase-locked gamma is thought to reflect early stages of sensory perception; wherein immediate synchronization of signals occurs from neuronal assemblies within short distances (Hong et al., 2004). The subsequent phase-locked beta oscillation is assumed to reflect a modification of neuronal circuitry for higher-level neural processing such as stimulus encoding and consolidation (Bibbig et al., 2001; Hong et al., 2004). In neurotypical individuals and individuals with schizophrenia, a stronger click 1 beta response but not gamma, as measured by evoked power, was associated with stronger P50 gating and click 2 P50 suppression (Hong et al., 2004). Thus, the different frequency oscillations provide information about distinct neural processes contributing to sensory processing and can be examined using both power and synchrony. To our knowledge, no studies have examined gamma and beta oscillations associated with sensory gating in children with ASD.

This study sought to determine whether children with ASD differ from typically-developing (TD) children on sensory gating, as measured by the P50, N1, P2, and N2 ERP components. In other clinical groups, it has been shown that sensory gating deficits may be attributed to decreased orientation (smaller amplitude of click 1) or decreased suppression of the click 2. In schizophrenia, one group has reported a reduction in the amplitude of click 1 and no difference on click 2 (Jansen et al., 2010), while other researchers have reported that sensory gating defects are due to decreased suppression of the click 2 amplitude (Clementz et al., 1998; Freedman et al., 1987). Thus, we examined the group differences in the amplitude to each click. We also examined the association between sensory gating and a behavioral measure of sensory processing, the Short Sensory Profile. Additionally, we performed time-frequency analyses to examine group differences in power and synchrony of neural oscillations.

The study’s research questions included:

  1. Do children with ASD exhibit less robust sensory gating than TD peers measured at P50, N1, P2, and N2 ERP components? We hypothesized that children with ASD will have significantly less robust gating than TD peers at all the ERP components.

  2. Are there group differences in the neural processing of the click stimuli? We hypothesized that neural processing of the first click stimuli, but not the second click will be different for the two groups.

  3. Are there group differences in the evoked power and phase-locking of neural oscillations of click stimuli? We hypothesized that there will be significantly lower gamma and beta evoked power and phase-locking activity in children with ASD compared to TD peers.

  4. Do brain measures of sensory gating as represented by P50 difference score and beta evoked power and beta phase-locking correlate with behavioral measures of sensory processing? We hypothesized that there will be a significant correlation between brain measures of auditory sensory processing and behavioral measures of sensory processing.

2. Methods

2.1. Participants

Thirty-six children, ages 5 to 12 years participated in the study. Eighteen children (M = 8.41 years, SD = 2.11; 13 males, 5 females) had a confirmed clinical/medical diagnosis of ASD (based on DSM-5), or Asperger’s disorder (based on DSM-IV-TR) and were recruited from local clinics and parent support groups. Prior to study participation, the participant’s primary caregiver completed the Asperger Syndrome Diagnostic Scale (Myles et al., 2001), which was used to confirm the diagnosis of ASD. Asperger Syndrome Quotient for the sample ranged from 90 – 127, with 8 children scoring “Very likely” and 10 children had scores in the “Likely” category. All children with ASD were verbal. The control group consisted of 18 age- and gender-matched TD (M = 8.03, SD = 1.35; 13 males; 5 females) children. TD children were recruited from schools and organizations in the local community and from a network of prior research participants. The TD children were included if they did not have a history of neurological injuries, disabilities, and family history of mental health disorders. Parent reports from all participants indicated no hearing deficits. Intelligence was measured using the Wechsler abbreviated scale of intelligence (Stano, 2004). The control group had a significantly higher IQ score (M = 111.39, SD = 11.09) compared to the ASD group (M = 98.72, SD = 16.92), t (34) = 2.66, p = .012. All the procedures performed in the research involving human participants were in accordance with the ethical standards of the institutional review committee at the local university and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

2.2. Data Collection

Upon volunteering, parents of the participants were mailed an information packet with consent forms and the Short Sensory Profile (Dunn, 1999). Parent permission and child assent were obtained from all participants and testing procedures were reviewed at the beginning of their visit to the lab and prior to the start of the child’s preparation for EEG recordings. Time was also allowed to build rapport with the children by answering any questions and sharing photos of previous child participants wearing the EEG cap and electrodes. The participants were seated in a comfortable high-back chair. All participants underwent a brief training to reduce eye blinks and muscle artifacts prior to the EEG recording. A brief click stimulus (3 ms) and a stepping procedure (Levitt, 1971) was used to determine the hearing threshold of the participants. The intensity of the click stimuli during the presentation of the gating paradigm was set at 60 dB above the threshold for each individual except for eight children with ASD who were unable to perform the threshold testing due to difficulty in understanding the instructions for the task. The stimulus intensity for the gating task was set to 85 dB SPL for these participants. To ensure that the intensity of the click used in the sensory gating paradigm would not produce a startle reaction, participants heard a pair of clicks prior to the EEG recording. Participants did not display a startle response at any time during the recordings. All participants were able to complete the EEG recording.

2.3. Sensory Gating paradigm

A modified sensory gating paradigm was used which consisted of presentations of 100 pairs of click stimulus while the participant watched a silent claymation movie. The click stimuli were presented binaurally through the ER-3A inserted earphones (Etymotic Research). Each click had a 3 ms duration. The paired click stimuli were presented with a stimulus onset asynchrony (SOA) of 500 ms and an 8-second inter-trial-interval between pairs. The stimuli were presented using E-Prime (Psychological Software Tools, Pittsburg, PA).

2.4. EEG/ERP data recording

EEG data were recorded using a 32-channel BioSemi ActiveTwo System (BioSemi, Amsterdam, Netherlands). The vertical and horizontal electro-oculograms (EOG) were recorded from four facial electrodes: two on the left supraorbital and infraorbital region, and two on the lateral aspect of the outer canthus of the right and left eye. Two flat electrodes on the right and left earlobes served as the offline reference. The online reference and ground consisted of a Common Mode Sense (CMS) active electrode and a Driven Right Leg (DRL) passive electrode (specific to the Biosemi system). Data were sampled at 1024 Hz (bandwidth: 0–268 Hz).

2.5. Data reduction for ERP analyses

Offline analyses were conducted using BrainVision Analyzer (Brain Products GmbH, Gilching, Germany). Peak-to-peak measures for the P50, N1, P2, and N2 components were obtained based on previously established procedures (Davies et al., 2009). Averaged ERPs were composed from the running EEG data. First, the four EOG channels were converted to a vertical and a horizontal bipolar EOG.

For the P50 component, data were filtered with a bandpass setting of 10 – 200 Hz (Chang et al., 2012) with a 24 dB/octave roll-off, run in both forward and reverse directions to avoid a phase shift. Data were segmented time-locked to the onset of the click into epochs representing either the click 1 or click 2 with a duration of −100 ms to 200 ms. A baseline correction was performed relative to a baseline of −100 ms to 0 ms.

For the N1, P2, and N2 components, data were filtered with a bandpass of .23 – 30 Hz (24 dB/octave). Data were segmented time-locked to the stimulus onset into epochs representing either the click 1 or click 2 with a duration of −100 to 500 ms. Baseline correction was performed relative to a baseline of −100 ms to 0 ms. Next, an eye regression technique was used to remove eye movement artifacts (Segalowitz, 1999). For the non-rejected segments of the N1, P2, and N2 components, another baseline correction was done relative to a baseline of −100 ms to 0 ms.

For all components, segments containing voltage deviations higher than ±100 microvolts (μV) on any of the EOG or EEG channels were removed. Segments were then averaged to create ERP waveforms for both the clicks to measure the P50, N1, P2, and N2 components for each participant. There was no significant difference between the final number of segments for the P50 component analyzed for the TD group (M = 99.58, SD = .60), and children with ASD (M = 99.5, SD = .64), t (34) = .40, p = .69. However, there was a significant difference in the number of segments analyzed for the N1-P2-N2 components for the TD group (M = 99.33, SD = 1.14) compared to children with ASD (M = 97.42, SD = 3.34), t (20.89) = 2.31, p = .032. The average number of segments was used as a covariate in the analyses of N1-P2-N2 components.

Peak-to-peak amplitude measurement was chosen compared to baseline-to-peak since the former is shown to be more reliable in children (Chang et al., 2012). Additionally, the topographical maps show the negativity of the ERP components which justify the use of peak-to-peak measurement. See Figure 1. The P50 component peak-to-peak amplitude was calculated as the difference between the P50 component peak and its preceding negative component - the N45. The P50 peak was defined as the maximum positive point between 40 to 80 ms post-stimulus onset. To identify the N45 peak, the P30 was defined as the maximum positive point between 25 to 40 ms post-stimulus onset, and the most negative peak between the P30 and P50 components was labeled as the N45. The peak-to-peak amplitude of the N1 component was identified as the difference in amplitude between the peak of the N1 component and its preceding positive peak. The N1 peak was scored as the most negative peak between 70 to 180 ms post-stimulus onset. The maximum positive peak preceding the N1 was scored between 40 to 100 ms and labeled as P1. The P2 component was scored as the most positive peak between 115 to 270 ms post-stimulus onset and peak-to-peak amplitude was defined as the difference in amplitude between the N1 peak and the P2 peak. The N2 component was defined as the most negative peak between 170 to 375 ms post-stimulus onset and the peak-to-peak amplitude was scored as the difference in amplitude between the P2 peak and the N2 peak. To ensure correct identification, two senior authors independently reviewed the component peaks. See figure 2 for individual ERP plots from children in each group showing that the component peaks are identifiable across the distribution of peak amplitudes.

Figure 1:

Figure 1:

Open in a new tab

Topographical map of the P50 component based on averaged event-related potentials of click 1 and click 2.

Figure 2:

Figure 2:

Open in a new tab

Event-related potential plots for three typically-developing children (depicted in blue) on the left and three children with autism spectrum disorder (depicted in green) on the right. The averaged brain response to the first click stimuli is shown with a solid line, while the second click is shown as a dashed line. Data were filtered with a bandpass setting of .23 to 30 Hz. The three children in each group represent the 75th, 50th, and 25th percentile of the respective group’s distribution (panel A, panel B, and panel C, respectively) of the Click 1 peak-to-peak N1 amplitudes. The N1 peak-to-peak amplitude was calculated as the difference in amplitude between the N1 peak and the preceding P1 peak.

The T/C ratio and difference scores were computed to quantify gating abilities. The T/C ratio approaching 0 is indicative of robust gating and the T/C ratio approaching 1 indicates less gating (Cromwell et al., 2008). Difference scores were calculated by subtracting peak-to-peak P50, N1, P2, or N2 amplitude of click 2 from the peak-to-peak P50, N1, P2, or N2 amplitude of click 1 respectively. For the positive components - P50 and P2, a large positive difference score indicates better gating. For the N1 and N2 (negative ERP components), a large negative difference indicates better gating. Of the 32 channels, the central site Cz was analyzed based on the topographical map of activation for the P50 component (See Figure 1). Additionally, the Cz site has been used in previous studies using the sensory gating paradigm.

2.6. Electrophysiological data reduction for time-frequency analyses

The four EOGs were transformed to a horizontal and a vertical bipolar EOG. Data were filtered with a bandpass of 0.1 – 80 Hz. Next, data segmentation time-locked to the click onset was done with a duration of −1400 ms to 2400 ms. Following this, a baseline correction procedure was conducted relative to a baseline of −200 ms to 0 ms. To remove eye movement artifacts, an eye regression technique was performed. For the non-rejected segments, another baseline correction was conducted relative to a baseline of −200 ms to 0 ms. Segments were excluded if they had voltage artifacts higher than ±100 μV on any EOG or EEG channels.

Custom written Matlab software were used for analyses. Time-frequency analyses consists of extracting frequency spectrum information such as magnitude and phase from event-locked EEG oscillations. Morlet wavelet transform was performed using Matlab software that converted waveforms into sinusoidal waves of different frequencies and phase angles. The frequency parameters examined included (1) evoked power, measuring signal intensity, which is calculated as the average of the time-frequency data of each segment, and (2) phase-locking factor, which depicted synchronization of signal phase across trials. Evoked power measures the event-related changes in EEG power that are phase-locked to the onset of the event. Phase-locking factor measures the consistency in the signal phase angle of the oscillations evoked by the onset of the event across single trials. In the time-frequency plots, the X-axis depicts time in milliseconds, and the Y-axis depicts the frequencies from 0 – 80 Hz. The intensity legend shows greater activity in warmer colors.

For descriptive purposes only, comparison of values from two time-frequency matrices was derived using t-tests (for ASD vs. TD differences) or paired t-tests (for within-group differences). For the t maps, alpha was set at .05. To control for inflation of Type I error, a value was depicted as significant if at least five consecutive significant comparisons were observed around the target value. This number was selected since it delivered the best visual representation of the differences in the regions of interest (ROI) between the experimental conditions defined in the statistical analysis section. In this study, only phase-locked evoked power and phase-locking factor plots were analyzed since they can be directly compared to the standard ERP-based analysis of sensory gating. ERPs represent stimulus-locked time-domain averaged waveforms highlighting phase-locked oscillations, while non-phase locked oscillations are averaged out (Klimesch et al., 1998).

2.7. Behavioral Measures

The Short Sensory Profile (SSP), a 38-item parent-rated screening questionnaire, was used to evaluate functional behaviors related to a child’s sensory processing abilities and was derived from its parent measure - the Sensory Profile (Dunn, 1999). Items measure functional behaviors that are indicative of sensory processing dysfunction in everyday contexts. Scores were derived for the following seven domains: i. tactile; ii. taste/smell, iii. movement, iv. visual/auditory sensitivity; v. under-responsive/seeks sensation; vi. auditory filtering; and vii. low energy/weak, and an overall score. Sample items include auditory filtering item 22: is distracted or has trouble functioning if there is a lot of noise around; visual/auditory sensitivity item 34: responds negatively to unexpected or loud noises. Higher scores indicate more typical performance, while lower scores indicate either a probable or definite difficulty in sensory processing abilities. The SSP has a high internal consistency of the subdomains and overall score (r = .70 - .90) and acceptable discriminative validity (> 95% in differentiating between children with and without sensory processing dysfunction).

The Asperger Syndrome Diagnostic Scale (ASDS) is a norm-referenced scale to assist in the diagnosis of Asperger Syndrome as defined in the DSM-IV-TR (Myles et al., 2001). The scale has been validated for use in children between 5 to 18 years of age. It includes 50 items grouped into five subscales, namely Social, Language, Maladaptive, Sensorimotor, and Cognitive. Items are based on the diagnostic criteria of Asperger Syndrome published in the DSM-IV, International Classification of Diseases - Tenth Edition, and a review of the literature. All items are added together to obtain an overall Asperger Syndrome Quotient that indicates the likelihood that a child has Asperger Syndrome. The scale has good reliability (alpha coefficient of .83 for the Asperger Syndrome Quotient) and criterion-prediction validity.

2.8. Statistical Methods

Prior to conducting the statistical analyses, the assumptions of homogeneity of variance and normalcy were tested. If Levene’s test of homogeneity of variance was significant, the degrees of freedom were appropriately adjusted. To examine the first hypothesis of group differences on P50, N1, P2, and N2 gating, multi-variate analysis of variance (MANOVA) was used. To compare the amplitude of the click 2 to the amplitude of click 1 for the P50, N1, P2, and N2 components, for each group separately, paired-samples t-tests were conducted. For the second hypothesis, a 2 (Clicks) by 2 (Group) ANOVA (for P50) or ANCOVA (for N1-P2-N2 covarying for average number of segments) was conducted and posteriori Tukey’s honestly significant difference (HSD) tests were conducted to test differences between cell means when there were significant interaction effects (Kirk, 1968, pp. 265–269).

For the third hypothesis, two ROIs were identified for evoked and phase-locked power, 1) Beta: 13 – 18 Hz from 60 to 90 ms and 2) Gamma: 30 – 50 Hz from 60 – 90 ms. Data from each ROI were analyzed using a 2 (Clicks) by 2 (frequency - beta and gamma) by 2 (Group) ANOVA which was followed up by a posteriori Tukey’s HSD tests. For the fourth hypothesis, simple linear regressions were used to examine the association between the components of the SSP and the P50 and N1 difference scores. Statistical analyses were conducted using the Statistical Package for Social Sciences, 24.0 version.

3. Results

3.1. Group differences in the ERP components

The means and standard deviations of the peak-to-peak amplitudes of the P50, N1, P2, and N2 ERP components, along with the T/C ratios and difference scores are shown in Table 1.

Table 1.

Descriptive data (Mean and standard deviation [SD]) of sensory gating event-related potential components between children with autism spectrum disorders (ASD) and typically-developing (TD) children.

Amplitude click 1 Amplitude click 2 Difference Score T/C Ratio
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
P50 Component
TD 5.65 (2.30) 3.43 (1.29) 2.22 (1.93) 0.65 (0.26)
ASD 2.92 (1.44) 2.86 (1.21) .06 (1.51) 1.18 (0.70)
N1 Component
TD −7.03 (3.68) −4.55 (2.97) −2.48 (3.97) 0.94 (1.45)
ASD −3.36 (2.33) −4.15 (2.05) .79 (2.35) 1.71 (1.06)
P2 Component
TD 8.35 (5.47) 3.26 (1.69) 5.10 (5.28) 0.61 (0.62)
ASD 5.87 (2.78) 3.86 (2.06) 2.01 (2.64) 0.77 (0.49)
N2 Component
TD −7.94 (4.62) −4.82 (2.52) −3.12 (5.61) 0.77 (0.54)
ASD −6.85 (3.98) −4.37 (2.68) −2.48 (4.35) 1.05 (1.51)
Open in a new tab

A univariate ANOVA for the P50 difference score indicated a main effect of group, F (1,34) = 13.95, p < .001, η2p = .29. A 2 (Group) by 3 (N1-P2-N2 ERP components - difference scores) MANCOVA (with average number of segments as a covariate) showed a significant group effect for the N1 difference score, F (1,33) = 5.29, p = .028, η2p = .14, but not for the P2 (p = .11) and N2 component (p = .58). Thus, children with ASD had less robust sensory gating than TD peers at the P50 and N1 components (See Figure 3). Often grand averages are difficult to interpret especially for children due to great variability in latency and amplitudes across the participants, see Figure 2 for individual ERP plots from children in each group.

Figure 3:

Figure 3:

Open in a new tab

Grand averaged event-related potentials for each group (typically-developing children [TD] and children with autism spectrum disorder [ASD]) and each click (click 1 and click 2) time-locked to the auditory stimuli at the Cz electrode site

3.1.1. Suppression of click 2 in both groups

A 2 (P50 clicks) by 2 (Group) ANOVA revealed a significant group by clicks interaction effect, F (1,33) = 13.95, p = .001, η2p = .29. Similarly, the ANCOVA for the N1 component revealed a significant group by clicks interaction effect, F (1,33) = 5.29, p = .028, η2p = .14. However, there was no main or interaction effect for the ANCOVAs for the P2 and N2 components.

Posteriori tests indicated that for TD children, click 2 was significantly smaller than click 1 at the P50 (q (1, 17) = 5.84, p < .01), N1 (q (1, 17) = 3.7, p < .05), P2 (q (1, 17) = 6.44, p < .01), and N2 (q (1, 17) = 3.71, p < .05) components. Thus, TD children demonstrated robust gating at all the ERP components. For children with ASD, there was no significant difference between click 1 and click 2 at the P50 (q (1, 17) = .16, p > .05), and N1 (q (1, 17) = 1.18, p > .05) components. However, click 2 was significantly smaller than click 1 at the P2 (q (1, 17) = 2.54, p < .05), and the N2 (q (1, 17) = 2.95, p < .05) components.

3.1.2. Group Differences on the click 1, but not click 2

To answer research question 2, posteriori tests were conducted based on the ANOVAs above. For click 1, compared to TD children, children with ASD had significantly smaller P50 amplitudes, q (1, 17) = 7.18, p < .01, smaller N1 amplitudes, q (1, 17) = 5.48, p < .01, and smaller P2 amplitudes, q (1, 17) = 3.14, p < .05. There was no significant difference in click 1 for the N2 component, q (1, 17) = 1.3, p > .05. This group difference in click 1 at the early components, suggests that children with ASD not only have difficulty in gating but also have difficulty with early neural orientation to the first click. There were no significant group differences for click 2 on any of the ERP components, P50 amplitude, q (1, 17) = 1.5, p > .05; N1 amplitude, q (1, 17) = .60, p > .05; P2 amplitude, q (1, 17) = .76, p > .05; and N2 amplitude, q (1, 17) = .54, p > .05.

3.2. Time-frequency Analyses

Visual inspection of the time-frequency evoked power plot indicates an increase in brain activity corresponding to the onset of the click sounds (See Figure 4). For both groups, click 1 has greater intensity than click 2. Additionally, as expected, children with ASD appear to have less evoked power than the TD group. Figure 4 A and B depicts increases in evoked power to the click stimuli in the theta, alpha, beta, and low gamma bands. Gamma activity is evident in the first 100 ms post-stimulus onset. Activity in the alpha and beta bands appear to be highest around 50 to 150 ms post-stimulus onset, while activity in the theta bands appears to be maximal from 100 ms to 350 ms post-stimulus onset. For the ROI for evoked power, a 2 (Clicks) by 2 (Frequency) by 2 (Group) repeated-measures ANOVA indicated a significant click by frequency by group interaction effect, F (1,34) = 17.99, p < .0005, η2p = .35 and a significant main effect of group, F (1,34) = 14.98, p < .0005, η2p = .31. The post-hoc ANOVA for beta evoked power indicated a significant group by click interaction effect, F (1,34) = 18.40, p < .0005, and a main effect of group, F (1,34) = 16.83, p < .0005 indicating reduced beta evoked power in both clicks in children with ASD compared to controls. However, for gamma evoked power, the group by click interaction effect, F (1,34) = .28, p =.60, and the main effect of group, F (1,34) = .47, p =.50, was not significant.

Figure 4:

Figure 4:

Open in a new tab

Time-frequency plot of evoked power from the sensory gating paradigm. The white line at 0 milliseconds on the time axis indicates the onset of click 1, and the second white line at 500 milliseconds depicts the onset of click 2. A. Evoked power for typically-developing (TD) children. B. Evoked power for children with autism spectrum disorders (ASD). C. T map comparing the evoked power for the two groups.

Figure 5 depicts the phase-locking factor for the two groups. In both groups, click 1 has higher phase synchronization compared to click 2, however, this suppression effect is more evident in the TD group. For the TD group, gamma activity around 0 – 50 ms is followed by highly synchronized activity in the beta bands around 50 – 150 ms post-stimulus onset. A similar gamma-beta shift is visible in click 2. In children with ASD, there appears to be a lack of gamma synchronization following both clicks, and a reduced presence of beta activity around 100 – 200 ms post-stimulus onset. For the ROI analysis for phase-locking factor, a 2 (Clicks) by 2 (Frequency) by 2 (Group) repeated-measures ANOVA indicated a significant click by frequency by group interaction effect, F (1,34) = 18.37, p < .0005, η2p = .35, and a significant main effect of group, F (1,34) = 14.98, p < .0005, η2p = .31. The post-hoc ANOVA for beta phase-locking indicated a significant group by click interaction effect, F (1,34) = 25.88, p < .0005, and a main effect of group, F (1,34) = 18.05, p < .0005 indicating reduced beta phase-locking in both clicks in children with ASD compared to controls. However, for gamma evoked power, the group by click interaction effect, F (1,34) = .19, p =.67, and the main effect of group, F (1,39) = 1.89, p =.18, was not significant.

Figure 5.

Figure 5.

Open in a new tab

Time-frequency plot depicting phase-locking or inter-trial coherence during the sensory gating paradigm. The white line at 0 milliseconds on the time axis indicates the onset of click 1, and the second white line at 500 milliseconds depicts the onset of click 2. A. Phase-locking plot for typically-developing (TD) children. B. Phase-locking for children with autism spectrum disorders (ASD). C. T map comparing Phase-locking for the two groups.

To further examine differences in neural oscillations associated with gating, a t-map comparing the evoked power plot of click 2 from the click 1 plot was generated. Figure 6 depicts the differences in evoked power between click 1 and 2 for each group separately. As expected, posteriori tests indicated that for TD children click 1 had greater evoked power in the beta frequency than click 2, q (1, 17) = 6.97, p < .01. (See Figure 6-A). Interestingly, in children with ASD, click 2 is not significantly smaller than click 1, 60 – 90 ms post-stimulus onset, q (1, 17) = .95, p > .05 (See Figure 6-B). Moreover, the ASD group had significantly lower click 1 beta power than the TD group, q (1, 17) = 8.03, p < .01. This suggests the lack of an orientation response in children with ASD to click 1 immediately following the stimulus, which is evident in TD children. This finding further strengthens the ERP findings of smaller click 1 P50, N1, and P2 amplitudes in children with ASD compared to TD peers; and the presence of gating in the mid-latency N2 and P2 components in children with ASD.

Figure 6:

Figure 6:

Open in a new tab

T map comparing evoked power between the two clicks. A. Comparison of clicks in typically-developing (TD) children. B. Comparison of clicks in children with autism spectrum disorders (ASD).

3.3. Sensory Processing and Sensory Gating

As shown in Table 2, there were significant group differences on every subscale and the total score of the SSP, the behavioral measure of sensory processing. A simple linear regression was used to predict the total SSP score from the P50 and N1 difference scores. A significant regression equation was found, F (2,33) = 7.36, p = .002, R = .56, R2 = .31. Thus, the early P50 and N1 gating accounted for 31% of the variance in the behavioral sensory processing scores. Additionally, across all participants, greater sensory processing issues on the Sensory Profile total score correlated with lower click 1 evoked beta power and phase synchronization F (2,33) = 6.94, p = .003, R = .54, R2 = .30. Thus, evoked and phase-locked beta activity accounted for 30% of the variance in the behavioral sensory processing scores.

Table 2.

Descriptive data (Mean and standard deviation [SD]) and Comparison of the Short Sensory Profile between children with autism spectrum disorders (ASD) and typically-developing (TD) children.

Tactile Sensitivity Taste Smell Sensitivity Movement Sensitivity Visual auditory sensitivity Underresponsive Seeks Sensation Auditory Filtering Low Energy Weak Total
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)
TD 32.83 (2.73) 17.39 (3.53) 13.11 (2.45) 19.83 (3.78) 27.39 (4.18) 23.67 (2.79) 27.94 (3.28) 162.17 (14.13)
ASD 25.33 (5.55) 10.72 (5.30) 10.50 (3.15) 15.39 (4.13) 18.33 (5.95) 14.06 (5.17) 19.56 (7.79) 113.89 (26.97)
t value 5.14*** 4.44*** 2.78* 3.37** 5.23*** 6.94*** 4.21*** 6.73***
Open in a new tab
*

p < .05;

**

p < .01;

***

p < .0001

4. Discussion

This study examined sensory gating in children with ASD compared with TD children. The first hypothesis was supported such that children with ASD had significantly less robust gating than TD peers in the early components. Our second hypothesis was also supported, indicating significant group differences on click 1 at the P50 and N1 components. Reduced amplitudes of click 1 suggest impaired early neural orientation to auditory stimuli. Additionally, regarding the third hypothesis, the time-frequency analysis provided additional support for the ERP findings. Specifically, children with ASD had significantly reduced evoked power and less phase synchronization at beta and gamma frequencies to both clicks. TD children had a greater difference between the two clicks than children with ASD. Moreover, there was no difference between the two clicks during the first 100 ms in children with ASD. These results suggest that the primary difference between the two groups could be attributed primarily to deficits in orientation to click 1, along with impaired filtering of redundant information. Our fourth hypothesis was also supported, such that neural measures of sensory gating significantly predicted sensory processing as measured using the SSP, the behavioral measure of sensory processing in everyday activities.

4.1. Reduced gating in children with ASD

Contrary to some previous studies (Kemner et al., 2002; Orekhova et al., 2008), and consistent with a recent study (Madsen et al., 2015), children with ASD had significantly reduced sensory gating abilities compared to TD peers. The inconsistency in findings related to gating in ASD may be due to different clinical characteristics of children across the spectrum, and/or methodological reasons, such as a small sample size, or differences in the paired click paradigm. For example, Kemner and colleagues examined 12 children with ASD and 11 children in the control group. Both Orekhova et al. and Madsen et al. and the current study had 18 or more participants in both groups. Kemner et al. had the children close their eyes during the EEG data collection, Madsen et al. had children stare at a fixation point, while Orekhova et al. and the current study had the children watch a silent movie. White noise bursts were used by Madsen et al. (80dB) and Orekhova et al. (90dB) and in contrast, click sounds were used by Kemner et al. (75dB) and the current study (60dB above threshold ~85dB). The filter settings used in data reduction were different for each of the studies. Differences in filter settings have been shown to impact P50 amplitude, noise power, and signal to noise ratio measures (Chang et al., 2012). All these factors have the potential to affect sensory gating and could account for some of the inconsistencies found across studies.

In the current study, gating deficits were seen not only in P50 suppression but also in N1 suppression. This finding of N1 gating deficits in children with ASD suggest that gating deficits persist in auditory processing beyond the P50 component in this clinical group. Similar findings of the persistence of gating deficits measured at the N1 have been reported in children with ASD (Orekhova et al., 2009) and also in individuals with schizophrenia (Boutros et al., 2004; Brockhaus-Dumke et al., 2008). Our current findings, for the P50 and N1 components, are also consistent with previous research examining sensory gating in children with SPD (Davies et al., 2009). Interestingly, at the P2 and N2 components, children with ASD had similar gating as compared to the control group in this current study. The presence of gating at the P2 and N2 components suggests that gating may occur later in the processing phase in children with ASD. Therefore, it is possible that children with ASD may compensate for initial orienting deficits at a later stage of cognitive processing.

4.2. Click 1 difference

Previous research has suggested that deficits in filtering out irrelevant information may lead children with ASD to avoid most external stimulation as a method to reduce disorganizing experiences (Kootz et al., 1982). Children with ASD had significantly reduced P50, N1, and P2 amplitudes to click 1 compared to TD peers; a finding consistent with existing literature (Madsen et al., 2015, 2014; Orekhova et al., 2009). Similar findings of reduced P50, N1 and N2 amplitude of click 1 have also been found in children with ASD compared to neurotypical controls (Orekhova et al., 2009). Persistence of deficits in the mid- and late-latency ERPs beyond the early P50 component has also been observed in studies using other EEG paradigms (Bomba and Pang, 2004). Researchers have proposed that gating deficits are associated with two mechanisms: reduced response to click 1 (orientation), and reduced suppression of click 2 (Jansen et al., 2010). Based on the results of our study sample, the decreased response to the first click is the primary difference between the groups. This should be examined in other studies to determine if children with ASD have difficulty with both sensory processing mechanisms namely orientation and gating/suppression.

4.3. Reduced power and phase synchronization in the gamma and beta frequencies in ASD

As expected, our results showed that TD children have a strong evoked response to the click stimuli in the gamma frequency followed by beta frequency. This gamma to beta shift to the clicks in the gating paradigm has been well-documented in neurotypical adults (Hong et al., 2004; Kisley and Cornwell, 2006). Children with ASD had significantly lower activity in the beta bands to both the click stimuli. Activity in the beta frequency has been associated with stimulus-driven salience (Kisley and Cornwell, 2006). A diminished salience response indicates a deficit in automatic orienting to novel information. In contrast to task-driven salience, which is the voluntary allocation of top-down selective attention, stimulus-driven salience is involuntarily determined by the physical features of a stimulus. A diminished beta response to click 1 in children with ASD, bolsters the ERP findings of significantly smaller click 1 amplitude in children with ASD, which collectively indicate a deficit in automatic orienting in children with ASD.

In our study, we did not find significant group differences in the gamma bands corresponding to both click 1 and 2. Contrary to our findings, reduced early gamma-band evoked power and phase-locking to an auditory stimulus were shown in children and adults with ASD (Gandal et al., 2010; Rojas et al., 2008). Further examination of gamma-band activity in ASD is required.

4.4. Relationship between neural and behavioral measures of sensory processing

In this study, children with ASD had significantly impaired sensory processing abilities on all domains of sensory processing as measured by the SSP compared to TD peers. Furthermore, our findings also indicate a significant relationship between early neural processing as measured by P50 and N1 gating, and beta oscillatory activity and behavioral measures of processing sensory information across a variety of everyday activities as measured by the SSP. These relationships strengthen the argument for a neurological basis for sensory processing dysfunction in children with ASD. Sensory information needs to be efficiently filtered pre-attentively to enable the brain to function properly and enable meaningful engagement with the environment. Our data have shown that children with ASD have difficulties with orienting to auditory stimuli rather than a deficit in suppressing redundant sounds. Reduced oscillatory power and synchrony of brain activity at click 1 was associated with greater sensory processing deficits, suggesting that behavioral sensory processing deficits may be associated with reduced early neural orientation. P50 and N1 suppression deficits have been theoretically linked to sensory overload and sensory processing deficits (Marshall et al., 2004). Previous research has found similar relationships between neural and behavioral measures of sensory processing in children with ASD (Marco et al., 2011).

4.5. Study limitations

This study examined 18 children with ASD or Asperger’s Syndrome. Besides confirming the diagnosis with the ASDS questionnaire, no other diagnostic tools were used, which is a limitation of the present study. However, parents reported that all children with ASD had been diagnosed clinically by psychology or medical personnel. Although the current sample of 18 children with ASD appears small, the findings of large effect sizes (η2p ranging from .16 to .60) and observed power (ranging from .90 - .99) indicate that the sample size is statistically sufficient for conducting this study. Nevertheless, the generalizability of the study finding to children across the spectrum should be done with caution. The children with ASD included in this current study had intellectual abilities within normal limits, as measured by the WASI, with only two children with ASD scoring below 85. Orekhova et al. (2008) found sensory gating was significantly reduced in children with ASD and intellectual impairment (IQ < 72), but not in children with ASD without intellectual impairment. Further research examining gating in the different subcategories of the autism spectrum is warranted. To ensure correct identification of the ERP component peaks, we used appropriate filter settings and senior authors independently reviewed the peaks. However, due to the small amplitude of the P50 component and susceptibility to noise contamination, researchers are encouraged to use blinding and set threshold limits (≥0.5 μV) on peak identification (Boutros, 2008).

As mentioned earlier, studies examining sensory gating in children with ASD have had different procedures of collecting the data - specifically what the child did during the paradigm (i.e., closed eyes, stared at a fixation point or watched a silent movie). It is possible that how the child is “occupied” during the paradigm could influence gating, specifically what the child “attends” to in the different procedures. Future research should examine the influence of the methodology employed during the paradigm to assess the effects of attention on gating. Understanding the contribution of neural oscillations may provide a novel approach to examine the basis for abnormal sensory processing. Neural oscillations provide information about the underlying mechanisms resulting in ERP abnormalities. Additional longitudinal research is required to examine the developmental trend of gating in children with ASD. As shown by a number of studies, sensory gating ability changes with age in children (e.g., Brinkman and Stauder 2007) and longitudinal studies would allow the possibility of examining individual differences across development.

5. Conclusion

This study shows that children with ASD display different brain processing mechanisms to auditory stimuli compared to TD children, specifically in orienting and filtering auditory stimuli. Moreover, neurological processing is strongly associated with sensory processing in everyday activities suggesting that the sensory processing deficits observed in children with ASD may arise from atypical neurophysiological functioning related to gating. The study findings can aid in the understanding of the neurophysiological basis of behavioral concerns in ASD, particularly for behaviors triggered by sensory experiences in daily activities. Understanding the nature of sensory processing challenges in children with ASD may provide guidance on potential treatment strategies.

Highlights.

  • Children with autism showed significantly reduced gating at P50, N1, and P2 event-related potential components.

  • Children with autism show reduced orientation to auditory stimuli compared to typically-developing children.

  • Time-frequency analysis show reduced neural synchronization of stimuli in children with autism.

Acknowledgments:

This study was funded in part by the National Institute of Health (R03HD049532) & Colorado State University College of Health and Human Sciences to PLD & WJG, and by a graduate student grant from the Organization of Autism Research to JC.

Footnotes

Data Accessibility

The EEG data and E-prime code used in the current study are available from the corresponding author on reasonable request.

Conflicts of Interest: The authors declare there are no conflicts of interest.

References

  1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing; 2013. [Google Scholar]
  2. Baker AEZ, Lane A, Angley MT, Young RL. The Relationship Between Sensory Processing Patterns and Behavioural Responsiveness in Autistic Disorder: A Pilot Study. J Autism Dev Disord 2008;38:867–75. 10.1007/s10803-007-0459-0. [DOI] [PubMed] [Google Scholar]
  3. Bibbig A, Faulkner HJ, Whittington MA, Traub RD. Self-Organized Synaptic Plasticity Contributes to the Shaping of γ and β Oscillations In Vitro. J Neurosci 2001;21:9053–67. 10.1523/JNEUROSCI.21-22-09053.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bomba MD, Pang EW. Cortical auditory evoked potentials in autism: a review. Int J Psychophysiol 2004;53:161–9. 10.1016/j.ijpsycho.2004.04.001. [DOI] [PubMed] [Google Scholar]
  5. Boutros N. Lack of blinding in gating studies. Schizophr Res 2008;103:336; author reply 337. 10.1016/j.schres.2008.02.017. [DOI] [PubMed] [Google Scholar]
  6. Boutros NN, Korzyukov O, Jansen B, Feingold A, Bell M. Sensory gating deficits during the mid-latency phase of information processing in medicated schizophrenia patients. Psychiatry Res 2004;126:203–15. 10.1016/j.psychres.2004.01.007. [DOI] [PubMed] [Google Scholar]
  7. Brinkman MJR, Stauder JEA. Development and gender in the P50 paradigm. Clin Neurophysiol 2007;118:1517–24. 10.1016/j.clinph.2007.04.002. [DOI] [PubMed] [Google Scholar]
  8. Brockhaus-Dumke A, Schultze-Lutter F, Mueller R, Tendolkar I, Bechdolf A, Pukrop R, et al. Sensory Gating in Schizophrenia: P50 and N100 Gating in Antipsychotic-Free Subjects at Risk, First-Episode, and Chronic Patients. Biol Psychiatry 2008;64:376–84. 10.1016/j.biopsych.2008.02.006. [DOI] [PubMed] [Google Scholar]
  9. Calkins ME, Dobie DJ, Cadenhead KS, Olincy A, Freedman R, Green MF, et al. The Consortium on the Genetics of Endophenotypes in Schizophrenia: Model Recruitment, Assessment, and Endophenotyping Methods for a Multisite Collaboration. Schizophr Bull 2007;33:33–48. 10.1093/schbul/sbl044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chang W-P, Gavin WJ, Davies PL. Bandpass filter settings differentially affect measurement of P50 sensory gating in children and adults. Clin Neurophysiol 2012;123:2264–72. 10.1016/j.clinph.2012.03.019. [DOI] [PubMed] [Google Scholar]
  11. Clementz BA, Geyer MA, Braff DL. Poor P50 Suppression Among Schizophrenia Patients and Their First-Degree Biological Relatives. Am J Psychiatry 1998;155:1691–4. 10.1176/ajp.155.12.1691. [DOI] [PubMed] [Google Scholar]
  12. Cromwell HC, Mears RP, Wan L, Boutros NN. Sensory Gating: A Translational Effort from Basic to Clinical Science. Clin EEG Neurosci 2008;39:69–72. 10.1177/155005940803900209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dalecki A, Croft RJ, Johnstone SJ. An evaluation of P50 paired-click methodologies. Psychophysiology 2011;48:1692–700. 10.1111/j.1469-8986.2011.01262.x. [DOI] [PubMed] [Google Scholar]
  14. Davies PL, Chang W-P, Gavin WJ. Maturation of sensory gating performance in children with and without sensory processing disorders. Int J Psychophysiol 2009;72:187–97. 10.1016/j.ijpsycho.2008.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Davies PL, Gavin WJ. Validating the Diagnosis of Sensory Processing Disorders Using EEG Technology. Am J Occup Ther 2007;61:176–89. 10.5014/ajot.61.2.176 [DOI] [PubMed] [Google Scholar]
  16. Dunn W. Sensory profile: User’s manual. Psychological Corporation; San Antonio, TX; 1999. [Google Scholar]
  17. Freedman R, Adler LE, Gerhardt GA, Waldo M, Baker N, Rose GM, et al. Neurobiological Studies of Sensory Gating in Schizophrenia. Schizophr Bull 1987;13:669–78. 10.1093/schbul/13.4.669. [DOI] [PubMed] [Google Scholar]
  18. Freedman R, Adler LE, Myles-Worsley M, Nagamoto HT, Miller C, Kisley M, et al. Inhibitory Gating of an Evoked Response to Repeated Auditory Stimuli in Schizophrenic and Normal Subjects: Human Recordings, Computer Simulation, and an Animal Model. Arch Gen Psychiatry 1996;53:1114–21. 10.1001/archpsyc.1996.01830120052009. [DOI] [PubMed] [Google Scholar]
  19. Robert Freedman, Adler LE, Waldo M. Gating of the Auditory Evoked Potential in Children and Adults. Psychophysiology 1987;24:223–7. 10.1111/j.1469-8986.1987.tb00282.x. [DOI] [PubMed] [Google Scholar]
  20. Freedman R, Waldo M, Bickford-Wimer P, Nagamoto H. Elementary neuronal dysfunctions in schizophrenia. Schizophr Res 1991;4:233–43. 10.1016/0920-9964(91)90035-P. [DOI] [PubMed] [Google Scholar]
  21. Gandal MJ, Edgar JC, Ehrlichman RS, Mehta M, Roberts TPL, Siegel SJ. Validating γ Oscillations and Delayed Auditory Responses as Translational Biomarkers of Autism. Biol Psychiatry 2010;68:1100–6. 10.1016/j.biopsych.2010.09.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hong L, Summerfelt A, McMahon RP, Thaker GK, Buchanan RW. Gamma/beta oscillation and sensory gating deficit in schizophrenia. NeuroReport 2004;15:155. [DOI] [PubMed] [Google Scholar]
  23. Iarocci G, McDonald J. Sensory Integration and the Perceptual Experience of Persons with Autism. J Autism Dev Disord 2006;36:77–90. 10.1007/s10803-005-0044-3. [DOI] [PubMed] [Google Scholar]
  24. Jansen BH, Hu L, Boutros NN. Auditory evoked potential variability in healthy and schizophrenia subjects. Clin Neurophysiol 2010;121:1233–9. 10.1016/j.clinph.2010.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kemner C, Oranje B, Verbaten MN, Van HE. Normal P50 gating in children with autism. J Clin Psychiatry 2002;63:214–7. [DOI] [PubMed] [Google Scholar]
  26. Kirk RE. Experimental design; procedures for the behavioral sciences. Belmont, Calif: Brooks/Cole Pub. Co; 1968. [Google Scholar]
  27. Kisley MA, Cornwell ZM. Gamma and beta neural activity evoked during a sensory gating paradigm: Effects of auditory, somatosensory and cross-modal stimulation. Clin Neurophysiol 2006;117:2549–63. 10.1016/j.clinph.2006.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Klimesch W, Russegger H, Doppelmayr M, Pachinger T. A method for the calculation of induced band power: implications for the significance of brain oscillations. Electroenceph Clin Neurophysiol 1998;108:123–30. 10.1016/S0168-5597(97)00078-6. [DOI] [PubMed] [Google Scholar]
  29. Kootz JP, Marinelli B, Cohen DJ. Modulation of response to environmental stimulation in autistic children. J Autism Dev Disord 1982;12:185–93. 10.1007/BF01531308. [DOI] [PubMed] [Google Scholar]
  30. Levitt H. Transformed Up-Down Methods in Psychoacoustics. J Acoust Soc Am 1971;49:467–77. 10.1121/1.1912375. [DOI] [PubMed] [Google Scholar]
  31. Madsen GF, Bilenberg N, Cantio C, Oranje B. Increased Prepulse Inhibition and Sensitization of the Startle Reflex in Autistic Children. Autism Res 2014;7:94–103. 10.1002/aur.1337. [DOI] [PubMed] [Google Scholar]
  32. Madsen GF, Bilenberg N, Jepsen JR, Glenthøj B, Cantio C, Oranje B. Normal P50 Gating in Children with Autism, Yet Attenuated P50 Amplitude in the Asperger Subcategory. Autism Res 2015;8:371–8. 10.1002/aur.1452. [DOI] [PubMed] [Google Scholar]
  33. Magnée MJCM, Oranje B, van Engeland H, Kahn RS, Kemner C. Cross-sensory gating in schizophrenia and autism spectrum disorder: EEG evidence for impaired brain connectivity? Neuropsychologia 2009;47:1728–32. 10.1016/j.neuropsychologia.2009.02.012. [DOI] [PubMed] [Google Scholar]
  34. Marco EJ, Hinkley LBN, Hill SS, Nagarajan SS. Sensory Processing in Autism: A Review of Neurophysiologic Findings. Pediatr Res 2011. 10.1203/PDR.0b013e3182130c54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Marshall PJ, Bar-Haim Y, Fox NA. The development of P50 suppression in the auditory event-related potential. Int J Psychophysiol 2004;51:135–41. 10.1016/j.ijpsycho.2003.08.004. [DOI] [PubMed] [Google Scholar]
  36. Miller LJ, Lane SJ. Toward a consensus in terminology in sensory integration theory and practice: Part 1: Taxonomy of neurophysiological processes. Sensory Integration Special Interest Section Quarterly 2000;23:1–4. [Google Scholar]
  37. Myles B, Simpson RL, Bock SJ, Pro-Ed (Firm) Asperger syndrome diagnostic scale. Austin, Tex.: Pro-Ed; 2001. [Google Scholar]
  38. Orekhova EV, Stroganova TA, Prokofiev AO, Nygren G, Gillberg C, Elam M. The right hemisphere fails to respond to temporal novelty in autism: Evidence from an ERP study. Clin Neurophysiol 2009;120:520–9. 10.1016/j.clinph.2008.12.034. [DOI] [PubMed] [Google Scholar]
  39. Orekhova EV, Stroganova TA, Prokofyev AO, Nygren G, Gillberg C, Elam M. Sensory gating in young children with autism: Relation to age, IQ, and EEG gamma oscillations. Neurosci Lett 2008;434:218–23. 10.1016/j.neulet.2008.01.066. [DOI] [PubMed] [Google Scholar]
  40. Rentzsch J, Jockers-Scherübl MC, Boutros NN, Gallinat J. Test-retest reliability of P50, N100 and P200 auditory sensory gating in healthy subjects. Int J Psychophysiol 2008;67:81–90. 10.1016/j.ijpsycho.2007.10.006. [DOI] [PubMed] [Google Scholar]
  41. Rojas DC, Maharajh K, Teale P, Rogers SJ. Reduced neural synchronization of gamma-band MEG oscillations in first-degree relatives of children with autism. BMC Psychiatry 2008;8:66 10.1186/1471-244X-8-66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Segalowitz SJ. EPRScore Program: Peak and Area Analysis of Event-Related Potentials. St. Catharines, ON: Brock University; 1999. [Google Scholar]
  43. Stano JF. Wechsler Abbreviated Scale of Intelligence. Rehabilitation Counseling Bulletin; Washington: 2004;48:56–7. [Google Scholar]
  44. Tomchek SD, Dunn W. Sensory processing in children with and without autism: a comparative study using the short sensory profile. Am J Occup Ther. 2007;61:190–200. 10.5014/ajot.61.2.190 [DOI] [PubMed] [Google Scholar]

RESOURCES