Review on experimental Studies in Autism: Does
separating sensory Symptoms into Age-Groups and
respective Senses reveal a Sensory Pattern?
By Leonie Dühlmeyer
Student ID: 10437312
Supervisor: Anne Geeke Lever
Co-Assessor: Dr. Cedric Koolschijn
Date: 19.06.2014
Master in Brain and Cognitive Sciences, University of Amsterdam
Track Cognitive Neuroscience
Abstract
Background: Abnormal sensory processing is prevalent in many patients with Autism Spectrum
Disorder. Due to the new edition of the DSM, this has been extensively researched. No consistent course of aberrant processing could be detected so far. Some modulating factors have been identified, but not applied to experimental studies. For this purpose, we review experimental studies researching sensory processing in autism separated into age-groups - to account for developmental trajectories - and senses – to account for psychophysical influence. Methods: All controlled experiments - applying non-social stimuli - from 1st January 2005 until 31st December 2013 were identified through systematic searches using Google Scholar and Pubmed search engines. A total of 72 experiments were included; thirty-three visual, sixteen auditory, six tactile, one gustatory, five olfactory and eleven multisensory. Results: Sensory symptoms are complex in Autism. They are spread across all sensory domains and comprise enhanced as well as decreased processing. A general relation to age could be determined for the gustatory and tactile domain. Conclusion: Experiments investigating sensory symptoms in autism should include modulating factors in their set-up. Age-ranges should be kept narrow and more longitudinal experiments need to be performed to control for developmental influence. Keywords: Autism, sensory processing, development, experimental.
Introduction
Autism Spectrum Disorders (ASD) are a range of complex developmental disorders that can lead to aberrations in thinking, feeling, communication (verbal & non-verbal) and other forms of social interaction (American Psychiatric Association [APA], 2013). In ASD, the effect on each person varies and the disorder is usually first diagnosed in childhood (APA, 2013). Of all children, around one in 88 children are diagnosed, with three to four times more boys than girls (APA, 2013). ASD is a lifelong condition; although, increasingly more patients eventually function to independently lead full lives (APA, 2013). Next to the above described behavioral features, ASD contains a strong genetic component, as well as abnormalities in neurophysiology, neuroanatomy (Tsermentseli, O’Brien & Spencer, 2008) and sensory processing (APA, 2013). Sensory processing describes the neurological procedure of organizing sensations and translating them into appropriate motor and behavioral responses (Koziol et al., 2011). These sensations can derive from inside the body or from the environment (Glanze, Anderson & Anderson, 1990). While inside the body, they stem from the relative position of body-parts to each other and the strength being employed in movement, referred to as proprioception, environmental sensations pass from the external world through the five sensory organs, which define them as being visual, auditory, olfactory, gustatory or tactile (Glanze et al., 1990). Common sensory symptoms in ASD are apparent indifference to pain or temperature, negative response to specific sounds or textures, excessive touching or smelling of objects and visual fascination by lights or movement (APA, 2013). The DSM-V lists sensory aberrations in ASD under restricted repetitive patterns of behavior indexing severeness in ASD (APA, 2013).
The purpose of this review is to investigate developmental changes in sensory symptoms in ASD and the variance in the existence or strength of sensory symptoms between the five senses. The relevance for this review derives from the insufficient role of sensory symptoms assorted by the DSM-5 and the lack of qualitative and quantitative information on sensory symptoms in the five senses during different developmental stages in the lives of patients with ASD (from here on referred to as ASDs). This information could not only aid in indexing a screening for ASD, but also develop a theory for its cause or a suitable therapy for patients.
The important role of sensory symptoms in ASD is not only demonstrated by patients reporting them (in addition to impairments in emotional regulation) to be the core symptoms of ASD (Chamak et al., 2008), but the initial appearance of abnormal sensory processing often predates diagnosis (Adrien et al., 1993; Baranek, 1999; Dahlgren & Gillberg, 1989; Lord, 1995). Moreover, 90% of the children with ASD (both low and high functioning) were found to have sensory symptoms (Leekam et al., 2007) - often in two or three sensory domains, regardless of age and IQ. In addition,
aberrant reactions to auditory stimuli are the single most typical feature of very young children with ASD, separating them from other children with developmental delay (Dahlgreen & Gillberg, 1989). However, a cause for those symptoms – that could explain their origin and relation of symptoms to each other - could not be established yet. Knowing the cause of sensory symptoms could allow treat them, which would further reduce social symptoms if these are secondary as some theories suggest (Hutt et al., 1964; DesLauriers & Carlson 1960; Frith & Happé, 1996). As a consequence, living with ASD would be ameliorated.
In order to reveal a cause for sensory symptoms, one review compared 48 empirical and 27 theoretical or conceptual studies on sensory symptoms from 1960 until 2005 (Rogers & Ozonoff, 2005). Next to investigating causal theories, the authors evaluated the studies for responsiveness. Responsiveness refers to hyper- and hypo-responsiveness to sensory stimuli, which are the two most prevalent sensory symptoms reported in ASD (Baranek, 2002; O’Neill & Jones, 1997). The broad spectrum of sensory symptoms is displayed in the main part of this review. Evaluation of hyper- and responsiveness to sensory stimuli led to mixed results, yet, somewhat more hypo-responsiveness was found - especially in the auditory domain. Hyper-hypo-responsiveness in one sense and hypo-responsiveness in another implies that each individual sense should be investigated separately, since otherwise results could be averaged out. Furthermore, heterogeneity in diagnostic criteria from 1960 to 2005 is reported as a major limitation of the review. For this purpose, this review will include experimental studies from 2005 on and cluster results according to their respective sense.
Another, meta-analytical, study presumed the heterogeneity in hyper- and hypo-responsiveness to lay in three factors: 1. chronological age (CA), 2. severeness of ASD, 3. type of control group (typically developing (TD) or developmental delay (DD)) (Ben-Sasson et al., 2009). The authors reviewed 14 studies of parent reports for hypo- and hyper-responsiveness to sensory stimuli and seeking for sensory input, with age groups comprising 0 to 3.4 years, 3.5 to 6.4 years, 6.5 to 9.4 years, and above 9.5 years. Unlike previous work (REF), the analysis displayed a non-linear course: the frequency of sensory behaviors overall, hyper-responsiveness and seeking rose in children until the age-group of 6.5 to 9.4 years and decreased thereafter. Yet, hypo-responsiveness exhibited no consistent course. Although the researchers discovered factors which explain variance in responsiveness measures, they averaged over sensory domains and reviewed only parent questionnaires, which are not objective (Cremeens, Eiser & Blades, 2006). In addition, standard questionnaires do not specify the character of sensory symptoms, but merely the intensity or frequency of the predefined symptoms. However, as reflected in these results, sensory symptoms can change according to developmental stage (Ben-Sasson et al., 2009). This is likely due to sensory organs developing until the teenage years (Moore, 2002; Moore & Linthicum, 2007). For these
purposes, the current review investigates experimental studies and results were reviewed according to different age-groups. This is to provide qualitative and objective insight into sensory symptoms and respecting developmental trajectories in ASD. Effects of severeness of ASD and type of control group were also found, however, we do not focus on high- or low-functioning ASDs, because we aim to reveal sensory symptoms valid for all ASDs. Moreover, controlling for IQ or mental age (MA) can account for sensory differences that result from a developmental delay (DD), but it has been argued that these developmental delays are part of ASD and should not be controlled for (Wing, Gould & Gillberg, 2011).
Although in the current edition of the DSM-5 (published in May 2013) diagnostic criteria for ASD are redefined (APA, 2013), leading to sensory symptoms being exhaustively investigated, yet, to our knowledge, no review has attempted to quantitatively accumulate evidence for the quality of sensory symptoms in ASD including all sensory domains, by means of reviewing experimental papers. For this purpose, this review focuses on controlled experiments on sensory perception in ASD, including implications from the above described reviews (Rogers & Ozonoff, 2005; Ben-Sasson et al., 2009). We aim to find quantitative differences in sensory symptoms between the senses, while elucidating their quality through reviewing experimental studies. We further explore variance in existence or strength of sensory symptoms between three age-groups.
Literature review search strategies
A computerized search in Google Scholar and PubMed was performed using the keywords: Sensory, visual, auditory, tactile, gustatory, olfactory, see, hear, taste, smell with ASD & autistic. Studies were included when adhering to the following criteria: 1) ASDs were diagnosed with reliable measures for ASD such as those of the DSM-4, ADOS, ASSQ or ADI-R (American Psychiatric Association, 1994; Lord et al., 1994; Ehlers, Gillberg, & Wing, 1999; LeCouteur & Lord, 2003);2) an experimental study design was used including a control group; 3) studies were published between 1st January 2005 and 31st December 2013 4) studies were published in English and in a peer-reviewed journal; 5) non-social stimuli were used.
Moreover, we interpreted results through different age-groups to account for developmental factors: younger than 6 years, 6 to 12 years, and older than 12 years. These ranges were chosen because 1) the highest difference in sensory symptoms between groups was reported for studies that included children ages 6.5 –9.4 years and these experiments reported homogeneous total and over-responsiveness scores (Ben-Sasson et al., 2009; Ermer & Dunn 1998; Gal 2006; Schaaf et al., 2006; Talay-Ongan and Wood, 2000); 2) the commonly used adolescent/adult sensory profile (Brown &
Dunn, 2002) measures from 11 years and above; 3) auditory structures and systems are fully developed by the age 12 years (Moore, 2002; Moore & Linthicum, 2007).
We included studies that are both controlling and not controlling for age and/or IQ and the respective tables state when studies were not controlled for age and/or IQ, since this could result in larger differences to the control group. Matched for IQ refers to matching for non-verbal IQ, performance IQ or MA. Studies are separated for mean age, if only a range is available, the experiment is sorted as mean of minimum and maximum age. In the case of multiple ASD groups, the study is assorted by youngest mean CA.
We included a total of 72 studies; thirty-three researching visual symptoms, seventeen auditory symptoms, seven tactile symptoms, one gustatory symptoms, five olfactory symptoms and ten multisensory symptoms.
Experimental studies
Visual
Multiple facets of visual perception in ASD where researched and this sense is the most extensively researched. The main aspects characterizing the experiments’ foci are: Visual search, global vs. detail processing, form and motion perception, visual acuity and neurophysiological aspects researched with EEG.
<6 years
Only one experiment was performed in young children with ASD, below the age of 6 years. This experiment revealed superior visual search skills. The toddlers with ASD detected the target twice as often as controls (Kaldy et al., 2011).
6-12 years
Global versus local processing is one aspect that reliably reveals sensory processing abnormalities in individuals with ASD (Happé & Frith, 2006). Global processing is considered the ability to integrate different pieces of information into a coherent whole, whereas local processing is the ability to focus and perceive local details of a certain object or situation (Happé & Frith, 2006). ASD is typically characterized by atypical global/local processing (Happé & Frith, 2006). Within the visual symptom domain, four studies investigated global versus local information processing among children with ASD, showing a consistent course. Children with ASD preferred global over local stimuli,
equally to individually verbal MA and CA matched controls (Deruelle et al., 2006; Iarocci et al., 2006). In addition, ASDs applied the same search strategy as controls, however, could almost not be diverted by eight distractors (Baldassi et al., 2009). Thus, the children with ASD exhibited stronger sensitivity towards local stimuli (Iarocci et al., 2006; Pellicano et al., 2005) without decreased sensitivity for global stimuli. Including DD controls suggests that this phenomenon appears to stem from ASD, rather than DD, since non-verbal MA controls were slightly more sensitive to local bias than verbal MA controls (Iarocci et al., 2006).
The findings for static global processing, however, differ from global motion processing results. Results from motion and form perception can reveal aberration in the two visual streams. Different stimuli can activate the lower (form) or higher (global motion) pathway of visual processing. Both performed experiments in this domain presented flicker light to estimate form processing and coherently upwards or downwards moving dots - among dots moving into random directions- to estimate global motion processing. Children with ASD processed form equally well as controls, suggesting a normally functioning lower level of the dorsal visual pathway. Yet, they exhibited higher global motion thresholds, reflecting impaired higher levels of the visual pathway (Pellicano et al., 2005). These results could be replicated, including a differentiation from dyslexic participants, which showed impairments in both pathways (Pellicano et al., 2008).
Processing in the lower level visual pathway was further investigated by means of EEG, where two homologous regions of the occipital cortex were recorded. Compared to the control group, the ASD group (controls differed in mean CA, IQ and gender) showed 50% less inter-hemispheric synchrony in and below the theta band, while no event-related inter-hemispheric activity above theta-band existed. This occurred despite bilaterally increased power. Wavelet power analysis revealed that the ASD group responded faster to stimulation, recovered slower, and alternated stronger at long latencies (Isler et al., 2010).
>12 years
Experiments investigating visual symptoms are only informative if ASDs vision is accurate. For this purpose, four experiments researched visual acuity (VA) in ASD. While one experiment reported ASDs VA laying in the region of VA reported for birds of prey (Ashwin et al. 2009), others found no difference to neither TD control (Bölte et al., 2012; Kéїta, Mottron & Bertone, 2010; Tavassoli et al., 2011) nor schizophrenics (Bölte et al., 2012). However, a crowding effect on VA from flanking stimuli occurred stronger for the TD group, supporting enhanced detail processing in ASD (Kéїta, Mottron & Bertone, 2010).
It has been speculated that enhanced visual search in ASD could account for some of the reported superior task performance (Plaisted, O'Riordan & Baron-Cohen, 1998). To elucidate the neural underpinnings of this phenomenon, ASDs performed a visual search task in an fMRI scanner (Keehn et al., 2008). Reaction times (RTs) were equal, but set size slopes indicated increased search efficiency in ASD. The most intriguing discovery, however, are the differential activated networks. While the ASD group activated parietal, frontal, and occipital networks, the TD group displayed less extensive activation, mostly limited to the occipito-temporal regions. The researchers concluded that search efficiency in ASD may be related to enhanced discrimination (displayed by occipital activation), as well as increased top-down modulation of visual attention (displayed with fronto-parietal activation) (Keehn et al., 2008). Another experiment presumed the cause for superior visual search to lay in enhanced memory for rejected distractor locations (Joseph et al., 2009). The ASD group reacted overall faster and did not disrupt in search efficiency in the dynamic condition. This reduces the likelihood of superior memory of rejected distractors affecting visual search. RT and set size relation (in both static and dynamic search) revealed lower intercepts for ASDs (correlating with increasing level of ASD in the static search condition). This suggests that superior search indeed derives from non-search processes. Furthermore, eye-movement analysis revealed shorter fixation duration of stimuli by ASDs. A third experiment found enhanced performance in simple visual detection tasks and enhanced discrimination of simple gratings, when compared to gifted controls with a visuospatial peak (Caron et al., 2006).
The latter result demonstrates how enhanced local processing can affect extending aspects of visual perception. Seven experiments on global vs. local processing were performed in ASD patients above the age of 12 years, making it the second most researched aspect of this age group (Bölte et
al., 2008; Caron et al., 2006; Damarla et al., 2010; Jarrold, Gilchrist & Bender, 2005; Manjaly et al.,
2007; Rondan & Deruelle, 2007; Wang et al., 2007). These experiments revealed preference for global stimuli in ASD (Rondan & Deruelle, 2007; Wang et al., 2007). However, all but two experiments found enhanced local processing (Bölte et al., 2008; Jarrold, Gilchrist & Bender, 2005; Rondan & Deruelle, 2007; Wang et al., 2007), even compared to a gifted control group with visuospatial peak (Caron et
al., 2006). One of the two experiments finding equal performance, however, revealed via the control
task that the ASD group performed better when local stimuli are present than without (Manjaly et al., 2007). The second experiment showing equal performance found other differences between the group: the HFA group activated their left dorsolateral prefrontal and inferior parietal areas less, while they stronger activated visuospatial areas (Damarla et al., 2010). Moreover, the HFA group’s higher-order working memory and executive areas were less functionally connected with visuospatial regions. Their size of the corpus callosum (an index of anatomical connectivity) was positively
correlated with frontal–posterior functional connectivity. Next to shorter RTs and higher accuracy, enhanced local processing in ASD becomes apparent through a number of other behavioral aspects. One aspect is ASDs latencies during the global vs. local feature detection task correlating with latencies in the condition, where the target was identified by a unique perceptual feature. Controls latencies, on the other hand, related to the condition where the target was distinguishable by a conjunction of features (Jarrold, Gilchrist and Bender, 2005). Also, ASD participants were stronger distracted by incongruent local features, interfering with correct responses, while controls expressed bidirectional interference (Wang et al., 2007). Interestingly, ASDs responded slower than controls during short exposure times of the stimuli and narrow visual angle (Wang et al., 2007). Through employing prior knowledge about activated brain regions, an fMRI study investigated whether enhanced detail processing in ASD results from a relative amplification of early perceptual processes (Manjaly et al., 2007). This would improve local processing, but not affect the global level as behavioral data report. The study revealed activations in right primary visual cortex and bilateral extrastriate areas during the search for local and global structures, whereas controls activated left parietal and premotor areas. Another fMRI experiment analyzed participants’ activity in striate and extrastriate visual cortex during a global vs. local processing compared to a colour counting control task (Bölte et al., 2008). The ASD group responded with lower hemodynamic changes to the global vs. local task in the right ventral quadrant of the visual area V2. Their overall findings indicate local processing being associated with altered responses of angle and grating-selective neurons. These contribute to shape representation, gestalt organization and figure-ground.
Psychophysical measures of global processing are thought to be reflected in form- and motion coherence thresholds. These can be investigated through pattern- and motion-sensitive mechanisms in the two visual streams (Tsermentseli, O’Brien & Spencer, 2008). While, in some experiments, ASD participants expressed significant form- and motion-coherence deficits (Davis et al., 2006; Sanchez-Marin & Padilla-Medina, 2008; Spencer & O’Brien, 2006), others found no difference after controlling for IQ (de Jonge et al., 2007; Jones et al., 2011; Koldewyn, Whitney and Rivera, 2010). In addition, one study found concurrent enhanced and decreased form perception in participants with HFA (Bertone et al., 2005). Here, participants displayed superior abilities for identifying the orientation of simple, luminance-defined (or first-order) gratings. However, they also displayed inferior performance for complex, texture-defined gratings. Other deficits in form- and motion processing were manifested by ASDs discriminating trajectories worse (Davis et al., 2006), perceiving high-spatial-frequency contrast maleficent (Davis et al., 2006) and tracking targets less successful (Takarae et al., 2008). Interestingly, ASDs without language delay (LD) required more time to track the targets compared to
ASDs with LD, failed to exhibit typical rightward directional advantage and their pursuit performance did not correlate with manual motor skills (like in ASD with delay). However, ASDs without LD exhibited only slight impairments in the left hemi field, while the group with LD was bilaterally impaired (Takarae et al., 2008). ASP participants did not differ from the controls in form- or motion perception (Spencer & O’Brien, 2006). However, both ASD and ASP participants express impaired object boundary detection. The timing of this deficient processing was determined via EEG as early as 120ms after stimulus presentation. ASDs further showed an increase in consecutive activity at lateral occipital sites (225ms). On the contrary, separating the figure from the background - associated with recurrent processing between higher and lower visual areas (around 260ms) - was typical (Vandenbroucke et al., 2009). Intriguingly, subjective visual hypersensitivity correlated with greater deficits across vision tests (Davis, 2006)
EEG is an excellent measure to assess neural dynamics, such as event-related potentials or oscillations, with high temporal precision (Senkowski et al., 2007). Two experiments researched early-stage visual processing abnormalities in ASD by investigating ERPs (Baruth et al., 2010), as well as spectral frequencies (Brown et al., 2005). ASDs expressed abnormally large cortical responses to task-irrelevant stimuli, over both frontal and parieto-occipital regions of interest. They furthermore displayed significantly higher rates of motor response error and overall stimulus discrimination disruption (Baruth et al., 2010). Regarding spectral analysis, ASDs displayed an increase of activity in all frequency bands. This included an early 100ms gamma peak, which is 50 to 70ms earlier then seen in the control group (Brown et al., 2005). Another EEG experiment evoked changes in the α- and γ-frequency bands, as well as visual evoked potentials (Milne et al., 2009). ERP processes varied between the groups, most variance was seen in the components spatially close to the striate or extrastriate cortex. Here, spatial frequency of the stimulus affected the induced increases in alpha-and gamma-balpha-and power in ASD. Further, stimulus spatial frequency reduced the time to peak α -band power in participants with ASD. Participants with ASD displayed an increase in induced α -band power of near the cingulate gyrus, whereas induced EEG signals near the parietal cortex did not differ.
Author Year Na CA Mean
CA Range
I
Q age Topic Task/Stimuli Results Ashwin et al. 2009 15 38.9 22-62 0 1
Visual
acuity Landolt-Cs ASD > control Baldassi et al. 2009 12 11.2 1 1
Global vs. Local
Detecting tilted among vertically striped
stim-uli ASD > control; ASD almost not influenced by distractors
Baruth et al. 2010 15 13.9 9-20 1 1 EEG Oddball task using illusory Kanisza squares
ASD = abnormally large cortical response to task-irrelevant stimuli over both frontal and parieto-occipital regions of interest -> differ-ences during early stages of visual processing. ASDs = higher rates of motor response error & overall stimulus discrimination disrup-tion
Bertone 2005 13 22.3 11-31 1 1 Form & Motion
processing Flicker contrast sensitivity task
ASD > control at identifying the orientation of simple, luminance defined gratings. ASD < at identifying complex, texture defined gratings
Bölte et al. 2008 7 27.7 1 1
Global vs.
Local fMRI + Wechsler Block Design test (BDT)
ASD < control in activating the right ventral quadrant of the visual area V2. In ASD, local processing in BDT related to altered re-sponses of angle- and grating-selective neurons. ASD > control in low-level processing causing superior performance in BDT Bölte et al. 2012 34 19.8 1 1
Visual
acuity Incl. schizophrenics: Landolt-Cs ASD = control & schizophrenics Brown et al. 2005 24 14.7 11-17 1 1 EEG
Oddball task using illusory Kanisza squares. Control=MA matched
ASD = overall increase in activity -> including early 100ms gamma peak, 50 to 70ms earlier than control
Caron et al. 2006 8 18.8 0 0
Global vs. Local
Modified BDT during various levels of percep-tual cohesiveness, global perception, visuomo-tor speed, visual search, speed of visual en-coding and visual memory
ASD < control in perceptual coherence. ASD > control in local pro-cessing, simple visual detection discrimination of simple grating. ASD = control in remaining tasks
Damarla et al. 2010 13 19 15-35 1 1
Global vs. Local
FMRI + Children's Embedded Figures task (CEFT)
ASD = controls in CEFT. ASD<control in activating DLPFC & inferior parietal areas. ASD > control in activating visiospatial areas. ASD < control in funtional connections between higher-order working memory and executive areas. ASDs size of corpus callosum = posi-tively correlated with frontal-posterior functional connectivity
Table 1.1: Overview of reviewed visual Papers. “Na” indicates the number of participants, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age”
Author Year Na CA Mean
CA
Range IQ age Topic Task/Stimuli Results Davis et
al. 2006 9 12.3 10-18 0 1 Form & Motion processing
Presentation of letters (form perception) & moving dots (motion perception) with varying amount of visual noise in both matching and discrimination paradigm
ASD < control in discriminating trajectories and perceiving high spatial frequency contrast. Subjective visual hypersensitivity in ASD correlated with greater deficit across vision tests
de Jonge
et al. 2007 varying varying 7-12; 13-18;
19-33 1 1
Form & Mo-tion pro-cessing, Vis-ual acuity
Longitudinal + parents: Contrast sensitivity with Vistech contrast sensitivity charts; Visual acuity (VA) with Landolt-Cs; Motion coherence task with Random Moving Dots paradigm (RMDP)(coherently moving dots among ran-domly moving dots); Moving-shape discrimina-tion task with moving shapes (composed of dots) against random-moving-dots back-ground; Form discrimination task with white
shapes against black background ASD = control in all tasks Derulle et
al. 2006 13 7.7 1 1
Global vs.
Local Preference for hierarichal Figures ASD = control in Preference for global level Iarocci et
al. 2006 12; 20 7.9;7.8 6-10 1 1
Global vs. Local
Dot pairs in local and global spatial relation with normal, gobal difficult and local difficult condition ; Hierarchical shapes (dots and
squares) ASD=control in preference for global level; ASD>control in sensitivity for local bias Isler et al. 2010 6 7.8 6-8 0 0 EEG Long latency flicker light
ASD responded faster, recovered slower and alternated stronger at long latencies. ASD bilater-ally increased power, but 50% less inter-hemispheric synchrony in and below theta and no event related activity above theta
Jarrold, Gilchrist,
Bender 2005 18 12.4 8-15 1 0
Global vs. Local
CEFT; Visual search task with feature and con-junction trial: Stimuli = clowns with varying col-ors and shapes
ASD > control in target detection. ASDs CEFT latencies corresponded to latencies in the feature condition, controls' to conjunction condition
Jones et
al. 2011 89/79 15.6 15-17 1 1 Form & Motion processing
RMDP; Form-form-motion coherence: Stimuli = dots forming a rectangle set against a back-ground; elements in the background moved coherently in one direction, while dots of the rectangle coherently moved in the opposite
di-rection. Tested across spectrum of IQ. ASD = control
Table 1.2: Overview of reviewed visual Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether
Author Year N a
CA Mean
CA
Range IQ age Topic Task/Stimuli Results Joseph et
al 2009 21 14.7 11-19 1 1 Visual search
Standard static & dynamic search task: Stimuli
= Letter "L" as target, "T"as distractor ASD > control at RT; ASD> control at set size slopes; ASD> control fixation duration Keehn et
al. 2008 9 15.1 1 1 Visual seach
FMRI + Detecting t-shaped characters,
dis-tracters = “t” with any abnormal rotation ASD = control at RT; ASD > control in set size slopes; ASD activated different brain areas Keita, Mottron
& Bertone 2010 19 19.6 13-33 1 1 Visual acuity
Landolt-C, VA during simultaneously presented
flanking stimuli ASD = control in VA. ASD > control in distraction through flanking stimuli Koldewyn,
Whitney &
Rivera 2010 30 15.1 11-20 0 1 Form & Motion
processing "Glass Pattern" (form perception) and RMDP
ASD = control in form perception. ASD < control in motion perception. But: ASD = control after controlling for IQ
Manjaly 2007 12 14.4 1 1
Global vs. Local
FMRI + Embedded Figures task (EFT), visuospa-tial control task with minimal local search
ASD = control in EFT; ASD slower control in control task; ASD activated right primary visual cortex & bilateral extrastriate areas during the EFT, whereas controls activated left parietal & premotor areas
Milne et al. 2009 20 12.1 8-18 0 0 EEG
Observing Gabor patches = black & white stripes of varying spatial frequencies
ASD's ERPs differed most spatially close to the striate or extrastriate cortex; Spatial frequency of the stimulus affected the induced increases in alpha- and gamma-band power. Stimulus spatial frequency reduced the time to peak α -band power in ASD. ASD =increase in induced α -band power of near the cingulate gyrus, induced EEG signals near the parietal cortex did not differ Pellicano et al. 2005 20 9.6 8-12 1 1
Form & Motion
processing Observing flicker light; RMDP; CEFT
ASD=control in flicker light detection thresholds; ASD<control in global motion detection thresh-olds; ASD > controls in CEFT latencies
Pellicano &
Gibson 2008 20 9.56 8-12 1 1 Form & Motion
processing Observing flicker light; RMDP
ASD=control in flicker light detection thresholds; ASD<control in global motion detection thresh-olds
Rondan &
Derulle 2007 26 26 18-43 0 1
Global vs.
Local Determining hierarichal geometrical shapes ASD = control in Preference for global level Sanchez-Marin & Padilla-Medina 2008 6 12.5 7-17 0 1 Form & Motion processing
Contrasting two images: one containing only Gaussian noise, one containing Gaussian noise
plus an additional signal ASD < control in both detecting still and moving additional signal
Table 1.3: Overview of reviewed visual Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether
Author Year N a
CA Mean
CA
Range IQ age Topic Task/Stimuli Results Spencer & O'Brien 2006 25 13.5; 12 1 1 Form & Motion processing
Detecting spatial form- and motion-coherence
thresholds using "Glass Patterns" ASD < control in both detecting form- and motion-coherence. ASP = control Takarae et al. 2008 77 15.3; 16.3 1 1 Form & Motion processing
Eye movement tasks, Motion perception tasks, Motion coherence tasks. ASD with and without language delay (LD)
ASD = control in 1st Eye Movement Tasks; ASD < control in 2nd Eye Movement task. ASD with LD < control (ASD w/o LD = control) in Motion perception.
Tavassoli et al 2011 20 30.4 1 1 Visual acuity
Landolt-Cs, Freiburg Visual Acuity test (FrACT), Early Treatment Diabetic Retinopathy Study
test (ETDRS) ASD = control in all tasks
Vandenbroucke
et al. 2009 13 20.8 16-28 1 1 Form & Motion processing
EEG + Detecting object boundaries of black squares in four orientations
ASD < control in object boundary detection already 120ms after stimulus presentation. A neural network model relates this to dysfuctional horizontal connections within early visual areas. ASD > control in activating occipital areas, possibly reflecting compensational mechanisms. ASD = con-trol in separating the figure from the background, associated with recurrent processing between higher and lower visual areas
Wang et al. 2007 15 14.7 8-21 1 1
Global vs. Local
Navon-type hierarichal numbers; free & forced choice condition with 3 visual angles & 3 expo-sure times
ASD > control in responding to local stimuli but preferred global, in free choice. ASD< control through local bias interferring stronger with correct responses in forced choice (controls = bidirec-tional). ASD < controls in RTs during short exposure time & effects of visual angle
Table 1.4: Overview of reviewed visual Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether
Auditory
Experiments in the auditory domain are less frequent than the ones in the visual domain and represented broader approaches. Thus, they could not be grouped for topics but for method of data acquisition i.e. behavioral, EEG, MEG, fMRI.
<6 years
Two experiments were performed in ASDs under the age of 6 years. Auditory brain stem responses were investigated in order to determine hearing status and the integrity of auditory brain stem pathways (Kwon et al., 2007). Analysis of auditory brain stem responses revealed prolonged inter-peak latency of wave V, wave I-V and wave III-V in the ASD group. The second experiment researched P50 sensory gating and its relation to IQ and EEG gamma waves (Orekhova et al., 2008). Sensory gating describes the inhibitory modulation of sensory input, of which a dysfunction may contribute to ASD phenotypes. Analysis revealed that a suppressed P50 response to a second click was significantly reduced in mentally retarded children with ASD, but normal in high-functioning ones. It was also noted that P50 gating improved with age in both groups. Lower P50 suppression also corresponded to higher ongoing gamma power in the children with ASD.
6-12 years
The capacity to discriminate the pitch of tones is a straightforward measure to assess the auditory capacity of patients with ASD without the need for imaging techniques. ASDs were found to detect pitch direction over small pitch distances superior to controls, whereas detection of musical contours was equal (Heaton, 2005; O'Riordan & Pasetti, 2006). Furthermore, ASD participants showed no difference in pitch sensitivity for music and speech stimuli (Järvinen-Pasley et al., 2007), while a significantly poorer performance during speech conditions was seen in the control group.
The mismatch negativity (MMN) is a component of an event-related Potential (ERP) and results from detecting an odd stimulus in a sequence of stimuli (Rinne et al., 2000). It is generated in the temporal and frontal lobe (Rinne et al., 2000). ERPs can be recorded with EEG or MEG, however, EEG represents the more practical method (Kaplan et al., 2005). MMN were investigated related to simple auditory stimuli (Dunn, Gomes and Gravel, 2008), duration, pitch and vowel changes while hearing speech stimuli as well as for changes in non-speech counterparts of the stimuli (Lepistö et al., 2005) and for similarity to parental traits (Jansson-Verkasalo et al., 2005). Regarding simple auditory stimuli, ASDs demonstrated a significantly smaller amplitude of the MMN during unattended conditions. Interestingly, its amplitude was equal between groups when attending to the stimuli
(Dunn, Gomes and Gravel, 2008). Investigating different stages of cortical auditory processing (sound encoding, discrimination and orienting) in children with ASD revealed impaired sound encoding, reflected by diminished amplitude in response to sound repetition. In addition, enhanced MMN was observed during pitch changes, as well as impaired discrimination between duration changes (Lepistö et al., 2005). Children with ASP have been known to share similar behavioral characteristics with their parents; the syndrome is known to run particularly in the male side of the family. In a third experiment, MMN was used to investigate if children with ASP share central auditory processing traits with their parents. Sound encoding by children with ASP was similarly abnormal than in both their mother and father. Additionally, abnormal cortical auditory discrimination resembled paternal traits in children with ASP (Jansson-Verkasalo et al., 2005).
Spatial source localization with MEG is somewhat superior to EEG (Cohen et al., 1990), for this purpose MEG is the superior method when specified regions are of interest. The auditory evoked field response at the superior temporal gyrus at 50ms (M50) and 100ms (M100) was investigated to elucidate its stimulus-frequency dependence (Roberts et al., 2010). The researchers suspected an effect of language impairment (LI) in ASD. For this purpose the ASD group was separated into two subgroups – with and without LI. No differences between groups were seen at M50. Yet, children with ASD (with and without LI) showed ~ 11ms delayed M100 response in the right hemisphere. Largest latency differences were observed during the 500Hz condition in the right hemisphere. A fundamental dilemma when studying ASD is the difficulty in modeling complex behavioral phenotypes (such as language) in rodents. Therefore, a study was conducted, assessing the translational potential of auditory evoked potentials (EP) endo-phenotypes seen in humans with ASD, into a mouse model for ASD (Gandal et al., 2010). Whole-cortex MEG was recorded during presentation of a pure auditory tone. Additionally, auditory EPs were recorded in mice that have been prenatally exposed to valproic acid (VPA mice = common ASD mouse model) and analyzed with analogous methods. Both the human group with ASD and the VPA exposed mice group exposed a similar 10% latency delay in the N1/M100 ERP. Also, both groups exhibited a reduction of the gamma frequency (30-50Hz) phase-locking factor. The delay is associated with ASD like behavioral deficits in the mice. In the VPA mice, gamma phase-locking factor was correlated with expression of a risk gene for ASD, neuroligin-3. As well as the neural deficits seen in the mice were modulated by the mGluR5-receptor antagonist MPEP.
> 12 years
Behavioral experiments in ASDs above 12 years of age investigated sound discrimination (Bonnel et al., 2010; Jones et al., 2009), as well as their capability to modulate noise (Alcántara et
ASDs performed equally well as adults (Jones et al., 2009). However, an ASD-subgroup (20%) characterized by average intellect and a delayed language onset, demonstrated increased frequency discrimination skills (1.65 SDs above the control mean). Participants in the ASD group who performed poorly on the intensity discrimination task reported more auditory symptoms, associated as a coping mechanism for loudness. Interestingly, more auditory symptoms were reported across the full range of measures for individuals who performed well during the discrimination task. A second study researched enhanced perception of simple tones, but diminished perception of complex tones in ASD stemming from additional regions involved in processing (Bonnel et al., 2010). The study comprised four auditory discrimination experiments targeting pitch, non-vocal and vocal timbre and loudness. Stimuli were designed to be temporally simple or complex requiring integration of multiple regions. The participants with ASD (but not ASP) displayed enhanced pitch discrimination for simple tones. However, no significant differences were seen in complex tasks designed to incorporate various auditory regions between participants with ASD and controls. The third experiment originated from previous reports of decreased performance in temporally modulated noise paradigms in ASD. Participants were exposed to sinusoidal amplitude modulations of a broad band noise, where ASDs exhibited a generally higher modulation rate in all conditions as well as temporal processing efficiency and differences of the temporal-envelope resolution (Alcàntara et al., 2012).
Noise modulation in ASD was further investigated by means of EEG under a slightly different aspect (Teder-Sälejärvi et al., 2005). The ability of ASDs to focus attention on the selecting a sound within a noisy environment was investigated. Opposed to the above stated finding, here analysis of ERP amplitude indicated that people with ASD have a lesser ability to focus their attention on a relevant sound source when compared with TD controls. Reported deficits in long range functional connectivity as well as decreased inter-regional interactions are hypothesized to contribute to phenotypical ASD symptoms. To find out whether deficient functional connectivity stems from aberrant long-range tracts, or a local deficit in neural connections, the integrity of local circuitry was investigated via gamma band activity in auditory cortices (Wilson et al., 2007). The ASD group exhibited significantly reduced left hemispheric 40 Hz power from 200-500ms post-stimulus onset. A second MEG project investigated aberrant arousal systems which may compromise ASD’s ability to regulate an optimal response (Orekhova et al., 2012). The P100m component of the response wave is the most prominent one of the field response in children; previous studies have shown that it may reflect pre-attentive arousal processes. TD participants displayed a rightward lateralized P100m not seen in the children with ASD. The children with ASD also exhibited a tendency to reduced P100m in the right hemisphere.
An fMRI design was implemented in order to explore the neural basis of complex non-social sound processing via modulation of the spectral and temporal complexity of auditory stimuli (Samson et al., 2011). Both groups performed similarly on the sound detection task. Yet, the control group displayed great activity in anterolateral superior temporal gyrus during increasing temporal complexity, while the ASD group activated the Heschl’s gyrus. In participants with ASD, temporally complex, but not spectrally complex sounds lead to increased activity in the primary auditory cortex and a reduced activity in the non-primary auditory cortex.
Author Year Na CA Mean
CA
Range IQ Age Task/Stimuli Results
Alcàntara
et al. 2012 6 12.8 10-14 1 1
Presentation of sinusoidal amplitude modulations of a broad band noise
ASDs exhibited a generally higher modulation rate in all conditions, plus significant differences of the temporal-envelope resolution and temporal processing efficiency
Bonnel et al 2010 15; 14 22.7; 24.2 14-36 1 1
Auditory discrimination experiments targeting pitch, non-vocal and vocal timbre and loudness: Stimuli were temporally simple or complex requiring integration of multiple regions.
ASD (but not ASP) > controls in pitch discrimination for simple tones. ASD = control in complex tasks designed to incorporate various auditory regions
Dunn, Gomes &
Gravel 2008 34 9.0 6-12 1 1
EEG + Oddball sound encoding: a deviant stimulus
is presented randomly in a series of standard stimuli
ASD < control in MMN amplitude during unattended conditions. ASD = control in MMN Amplitude during attended conditions
Gandal et
al. 2010 25 10.2 0 1
MEG + presentation of pure auditory tone. Subjects were ASDs, TDs & VPA mice
ASD = VPA mice in 10% latency delay of N1/M100 ERP. ASD = VPA mice reduction of gamma phase-locking factor. In the VPA mice, gamma phase-locking factor was correlated with expression of risk gene for ASD, neuroligin-3. Neural deficits seen in the mice were modulated by the mGluR5-receptor antagonist MPEP
Heaton 2005 15 10.0 7-15 1 1
Discriminating pitch intervals and musical contours
HFA> control in pitch direction over small pitch distances. ASD = controls in contour discrimination
Jansson-Verkasalo
et al. 2005 19 10.6 0 1 EEG + Oddball sound encoding
ASP sound encoding MMN = similarly abnormal than both their mother's and father's. ASPs abnormal cortical auditory
discrimination resembled paternal traits.
Järvinen-Pasley et
al. 2007 19 11.6 8-17 1 1
Pitch sequence discrimination task: music or speech
ASD = control in music pitch discrimination. ASD > control in speach pitch discrimination. ASD pitch discrimination is less domain-specific.
Table 2.1: Overview of reviewed auditory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states
Author Year Na CA
Mean CA Range IQ Task/Stimuli Results
Jones et
al. 2009 72 15.5 1 1
Auditory discrimination capacity for frequency, intensity and duration differences in pairs of sounds
ASD = control. However, ASD-subgroup (20% ) with average intellect & delayed language onset > controls in frequency discrimination (1.65 SDs). ASDs who performed poorly on the intensity discrimination task reported more auditory symptoms. More auditory symptoms across full range of measures in individuals who performed well during the discrimination task.
Kwon et
al. 2007 71 3.5 0 1
Determining hearing status and the integrity of auditory brainstem pathways via auditory brainstem
responses ASD = control difference in pitch sensitivity across conditions
Lepistö et
al. 2005 15 9.4 7-11 0 1
EEG + sound encoding, discrimination and orienting: children were
instructed to watch silent videos and ignore the auditory stimuli
ASD < control in sound encoding, reflected by diminished amplitude in
response to sound repetition. ASD > control in MMN during pitch changes. ASD < control in discriminating duration changes.
Orekhova
et al. 2012 14 12.7 0 1
MRI & MEG + Stimuli were paired white noise clicks 1000ms intervals within & randomly varying 8–11 sec
intervals between the ASD did not display rightward lateralized P100m.
Orekhova et al. 2008
21 5.9 3-8 0 1
EEG + Paired clicks paradigm
Mentally retarded ASD < control P50 response to second click (HFA ASDs = control). P50 gating improved with age in both groups. Lower P50 suppression corresponded to higher ongoing gamma power.
O'Riordan & Passetti
2006 12 8.0 1 1
Pitch discrimination ASD > control in auditory discrimination.
Roberts et al.
2010 25 10.2 1 1
MEG + sinusoidal tones binaurally presented
ASD = control at M50. ASD showed 11ms delayed M100 latency int he right hemisphere. Largest delay of M100 in right hemisphere at 500Hz stimulus. M100 delay = unrelated to language delay.
Table 2.2: Overview of reviewed auditory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states
Author Year Na CA Mean
CA
Range IQ Age Task/Stimuli Results
Samson et al.
2010 15 24.4
14-35 1 1 fMRI + Presentation of pure or harmonic tones in different frequencies
ASD = control in the detection task.Controls displayed great activity in anterolateral superior temporal gyrus during increasing temporal complexity, while the ASD group activated the Heschl’s gyrus. In ASDs, temporally complex, but not spectrally complex sounds lead to increased activity in the primary auditory cortex & reduced activity in the non-primary auditory cortex.
Teder-Sälejärvi 2005 7 33.3
29-39 0 1 EEG + Focusing attention on selecting
a sound within a noisy environment ASD < controls in ability to focus their attention on a relevant sound source Wilson et
al. 2007 10 12.4 7-17
0;
covariate 1
MEG + Presentation of 500ms monaural click trains with 25ms inter-click intervals
ASDs exhibited significantly reduced left hemispheric 40 Hz power from 200-500ms post-stimulus onset.
Table 2.3: Overview of reviewed auditory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states
Olfactory
Olfactory processing is an interesting aspect of sensory symptoms in ASD, because all sensory processing except the olfactory passes through the thalamus and thalamic dysfunctions have been suggested as the cause for sensory symptoms in ASD (Hardan et al., 2008).
6-12 years
Olfactory identification (OI) was the main aspect of interest (Brewer et al., 2008; May et al., 2011), but also odor detection thresholds (Dudova et al., 2011) and pleasantness of smells was investigated (Hrdlicka et al., 2011). OI deficits are associated with orbitofrontal cortex (OFC) dysfunction (May et al., 2011), as well as neurodevelopmental arrest of the limbic-prefrontal networks; these same networks are affected in the development of ASD and are observed in some of the ASD population (Brewer et al., 2008). In a 5 year follow-up study on children with HFA who previously exhibited an abnormal age-OI association, all groups displayed improved OI through both nostrils during the follow up exam (May et al., 2011). However, some HFA participants expressed reduced OI, whereas ASP did not. Additionally, IQ-OI relationships were exhibited between the HFA and ASP groups. Two other experiments, however, found no OI impairment identified in ASD participants (Brewer et al, 2008; Dudova et al., 2011). Yet, OI was negatively associated with age in the participants with HFA. Odor detection thresholds were significantly impaired in both HFA and ASP (Dudova et al., 2011). ASDs were better at identifying the smell of an orange, but worse for cloves. No other identification differences were seen between the 14 other smells. Altogether, ASDs did not differ on total olfactory identification score. OI ability, as expressed by the total OI score, correlates with age in the control group, but not in the ASD group. The authors concluded both impaired as well as essentially normal OI detection in the ASD population studied. Furthermore, ASDs experienced the smell of cinnamon and pineapple as significantly less pleasant when compared to the controls (Hrdlicka et al., 2011).
>12 years
Author Year Na CA Mean CA
Range IQ Age Task/Stimuli Results
Brewer et al. 2008 15 7.1 5-9 1 1
Olfactory Identification (OI) with Univesity of Pennsylvania
Smell Identification Test (UPSIT)
ASD = control in olfactory identification (OI). OI was negatively associated with age in ASD.
Dudova et al. 2011 35 10.8 0 1 OI with Sniffin' Sticks
ASD < control in odor detection threshold. ASD > control at identifying smell of orange, but < control at identifying smell of clove. ASD = control at identifying remaining 14 smells
Hrdlicka et al. 2011 35 10.8 0 1 Pleasentness of odors with Sniffin' Sticks
ASDs experience the smell of cinnamon & pineapple less pleasant
May et al. 2011 9 HFA; 9 Asp 7; 12 1 1 Longitidunal. OI with UPSIT
ASD < control in OI. ASP = control in OI. HFA & ASP exhibited a IQ-OI relationship
HFA = control in improvement over age.
Tavassoli & Baron-Cohen 2012 38; 19 35.9; 28 1; 1 0; 1 OI and olfactory adaptation with Sniffin' Sticks ASD = control
Table 3: Overview of reviewed olfactory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states
Gustatory
Extreme food selectivity has been reported in ASD (Paul et al., 2007). Thus, investigating gustatory processing can aid identifying the causal mechanisms and adjust ASDs diet in a way that is both healthy and pleasant.
>12 years
Only one experiment met the inclusion criteria in the gustatory domain. Here, altered taste perception in ASD was explored (Tavassoli et al., 2012) by using ”Taste strips” for various flavors. ASDs displayed overall lower taste scores, as well as lower scores for bitter, sour and sweet. Only the perception of saltiness was normal.
Author Year Na CA
Mean IQ Age Task/Stimuli Results
Tavassoli &
Baron-Cohen 2012 23 35.8 1 0
Taste identification overall, as well as bitter, sour, sweet and
salty using Taste Strips ASD < control in all taste identifications except salty
Table 4: Overview of the reviewed gustatory Paper. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether control subjects were matched (1) or not matched (0) for IQ and age, respectively.
Tactile
Unusual tactile sensitivity is heavily associated with ASD (American Psychiatric Association, 2013). However, only little psychophysical proof for it exists. The following studies aimed to elucidate the characteristics of tactile sensation in ASD.
6-12 years
In this age group, ASD patients did not express lower selectivity for tactile stimuli per se (O’Riordan & Passetti, 2006; Ploog & Kim, 2007). They were only slightly less selective during a condition, where two previously positive associated stimuli were presented (Ploog and Kim, 2007). Sensitivity to wet or dry sandpaper, as well as psychophysical vibro-tactile threshold revealed no abnormalities in ASD (Güclü et al., 2006; O’Riordan & Passetti, 2006).
> 12 years
The experiments above 12 years of age were all performed in adults. Here, sensitivity to tactile information (Blakemore et al., 2006; Cascio et al., 2008), spatial localization (Tommerdahl et
al., 2007), as well as temporal order judgment (TOJ) and temporal discriminative threshold (TDT)
(Tommerdahl et al., 2008) were inspected. Self-activated tactile stimulation and external stimulation were both rated as more intense by ASD participants, while both groups preferred self-touch (Blakermore et al., 2006). Moreover, ASDs expressed increased sensitivity to vibration (Cascio et al., 2008). Regarding temporal aspects of tactile perception, ASDs did not demonstrate decreased TOJ to the same extent as controls, while experiencing synchronized conditioning stimuli. Additionally, the TOJ threshold of the ASD group is well above the TD controls (Tommerdahl et al., 2008). In judging spatial localization, ASDs performed superior when the adaptive stimulation is short (<.05), but equal performance during longer stimulation (Tommerdahl et al., 2007). Furthermore, they showed lower tactile threshold only at the 200Hz frequency, demonstrating hypersensitivity at 200Hz but not 30Hz (Blakermore et al., 2006). In contrast to finding displaying different perception, ASDs were equally sensitive to stimulation on the palm of the forearm- this area is especially sensitive through unmyelinated afferents. ASDs detected light touch, warmth and cool equally and reported similar pleasantness of textures.
Author Year Blakemore et al. 2006 Cascio et al. 2008 Güclü et al 2006 O'Riordan & Passetti 2006
Ploog & Kim 2007
Tommerdahl 2008
Tommerdahl et
al. 2007
Table 5: Overview of reviewed tactile Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the
chronological age-range of autistic participants. “IQ” and “Age” states whether control subjects were matched (1) or not matched (0) for IQ and age, respectively
Multisensory
A problem with multisensory integration of sensory information is one theory for the cause of sensory symptoms in ASD (Waterhouse, Fein & Modahl, 1996). Abnormalities in the hippocampus result in failed binding of all incoming sensory information of one event, which causes unusual reactions (Waterhouse, Fein & Modahl, 1996). Most studies applied audio-visual stimuli to research reactions to multisensory stimulation.
<6 years
Below the age of 6 years, toys could be used to observe interest in or aversion to multisensory stimulation. While all groups (ASD, TD and DD) displayed MA as a predictor of hyper-responsiveness, autistic children, higher levels of sensory aversion than TD children; aversion to multisensory toys increased as MA decreased. Additionally, non-responders to auditory stimuli were more prevalent in ASD (Baranek et al., 2007).
6-12 years
A standard metric EEG test of multisensory integration (MSI) - to determine the integrity of audio-somatosensory integration in children with ASD - revealed a difference in the integration between auditory and somatosensory information. TD children displayed a MSI at around 175ms, which was not seen in the children with ASD (Russo et al., 2010).
>12 years
Above the age of twelve years, multisenssory illusion (Foss-Feig et al., 2010; van der Smagt, Engeland and Kemner; 2007), audiovisual integration (Keane et al., 2010; Mongillo et al., 2008; Taylor, Isaac and Milne, 2010), as well as ERPs were researched (Maekawa et al., 2011; Magnee et al., 2009). Multisensory temporal functions in children with ASD were investigated by applying a temporally dependent low-level multisensory illusion (Foss-Feig et al., 2010; van der Smagt, Engeland and Kemner, 2007). The illusion was evoked through one light flash followed by or simultaneously with audio stimulation (single or multiple), resulting in a potential extra flash hallucination. Times between the flash and audio stimulation were altered to elucidate a temporal time-window of hallucinatory effect. While one experiment demonstrated participants with ASD seeing the hallucinatory flash over an extended range of flash/audio asynchronies (Foss-Feig et al., 2010), another one showed both groups increase in perceived flashes coincided with an increase in auditory stimuli (van der Smagt, Engeland and Kemner, 2007).
Audiovisual integration of speech, spatio-temporal relations, and temporal numerosity in adults with HFA revealed that ASDs and TDs were similarly affected by stimuli in congruency, and that audiovisual congruency did not affect one group more than the other (Keane et al., 2010). Furthermore, effects of altered uni-sensory temporal processes on multisensory abnormalities showed no differences between thresholds in the visual TOJ task between groups (Kwakye et al., 2010). However, children with ASD expressed higher thresholds for auditory TOJ tasks, indicating auditory impairment in temporal processing. On the multisensory TOJ task, children with ASD also showed a wider range of temporal intervals, displaying an extended temporal window for multisensory integration.
Regarding non-temporal audiovisual integration tasks, six tasks revealed normal levels in ASD for tasks with non-social stimuli (Mongillo et al., 2008), another group found a delay in visual accuracy and audiovisual integration (Taylor, Isaac and Milne, 2010). Additionally, audiovisual integration improved with age in the ASD group (Taylor, Isaac and Milne, 2010).
ERP studies elucidated bottom-up and top-down visual processing in people HFAs (Maekawa et al., 2011), as well as cross-sensory P50 suppression (Magnee et al., 2009). Participans with HFA detected visual targets faster, while focusing on a story, than their TD controls (Maekawa et al., 2011). P1 and P300 amplitudes, as well as a delayed P300, were significantly decreased in participants with HFA. Thus, bottom-up attention was normal, despite abnormal low-level (P1) and top-down (P300) visual information processing. Low-level audiovisual interaction was further investigated with EEG by means of cross-sensory P50 suppression (Magnee et al., 2009). No differences were found in neither auditory nor cross-sensory P50 suppression in people with HFA when compared with the TD controls. Differences were, however, seen between the control group and schizophrenic group, elucidating possible integrity issues involved in low-level cross-sensory processing.
Author Year Sense Na CA Mean
CA
Range IQ Age Task/Stimuli Results
Baranek
et al 2007 multisensory 56 3.7 0 0
Determining hyper-responsiveness via Sensory Processing Assessment for young children (SPA)
ASD & DD > control in sensory aversion. Aversion to multisensory toys increased as MA decreased. ASD > DD & control in nr. of non-responders to auditory stimuli
In all groups MA = predictor of hyper-responsiveness.
Foss-Feig
et al. 2010 multisensory 21 12.6 1 1
Determining multisensory temporal function via temporally dependent low-level Multisensory Illusion: Stimuli = light flash followed by audio stimulation (single
or multiple). ASD saw hallucinatory flash over an extended range of flash/audio asynchronies
Keane et
al. 2010 audiovisual 9 30.0 18-49 1 1
Audiovisual integration of speech, spatio-temporal relations, and spatio-temporal numerosity: Stimuli = vowels; grey discs, briefly flashing white plus 2 tones; visual
flash plus tone ASD = control
Kwakye
et al. 2010 35 12.2 1 1
Visual temporal order judgement (TOJ); auditory TOJ; audiovisual TOJ
ASD = control in visual TOJ; ASD < controls in threshold of auditory TOJ; ASD > controls in performance improvements over a wide range of temporal intervals
Maekawa
et al. 2011 audiovisual 11 28.0 18-40 0 1
Recording P1 & P300 ERPs, while responding to target stimuli on a screen & focusing on an auditory story
ASD > controls in target detection. P1 & P300 amplitudes were decreased in ASD & P300 was delayed.
Magnee
et al. 2009 audiovisual 13 22.9 1 1
EEG + P50 suppresion paradigm: Stimuli = white oval followed by click
Mongillo
et al. 2008 audiovisual 15 13.7 8-19 1 1
McGurk Task; Gender, Vowel, Ball Size and
Ball Composition Match/Mismatch Tasks ASD = control in non-social tasks Russo et
al. 2010
auditory-somatosensory 17 10.4 6-16 1 1
EEG + standard metric test of multisensory integration (MSI)
ASD integrate auditory and somatosensory information different than controls. ASD did not display MSI at 175ms as control. Overall MSI is less extensive in ASD.
Table 6.1: Overview of reviewed multisensory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether control subjects were matched (1) or not matched (0) for IQ and age, respectively.
Author Year Sense Na CA Mean
CA
Range IQ Age Task/Stimuli Results
Taylor, Isaac &
Milne 2010 audiovisual 24 12.6 7-17 1 1
audiovisual integration via audiovisual integrative McGurk exam
ASD showed delay in visual accuracy and audiovisual integration. Audiovisual integration improved with age in the ASD.
van der Smagt, Engeland &
Kemner 2007 audiovisual 15 20.7 1 1
Multisensory Illusion: Stimuli = light flash
followed by audio stimulation. ASD = control
Table 6.2: Overview of reviewed multisensory Papers. “Na” indicates the number, “CA Mean” the mean chronological age, and “CA Range” the chronological age-range of autistic participants. “IQ” and “Age” states whether control subjects were matched (1) or not matched (0) for IQ and age, respectively.
Discussion
This review aimed to find quantitative differences in sensory symptoms between the senses, while elucidating their quality through reviewing experimental studies. Moreover, variance in existence or strength of sensory symptoms between three age-groups was explored.
Differences in existence or intensity of sensory symptoms between the five senses
Due to the variable approaches and lack of replication, it is hardly possible to determine a relation of strength for sensory symptoms to each other (e.g. superior visual processing, inferior gustatory processing). After reviewing 72 controlled experimental studies, it becomes apparent that aberrant sensory processing in ASD is a complex phenomenon and varies within the disorder. Even within the same age group, results (e.g. see Visual Acuity) differ conspicuously. Yet, enhanced local processing (Iarocci et al., 2006; Pellicano et al., 2005) as well as abnormal form and motion perception deficits (Davis et al., 2006; Sanchez-Marin & Padilla-Medina, 2008; Spencer & O’Brien, 2006) in the visual domain can safely be attributed as sensory characteristic in ASD.
Additionally, both visual and auditory functional imaging experiments reveal abnormal functional pathways (visual: Vandenbroucke et al, 2008), regional functional activation (auditory: Samson et al., 2011. Visual: Baruth et al., 2010; Bölte et al., 2008; Damarla et al., 2010; Keehn at al., 2008; Manjaly et al., 2007), ERPs (auditory: Dunn, Gomes and Gravel, 2008; Jansson-Verkasalo et al., 2005; Gandal et al., 2010; Lepistö et al., 2005; Orekhova et al., 2008; Orekhova et al., 2012; Roberts et al., 2010; Teder-Sälejärvi et al., 2005. Visual: Milne et al., 2009) as well as use of oscillatory frequencies (Visual: Brown et al., 2005; Isler et al., 2010; Milne et al., 2009; Wilson et al., 2007).
Research in the auditory domain revealed several abnormal aspects in ASD processing, yet replication of the specific approaches disenables determining these findings as characteristic. Abnormal processing aspects include aberrant brainstem responses (Kwon et al. ,2007), sensory gating (Orekhova et al., 2008), sound encoding (Jansson-Verkasalo et al., 2005), MMN (Dunn, Gomes and Gravel, 2008; Gandal et al., 2010), noise modulation efficiency (Alcántra et al., 2012; Teder-Sälejärvi et al., 2005) and sound discrimination (Bonnel et al., 2010; Jones et al., 2009).
Olfactory investigation revealed decreased OI (May et al., 2011), as well as normal OI with an inverse correlation with age (Brewer et al., 2008; Dudova et al., 2011). Further patients with ASD enjoyed certain smells differently (Hrdllicka et al., 2011), while olfactory detection and adaptation were equal (Tavassoli & Baron-Cohen, 2012). Moreover, they expressed decreased taste perception