Chapter 1



NEURAL CORRELATES OF REPETITIVE BEHAVIOUR IN AUTISM SPECTRUM DISORDERAN INVESTIGATION OF THE BEHAVIOURAL AND NEURAL CORRELATES OF REPETITIVE BEHAVIOUR IN AUTISM SPECTRUM DISORDERByJenna Traynor, BAA Thesis Submitted to the School of Graduate Studies in Partial Fulfillment of the Requirements for a Degree of Doctor of PhilosophyMcMaster University ? Copyright by Jenna Traynor, Sept 2018DOCTOR OF PHILOSOPHY (2018)PSYCHOLOGY, NEUROSCIENCE, & BEHAVIOURMcMaster UniversityHamilton, Ontario, CanadaTITLE: An investigation of the behavioural and neural correlates of repetitive behaviour in autism spectrum disorderAUTHOR: Jenna Traynor, BA (York University)SUPERVISOR: Dr. Geoffrey B.C. Hall, Ph.D.NUMBER OF PAGES: xi, 229AbstractThe existing research on repetitive behaviour in Autism Spectrum Disorder (ASD) is sparse. The identification of biomarkers through the use of brain imaging is an important area of research that can provide information about the etiology and function of repetitive behaviour, which is poorly understood. To date, results from brain imaging studies have implied that there may be a common neurobiological substrate involved in social deficits and repetitive behaviour in ASD. At the same time, it is becoming increasingly apparent that ASD may be described as a spectrum of multiple ‘autisms,’ that vary in their behavioural presentation and genetic loci. As such, the identification of biomarkers of distinct repetitive behaviour subtypes across the autism spectrum is a priority. To date, several gaps in the literature remain, and the current body of research was completed to address these gaps. In this body of work, Chapter 1 offers an introduction to the subject matter. Chapter 2 presents a review of neuroimaging studies on repetitive behaviour in ASD and creates an organized framework for future research. In Chapter 3, the results of an eye-tracking study, which investigated the common neurobiological substrate in social deficits and repetitive behaviour, are presented. Chapter 4 outlines the results of a brain imaging study that investigated the association between repetitive behaviour and patterns of resting-state functional connectivity in ASD. In Chapter 5, the results of a pilot study that examined the association between inhibitory control and cortical gamma-amino-butyric acid (GABA) in ASD are presented. Last, chapter 6 offers a summary of these original research studies, and a discussion of implications and future directions. PrefaceTo my committee members, Louis Schmidt and Mel Rutherford, thank you for your continued guidance and support throughout my doctoral training. You are both wonderful mentors.To my supervisor, Geoffrey Hall, I am so grateful for all of the professional and personal guidance you have offered to me throughout my training. You are such a genuine and creative supervisor, and I have learned so much from you that I will take with me as I begin my exciting research career.To my parents, Brian and Yvonne, I would not be where I am today without your unwavering love and support. You have provided me with the foundation necessary to reach my dreams. I am so proud to have you both as parents, and I cannot express in words how much I love you both. To my grandparents, to have you so involved and supportive of my graduate work was a true blessing, and it made all the difference. You are my greatest role models. I love you.To my sister, Sarah, your formidable strength and courage consistently motivate me to work hard, acquire knowledge, and to explore this life in a way that is most authentic to who I am. I also appreciate all of the laughter we bring to our relationship, and it has helped to carry me through my graduate work. You inspire me and I love you.And finally, to Otis and Zumi, thank you for bringing me a type of indescribable love that can be felt between a human and her dogs. Thank you for keeping me company from apartment to apartment, for bringing me laughter when the sky was grey, and for walking with me, always. Table of Contents TOC \o "1-3" \h \z \u Chapter 1 PAGEREF _Toc399614763 \h 1Introduction PAGEREF _Toc399614764 \h 1References PAGEREF _Toc399614765 \h 18Chapter 2 PAGEREF _Toc399614766 \h 35Structural and Functional Neuroimaging of Restricted and Repetitive Behavior in Autism Spectrum Disorder PAGEREF _Toc399614767 \h 35Abstract and Key Words PAGEREF _Toc399614768 \h 36Introduction PAGEREF _Toc399614769 \h 37Methodology PAGEREF _Toc399614770 \h 41Structural MRI studies and RRB PAGEREF _Toc399614771 \h 41Functional MRI and RRB subtypes PAGEREF _Toc399614772 \h 47Resting-state functional connectivity and RRB PAGEREF _Toc399614773 \h 56Diffusion tensor imaging and RRB PAGEREF _Toc399614774 \h 57Future directions PAGEREF _Toc399614775 \h 59Tables and Figures PAGEREF _Toc399614776 \h 63Acknowledgements PAGEREF _Toc399614777 \h 69References PAGEREF _Toc399614778 \h 69Chapter 3 PAGEREF _Toc399614779 \h 77Eye Tracking Effort Expenditure and Autonomic Arousal to Social and Circumscribed Interest Stimuli in Autism Spectrum Disorder PAGEREF _Toc399614780 \h 77Abstract and Key Words PAGEREF _Toc399614781 \h 78Introduction PAGEREF _Toc399614782 \h 79Methods PAGEREF _Toc399614783 \h 84Results PAGEREF _Toc399614784 \h 92Discussion PAGEREF _Toc399614785 \h 96Tables and Figures PAGEREF _Toc399614786 \h 106References PAGEREF _Toc399614787 \h 111Supplementary Material PAGEREF _Toc399614788 \h 119Chapter 4 PAGEREF _Toc399614789 \h 124Indices of Repetitive Behaviour are Correlated with Patterns of Intrinsic Functional Connectivity in Youth with Autism Spectrum Disorder PAGEREF _Toc399614790 \h 124Abstract and Key Words PAGEREF _Toc399614791 \h 125Introduction PAGEREF _Toc399614792 \h 126Results PAGEREF _Toc399614793 \h 131Discussion PAGEREF _Toc399614794 \h 134Conclusions PAGEREF _Toc399614795 \h 144Experimental Procedure PAGEREF _Toc399614796 \h 147Tables and Figures PAGEREF _Toc399614797 \h 155Acknowledgments PAGEREF _Toc399614798 \h 161References PAGEREF _Toc399614799 \h 162Chapter 5 PAGEREF _Toc399614800 \h 175Frontostriatal functional connectivity during inhibitory control task performance is atypically correlated with cortical GABA concentration in autism spectrum disorder: A pilot study PAGEREF _Toc399614801 \h 175Abstract and Keywords PAGEREF _Toc399614802 \h 176Introduction PAGEREF _Toc399614803 \h 177Methods PAGEREF _Toc399614804 \h 182Tables and Figures PAGEREF _Toc399614805 \h 200References PAGEREF _Toc399614806 \h 204Chapter 6 PAGEREF _Toc399614807 \h 211General Discussion PAGEREF _Toc399614808 \h 211References PAGEREF _Toc399614809 \h 224List of Tables and Figures TOC \o "1-3" \h \z \u Chapter 2Structural and Functional Neuroimaging of Restricted and Repetitive Behavior in Autism Spectrum DisorderTable 1. Neuroimaging results presented in this review63Figure 1. a frontal and striatal developmental trajectory of rrb in asd67Figure 2. a neurocognitive model of rrb in asd68Chapter 3Eye Tracking Effort Expenditure and Autonomic Arousal to Social and Circumscribed Interest Stimuli in Autism Spectrum Disordertable 1. demographic informaton and descriptive statistics106table 2. descriptive statsitics for control participants' self-identified interests107table 3. list of participant interests107figure 1. example of stimuli in the three different conditions108figure 2. results of a split plot anova investigating effort expenditure to social, neutral, and interest stimuli in asd and control subjects108figure 3. results of within-subjects repeated measures anovas investigating percent change in pupil size to social, neutral, and interest stimuli in asd and control subjects109figure 4. results of within-subjects repeated measures anovas investigating change in blink rate to social, neutral, and interest stimuli in asd and control subjects110supplementary table 4. descriptive statistics for raw, observed pupil size data used in a supplementary repeated measures anova model122supplementary FIGURE 5. RESULTS OF WITHIN-SUBJECTS REPEATED MEASURES ANOVAS INVESTIGATING CHANGES IN RAW, OBSERVED PUPIL SIZE TO SOCIAL, NEUTRAL, AND INTEREST STIMULI IN ASD AND CONTROL SUBJECTS123 Chapter 4Indices of Repetitive Behaviour are Correlated with Patterns of Intrinsic Functional Connectivity in Youth with Autism Spectrum Disordertable 1. significant differences in seed-to-voxel resting-state functional connectivity (asd > td, 2-sided contrast)155table 2. significant differences in roi-to-roi resting-state functional connectivity (asd > td, 2-sided contrast) 156table 3. significant results from a bivariate correlation analysis demonstrating the relationship between two pca-derived factors from the rbs-r and connectivity within the asd group157table 4. demographic data157table 5. normalized vectors for each pca-derived component158figure 1. clusters from a seed-to-voxel analysis (asd > td, 2-sided contrast) showing significant connectivity with each roi159figure 2. significant patterns of roi-to-roi functional connectivity (asd > td, 2-sided contrast) 160supplementary table 6. within-subject differences in roi-to-roi resting-state connectivity (td male > td female, 2-sided contrast)173Chapter 5Frontostriatal functional connectivity during inhibitory control task performance is atypically correlated with cortical GABA concentration in autism spectrum disorder: A pilot studyTABLE 1. DEMOGRAPHIC INFORMATION AND DESCRIPTIVE STATISTICS FOR ASD AND CONTROL SUBJECTS 200TABLE 2. SIGNIFICANT BETWEEN-GROUP DIFFERENCES IN ROI-TO-ROI FUNCTIONAL CONNECTIVITY 200 TABLE 3. SIGNIFICANT BETWEEN-GROUP DIFFERENCES IN THE EFFECT OF GABA+ ON ROI-TO-ROI FUNCTIONAL CONNECTIVITY 201FIGURE 1. EXAMPLE OF VOXEL-OF-INTEREST PLACEMENT AND FIT OF GABA SPECTRA 201FIGURE 2. BEHAVIOURAL PERFORMANCE ON THE GO/NO-GO TASK 202FIGURE 3. SIGNIFICANT ROI-TO-ROI UNDERCONNECTIVITY BETWEEN THE RIGHT CAUDATE NUCLEUS AND THE RIGHT PUTAMEN DURING NO-GO BLOCKS IN ASD SUBJECTS, RELATIVE TO CONTROLS 202 FIGURE 4. SIGNIFICANT NEGATIVE CORRELATION BETWEEN CONNECTIVITY OF THE RIGHT INFERIOR FRONTAL GYRUS, PARS OPERCULARIS, AND THE LEFT PUTAMEN DURING NO-GO BLOCK AND GABA+, IN ASD SUBJECTS RELATIVE TO CONTROLS 203List of AbbreviationsACC……………………………………………………………………………anterior cingulate cortexADHD………………………………………………..Attention-Deficit-Hyperactivity DisorderADI-R……………………………………………………Autism Diagnostic Interview - RevisedADOS………………………………………………...Autism Diagnostic Observation ScheduleANOVA………………………………………………...…………………………… analysis of varianceAQ…………………………………………………………………………………………..Autism QuotientASD………………………………………………………………………… autism spectrum disorderBA……………………………………………………………………………………………Brodmann areaBOLD signal……………………………………………blood-oxygen-level dependent signalCIs………………………………………………………………………………circumscribed interestsCPRS-R………………………………………………...Connor’s Parent Rating Scale - RevisedCPT…………………………………………………………………..Continuous Performance TaskCSF…………………………………………………………………………………….cerebrospinal fluiddACC………………………………………………………………dorsal anterior cingulate cortexDLPFC……………………………………………………………… dorsolateral prefrontal cortexDMN………………………………………………………………………………default mode networkDTI…………………………………………………………………………….diffusion tensor imagingEF…………………………………………………………………………………...executive functioningFA……………………………………………………………………………………fractional anisotropyFDR………………………………………………………………………………….False Discovery Rate fMRI……………………………………………………functional magnetic resonance imagingFOV………………………………………………………………………………………………..field of viewFEW……………………………………………………………………………………..Family Wise ErrorGABA………………………………………………………………………gamma-aminobutyric acidGG……………………………………………………………………………………..Greenhouse-GeisserGlx…………………………………………………………………………….glutamate and glutamineGM………………………………………………………………………………………………….grey matterHAI………………………………………………………………………………….. high autism interest HFA………………………………………………………………...…………..high-functioning autismHSD………………………………………………………………….Honestly Significant DifferenceIFG…………………………………………………………………………………..inferior frontal gyrusIPL…………………………………………………………………………………...inferior parietal lobeIPS……………………………………………………………………………………..intraparietal sulcusIS…………………………………………………………………………………insistence on samenessMEG…………………………………………………………………………magnetoencephalographyMFG…………………………………………………………………………………..medial frontal gyrusMNI……………………………………………………………………Montreal Neurologic InstituteMPFC………………………………………………………………………….medial prefrontal cortexMRI…………………………………………………………………… magnetic resonance imagingMRS……………………………………………………………..magnetic resonance spectroscopyNAcc…………………………………………………………………………………...nucleus accumbensOCD…………………………………………………………………Obsessive Compulsive DisorderOFC……………………………………………………………………………………orbitofrontal cortexPC…………………………………………………………………………………….principal componentPCA……………………………………………………………………Principal Component AnalysisPCC…………………………………………………………………………..posterior cingulate cortexPFC………………………………………………………………………………………...prefrontal cortexPI……………………………………………………………………………………………Padua InventoryRMB………………………………………………………………….……repetitive motor behaviourROI………………………………………………………………………………………..region of interestRB………………………………………………………………….…………………repetitive behaviourRBS-R……………………………………………………...Repetitive Behaviour Scale - RevisedRRB……………………………………………………….……….. restricted repetitive behaviourSFG…………………………………………………………….…………………..superior frontal gyrusTD……………………………………………………………….…………………….typically developingTE……………………………………………………….……………………………………………..echo timeTR……………………………………………………….…………………………………….repetition timeV1………………………………………………………….………………………..primary visual cortexVMPFC…………………………………………….……………….ventromedial prefrontal cortexVOI………………………………………………………………………………………….voxel of interestWASI-II……………………Wechsler Abbreviated Scale of Intelligence – 2nd EditionWM………………………………………………………………………………………………white matterChapter 1IntroductionAutism Spectrum Disorder (ASD) is the fastest growing childhood neurobiological disorder worldwide (Centers for Disease Control and Prevention, 2017). Currently, 1 in 66 children in Canada are diagnosed with an ASD (National Autism Spectrum Disorder Surveillance System, 2018), which is a lifelong condition with no cure. Additionally, the ratio of males to females with a diagnosed ASD is approximately 3 to 1 (Loomes, Hull, & Mandy, 2017), which suggests that males are at an elevated risk for the disorder. ASD has a pervasive impact on individuals, families, and the healthcare system. Persons with ASD experience a significantly lower quality of life across the lifespan, with high levels of social isolation, low rates of employment and financial independence, fewer friends, less affective and sexual relationships, and poorer physical health (van Heijst & Geurts, 2015). Moreover, psychiatric comorbidities are extremely common; 72% of individuals with ASD have at least one comorbid psychiatric diagnosis (Leyfer et al., 2006), and rates of anxiety and depression far exceed the rate of these disorders in the general population (Strang et al., 2012). Additionally, the lifespan value of caregiver time needed to support an individual with ASD is about $5.5 million above the cost for a neurotypical (Dudley & Emery, 2014), and private behavioural interventions cost families upwards of $60,000 per year (Autism Program of Eastern Ontario, 2015). Societally, the cost of ASD in Ontario in 2013 was estimated at $186,134,100 (Autism Program of Eastern Ontario, 2015). ASD is characterized by two core diagnostic features; deficits in social communication and restricted, repetitive behaviour (American Psychiatric Association, 2013), and both symptoms present with vast heterogeneity across the autism spectrum. Briefly, social communication deficits include a range of prominent impairments that are often present by the first year of life and persist throughout the lifespan, including impaired social orientating, responding, joint attention, mentalizing, and social engagement (Zwaigenbaum et al., 2005). In childhood, these deficits can present as disinterest in social play or difficulty with imaginative play, followed by impaired social-emotional reciprocity across development (American Psychiatric Association, 2013). Although these impairments are often conspicuous in both males and females, there is recent evidence to suggest a sex-differentiated presentation of social deficits, with a less visible clinical presentation in women with high-functioning ASD (Kirkovsky, Enticott, & Fitzgerald, 2013).Repetitive behaviour, which is the second core symptom of ASD, is exhibited across the spectrum, often emerges by 12 months of age and persists throughout the lifespan, and is often severe (Dudley & Emery, 2014). Repetitive behaviour includes a wide range of behaviors including vocal stereotypies, stereotyped motor behaviour, insistence on sameness in daily routine or marked difficulty with change in routine, preoccupations with unusual objects or parts of objects (e.g., fan blades), unusual sensory interests, repetitive self-injurious behaviour, and intense, restricted interests (Bodfish, Symons, Parker, & Lewis, 2000). Factor and independent component analyses of gold standard assessment measures such as the Autism Diagnostic Interview- Revised (ADI-R; Lord, Rutter, & LeCouteur, 1994) and the Autism Diagnostic Observation Schedule (ADOS; Lord, Rutter, DeLavore, & Risi, 1999) have identified both a two-factor (Cuccaro et al., 2003; Szatmari et al., 2006) and a three-factor solution (Lam, Bodfish & Piven, 2008) for repetitive behaviour in ASD. The two-factor solution includes repetitive motor behaviour and insistence on sameness, and the three-factor solution additionally includes circumscribed interests. Alternatively, repetitive behaviour has been described as either lower- or higher-order (Turner, 1999). Lower-order behaviour possesses a distinct motor component (Lewis, Tanimura, Lee, & Bodfish, 2007), and is more rudimentary in nature (Turner, 1999). Higher-order behaviour has a distinct cognitive component (Turner, 1999) and includes behaviour such as insistence on sameness and circumscribed interests (Lewis et al., 2007). Brain imaging is a powerful tool that can be used to examine the neural correlates of ASD across development. On a macroscopic level, magnetic resonance imaging (MRI) and functional MRI (fMRI) have been used to identify structural and functional biomarkers of ASD, respectively. Structural approaches include MRI and diffusion tensor imaging (DTI), and are used to quantify the volume of brain structures and identify anatomically connected regions of the brain. Neuroanatomical volume and structural connectivity variances throughout the lifespan have been used as markers of structural growth rate and neuroplasticity in ASD, respectively (Mohammed-Rezazadeh, Frohlich, Loo, & Jeste, 2016). fMRI characterizes patterns of regional activation and indexes functional connectivity via the quantification of oscillations in the blood-oxygen-level dependent (BOLD) signal, which is derived from the ratio of oxygenated and deoxygenated hemoglobin in the brain (Di & Rao, 2007). BOLD signal can be quantified in the context of cognitive or social task performance (i.e., task-based fMRI) or in the absence of a task (i.e., resting-state fMRI). Task-based fMRI offers the advantage of describing neural activity in the context of specific behavioural performance, and fMRI paradigms have been developed to examine both social (e.g., Sabatino et al., 2013) and repetitive behaviour in ASD (review in Anagnostou & Taylor, 2011). On the other hand, resting-state fMRI identifies spontaneous, lower frequency oscillations in the BOLD signal that are correlated across functionally related regions (Biswal et al., 2010), and offers an advantage by characterizing more intrinsic brain processes. Further, magnetic resonance spectroscopy (MRS) has been used to quantify various brain metabolites and neurotransmitters in ASD, such as N-acetylaspartate, glutamate and glutamine (Glx), creatine and phosphocreatine (DeVito et al., 2007), and most recently, gamma-aminobutyric acid (GABA; review in Ford & Crewther, 2016). MRS capitalizes on the different frequency shifts of molecules, which result from the various chemical environments in which they are surrounded, and offers a detailed metabolite characterization of ASD (Bertholdo, Watcharakorn, & Castillo, 2013). Together, these imaging methods have been used to investigate potential biomarkers of ASD, and a number of structural and functional abnormalities, relative to neurotypicals, have been identified. Biomarker research can advance our understanding of the inherent complexity of ASD, and ultimately support earlier identification of symptoms, the prediction of risk and prognosis, treatment responsiveness, the identification of subgroups, and the development of targeted interventions (Anderson, 2015). Thus far, biomarker research has contributed to the development of key neurobiological theories of ASD. However, compared to social deficits, there has been markedly less biomarker research on repetitive behaviour in ASD (Boyd, McDonough, & Bodfish, 2012), despite the extremely impairing nature of repetitive behaviour in areas such as learning (Pierce & Courchesne, 2001), socialization (Nadig, Lee, Singh, Bosshart, & Ozonoff, 2010), and family functioning (Bishop, Richler, Cain, & Lord, 2007; Greenberg, Seltzer, Krauss, Chou, & Orsmond, 2006).For example, as a result of the extensive amount of imaging work conducted on social cognition and language in ASD (review in Barak & Feng, 2016; Uddin, 2011), several neurobiological theories have been proposed to better understand the social communication deficits in ASD. Examples of well-known neurobiological theories include the amygdala theory of autism (Baron-Cohen et al., 2000), the social motivation hypothesis (Chevallier, Kohls, Troiani, Brodkin, & Schultz, 2012), and hypotheses of mirror neuron dysfunction (Dapretto & Iacoboni, 2006; Gallese, Rochat, Cossu, & Sinigaglia, 2009). Although findings have been mixed (Hamilton, 2013), they have collectively advanced our understanding of the complex social presentation of ASD, and a number of behavioural interventions for social symptoms have been developed that capitalize on research findings (Boyd et al., 2012; Ospina et al., 2008).Briefly, the amygdala hypothesis posits that the etiology of social deficits in ASD is rooted in inherent amygdala damage (Baron-Cohen, 2000). However, due to a lack of evidence of social deficits in persons with known amygdala lesions (Birmingham, Cerf, & Adolphs, 2011), and contradictory findings pertaining to altered amygdala volume (Abbell et al., 1999; Aylward et al., 1999, Haznedar et al., 2000) and function in ASD (Baron-Cohen et al., 2000; Dalton et al., 2005; Kliemann, Dziobek, Hatri, Baudewig, & Heekeren, 2012), the amygdala theory has been revised to include a broader disruption of inter- and intra-limbic system circuitry in ASD (Birmingham et al., 2011), which is relevant not only to social cognition deficits, but also to reward contingency learning (Gaigg, 2012), and attentional allocation to environmental stimuli (Zalla & Sperduti, 2013). Relatedly, the social motivation hypothesis suggests that there is an early intrinsic motivational deficit in ASD, which originates in neural circuits that subserve salience and reward processing; namely, limbic and frontostriatal circuits (Chevallier et al., 2012). Early functional abnormalities in these areas are thought to result in reduced motivation to socially engage in ASD, and have downstream effects on behaviour throughout development (Chevallier et al., 2012). Finally, mirror neuron hypotheses postulate that ASD is characterized by an inherent dysfunction in parietal brain regions that contain mirror neurons (Dapretto & Iacoboni, 2006; Gallese et al., 2009). These regions activate both when performing an action and while watching another person perform the same action (Rizzolatti & Craighero, 2004). Such dysfunction is hypothesized to extend to difficulties simulating emotions and mental states, leading to the significant impairment in theory of mind and empathy seen in ASD (Dapretto & Iacoboni, 2006; Hamilton, 2013). In contrast, the existing literature on the neural substrata of repetitive behaviour is sparse. Despite growing interest in repetitive behaviour research in the past 15 years, findings have varied considerably and remain unintegrated (Boyd et al., 2012). This is likely due in part to the extensive heterogeneity in repetitive behaviour in ASD, whereby different subtypes are rooted in various neural circuits, but the relative paucity of imaging studies focused on repetitive behaviour has significantly contributed to the absence of pointed neurobiological theories. Consequently, umbrella theories pertaining more to the collective ASD symptom presentation, such as the theory of executive dysfunction (Ozonoff, 1995; Russel, 1997) and disrupted connectivity theory of ASD (e.g., Just, Cherkassky, Keller, & Minshew, 2004) have been used as a basis with which to more specifically investigate repetitive behaviour. Briefly, the theory of executive dysfunction (Ozonoff, 1995; Russel, 1997) states that a disturbance at the level of executive function plays a significant role in the manifestation of ASD symptoms. Neurobiologically, prefrontal cortical function supports the planning, organization, flexibility, and goal-directed behaviour required to carry out everyday tasks and socially engage (Miller & Cohen, 2001). Deficits in several areas of executive function have been consistently identified in ASD, including deficits in inhibitory control (Kleinhans, Akshoomoff, & Delis, 2005; Luna, Doll, Hegedus, Minshew, & Sweeney, 2006; Ozonoff, Strayer, McMahon, & Filloux, 1994), planning and cognitive flexibility (Hughes, Russell, & Robbins, 1994), set-shifting (Courchesne et al., 1994; Pascualvaca, Fantie, Papageorgiou, & Mirsky, 1998), working memory (Wang et al., 2017) and selective and sustained attention (Noterdaeme, Mildenberger, Minow, & Amorosa, 2002). Additionally, associations between repetitive behaviour and cognitive flexibility (Lopez, Lincoln, Ozonoff, & Lai, 2005; South et al., 2007), working memory (Lopez et al., 2005; Wolf, Chmielewski, Beste, & Roessner, 2017), and inhibitory control (Lopez et al., 2005; Mosconi et al., 2009; South, Ozonoff, & McMahon, 2007) have been found in ASD, which suggests that executive dysfunction plays a role in repetitive behaviour. As such, task-based fMRI has been used to examine the neural substrata of repetitive behaviour in the context of performance during tasks of executive function, which are commonly referred to as “repetitive behaviour proxies” (Anagnastou & Taylor, 2011).Relatedly, patterns of neural connectivity observed in ASD have led to the conceptualization of ASD as a disorder of disrupted connectivity, characterized by abnormal functional synchronization among various brain regions (Vasa, Mostofsky, & Ewen, 2016). For example, the under-connectivity theory of ASD (Just et al., 2004) posits that ASD is characterized by less synchronization between frontal and posterior regions during task performance, which is hypothesized to contribute to information processing deficits (Just, Keller, Malave, Kana, & Varma, 2012). Relatedly, findings of increased volume in radiate white matter tracts that connect neighbouring regions of the brain (Herbert et al., 2003) have led to hypotheses of local, short-range over-connectivity in ASD. As such, it has been suggested that ASD may be characterized by both global under-connectivity and local over-connectivity (Belmonte, Cook, Andersen, Greenough, & Beckel-Mitchener, 2004). However, connectivity findings have not consistently supported these hypotheses; some studies have found reduced long-range connectivity in ASD (Assaf et al., 2010; Ebisch et al., 2011; Just, Cherkassky, Keller, Kanan, & Minshew, 2007; Kana, Keller, Minshew, & Just, 2007), while others have demonstrated a mixed pattern of global over- and under-connectivity (Di Martino et al., 2011; Mizuno, Villalobosa, Daviesa, Dahla, & Muller, 2006; Shih et al., 2010). Similarly, although short-range over-connectivity has been identified in ASD (Supekar et al., 2013), mixed local connectivity patterns have also been found (Dajani & Uddin, 2015; Paakki et al., 2010). Moreover, repetitive behaviour has been associated with both over- and under-connectivity patterns (Abbott et al., 2018; Delmonte, Gallagher, O’Hanlon, McGrath, & Balsters, 2013; Di Martino et al., 2011). As such, although it is hypothesized that repetitive behaviour is associated with an overall disrupted connectivity profile in ASD, research is now focused on whether more specific connectivity profiles can be identified.By using these theories as a foundation for further research, the relationship between repetitive behaviour in ASD and i) executive dysfunction and ii) functional connectivity are being examined. However, in comparison to social symptoms, this research has only modestly advanced our understanding of repetitive behaviour, and a number of gaps in our understanding of repetitive behaviour remain. These gaps are outlined in detail below.First, a number of MRI and DTI studies have associated cortical and subcortical structural abnormalities with several subtypes of repetitive behaviour, including atypical environmental exploration (Pierce & Courchesne, 2001), insistence on sameness (Langen et al., 2013), repetitive self-injurious behaviour (Duerden et al., 2013), perseverative and compulsive behaviour (Hollander et al., 2005; Sears et al., 1999), and circumscribed interests (Hardan et al., 2005). Additionally, fMRI studies have identified altered functional circuitry in cognitive and motor control, attention, salience, and reward processing networks (Agam, Joseph, Barton, & Manoach, 2010; Bolling et al., 2011; Casico et al., 2013; Dichter et al., 2012; Goldberg et al., 2011; Kana et al., 2007; Mizuno et al., 2006; Sabatino et al., 2013; Schmitz et al., 2006; Shafritz, Dichter, Baranek, & Belger, 2008; Solomon et al., 2009; Turner, Frost, Linsencardt, McIlroy, & Muller, 2006). These patterns of circuitry have been elicited by several cognitive paradigms that have been specially designed to examine repetitive behaviour substrata (i.e., “repetitive behaviour proxies;” Anagnostou & Taylor, 2011), although it remains unclear which paradigms most consistently evoke patterns of connectivity in specific networks. We suggest that an organized framework of these structural and functional markers of repetitive behaviour across development is needed. Additionally, a succinct outline of the executive function paradigms used to investigate repetitive behaviour would offer an initial starting point with which to proceed further.Moreover, despite being categorically separate features, repetitive behaviour and social deficits are not observed in isolation; individuals are diagnosed with ASD if in addition to social deficits, at least two repetitive behaviour symptoms are observed (American Psychiatric Association, 2013). Understanding potential links between these two core symptoms is an ongoing process (South et al., 2007). Recently, research has suggested that there may be a common neurobiological substrate underlying these core symptoms, which is rooted in frontostriatal, salience- and reward-attribution circuitry (Benning et al., 2016; Cascio et al., 2014; Dichter et al., 2012; Foss-Feig et al., 2016; Pierce et al., 2015; Sabatino et al., 2013; Sasson, Turner-Brown, Holtzclaw, Lam, & Bodfish, 2008; Sasson, Elison, Turner-Brown, Dichter, & Bodfish, 2011; Watson et al., 2015). Indeed, the social motivation hypothesis posits that early abnormalities in these functional circuits may have detrimental and cascading effects on social behaviour in ASD (Chevallier et al., 2012). Further, animal (review in Lewis et al., 2007) and human studies (Hardan et al., 2005; Hollander et al., 2005; Langen, Durston, Kas, van Engeland, & Staal, 2011; Sears et al., 1999; Shafritz et al., 2008; Uddin et al., 2013) have presented strong evidence to suggest involvement of salience- and reward-related circuitry in repetitive behaviour. As salience and reward circuitry support the allocation of attention to environmental stimuli, (Corbetta & Shulman, 2002; Gaigg, 2012; Zalla & Sperduti, 2013), reward-related decision making and effort expenditure (Delgado, 2007; Balleine, Delgado, & Hikosaka, 2007) and autonomic arousal (Costafreda, Brammer, David, & Fu, 2008; Gaigg, 2012), the quantification of these processes has been initiated to further investigate the neurobiological underpinnings of the ASD symptom dyad.Thus far, eye-tracking paradigms have been used to record responses to social and repetitive behaviour stimuli in ASD. To date, differences in attentional gaze to social and repetitive behaviour stimuli in ASD have been identified (Pierce et al., 2015; Sasson et al., 2008; 2011), with some inconsistency in the direction of findings (Falck-Ytter, 2008; Sabatino DeCriscio et al., 2016). Additionally, the presence of repetitive behaviour has been found to significantly predict reward-related effort expenditure in ASD (Damiano, Aloi, Treadway, Bodfish, & Dichter, 2012). Further, differences in arousal-related pupillary response (Bradley, Miccoli, Escrig, Lang, 2008; Hess & Polt, 1960) to social images has been found in ASD, relative to controls (Anderson, Colombo, & Shaddy, 2006; Sepeta et al., 2014), although pupillary response to repetitive behaviour stimuli has not been examined. In fact, no study has objectively and simultaneously investigated attention, effort expenditure, and pupillary response to social and repetitive behaviour stimuli, which precludes an investigation of the collective role of these processes in the ASD symptom dyad. Such an investigation would also provide evidence to support or challenge the social motivation hypothesis (Chevallier et al., 2012) and the proposition that there may be a common neurobiological substrate across symptom categories. Further, in addition to MRI and behavioural paradigms, resting-state fMRI has been employed to examine intrinsic functional connectivity processes in ASD. Due to the absence of a task-performance requirement, resting-state fMRI shows promise for biomarker identification in ASD participants with a broad range of functioning (Allison, 2018). Recently, a handful of resting-state fMRI studies have been carried out to examine repetitive behaviour in ASD (Emerson et al., 2017; Glerean et al., 2016; Monk et al., 2009; Uddin et al., 2013; Weng et al., 2010). However, most have focused on a small number of ROIs in distinct neural circuits and unsurprisingly, results remain mixed (Muller et al., 2011). For example, resting-state connectivity patterns in the salience network (Uddin et al., 2013), default mode network (Monk et al., 2009; Weng et al., 2010) and ventro-temporal-limbic network (Glerean et al., 2016) have been correlated with total, overall repetitive behaviour scores on the ADI-R (Lord et al., 1994). However, given the variable presentation of repetitive behaviour across the spectrum (Szatmari et al., 2006; Turner, 1999), it is likely that more specific behavioural subtypes may be correlated with widespread patterns of intrinsic functional connectivity that are not easily identified through the examination of small number of network-specific ROIs. Further, use of the ADI-R, although sufficient for the quantification of overall, total repetitive behaviour, prevents a sound analysis of the potential associations between rest-connectivity and behavioural subtypes; the ADI-R was designed to aid clinicians in making a binary (yes/no) decisions pertaining to diagnosis (Lord et al., 1994), but it does not provide a continuous measure of the existing span of repetitive behaviour. Research that is focused on larger-scale, whole brain intrinsic neural circuitry, and which uses a more continuous measure of repetitive behaviour subtypes, would offer a comprehensive picture of the intrinsic underpinnings of repetitive behaviour. Additionally, task-based fMRI has been used to examine the neural substrata of repetitive behaviour in the context of performance during tasks of executive function. There is now a general consensus that motor and cognitive control-related, frontal-basal ganglia circuitry (Langen et al., 2011; Mizuno et al., 2006; Shafritz et al., 2008), as well as attention-related, frontoparietal circuitry (Kana et al., 2007; Shafritz et al., 2008) are disrupted during performance across tasks of executive function in ASD, and connectivity of these areas has been correlated with measures of repetitive behaviour collected outside of the scanner (Shafritz et al., 2008).Given that frontal and basal ganglia structures (i.e., striatum) contain high concentrations of the inhibitory neurotransmitter GABA (Akil et al., 2003; Koos & Tepper, 1999; Perlman, Weickert, Akil, Kleinman, 2004; Tremblay, Lee, & Rudy, 2016) and that associations between inhibitory control and repetitive behaviour have been identified (Lopez et al., 2005; Mosconi et al., 2009; South et al., 2007), it has recently been hypothesized that disruptions in the concentration of GABA may contribute to the pathophysiology of ASD (Kim, Lim, & Kaang, 2016; Presti, Watson, Kennedy, Yang, & Lewis, 2004). Further, animal models of ASD have provided strong evidence to suggest GABAergic involvement in repetitive behaviour, specifically (Han, Tai, Jones, Scheuer, & Catterall, 2014; Silverman et al., 2015). As such, MRS-quantified cortical GABA in ASD has recently emerged as a novel research method.Although in its infancy, this body of research has found either unchanged or decreased cortical GABA in childhood ASD, relative to controls (Brix et al., 2015; Carvalho Pereira, Violante, Mouga, Oliveria, & Castelo-Branco, 2018; Drenthen et al., 2016; Gaetz et al., 2014; Goji et al., 2107; Harada et al., 2011; Ito et al., 2017; Port et al., 2017; Puts et al., 2016; Rojas, Singel, Steinmetz, Hepburn, & Brown, 2014), but no differences in cortical GABA between adult ASD and control samples (Ajram et al., 2017; Port et al., 2017; Robertson, Ratai, & Kanwisher, 2015). These findings indicate that GABA may follow a developmental trajectory in ASD, which may have functional consequences. However, the quantification of cortical GABA alone does not provide information about GABA function. In fact, only one study has examined the association between cortical GABA and brain function, as measured by magnetoencephalography (MEG), and found an abnormal developmental trajectory of gamma-band coherence and cortical GABA coupling in ASD (Port et al., 2017). Accordingly, the field would benefit from additional work focused on the relationship between GABA and brain function in ASD, and more specifically, repetitive behaviour.Overall, as there is a lack of reliable diagnostic biomarkers of repetitive behaviour in ASD, there are no efficacious, evidence-based behavioural or pharmacological treatments for this core symptom (Brondino et al., 2015). Therefore, the primary focus of this research program was to investigate biomarkers of repetitive behaviour, with a focus on addressing the aforementioned gaps in the research literature. As there is an urgent need for intervention development, biomarker research can help to identify individuals who are likely to benefit from tailored interventions for repetitive behaviour, ultimately helping to increase quality of life in individuals with ASD, and reduce the large familial and societal impact of this disorder. To provide a comprehensive understanding of the behavioural and neural correlates of repetitive behaviour in ASD, the current research program employed multiple methods of investigation, which are briefly outlined below.Chapter 2 presents a literature review of structural and functional neuroimaging studies on repetitive behaviour in ASD and creates an organized framework of i) the neural correlates of repetitive behaviour and ii) the experimental paradigms used to image these correlates. Chapter 3 presents the results of an eye-tracking study, which investigated the hypothesized, common neurobiological substrate involved in social deficits and repetitive behaviour. In Chapter 4, the results of a resting-state fMRI study that investigated the association between indices of repetitive behaviour and widespread patterns of intrinsic functional connectivity in ASD are presented. In Chapter 5, the results of a pilot study that combined task-based fMRI with MRS to examine the neural correlates of inhibitory control in ASD, and the association between these correlates and cortical GABA, are presented. Last, chapter 6 offers a summary of these original research studies, and a discussion of implications and future directions. ReferencesAbell, F., Krams, M., Ashburner, J., Passingham, R., Friston, K., Frackowiak, R., . . . Frith, U. (1999). The neuroanatomy of autism: a voxel-based whole brain analysis of structural scans. Neuroreport, 10, 1647 – 1651.Abbott, A.E., Linke, A.C., Nair, A., Jahedi, A., Alba, L.A., Keown, C.L., . . . Muller, R. (2018). Repetitive behaviors in autism are linked to imbalance of corticostriatal connectivity: a functional connectivity MRI study. Social Cognitive and Affective Neuroscience, 32 – 42.Agam, Y., Joseph, R.M., Barton, J.J.S., & Manoach, D.S. (2010). Reduced cognitive control of response inhibition by the anterior cingulate cortex in autism spectrum disorders. NeuroImage, 52, 336-347. Ajram, L.A., Horder, J., Mendez, M.A., Galanopolous, A., Brennan, L.P., Wichers, R.H., . . . McAlonan, G.M. (2017). Shifting brain inhibitory balance and connectivity of the prefrontal cortex of adults with autism spectrum disorder. Translational Psychiatry, 7, e1137.Akil, M., Kolachana, B.S., Rothmond, D.A., Hyde, T.M., Weinberger, D.R., & Kleinman, J.E. (2003). Catechol-O-methyltransferase genotype and dopamine regulation in the human brain. Journal of Neuroscience, 2008-2013.Allison, J. (2018). Neuroimaging in neurodevelopmental disorders: focus on resting-state fMRI analysis of intrinsic functional brain connectivity. Current Opinion in Neurology, 31(2), 140 – 148. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.Anagnostou, E., & Taylor, M. (2011). Review of neuroimaging in autism spectrum disorders: what have we learned and where do we go from here. Mol Autism, 2(1), 4. Anderson, C.J., Colombo, J., & Shaddy, J.D. (2006). Visual scanning and pupillary responses in young children with Autism Spectrum Disorder. J Clin Exp Neuropsychol, 28, 1238–1256. Anderson, G.M. (2015). Autism biomarkers: challenges, pitfalls, and possibilities. J Autism Dev Disord, 45, 1103 – 1113.Assaf, M., Jagannathan, K., Calhoun, V.D., Miller, L., Stevens, M.C., Sahl, R., . . . Pearlson, G.D. (2010). Abnormal functional connectivity of default mode sub-networks in autism spectrum disorder patients.?Neuroimage, 53(1), 247–256.Autism Program of Eastern Ontario (2015). Retrieved from The Children’s Hospital of Eatern Ontario, url: , E.H., Minshew, N.J., Goldstein, G., Honeycutt, N.A., Augustine, A.M., Yates, K.O., . . . Pearlson, G.D. (1999). MRI volumes of amygdala and hippocampus in non-mentally retarded autistic adolescents and adults. Neurology, 53, 2145 – 2150.Balleine, B.W., Delgado, M.R., & Hikosaka, O. (2007). The role of the dorsal striatum in reward and decision-making.?J Neurosci.?27, 8161–8165.Barak, B., & Feng, G. (2016). Neurobiology of social behavior abnormalities in autism and williams syndrome. Nat Neurosci, 19(6), 647 – 655.Baron-Cohen, S., Ring, H.A., Bullmore, E.T., Wheelwright, S., Ashwin, S., & Williams, S.C. (2000). The amygdala theory of autism. Neurosci Biobehav Rev, 24, 355 – 364.Belmonte, M.K., Cook, E.H., Andersen, G.M., Greenough, W.T., & Beckel-Mitchener, A. (2004). Autism as a disorder of neural information processing: directions for research and targets for therapy. Mol Psychiatry, 9(7), 646 – 663.Benning, S.D., Kovac, M., Campbell, A., Miller, S., Hanna, E.K., Damiano, C.R., . . . Dichter, G.S. (2016). Late positive potential ERP responses to social and nonsocial stimuli in youth with autism spectrum disorder. J Autism Dev Disord, 46(9), 3068 – 3077.Bertholdo, D., Watcharakorn, A., & Castillo, M. (2013). Brain proton magnetic resonance spectroscopy: introduction and overview. Neuroimaging Clin N Am, 23, 359–380. Birmingham, E., Cerf, M., & Adolphs, R. (2011). Comparing social attention in autism and amygdala lesions: effects of stimulus and task condition. Soc. Neurosci, 6, 420 – 435.Bishop, S.L., Richler, J., Cain, A.C., & Lord, C. (2007). Predictors of perceived negative impact in mothers of children with autism spectrum disorder. Am J of Mental Retardation, 112(6), 450–461. Biswal, B.B., Mennes, M., Zuo, X., Gohel, S., Kelly, C., Smith, S.M., . . . Milham, M.P. (2010). Toward discovery science of human brain function. PNAS, 107(10), 4734-4739. Bodfish, J.W., Symons, F.J., Parker, D.E., & Lewis, M.H. (2000). Varieties of repetitive behavior in autism: Comparisons to mental retardation. J Autism Dev Disord. 30, 237–243. Bolling, D.Z., Pitskel, N.B., Deen, B., Crowley, M.J., McPartland, J.C., Kaiser M.D., . . . Pelphrey, K.A. (2011). Enhanced neural responses to rule violation in children with autism: A comparison to social exclusion. Dev Cogn Neurosci, 1(3), 280-294. ?Boyd, B.A., McDonough, S.G., & Bodfish, J.W. (2012). Evidence-based behavioral interventions for repetitive behaviors in autism. J Autism Dev Disord, 42(6), 1236 – 1248. Bradley, M.M., Miccoli, L., Escrig, M.A., & Lang, P.J. (2008). The pupil as a measure of emotional arousal and autonomic activation. Psychophysiology, 45, 602–607. ?Brix, M.K., Ersland, L., Hugdahl, K., Grüner, R., Posserud, M.B., Hammar, ?, . . . Beyer, M.K. (2015). Brain MR spectroscopy in autism spectrum disorder—the GABA excitatory/inhibitory imbalance theory revisited. Front Hum Neurosci, 9, 1–12. Brondino, N., Fusar-Poli, L., Panisi, C., Damiani, S., Barale, F., Politi, P. (2015). Pharmacological modulation of GABA function in autism spectrum disorders: a systematic review of human studies. J Autism Dev Disord, 46(3), 825 – 839.Carvalho Pereira, A., Violante, I.R., Mouga, S., Oliveria, G., & Castelo-Branco, M. (2018). Medial frontal love neurochemistry in autism spectrum disorder is marked by reduced n-acetylaspartate and unchanged gamma-aminobutyric acid and glutamate + glutamine levels. J Autism Dev Disord, 48(5), 1467 – 1482.Cascio, J.C., Foss-Feig, J.H., Heacock, J., Schauder, K.B., Loring, W.A., Rogers, B.P., & Bolton, S. (2014). Affective neural response to restricted interests in autism spectrum disorders. J Child Psychol Psychiatry, 55(2), 162-171. ?Centres for Disease Control and Prevention. (2017). Autism spectrum disorder. Retrieved from: , C., Kohls, G., Troiani, V., Brodkin, E.S., & Schultz, R.T. (2012). The social motivation theory of autism. Trends Cogn Sci, 16(4), 231–239. Corbetta, M., & Shulman, G.L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neurosci, 3(3), 201–215. Costafreda, S.G., Brammer, M.J., David, A.S., & Fu, C. H. (2008). Predictors of amygdala activation during the processing of emotional stimuli: a meta-analysis of 385 PET and fMRI studies. Brain Res Rev, 58, 57–70. Courchesne, E., Townsend, J., Akshoomoff, N.A., Saitoh, O., Yeung-Courchesne, R., Lincoln, A.J., . . . Lau, L. (1994). Impairment in shifting attention in autistic and cerebellar patients. Behav Neurosci, 108(5), 848 – 865.Cuccaro, M.L., Shao, Y., Grubber, J., Slifer, M., Wolpert, C.M., Donnelly, S.L., . . . Vance, M.A. (2003). Factor analysis of restricted and repetitive behaviors in autism using the Autism Diagnostic Interview-R. Child Psychiatry Hum Dev, 34, 3-17. Dajani, D.R., & Uddin, L.Q. (2015). Local brain connectivity across development in autism spectrum disorder: A cross-sectional investigation.?Autism Res, 9(1), 43 – 54.Dalton, K.M., Nacewicz, B.M., Johnstone, T., Schaefer, H. S., Gernsbacher, M. A., Goldsmith, H. H., . . . Davidson, R.J. (2005). Gaze fixation and the neural circuitry of face processing in autism.?Nat. Neurosci,?8, 519–526.Damiano, C.R., Aloi, J., Treadway, M., Bodfish, J.W., Dichter, G.S. (2012). Adults with autism spectrum disorders exhibit decreased sensitivity to reward parameters when making effort-based decisions. J Neurodev Disord, 4, 13.Dapretto, M., Iacoboni, M. (2006). The mirror neuron system and the consequences of its dysfunction. Nature Reviews Neuroscience, 7(12), 942 – 951.Delgado, M.R. (2007). Reward-related responses in the human striatum.?Ann N Y Acad Sci,?1104, 70–88.Delmonte, S., Gallagher, L., O’Hanlon, E., McGrath, J., & Balsters, J.H. (2013). Functional and structural connectivity of frontostriatal circuitry in autism spectrum disorder. Front Hum Neurosci, 7(430), 1–14.?DeVito, T.J., Drost, D.J., Neufeld, R.W.J., Rajakumar, N., Pavlosky, W., Williamson, P., Nicolson, R. (2007). Evidence for cortical dysfunction in autism: a proton magnetic resonance spectroscopy imaging study. Biol Psychiatry, 61(4), 465 – 473.Di, X., & Rao, H. (2007). Progress in functional connectivity analysis. Progress in Biochemistry and Biophysics, 1, 34-35. Dichter, G.S, Felder, J.N., Green, S.R., Rittenberg, A.M., Sasson, N.J., Bodfish, J.W. (2012). Reward circuitry function in autism spectrum disorders. SCAN, 7, 160-172. ?Di Martino, A., Kelly, C., Grzadzinski, R., Zuo, X.N., Mennes, M., Mairena, M.A., . . . Milham, M.P. (2011). Aberrant striatal functional connectivity in children with autism.?Biol Psychiatry, 69(9), 847–856.?Drenthen, G.S., Barendse, E.M., Aldenkamp, A.P., van Veenendaal, T.M., Puts, N.A.J., Edden, R.A.E., . . . Jansen, J.F. (2016). Altered neurotransmitter metabolism in adolescents with high-functioning autism. Psychiatry Research Neuroimaging, 256, 44–49.Dudley, C., & Emery, H. (2014). Costs of support and care for individuals living with autism spectrum disorder. University of Calgary School of Public Policy Research Papers, 7(1), 1 – 49. Duerden, E.G., Card, D., Roberts, W., Mak-Fan, K.M., Chakravarty, M., Lerch, J.P., & Taylor, M.J. (2013). Self-injurious behaviors are associated with alterations in the somatosensory system in children with autism spectrum disorder. Brain Struct Funct, 219(4), 1251-1261. Ebisch, S.J.H., Gallese, V., Willems, R.M., Mantini, D., Groen, W.B., Romani, G.L., . . . Bekkering, H. (2011). Altered intrinsic functional connectivity of anterior and posterior insula regions in high-functioning participants with autism spectrum disorder.?Hum Brain Mapp, 32, 1013–1028.Emerson, R.W., Adams, C., Nishino, T., Hazlett, H.C., Wolff, J.J., Zwaigenbaum, L., . . . Piven, J. (2017). Functional neuroimaging of high-risk 6-month-old infants predicts a diagnosis of autism at 24 months of age. Sci Transl Med, 9, eaag2882. Falck-Ytter, T. (2008). Face inversion effects in autism: a combined looking time and pupillometric study. Autism Res, 1, 297–306. ?Ford, T.C., & Crewther, D.P. (2016). A comprehensive review of the 1H-MRS metabolite spectrum in autism spectrum disorder. Frontiers in Molecular Neuroscience, 9, 1 – 27.Foss-Feig, J.H., McGugin, R.W., Gauthier, I., Mash, L.E., Ventola, P., & Cascio, C.J. (2016). A functional neuroimaging study of fusiform response to restricted interests in children and adolescents with autism spectrum disorder. J Neurodev Disord, 8, 15.Gaetz, W., Bloy, L., Wang, D.J., Port, R.G., Blaskey, L., Levy, S.E., & Roberts, T.P.L. (2014). GABA estimation in the brains of children on the autism spectrum: Measurement precision and regional cortical variation. NeuroImage, 86, 1–9. Gaigg, S.B. (2012). The interplay between emotion and cognition in autism spectrum disorder: implications for developmental theory. Front Integr Neurosci, 6(113), 1 – 35.Gallese, V., Rochat, M., Cossu, G., & Sinigaglia, C. (2009). Motor cognition and its role in the phylogeny and ontogeny of action understanding. Developmental Psychology 45(1), 103–113.Glerean, E., Pan, R.K., Salmi, J., Kujala, R., Lahnakoski, J.M., Roine, U., . . . Jaaskelainen, I.P. (2016). Reorganization of functionally connected brain subnetworks in high-functioning autism. Hum Brain Mapp, 37(3), 1066–1079.Goji, A., Ito, H., Mori, K., Harada, M., Hisaoka, S., Toda, Y., . . . Kagami, S. (2017). Assessment of anterior cingulate cortex (ACC) and left cerebellar metabolism in asperger’s syndrome with proton magnetic resonance spectroscopy (MRS). Plos ONE, 12(1), e0169288 Goldberg, M.C., Spinellia, S., Joela, S., Pekara, J.J., Dencklaa, M.B., & Mostofskya, S.H. (2011). Children with high functioning autism show increased prefrontal and temporal cortex activity during error monitoring. Dev Cogn Neuroscience, 1(1), 47-56.Greenberg, J.S., Seltzer, M.M., Krauss, M.W., Chou, R.J., Orsmond, G. (2006). Bidirectional effects of expressed emotion and behavior problems and symptoms in adolescents and adults with autism. Am J Mental Retardation, 111, 229–249. Hamilton, A.D.C. (2013). Reflecting on the mirror neuron system in autism: A systematic review of current theories. Dev Cog Neurosci, 3, 91 – 105.Han, S., Tai, C., Jones, C.J., Scheuer, T., & Catterall, W.A. (2014). Enhancement of inhibitory neurotransmission by GABAA receptors having A2,3- subunits ameliorates behavioural deficits in a mouse mode of autism. Neuron, 81(6), 1282 – 1289.Harada, M., Taki, M. M., Nose, A., Kubo, H., Mori, K., Nishitani, H., & Matsuda, T. (2011). Non-invasive evaluation of the GABAergic/glutamatergic system in autistic patients observed by MEGA-editing proton MR spectroscopy using a clinical 3 T instrument. J Autism Dev Disord, 41(4), 447–454. Hardan, A.Y., Girgis, R.R., Lacerda, A.L.T., Yorbik, O., Kilpatrick, M., Keshavan, M., & Minshew, N.J. (2005). Magnetic resonance imaging study of the orbitofrontal cortex in autism. J Child Neurol, 21, 866- 871. ?Haznedar, M.M., Buchsbaum, M.S., Wei, T.C., Hof, P.R., Cartwright, C., Bienstock, C.A., . . . & Hollander, E. (2000). Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging.?Am. J. Psychiatry?157, 1994–2001.Herbert, M.R., Ziegler, D.A., Deutsch, C.K., O’Brien, L.M., Lange, N., & Bakardjiev, A. (2003). Dissociations of cerebral cortex, subcortical and cerebral white matter volumes in autistic boys. Brain, 126, 1182 – 1192.Hess, E.H., & Polt, J.M. (1960). Pupil size as related to interest value of visual stimuli. Science, 132, 349–350. Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M., Licalzi, E., . . . Buchsbaum, M. (2005). Striatal volumes on magnetic resonance imaging and repetitive behaviors in autism. Biol Psychiatry, 58(3), 226-232. Hughes, C., Russell, J., & Robbins, T.W. (1994). Evidence for executive dysfunction in autism.?Neuropsychologia, 32(4), 477–492.Ito, H., Mori, K., Harada, M., Hisaoka, S., Toda, Y., Mori, T., . . . Kagami, S. (2017). A proton magnetic resonance spectroscopic study in autism spectrum disorder using a 3-tesla clinical magnetic resonance imaging (MRI) System: The anterior cingulate cortex and the left cerebellum. Journal of Child Neurology, 32(8), 731 – 739.Just, M.A., Cherkassky, V.L., Keller, T.A., Minshew, N.J. (2004). Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity. Brain, 127, 1811 – 1821.Just, M.A., Cherkassky, V.L., Keller, T.A., Kana, R.K., Minshew, N.J. (2007). Functional and anatomical cortical underconnectivity in autism: Evidence from an fMRI study of an executive function task and corpus callosum morphometry.?Cereb Cortex, 17, 951–961.Just, M.A., Keller, T.A., Malave, V.L., Kana, R.K., & Varma, S. (2012). Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity. Neurosci Biobehav Rev, 36(4), 1292 – 1313.Kana, R.K., Keller, T.A., Minshew, N.J., & Just, M.A. (2007). Inhibitory control in high-functioning autism: Decreased activation and underconnectivity in inhibition networks. Biol Psychiatry, 62(3), 198-206. Kim, H., Lim, C., & Kaang, B. (2016). Neuronal mechanisms and circuits underlying repetitive behaviors in mouse models of autism spectrum disorder. Behavioral and Brain Functions, 12, 3. Kirkovsky, M., Enticott, P.G., & Fitzgerald, P.B. (2013). A review of the role of female gender in autism spectrum disorders. J Autism Dev Disord, 43, 2584 – 2603.Kliemann, D., Dziobek, I., Hatri, A., Baudewig, J., & Heekeren, H. R. (2012). The role of the amygdala in atypical gaze on emotional faces in autism spectrum disorders.?J Neurosci,?32, 9469–9476.Kleinhans, N., Akshoomoff , N., & Delis, D. C. (2005). Executive functions in autism and asperger's disorder: flexibility, fluency, and inhibition. Developmental Neuropsychology, 27(3), 379-401.Koos, T., & Tepper, J.M. (1999). Inhibitory control of neostriatal projection neurons by GABAergic interneurons. Nature Neuroscience, 2, 467 – 472.Lam, K.S.L., Bodfish, J.W., & Piven, J. (2008). Evidence for three subtypes of repetitive behavior in autism that differ in familiality and association with other symptoms. J Child Psychol Psychiatry, 49(11), 1193-1200. Langen, M., Bos, D., Noordermeer, S.D., Nederveen, H., van Engeland, H., & Durston, S. (2013). Changes in the development of the striatum are involved in repetitive behaviors in autism. Biol Psychiatry, 75(5), 405-411. Langen, M., Durston, S., Kas, M.J., van Engeland, H., & Staal, W.G. (2011). The neurobiology of repetitive behavior:…and men. Neurosci Biobehav Rev. 35(3), 356–265. Lewis, M.H., Tanimura, Y., Lee, L.W., & Bodfish, J. (2007). Animal models of restricted repetitive behavior in autism. Behav Brain Res, 176, 66-74.?Leyfer, O.T., Folstein, S.E., Bacalman, S., Davis, N.O., Dinh, E., Morgan, J., . . . Lainhart, J.E. (2006). Comorbid psychiatric disorders in children with autism: interview development and rates of disorders.?J Autism Dev Disord, 36, 849–861.?Loomes, R., Hull, L., & Mandy, W.P.L. (2017). What is the male-to-female ratio in autism spectrum disorder? A systematic review and meta-analysis. J Am Acad Child Adolesc Psychiatry, 56(6), 466 – 474.Lopez, B.R., Lincoln, A.J., Ozonoff, S., & Lai, Z. (2005). Examining the relationship between executive functions and restricted, repetitive symptoms of autistic disorder. J Autism Dev Disord, 35, 445–460.Lord, C., Rutter, M., DeLavore, P.C., & Risi, S .(1999).?Autism Diagnostic Observation Schedule.?Los Angeles: Western Psychological Services.Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord, 24, 659–685. Luna, B., Doll, B.S., Hegedus, S.J., Minshew, N.J., & Sweeney, J.A. (2006). Maturation of executive function in autism. Biological Psychiatry, 61(4), 474-481.Miller, E.K., & Cohen, J.D. (2001). An integrative theory of prefrontal cortex function. Annu Rev Neurosci, 24, 167 – 202.Mizuno, A., Villalobosa, M.E., Daviesa, M.M., Dahla, B.C., & Muller, R. (2006). Partially enhanced thalamocortical functional connectivity in autism. Brain Res, 1104, 160-174. Mohammad-Rezazadeh, I., Frohlich, J., Loo, S.K., & Jeste, S.S. (2016). Brain connectivity in autism spectrum disorder. Curr Opin Neurol, 29(2), 137 – 147. Monk, C.S., Peltier, S.J., Wiggins, J.L., Weng, S., Carrasco, M., Risi, S., & Lord, C. (2009). Abnormalities of intrinsic functional connectivity in autism spectrum disorders. Neuroimage, 47, 764–772. Mosconi, M.W., Kay, M., D’Cruz, A.M., Seidenfeld, A., Guter, S., Standford, L.D., & Sweeney, J.A. (2009) Impaired inhibitory control is associated with higher-order repetitive behaviors in autism spectrum disorders. Psychological Medicine, 39(9), 1559 – 1566.Muller, R.A., Shih, P., Keehn, B., Deyoe, J.R., Leyden, K.M., & Shukla, D.K. (2011). Underconnected, but how? A survey of functional connectivity MRI studies in autism spectrum disorders. Cereb Cortex, 21, 2233–2243.Nadig, A., Lee, I., Singh, L., Bosshart, K., & Ozonoff, S. (2010). How does the topic of conversation affect verbal exchange and eye gaze? A comparison between typical development and high-functioning autism. Neuropsychologia, 48, 2730–2739.National Autism Spectrum Disorder Surveillance System (2018). Autism spectrum disorder among children and youth in canada 2018. Retrieved from: , M., Mildenberger, K., Minow, F., Amorosa, H. (2002). Evaluation of neuromotor deficits in children with autism and children with a specific speech and language disorder.?Eur Child Adolesc Psychiatry, 11(5), 219–225Ospina, M.B., Krebs Seida, J., Clark, B., Karkhaneh, M., Hartling, L., Tjosvold, L., . . . Smith, V. (2008). Behavioral and developmental interventions for autism spectrum disorder: a clinical systematic review. PLoS One, 3(11), e3755.Ozonoff, S. (1995). Executive functions in autism. In: Schopler, E., & Mesibov, G. (Eds.)?Learning and Cognition in Autism.?New York: Plenum Press, 199 – 219.Ozonoff, S., Strayer,D.L., McMahon, W.M., & Filloux, F. (1994). Executive function abilities in autism and Tourette syndrome: an information-processing approach. Journal of Child Psychology and Psychiatry, 35, 1015–1032.Paakki, J.J., Rahko, J., Long, X., Moilanen, I., Tervonen, O., Nikkinen, J., . . . Kiviniemi, V. (2010). Alterations in regional homogeneity of resting-state brain activity in autism spectrum disorders.?Brain Res, 1321, 169–179.Pascualvaca, D.M., Fantie, B.D., Papageorgiou, M., & Mirsky, A.F. (1998). Attentional capacities in children with autism: is there a general deficit in shifting focus??J Autism Dev Disord, 28(6), 467–478.Perlman, W.R., Weickert, C.S., Akil, M., & Kleinman, J.E. (2004). Postmortem investigations of the pathophysiology of schizophrenia: the role of susceptibility genes. J Psychiatry Neurosci, 29, 287 – 293.Pierce, K., & Courchesne, E. (2001). Evidence for a cerebellar role in reduced exploration and stereotyped behavior in autism. Biol Psychiatry, 49, 655–664.Pierce, K., Marinero, S., Hazin, R., McKenna, B., Carter Barnes, C., & Malige, A. (2015). Eye tracking reveals abnormal visual preference for geometric images as an early biomarker of an autism spectrum disorder subtype associated with increased symptom severity. Biol Psychiatry, 79, 657 – 666.Port, R.G., Gaetz, W., Bloy, L., Wang, D.J., Blaskey, L., Kuschner, E.S., . . . Roberts, T.P.L (2017). Exploring the relationship between cortical GABA concentrations, auditory gamma-band responses and development in ASD: Evidence for an altered maturational trajectory in ASD. Autism Research, 10(4), 593 – 607.Presti, M.F., Watson, C.J., Kennedy, R.T., Yang, M., & Lewis, M.H. (2004). Behavior-related alterations of striatal neurochemistry in a mouse model of stereotyped movement disorder. Pharmacol Biochem Behav, 77(3), 501–507.Puts, N.A.J., Wodka, E.L., Harris, A.D., Crocetti, D., Tommerdahl, M., Mostofsky, S.H., & Edden, R.A.E. (2016). Reduced GABA and altered somatosensory function in children with autism spectrum disorder. Autism Research, 10, 608–619.Rizzolatti, G., & Craighero, L. (2004). The mirror neuron system. Annual Review of Neuroscience, 27, 169 – 192.Robertson, C.E., Ratai, E.M., & Kanwisher, N. (2015). Reduced GABAergic action in the autistic brain. Current Biology, 26(1), 80–85.Rojas, D.C., Singel, D., Steinmetz, S., Hepburn, S., & Brown, M.S. (2014). Decreased left perisylvian GABA concentration in children with autism and unaffected siblings. NeuroImage, 86, 28–34.Russel, J. (Ed.). (1997). Autism as an executive disorder. New York, NY: Oxford University Press.Sabatino, A., Rittenberg, A., Sasson, N.J., Turner-Brown, L., Bodfish, J.W., & Dichter, G.S. (2013). Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord, 43 (12), 2903–2913. Sabatino DiCriscio, A., Miller, S.J., Hanna, E.K., Kovac, M., Turner-Brown, L., Sasson, N.J., . . . Dichter, G.S. (2016). Brief report: cognitive control of social and non-social visual attention in autism. J Autism Dev Disord, 46, 2797 – 2805.Sasson, N.J., Turner-Brown, L.M., Holtzclaw, T.N., Lam, K.S.L., & Bodfish, J.W. (2008). Children with autism demonstrate circumscribed attention during passive viewing of complex social and non-social picture arrays. Autism Res, 1: 1.Sasson, N.J., Elison, J.T., Turner-Brown, L.M., Dichter, G.S., & Bodfish, J.W. (2011). Brief report: circumscribed attention in young children with autism. J Autism Dev Disord, 41(2), 242 – 247.Schmitz, N., Rubia, K., Daly, E., Smith, A., Williams, S., & Murphy, D.G.M. (2006). Neural correlates of executive function in autistic spectrum disorders. Biol Psychiatry, 59, 7-16 Sears, L.L., Vest, C., Mohammed, S., Bailey, J., Rason, B.J., & Piven, J. (1999). An MRI study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol Psychiatry, 23(4), 613- 624. Sepeta, L., Tsuchiya, N., Davies, M.S., Sigman, M., Bookheimer, S.Y., & Dapretto, M. (2014). Abnormal social reward processing in autism as indexed by pupillary responses to happy faces. J Neurodevl Disord, 4, 17.Shafritz, K.M., Dichter, G.S., Baranek, G.T., & Belger, A. (2008). The neural circuitry mediating shifts in behavioral response and cognitive set in autism. Biol Psychiatry, 63(10), 974- 980. ?Shih, P., Shen, M., Ottle, B., Keehn, B., Gaffrey, M.S., & Muller, R.A. (2010). Atypical network connectivity for imitation in autism spectrum disorder.?Neuropsychologia, 48(10), 2931–2939.?Silverman, J.L., Pride, M.C., Hayes, J.E., Puhger, K.R., Butler-Struben, H.M., Baker, S., Crawley, N.J. (2015). GABA B receptor agonist R-baclofen reverses social deficits and reduces repetitive behavior in two mouse models of autism. Neuropsychopharmacology, 40(9), 2228–2239.Solomon, M., Ozonoff, S.J., Ursu, S., Ravizza, S., Cummings, N., Ly, S., & Carter, C.S. (2009). The neural substrates of cognitive control deficits in autism spectrum disorder. Neuropsychologia, 47, 2515-2526. ?South, M., Ozonoff, S., & McMahon, W.M. (2007). The relationship between executive functioning, central coherence, and repetitive behaviors in the high-functioning autism spectrum. Autism, 11(5), 437 – 451.Strang, J.F., Kenworthy, L., Daniolos, P., Case, L., Wills, M.C., Martin, A., & Wallace, G.L. (2012). Depression and anxiety symptoms in children and adolescents with autism spectrum disorders without intellectual disability. Res Autism Spectr Disord, 6(1), 406-412.Szatmari, P., Georgiades, S., Bryson, S., Zwaigenbaum, L., Roberts, W., Mahoney, W., . . . Tuff, L. (2006). Investigating the structure of the restricted and repetitive behaviors and interests domain of autism. J Child Psychol Psychiatry, 47(6), 582- 590.Supekar, K., Uddin, L.Q., Khouzam, A., Phillips, J., Gaillard, W.D., Kenworthy, L.E., . . . Menon, V. (2013). Brain hyperconnectivity in children with autism and its links to social deficits.?Cell Rep, 5(3), 738–747.Tremblay, R., Lee, S., & Rudy, B. (2016). GABAergic interneurons in the neocortex: from cellular properties to circuits. Neuron, 91(2), 260 – 292.Turner, M.A. (1999). Annotation: repetitive behavior in autism: A review of psychological research. J Child Psychol Psychiatry, 40(6), 839-849.Turner, K.C., Frost, L., Linsencardt, D., McIlroy, J.R., & Muller, R. (2006). Atypical diffuse functional connectivity between caudate nuclei and cerebral cortex in autism. Behav Brain Funct, 2(1), 34. Uddin, L.Q. (2011). The self in autism: an emerging view from neuroimaging. Neurocase, 17(3), 201 – 208.Uddin, L.Q., Supekar, K., Lynch, C.J., Khouzam, A., Phillips, J., Feinstein, C., . . . Menon, V. (2013). Salience network-based classification and prediction of symptom severity in children with autism. JAMA Psychiatry, 70(8), 869 – 879.van Heijst, B.F.C., & Geurts, H.M. (2015). Quality of life in autism across the lifespan: A meta-analysis. Autism, 19(2), 158 – 167.Vasa, R.A., Mostofsky, S.H., & Ewen, J.B. (2016). The disrupted connectivity hypothesis of autism spectrum disorders: time for the next phase of research. Biol Psychiatry Cogn Neurosci Neuroimaging, 1(3), 245 – 252.Wang, Y., Zhang, Y.B., Liu, L.L., Cui, J.F., Wang, J., Shum, D.H., can Amelsvoort, T., & Chan, R.C. (2017). A meta-analysis of working memory impairments in autism spectrum disorders. Neuropsychological Rev, 27(1), 46 – 61.Watson, K.K., Miller, S., Hannah, E., Kovac, M., Damiano, C.R., Sabatino-DiCrisco, A., . . . Dichter, G.S. (2015). Increased reward value of non-social stimuli in children and adolescents with autism. Frontiers in Psychology, 6, 1026.Weng, S.J., Wiggins, J.L., Peltier, S.J., Carrasco, M., Risi, S., & Lord, C., 2010. Alterations of resting state functional connectivity in the default network in adolescents with autism spectrum disorders. Brain Res, 1313, 202–214. Wolff, N., Chmielewski, W.X., Beste, C., & Roessner, V. (2017). Working memory load affects repetitive behaviour but not cognitive flexibility in adolescent autism spectrum disorder. World J Biol Psychiatry, 1 – 12.Zalla, T., & Sperduti, M. (2013). The amygdala and the relevance detection theory of autism: an evolutionary perspective. Front Hum Neurosci, 7, 894.Zwaigenbaum, L., Bryson, S., Rogers, T., Roberts, W., Brian, J., & Szatmari, P. (2005). Behavioral manifestations of autism in the first year of life. Int J Devl Neurosci, 23, 143–152.Chapter 2Structural and Functional Neuroimaging of Restricted and Repetitive Behavior in Autism Spectrum DisorderJenna M. Traynor1 and Geoffrey B.C. Hall21 PhD Candidate, Department of Psychology, Neuroscience & Behavior, McMaster University, Hamilton, Ontario, Canada2 Associate Professor, Department of Psychology, Neuroscience & Behavior, McMaster University, Hamilton, Ontario, CanadaTraynor, J.M., & Hall, G.B.C. (2015). Structural and functional neuroimaging of restricted and repetitive behavior in autism spectrum disorder. Journal of Intellectual Disability – Diagnosis and Treatment, 3, 21 – 34.Abstract and Key WordsA prominent symptom of Autism Spectrum Disorder includes restricted and repetitive behaviours. This symptom has been divided into three subtypes: repetitive motor behaviour, insistence on sameness and circumscribed interests. In the past, the neural correlates of these behaviours have been largely understudied. More recently, neuroimaging studies have pointed to a number of neural networks that may underlay these behaviours. However, results from this work have been varied and remain difficult to integrate. The purpose of this review is to summarize recent neuroimaging studies on restricted and repetitive behaviours in autism, and to provide an organized framework that will permit a clearer understanding of the neural correlates of these behaviours. Using a developmental perspective, this review will identify that there are distinct and overlapping neural networks that are associated with repetitive motor behaviour, insistence on sameness and circumscribed interests. In addition, this review will identify a series of executive and affective function tasks that have proven efficacious in the study of repetitive behaviour. KEYWORDS: autism spectrum disorder, neuroimaging, repetitive behaviour, repetitive motor behaviour, insistence on sameness, circumscribed interestsIntroductionAutism Spectrum Disorder (ASD) is a neurobiological condition characterized by deficits in social communication and restricted, repetitive behavior (RRB) [1]. It is a complex disorder associated with diverse behavioral symptoms and complex neural substrata. Recently, a growing body of neuroimaging research focused on examining RRB in ASD has emerged. RRB encompasses a broad range of heterogeneous behaviors that have proven difficult to quantify [2]. These behaviours likely reflect complex genetic and environmental interactions across development [3]. Given this heterogeneity, the conceptualization of RRB as one broadly inclusive category has been replaced by a framework that stratifies RRB into three distinct subtypes or factors. These include Repetitive Motor Behavior (RMB), Insistence on Sameness (IS) and Circumscribed Interests (CI) [4]. These three behavioral subtypes have been determined by factor analyses and independent component analyses of diagnostic measures such as the Autism Diagnostic Interview- Revised (ADI-R) and the Autism Diagnostic Observation Schedule (ADOS) [4-6]. The identification of these subtypes has allowed for a more fine-tuned behavioral profile of ASD. For example, RMB is associated with a distinct motor component [7], and has been described as more rudimentary, or “lower order,” in nature [8]. Some examples of RMB include hand flapping, body rocking and other self-stimulatory behaviors. On the other hand, IS and CI have been described as “higher order” [8] and are linked to distinctly cognitive components [7]. Rigid adherence to routines and unusually intense preoccupations are examples of IS and CI, respectively. Recently, a number of neuroimaging research studies have examined the relationship between these distinct RRB subtypes and underlying neural circuitry [9], bridging the gap between behavioral and neural findings in ASD. However, the results of these studies have been mixed. For example, by using structural magnetic resonance imaging (MRI), some studies have revealed patterns of morphological brain changes that are significantly correlated with different RRB subtypes [10-12], whereas other studies have found that no correlations hold across RRB categories [24]. These studies have been additionally complicated by the use of different neuroimaging methods and sample demographics. For example, in addition to MRI, many studies have used functional magnetic resonance imaging (fMRI) to examine RRB indirectly, by analyzing brain activation during the performance of executive functioning (EF) tasks. These EF tasks have been used as alternative RRB measures, as a means to examine RRB using task-based paradigms [14]. Recently, a number of EF tasks have been proposed to study RRB in ASD. Findings have linked performance and neural activation during these EF tasks to ASD scores on diagnostic measures [15-18]. In addition, a handful of studies have drawn correlations between neural activation on some of these EF paradigms and specific RRB scores on diagnostic measures. These studies demonstrate the specific efficacy of these paradigms in the study of RRB. With some studies establishing direct links between RRB and brain function, and other studies proposing new paradigms to investigate RRB in ASD, a review and organization of the literature is needed. These findings hold great clinical promise, as they may contribute to a more precise neural understanding of RRB. They also offer potential for the development of targeted interventions and novel pharmacological therapies. With this said, this review has two purposes. First, this review is intended to synthesize recent structural MRI studies into an organized framework to permit a clearer understanding of the morphological abnormalities associated with RMB, IS and CI. This summary will demonstrate that these distinct RRB subtypes are associated with both discrete and overlapping regional abnormalities that are best considered from a developmental framework. For example, discrete regional abnormalities of the frontal cortex and cerebellum have been associated with RMB [19]. In contrast, overlapping abnormalities in frontal and striatal regions appear as key contributors to all three subtypes (i.e., RMB, IS and CI) across development [11, 19-20]. This review presents a broad frontal striatal developmental trajectory of RRB in ASD (Figure 1), and indicates underlying motor and cognitive control deficits in the development of RRB.The second purpose of this review is to synthesize recent fMRI studies, focusing on network abnormalities and differences in RRB neural circuitry. Inherent in this synthesis is an evaluation of several EF tasks that have been used as RRB proxies in a number of fMRI studies. EF tasks that examine visual motor coordination, motor and cognitive inhibition, rule violation, cognitive set shifting, target detection, and delay incentive as they relate to RRB, are included in this review. An organized framework of EF tasks that have been used as RRB proxies is presented. This framework outlines which EF tasks elicit brain activation, which is correlated with RRB scores on diagnostic measures. This comparison will allow for an initial organization of findings and pinpoint areas of interest for future research. This comparison identifies that there are overlapping neural network abnormalities that are elicited by certain EF tasks. This overlap is consistent with theories identifying ASD as a brain connectivity disorder. These findings are presented in a neurocognitive model of RRB (Figure 2), and identify EF tasks that may be of value in future RRB research. More specifically, this model shows that abnormalities in neural networks implicated in cognitive and motor control, attention, salience attribution, and reward function, may underlay RRB in ASD. Differences in sample age, diagnosis, and methodology between studies are also discussed in light of their impact on results.The structure of this review is as follows: first, structural neuroimaging (MRI) studies are discussed. Then, functional neuroimaging (fMRI) studies are discussed. This discussion of fMRI studies includes studies that have employed tasks of EF as indirect measures of RRB. Within each structural and functional section, RMB, IS and CI behaviors are discussed. Following this, resting-state functional connectivity and diffusion tensor imaging methods are briefly discussed. Resting-state and diffusion tensor imaging studies are examined separately because they are relatively new approaches to studying RRB in ASD, and do not yet fit the stratified framework of subtypes outlined above. Last, future directions are considered.MethodologyReported in the current review are all studies that were identified based on a comprehensive literature search in PubMed, PsychINFO and Google Scholar. Searches were conducted for neuroimaging studies conducted between 1999 and 2013 that examined repetitive behaviours in children and adults with an ASD (autism, high functioning autism Asperger syndrome, and pervasive developmental disorder not otherwise specified). Only peer-reviewed articles for which English text was available were included in the review. Key words used in the literature search included “autism”, “neuroimaging”, “[functional] magnetic resonance imaging”, “diffusion tensor imaging”, “restricted [and] repetitive behaviour”, “circumscribed interests”, “insistence on sameness”, “motor stereotypies”, “repetitive motor behaviour”, “repetitive cognitive behaviour”, and “executive function”. See Table 1 for a list of papers reviewed, including number of participants, age range, diagnosis, type of neuroimaging method used and a summary of findings from each study included in this review.Structural MRI studies and RRBLower order, repetitive motor behavior (RMB)The importance of studying RMB is underscored by the observation that it is the only RRB subtype that is present as early as 15 months [21], and observed across different levels of functioning across the autism spectrum [22]. The expression of RMB is also quite heterogeneous and includes abnormal gait, somatic movement, and vocal production. This heterogeneity suggests potentially widespread early neural candidates. Therefore, the study of neural circuitry underlying RMB offers potential for the identification of early biological markers of ASD that may serve as the initial basis for a cascade of further deficits.However, structural neuroimaging studies investigating lower order RMB in ASD have been few. One factor that may contribute to this paucity of imaging studies is the requirement for minimized motion while in the MRI scanner. This requirement conflicts with the excessive somatic movement characteristic of RMB, rendering this subtype particularly difficult to investigate using MRI. Currently, the few available studies relating structural brain abnormalities to RMB show little consensus in results. Pediatric studies have linked increases in entire frontal cortex volume and cerebellar hypoplasia with the atypical exploration of objects in the environment [19]. In addition, self-injurious behaviors have been negatively correlated with thalamic volume, as well as with superior parietal and somatosensory cortical thickness [10]. Abnormal parietal folding has also been linked with measures of RRB in a sample of children with Asperger’s Syndrome. Interestingly, this correlation did not hold when examining both high and low-functioning autism samples [23]. This suggests that the biological bases of RRB may be diagnosis dependent. Finally, in contrast to the above findings, a recent study failed to find any significant association between morphometric alterations in child ASD brains and RMB scores [24]. Clearly more work is needed to identify the relation between structural brain abnormalities in childhood ASD and RMB. Higher order, insistence on sameness, and circumscribed interestsIn addition to lower order behaviors (i.e., RMB), research has also been conducted on higher-order behaviors such as IS and CI. These higher-order subtypes have been studied more extensively, and structural MRI research has revealed substantial cortical and subcortical associations.Thus far, evidence from structural MRI studies investigating IS and CI behaviors suggest that the neural underpinnings of these two subtypes are not completely separable, and may overlap considerably. In addition, data collected from pediatric and adult samples has revealed an early developmental progression of broad frontal and striatal involvement in these higher-order behaviors [9, 11, 19-20].Although it has been found that higher order IS and CI exist in very young ASD samples [25-26], associations between these subtypes and striatal abnormalities show a complex pattern across development. For example, in 3-4 year old ASD samples, findings of uncorrected striatal enlargement in ASD children relative to controls have failed to meet statistical significance [27]. However, longitudinal data have revealed faster striatal growth rate (as opposed to enlargement) in school aged ASD subjects relative to controls [9]. Further, this increased growth rate has been both positively and negatively correlated with IS scores on the ADI-R [9, 28]. In this age group specifically, increased growth rate of the caudate nucleus has shown a significant negative association with IS [28]. In the same age group, growth rate of the putamen has shown a significant positive correlation with IS [9]. In studies employing older ASD samples between 7 and 12 years of age, findings of increased caudate [29], putamen and globus pallidus [30] volumes relative to controls, have been reported, and increased volume of the caudate has been correlated with measures of impulsivity [31]. These data suggest that abnormal growth rate of the striatum in early childhood may be a precipitating factor in the development of higher-order RRB later in life. As well, enlargement of striatal structures in later childhood and adolescence have been found. However, associations between striatal enlargement and RRB have yet to emerge in this age group.Continuing this developmental trajectory of the striatum, several adult studies have reported significant correlations between striatal enlargement and higher-order RRB scores. For example, caudate [11-12] and putamen enlargement [11] have been correlated with “perseverations” and “obsessions/compulsions” scores on the Autism Diagnostic Interview. Similar to pediatric findings, these correlations have also been both positive and negative in direction, speaking to the complexity of IS and CI across development. Additionally, enlargement of the lateral orbitofrontal cortex (OFC) has been positively correlated with CI in adults, but not in children [20]. Complementing this correlational pattern, behavioral research on adolescent ASD subjects has demonstrated that as age increases, the severity of CI behavior becomes more intense [32]. Structural MRI and RRB subtypes – a summary and comparisonThus far, data on lower order RMB point to enlargement of frontal cortical structures, as well as cerebellar hypoplasia in childhood. In addition, reduced cortical thickness and abnormal folding in post central, parietal and subcortical structures, has been found. However, greater methodological consistency is needed in future RMB studies to substantiate and replicate these findings, as studies have also demonstrated null findings [24]. Given that RMB is of the earliest subtype to occur, findings could aid as structural biomarkers for high-risk ASD children. Taken together, the above data indicate widespread structural abnormality in RMB. The findings are so varied, that it may be of value to further parse these qualitatively different motor behaviors in order to arrive at more discernable RMB subtypes.Alternatively, IS and CI subtypes involve irregularities of the OFC and striatum (i.e., putamen and caudate). The OFC plays a role in reward-based decision making, reinforcement learning, and emotion processing [33]. Irregularities of OFC function are also found in obsessive-compulsive disorder [34] indicating that OFC abnormality may broadly underlay narrow-ranged perseverative behaviour, trans-diagnostically. In addition, the putamen and caudate together form the dorsal striatum, which is implicated in a wide range of functions including inhibition, motor initiation [35-36], salience assessment and reward expectancy in decision making [37]. Taken together, it is suggested that the structural brain regions underlying IS and CI stem from frontal-dorsal striatal abnormalities, and that these abnormalities may be associated with deficits in both reward-based or emotional decision making, as well as motor inhibition. Collectively, an examination of the regions putatively implicated in lower and higher order RRB suggests that there are commonalities in frontal-striatal irregularities across all three subtypes (i.e., RMB, IS, and CI). Moreover, abnormalities in specific structures of the frontal-striatal pathway show associations with specific types of RRB at different developmental time points (Figure 1).Concerning frontal cortical areas, both RMB and CI are reflected by enlargement of the frontal cortex, with CI behavior specifically localized to abnormalities of the OFC. In addition, striatal abnormalities that are localized to the caudate nucleus appear important to higher order RRB. Other parts of the brain that are closely tied to the striatum (i.e., thalamus) are implicated in self-injurious behavior.Recent research on the role of the striatum in animal and human studies has revealed its importance in both motor control [35-36] and more cognitive, obsessive-compulsive behaviors [38]. The involvement of the striatum in both lower and higher order behaviour suggests that abnormalities in this group of structures may play a role in motor deficits in early childhood (i.e., RMB), which leads to higher-order, cognitive deficits across development (i.e., IS and CI). This literature may suggest that frontal-striatal abnormality serves as an early biomarker for ASD. The utility of this suggestion will require future research.Functional MRI and RRB subtypesThe structural architecture of the brain plays an important role in the integration of information across functional networks [39]. Thus, a number of other studies have employed functional MRI (fMRI), in addition to morphometric analyses of MRI data. These studies have used fMRI to examine RRB subtypes indirectly, by analyzing brain activation during the performance of affective and executive functioning (EF) tasks. Both traditional fMRI as well as functional connectivity analyses have been used. fMRI uses the blood oxygen level dependent (BOLD) signal to identify regional changes in neuronal activity. These changes are associated with the presentation of a particular stimuli or the performance of a task. Functional connectivity analyses examine the synchronous coupling of activity across localized brain regions in order to provide information about functional integration (i.e., networks) [40].It is important to consider how EF paradigms may be used in neuroimaging studies to evaluate RRB, as a greater number of studies are establishing relationships between RRB subtypes and neural circuitry. To date, a handful of studies have linked neural activity during EF tasks to distinct RRB subtypes. Many studies have also correlated findings of neural abnormality during EF tasks with total RRB scores on the ADI-R. In the following sections, EF tasks that are both rudimentary and/or motor in nature, as well as those that require higher-order, cognitive demand, have been summarized. EF tasks that have recruited similar neural networks are summarized together, and preliminary links to RRB subtypes are presented.Lower order executive functioning tasks and RRBA few fMRI studies have employed EF paradigms that are primarily motor in nature, and these paradigms have been used to explore lower-order RRB in ASD. Results have been correlated with total RRB scores on the ADI-R. These studies have been conducted using older adolescent and adult ASD samples, and these studies need to employ younger samples for better generalization of results. EF paradigms using anti-saccade and Go/No Go tasks (assessing motor inhibition), as well as simple visual motor coordination tasks (assessing motor planning and coordination), have been employed to explore the neural circuitry related to RRB. The anti-saccade task has revealed hypoactivation in the frontal eye fields and the dorsal anterior cingulate cortex (dACC), in ASD subjects relative to controls. The task has also shown reduced functional connectivity between these neural regions [41]. Despite overall hypoactivation in the frontal eye fields in ASD subjects, greater frontal eye field activation during the anti-saccade task has been correlated with more severe total RRB scores on the ADI-R [41]. Agam and colleagues have suggested that this correlation reflects greater cognitive effort in ASD subjects, which is required to successfully inhibit motor behavior. Hypoactivation of the ACC has also been shown during other motor inhibition tasks such as the Go/No Go [42]. The Go/No Go has revealed hyperactivation in the left inferior and orbital frontal gyrii in ASD subjects relative to controls [43].Aside from the anti-saccade and Go/No Go, simple visual motor coordination tasks have been employed. These tasks have revealed heightened functional connectivity in ASD across vast thalamocortical and caudate-cortical networks. Hypotheses for these aberrant cortical-subcortical connectivity patterns include reduced, early synaptic pruning, abnormal white matter maturation, and thalamic gate dysfunction during development [44-45]. These vast abnormalities in cortical-subcortical networks may suggest an array of several neural mechanisms that support RRB. Although these studies did not attempt to draw correlations between BOLD activation and RRB scores, the rudimentary and motor nature of these tasks identify them as plausible RRB proxy measures. The results are distinguished as relevant to stereotypic behavior in ASD [45] indicating that thalamic-cortical and caudate-cortical connectivity should be investigated in future work.Higher order executive functioning tasks and RRBA number of cognitive EF tasks have been employed in order to deepen the understanding of neural involvement in higher order RRB. These EF tasks have investigated BOLD activation during particular cognitive demands. These cognitive demands have been linked to IS and CI. Some studies have also directly correlated neural activation during these tasks with IS and CI scores on the ADI-R. EF tasks requiring cognitive interference inhibition, rule violation processing, simple target detection, and cognitive set shifting have been employed to investigate the neural circuitry of higher order RRB. Novel paradigms specific to CI behavior have also been used. These paradigms are examined in the following sections.Behavioral inhibition, rule violation and the salience network. Behavioral evidence has demonstrated correlations between deficits in inhibition and higher-order RRB scores on diagnostic measures [17, 46]. This evidence has justified using inhibition paradigms in RRB imaging studies. Interestingly, fMRI studies using inhibition, rule violation processing, and error monitoring paradigms have demonstrated significant overlap in neural abnormalities displayed by ASD subjects, compared to controls. Results from these paradigms have revealed abnormal activation of the insula [13, 43], abnormal activation in extensive frontal control regions [42-43, 47], aberrant connectivity of the insulii with frontal structures [42, 47], abnormal activation in the parietal cortex [43], and reduced synchronization and under connectivity in frontal-parietal networks [42, 48]. In one study, this lack of frontoparietal synchronization was correlated with an increase in attention-deficit hyperactivity symptoms [48]. This correlation suggests that deficits in attention may possibly underlie differences in cognitive performance between ASD and control subjects. Future studies employing these paradigms should investigate the correlation between RRB scores on diagnostic measures and neural activation. These studies would strengthen the understanding of neural contributors to RRB given the behavioral evidence linking inhibition and RRB [17, 46], and the strong neural overlap during inhibition, rule violation, and error monitoring task performance. Further, these studies indicate the salience network in RRB. The salience network includes the ACC, and the insula [49]. Closely connected with the salience network are neural circuits sub serving inhibition and attentional processes (including the ACC, cingulate gyrus, insula and parietal regions). The insula plays a role in attributing negative internal emotional states to decision making [50], and strong insular involvement in the above studies is posited to indicate heightened sensitivity to errors/violations in ASD subjects compared to controls [13, 43]. Also seen in the above studies is strong frontoparietal network involvement. The frontoparietal network plays a large role in attention and the selection of sensory input from the environment [51].Taken together, these results suggest that ASD deficits in higher-order executive functioning may involve neural abnormalities in salience attribution, inhibition, and attention networks. Future research investigating correlations between these network abnormalities and RRB scores on diagnostic measures will be essential. These studies may solidify connections between RRB subtypes and salience, inhibitory, and attentional processes. Recent ASD research has revealed neural deficits in attentional alerting, orienting, and executive control networks [52]. This further demonstrates that attentional dysfunction may be a core contributor to ASD symptomology. Additionally, initial confirmation of salience network involvement in RRB comes from a recent resting-state fMRI study. This study found that hyper-connectivity of the salience network in ASD children was specifically predictive of RRB scores on the ADI-R [49].In addition to the salience network, evidence of frontoparietal (attention) network involvement in RRB has been shown during other types of EF tasks, such as target detection and cognitive flexibility. This work is described in the following section.Target detection, cognitive flexibility, and the frontoparietal network. In addition to inhibition and rule violation tasks, target detection and cognitive flexibility tasks have revealed abnormal frontoparietal activation in ASD subjects, relative to controls. Justification for the use of cognitive flexibility paradigms comes from behavioral evidence showing a relationship between cognitive flexibility and RRB [16]. Difficulty with change in the environment that is inherent in certain types of RRB (such as IS) [53] also contributes to the face validity of target detection and flexibility tasks as RRB proxies.Both cognitive flexibility and target detection tasks have shown abnormal patterns of activation in the ACC [53-54], as well as complex parietal dysfunction [43, 54]. These altered patterns of activation in the ACC and parietal lobe during target detection have been negatively correlated with the severity of RRB scores on the ADI-R [54]. This correlation further indicates attentional (i.e., frontoparietal) processes in the maintenance of RRB. Moreover, hypoactivation of the basal ganglia during poor cognitive set shifting performance has been identified in ASD [54].To summarize, behavioral evidence linking cognitive flexibility and RRB, and correlations between neural activation during target detection performance and RRB scores, indicate that both of these executive functions underlay RRB in ASD, and should be explored further.Circumscribed interests. Lastly, researchers have employed EF paradigms to specifically investigate the higher-order, CI subtype [55-57]. These novel approaches use fMRI to investigate BOLD activation during the visual presentation of CI stimuli. For example, pictures of common CI in ASD, such as planes, trains and other mechanical objects, have been used. BOLD activation during the presentation of CI stimuli has been compared to activation during the presentation of other types of visual stimuli, such as social images (e.g., faces), or non-CI stimuli (i.e., control objects or shapes). Results have revealed that ASD individuals show hyper activation to CI stimuli, but not to social or non-CI stimuli, in the ventromedial prefrontal cortex (VMPFC)- nucleus accumbens (NAcc) reward network [56] and in insular, salience attribution networks [55]. Hyperactivation of the insula to CI stimuli in ASD subjects was also correlated with the intensity of everyday CI behavior, as recorded in a parent-report measure [55].Interestingly, one other study used similar CI stimuli in an inhibition paradigm, but found hyperactivation of the superior frontal gyrus and right insular cortex to social stimuli, and not to CI stimuli [57]. This study also found hypoactivation of the caudate nucleus to CI stimuli in ASD subjects [57]. Finally, this study found that higher RRB scores in ASD subjects were correlated with a decrease in left inferior and right middle frontal gyrus activation. Although in contrast to the above results of Casico and colleagues [55], these findings have been interpreted as heightened recruitment of cognitive control areas needed for successful inhibition of social stimuli [57]. This suggests that it may be more difficult for those with ASD to inhibit social stimuli, compared to controls [57].Taken together, these results suggest that salience (i.e., insula) and reward network dysfunction (i.e., nucleus accumbens), as well as defective cognitive control (i.e., caudate nucleus), may be involved in the expression of CI behavior. However, increased insular activation to both social and CI stimuli indicate a complex role of the salience network in CI behavior. The overlapping neural activation during the presentation of both social and CI stimuli also indicate potential dysfunction of the reward network in both social and repetitive behavior symptoms of ASD. This suggestion parallels ASD studies that have shown early neuronal disorganization as a factor in the later devaluing of social interaction [58]. This suggestion is also similar to studies that have shown reduced activation of the dorsal striatum to social rewards in ASD [59]. Given these findings, future studies should further explore the relationship between social and RRB deficits, and their underlying neural counterparts.Executive functioning tasks and RRB subtypes – a neurocognitive model Taken together, the above affective and executive functioning paradigms have revealed important information about the neural circuitry underlying RRB in ASD. Lower order EF tasks investigating motor inhibition and visual motor coordination have revealed reduced frontal eye field-dACC connectivity. These tasks have also revealed increased connectivity in an array of cortical-subcortical networks. These studies have implicated several neural mechanisms that may underlay lower order RMB, however, future work is needed to solidify connections between specific RMB subtypes and abnormal neural circuitry. Given the vast heterogeneity in RMB across the spectrum, it is likely that the above results may reflect specific forms of RMB, while failing to identify the neural mechanisms of other forms. Alternatively, higher order EF paradigms have identified strong frontoinsular and frontoparietal involvement in RRB. In addition, reward network abnormalities, as well as cognitive control deficits, have been shown during performance in CI-specific paradigms that encompass affective elements.Collectively, neural abnormalities in the studies discussed above reveal that RRB in ASD may involve particular deficits in cognitive control, motor control, attention, salience attribution, and reward processing. Importantly, overlap in neural circuitry involvement during distinct EF task performance is shown. This underscores the complexity of neural involvement in the broader RRB subtype. This overlap is not surprising. Estimates suggest that between 1 and 6 independent neural components may overlap at a given voxel in the brain [60]. Viewed in this way, RRB in ASD appears as neurocognitively dimensional (Figure 2.) More specifically, abnormalities in a given neural network are identified during performance in multiple EF tasks. For example, abnormalities in the frontoparietal attention network appear as important to consider during target detection, rule violation, cognitive flexibility, and inhibition tasks. This neurocognitive dimensionality in RRB will be important to consider in future work. As there is a lack of intervention practices targeted toward RRB in ASD [61], these findings may provide a foundation for the development of reinforcement-based interventions that target specific neurocognitive deficits. They may also be useful for individuals displaying unique RRB profiles. Additionally, EF paradigms employing anti-saccade, target detection, and CI-specific, affective elements, have demonstrated efficacy in the study of RRB. These paradigms have drawn direct correlations between brain activation and RRB scores. EF paradigms employing cognitive and motor inhibition, rule violation, and cognitive flexibility, also show promise in the study of RRB. Further investigation will be required in order to use these paradigms as evidence-based RRB measures.Resting-state functional connectivity and RRBAnother imaging method known as resting-state functional connectivity MRI, has revealed an interesting relationship between social communication and RRB symptoms in ASD. Resting state studies (i.e. scans with an absence of any task) make use of spontaneous, low frequency oscillations in the BOLD signal that are correlated across functionally related regions [62]. These studies have identified a relationship between default mode network connectivity and the broader RRB class [63-64]. Interestingly, aberrant resting state connectivity of the posterior cingulate cortex (PCC) with both frontal and temporal cortices, and with the parahippocampal gyrus, has been correlated with poor verbal communication scores, poor social functioning scores, and an increase in RRB scores on the ADI-R [63-64]. These studies suggest a functional overlap between social and RRB symptoms in ASD, which can be observed at a basic, resting-state level. These findings also compliment the results of functional scanning paradigms in the above section (Circumscribed interests) that reveal overlap in neural responses to social and RRB (i.e., CI) stimuli. Aside from the default mode network, one other study also found a positive correlation between right medial frontal gyrus and caudate nucleus resting-state connectivity and RRB scores on the ADI-R [59]. This finding further demonstrates that abnormalities in the frontostriatal system are involved in RRB. Finally, hyper connectivity of the resting-state salience network has been correlated with RRB scores on the ADI-R [49]. In future, continued use of resting-state functional connectivity methods should provide useful information about RRB networks, and begin to draw correlations between specific RRB subtypes and resting-state brain function.Diffusion tensor imaging and RRBFinally, Diffusion Tensor Imaging (DTI) is an imaging method that uses the directional diffusion of water molecules in the brain to make inferences about the structural integrity of white matter fibres. Fractional anisotropy (FA) is a quantitative DTI value that identifies the degree of coherence between the direction of water molecules, indicating high or low connectivity of white matter microstructure [65]. Compared to other neuroimaging methods, there has been limited RRB research using DTI. However, there is a need for research using this method because it is important to investigate whether areas of altered brain activation are also associated with altered connectivity and changes to the microstructure of white matter.DTI studies have identified both cortical and subcortical white matter abnormalities. Reduced volume of the forceps minor – a white matter tract linking the lateral and medial surfaces of the frontal lobes [66] and reduced integrity of rostral ACC white matter [67] have been correlated with RRB scores on the ADI-R. An excess of short-range connections in white matter inferior to the ACC [68] have also been hypothesized to contribute to RRB expression and the inability of individuals with autism to disengage from a task or stimulus. Abnormalities in white matter surrounding the basal ganglia have also been correlated with behavioral inhibition performance [69]. This provides further support for the use of inhibition EF tasks to explore the architecture underlying RRB.Altogether, these findings indicate that structural white matter abnormalities may be associated with frontal, cingulate, and striatal alterations in ASD subjects, all of which are implicated in repetitive behavior. Yet, neural patterns associated with specific RRB subtypes are not identified from these DTI studies. Future work would benefit from an approach that stratifies RRB subtypes, and correlates these subtypes with specific white matter abnormalities.Future directionsThis review has identified that RRB subtypes in ASD encompass both distinct and overlapping neural networks that follow important developmental trajectories. The identification of distinct systems may be useful in the creation of targeted treatments for various individual RRB profiles, as well as in the identification of RRB biomarkers. Overlap in neural circuitry across subtypes is reflective of ASD as a heterogeneous condition, and reflects the complexity of brain-behavior relationships in RRB.A comparison of the neural networks underlying RMB, IS and CI has yielded common frontostriatal structural and functional deficits, implicating motor and cognitive control mechanisms across a broader RRB grouping. Importantly, this review has highlighted an initial developmental trajectory of frontal-striatal morphology and RRB. Although the association between RRB and striatal abnormality is quite complex, early striatal abnormalities may serve as a biomarker for both lower and higher order RRB profiles throughout development. However, correlations between striatal abnormality and RRB are largely unconsolidated, specifically during adolescence. Clearly there are still unknown neural mechanisms supporting higher-order RRB throughout development, and more studies investigating RRB subtypes in adolescent samples are needed. These studies would pinpoint critical periods in development where neural changes shape behavior.In addition to frontal striatal abnormality, analyses of fMRI data have highlighted potential motor and cognitive control, attention, salience, and reward processing deficits in RRB. Correlations between these neural network abnormalities and RRB scores are informative, and indicate the efficacy of using certain EF tasks to examine RRB networks. Specifically, EF tasks encompassing motor inhibition, target detection, and reward processing, reveal neural activation that is correlated with RRB scores. This demonstrates that these tasks are good RRB proxies. Additionally, preliminary data has indicated that cognitive flexibility and inhibition tasks, as well as visual motor coordination tasks, should be examined in future work. However, given that the examination of RRB subtypes is a relatively new approach, there is significant methodological inconsistency across studies. This makes quantitative comparison between RRB subtypes difficult. Future work would benefit from combining a variety of imaging methods to examine a given RRB subtype. This practice would contribute to more detailed neural profiling of RRB in ASD. For example, data driven or network based analysis methods like Independent Component Analyses could be employed to compliment a general linear model analyses. This would help with further investigation of irregularities in the functional networks underlying RRB.In addition, an interesting relationship between social and RRB symptoms has emerged, both in task-based fMRI studies, as well as in resting-state studies. The literature suggests that changes in neural reward function in ASD may play a role in the devaluing of social stimuli, while increasing the salience of CI stimuli. Further, abnormal resting state connectivity in overlapping default mode network regions is correlated with both social and RRB symptoms in ASD, indicating that these deficits may share common and intrinsic neural abnormalities.Finally, both resting-state fMRI and DTI studies provide preliminary evidence for underlying resting-state and white matter abnormalities in RRB, respectively. DTI studies have been important in the confirmation of frontostriatal irregularity in RRB at a white matter level. However, in order to further pinpoint the role of neural deficits in ASD symptoms, future work in these areas should build upon previous studies, and stratify RRB into subtypes. Data driven resting-state studies looking at other large scale resting state networks should also be investigated in the context of RRB. For example, functional abnormalities in thalamocortical networks have been identified during previous task-based paradigms (Mizuno et al., 2006), and future resting-state studies investigating thalamocortical circuitry and RRB could further identify the role of this circuit in RRB subtypes.Overall, this review has demonstrated the utility of stratifying RRB when conducting neuroimaging research on ASD. By identifying important neural networks that underlay RRB in ASD, this review has aided in the initial neural mapping of symptoms. This work is important, as RRB in ASD has traditionally been defined behaviourally. This review has also highlighted affective and executive functioning tasks that demonstrate efficacy in the study of RRB. In future, this knowledge may be of value in the development of interventions and treatments for individuals displaying unique RRB profiles.Tables and FiguresTable 1. Neuroimaging results presented in this review.Author(s)Sample Age Range (yMethodNeural region(s)Results (ASD vs. comparison group)Langen et al. (2013) [9]ASD (n=49) vs. TD (n=37)9 – 12 yrsMRIStriatum (caudate nucleus and putamen)Growth rate of striatum correlated with insistence on sameness behaviour on the ADI-R at preschool ageDuerden et al. (2013) [10]ASD (n=30) vs. TD (n=30)7 – 15 yrsMRI Thalamus, superior parietal lobe, somatosensory cortexVolume of thalamus and cortical thickness of superior parietal and somatosensory cortex correlated with self-injurious behaviourHollander et al. (2005) [11]ASD (n=17) vs. TD (n=17)17 – 57 yrsMRIStriatum (caudate nucleus and putamen)Enlargement of striatum correlated with ‘obsessions/ compulsions’ and ‘perseverations’ scores on the ADISears et al. (1999) [12]ASD (n=35) vs. TD (n=37)12 – 29 yrsMRICaudate NucleusEnlargement of caudate nucleus correlated with ‘obsessions/ compulsions’ and ‘perseverations’ scores on the ADIGoldberg et al. (2011) [13]HFA (n=11) vs. TD (n=15)8 – 12 yrsfMRI (using a Go/No Go task)InsulaHyperactivation of insula Pierce & Courchesne (2001) [19]ASD (n=14) vs. TD (n=14)3 – 5 yrsMRIFrontal cortex, cerebellumVolume of entire frontal cortex and cerebellum correlated with the atypical exploration of objects in the environmentHardan et al. (2005) [20]ASD (n=40) vs. TD (n=41)8 – 46 yrsMRIOFCEnlarged OFC correlated with circumscribed interest scores on the ADI-RNordahl et al. (2007) [23]Asperger’s (n=15) vs. TD (n=29)7 – 18 yrsMRIParietal CortexAbnormal parietal folding correlated with repetitive behaviour scores on the ADI-RGoldman et al. (2013)[24]ASD (n=31) vs. TD (n=30)mean ASD age 9 yrs (range N/A)MRISupplementary motor cortex, OFC, DLPFC, ACC, caudate nucleus, globus pallidus, thalamus, hypothalamusNo correlation found between volume of cortical or subcortical structures and video coded motor stereotypies Estes et al. (2011)[27]ASD (n=45) vs. TD (n=25)3 – 4 yrsMRIStriatum (caudate nucleus and putamen)No significant correlation found between uncorrected striatal enlargement and repetitive behaviour scores on the ADOSLangen et al. (2009)[28]HFA (n=99) vs. TD (n=89)6 – 25 yrsMRICaudate nucleusIncreased growth rate of the caudate nucleus correlated with insistence on sameness scores on the ADI-RLangen et al. (2007) [29]ASD (n=21) vs. TD (n=21)7 – 14 yrsMRICaudate nucleusIncreased volume of caudate nucleusHerbert et al. (2003) [30]ASD (n=17) vs. TD (n=15)7 – 11 yrsMRIPutamen, globus pallidusIncreased volume of putamen and globus pallidusVoelbel et al. (2006) [31]ASD (n=38) vs. TD (n=13)7 – 13 yrsMRICaudate nucleusIncreased volume of caudate nucleus predicted measures of impulsivity on the CPTAgam et al. (2010) [41]ASD (n=11) vs. TD (n=14)18 – 38 yrsfMRI (using an anti-saccade task)FEF, dorsal ACCHypoactivation of the FEF and dorsal ACC, reduced functional connectivity between the FEF and dorsal ACC, functional connectivity between the FEF and dorsal ACC correlated with repetitive behaviour scores on the ADI-RKana et al. (2007) [42]HFA (n=12) vs. TD (n=12)19 – 33 yrsfMRI (using an N-back inhibition task)ACCHypoactivation of the ACC, atypical connectivity of the insula with the frontal cortex, reduced synchronization and hypoconnectivity in the frontoparietal networkSchmitz et al. (2006) [43]ASD (n=10) vs. TD (n=12)18 – 52 yrsfMRI (using a Go/ No Go, a spatial Stroop, and a Switch Task)Inferior and orbitofrontal gyrii, insula, parietal cortexHyperactivation of the inferior and orbitofrontal gyrii, hyperactivation of the insula and of the parietal corticesMizuno et al. (2006) [44]HFA (n=8) vs. TD (n=8)15 – 43 yrsfMRI (using a visuo-motor coordination task)Frontal cortex, thalamusHeightened thalamocortical functional connectivityTurner et al. (2006) [45]HFA (n=8) vs. TD (n=8)15 – 43 yrsfMRI (using a visuo-motor coordination task)Frontal, parietal and occipital cortex, caudate nucleusBoth reduced and increased functional connectivity of the caudate nucleus with various regions in the frontal, parietal and occipital corticesBolling et al. (2011) [47]ASD (n=23) vs. TD (n=24)7 – 17 yrsfMRI (using a rule violation task)Frontal cortical regions, insulaAbnormal activation of the insula and DLPFC, hyperconnectivity in frontoinsular networksSoloman et al. (2009) [48]ASD (n=22) vs. TD (n=23)12 – 18 yrsfMRI (using a Preparing to Overcome Prepotency-cy Task)Frontal and parietal corticesReduced integration and hypoconnectivity in frontoparietal networks, hypoconnectivity in frontoparietal networks was correlated with ADHD scores on the CPRS-RUddin et al. (2013) [49]ASD (n=20) vs. TD (n=20)7 – 12 yrsResting-state fMRISalience network (ACC and insula)Hyperconnectivity of the salience network predicted repetitive behaviour scores on the ADI-RClery et al. (2013)[53]HFA (n=12) vs. TD (n=17)mean ASD age 28 yrs (range N/A)fMRI (using a target detection task)ACC, sensory cortexHyperactivity of the ACC, greater functional connectivity of the ACC with sensory cortical regionsShafritz et al. (2008) [54]HFA (n=18) vs. TD (n=15)mean ASD age 22 yrs (range N/A)fMRI (using a target detection task)ACC, IPS, basal gangliaAbnormal activation in the ACC and IPS correlated with repetitive behaviour scores on the ADI-R, hypoactivation of the basal gangliaCasico et al. (2013)[55]ASD (n=19) vs. TD (n=18)mean ASD age 12.5 years (range/N/A)fMRI (using a repetitive behaviour task)insulaHyperactivation of the insula to stimuli of circumscribed interests correlated with CI scores on a parent-report measureDichter et al. (2012)[56]ASD (n=15) vs. TD (n=16)mean ASD age 30 yrs (range N/A)fMRI (using a repetitive behaviour task)VMPFC, NAccHyperactivation of the VMPFC and NAcc to CI stimuliSabatino et al. (2013) [57]ASD (n=13) vs. TD (n=17)16 – 45 yrsfMRI (using a repetitive behaviour task)SFG, insula, caudate nucleusHyperactivation of SFG and insula to social stimuli, hypoactivation of caudate nucleus to CI stimuliDelmonte et al. (2013) [59]ASD (n=28) vs. TD (n=27)mean ASD age 17 yrs (range N/A)Resting state fMRIMFG, caudate nucleusResting state functional connectivity of the MFG and caudate nucleus correlated with repetitive behaviour scores on the ADI-RMonk et al. (2009) [63]ASD (n=12) vs. TD (n=12)mean ASD age 26 yrs (range N/A)Resting state fMRIDMNAbnormal resting state functional connectivity between structures in the DMN correlated with RRB scores on the ADI-RWeng et al. (2010) [64]ASD (n=16) vs. TD (n=15)13 – 18 yrsResting state fMRIDMNAbnormal resting state functional connectivity between structures in the DMN correlated with RRB scores on the ADI-RThomas et al. (2011)[66]HFA (n=12) vs. TD (n=18)20 – 49 yrsDTIForceps minorReduced volume of the forceps minor correlated with repetitive behaviour scores on the ADI-RThakkar et al. (2008)[67]ASD (n=12) vs. TD (n=12)mean ASD age 30 yrs (range N/A)DTIACCReduced integrity of rostral ACC white matter correlated with repetitive behaviour scores on the ADI-RTable 1. Legend: HFA – high functioning autism, TD – typically developing controls, ADI-R – Autism Diagnostic Interview Revised, CI – circumscribed interests, CPT – Continuous Performance Task, ADHD – Attention Deficit Hyperactivity Disorder, CPRS-R – Connor’s Parent Rating Scale – Revised, DMN – Default Mode Network, OFC – orbitofrontal cortex, DLPFC – dorsolateral prefrontal cortex, VMPFC – ventromedial prefrontal cortex, SFG – superior frontal gyrus, MFG – medial frontal gyrus, ACC – anterior cingulate cortex, IPS – intraparietal sulcus, NAcc – nucleus accumbens-571500-22860000-5480050120650Structural Abnormalities in Frontal and Striatal Regions00Structural Abnormalities in Frontal and Striatal RegionsFigure 1. A Frontal and Striatal Developmental Trajectory of RRB in ASD From childhood to adulthood, the development of RRB is correlated with several structural abnormalities in frontal and striatal regions. During childhood, enlargement of the entire frontal cortex is correlated with atypical exploration of objects in the environment [19]. Also during childhood, abnormal growth rate of striatal structures (i.e., caudate nucleus and putamen) is correlated with insistence on sameness [9] (top). During older childhood and adolescence, abnormal thalamic volume is correlated with self-injurious behaviour [10] and enlarged caudate nucleus volume is correlated with measures of impulsivity [31] (middle). In adulthood, abnormal caudate nucleus [11-12] and putamen [11] volumes are correlated with perseverative and compulsive behaviour, and orbitofrontal cortex enlargement is correlated with circumscribed interests [20] (bottom).Figure 2. A neurocognitive model of RRB in ASD. Abnormalities in the cognitive control, motor control, attention, salience attribution, and reward processing networks, are identified during tasks of executive function (EF) employed to study RRB. Moreover, the majority of these network abnormalities appear as dimensional. Several EF tasks reveal abnormal activation in the same neural network. Adapted from findings in [13, 41-45, 47-48, 54-57]. Legend: RRB – Restricted and Repetitive Behaviours. EF Tasks – Executive Function Tasks. CI-Specific – Circumscribed Interest-Specific.AcknowledgementsThe authors would like to thank Samantha Daniel for her generous contributions.ReferencesAmerican Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Washington, DC: Author 2013. M, Kas M, Staal W, vanEngeland H, Durston S. The neurobiology of repetitive behavior: of mice….Neurosci Biobehav Rev 2010; 35 (Pt 3): 345 – 355. M, Kim S. The pathophysiology of restricted and repetitive behavior. J Neurodev Disord 2009; 1: 114 – 132. KSL, Bodfish JW, Piven J. Evidence for three subtypes of repetitive behavior in autism that differ in familiality and association with other symptoms. The J Child Psychol Psychiatry 2008; 49 (Pt 11): 1193 – 1200. ML, Shao Y, Grubber J, Slifer M, Wolpert CM, Donnelly SL, et al. Factor analysis of restricted and repetitive behaviors in autism using the Autism Diagnostic Interview-R. Child Psychiatry Hum Dev 2003; 34: 3 – 17. Szatmari P, Georgiades S, Bryson S, Zwaigenbaum L, Roberts W, Mahoney W, et al. Investigating the structure of the restricted and repetitive behaviors and interests domain of autism. J Child Psychol and Psychiatry 2006; 47 (Pt 6): 582 – 590. MH, Tanimura Y, Lee LW, Bodfish J. Animal models of restricted repetitive behavior in autism. Behav Brain Res 2007; 176: 66 – 74. MA. Annotation: repetitive behavior in autism: A review of psychological research. J Child Psychol Psychiatry 1999; 40(Pt 6): 839 – 849. Langen M, Bos D, Noordermeer SD, Nederveen H, van Engeland H, Durston S. Changes in the development of the striatum are involved in repetitive behaviors in autism. Biol Psychiatry 2013; 75(Pt 5), 405 – 411. 10.1016/j.biopsych.2013.08.013Duerden EG, Card D, Roberts W, Mak-Fan KM, Chakravarty M, Lerch JP, et al. Self-injurious behaviors are associated with alterations in the somatosensory system in children with autism spectrum disorder. Brain Struct Funct 2013; 219 (Pt 4), 1251 – 1261. E, Anagnostou E, Chaplin W, Esposito K, Haznedar M, Licalzi E, et al. Striatal volumes on magnetic resonance imaging and repetitive behaviors in autism. Biol Psychiatry 2005; 58(Pt 3): 226 – 232. LL, Vest C, Mohammed S, Bailey J, Rason BJ, Piven J. An MRI study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol Psychiatry 1999; 23 (Pt 4): 613 – 624. (99)00020-2Goldberg MC, Spinellia S, Joela S, Pekara JJ, Dencklaa MB, Mostofskya SH. Children with high functioning autism show increased prefrontal and temporal cortex activity during error monitoring. Dev Cogn Neuroscience 2011; 1(Pt 1): 47-56. E, Taylor M. Review of neuroimaging in autism spectrum disorders: what have we learned and where do we go from here. Mol Autism 2011; 2(Pt 1): 4. MA, Cherkassky VL, Keller TA, Kana RK, Minshew NJ. Functional and anatomical cortical underconnectivity in autism: Evidence from an FMRI study of an executive function task and corpus callosum morphometry. Cereb Cortex 2007; 17(Pt 4): 951–961.Lopez BR, Lincoln AJ, Ozonoff S, Lai Z. Examining the relationship between executive functions and restricted, repetitive symptoms of autistic disorder. J Autism Dev Disord 2005; 35: 445–460. Mosconi MW, Kay M, D’Cruz AM, Seidenfeld A, Guter S, Standford LD, et al. Impaired inhibitory control is associated with higher-order repetitive behaviors in autism spectrum disorders. Psychol Med 2009; 39(Pt 9): 1559 – 1566. M, Ozonoff S , McMahon WM. The relationship between executive functioning, central coherence and repetitive behaviors in the high-functioning autism spectrum. Autism 2007; 11: 437. K, Courchesne E. Evidence for a cerebellar role in reduced exploration and stereotyped behavior in autism. Biol Psychiatry 2001; 49: 655-664. (00)01008-8Hardan AY, Girgis RR, Lacerda ALT, Yorbik O, Kilpatrick M, Keshavan M, et al. Magnetic resonance imaging study of the orbitofrontal cortex in autism. J Child Neurol 2005; 21: 866 – 871. B, McConachie H, Meins E, Fernyhough C, Couteur AL, Turner M, et al. The frequency of restricted and repetitive behaviors in a community sample of 15-month-old-infants. J Dev Behav Pediatr 2010; 3: 223-229. G, Pasca SP. Motor abnormalities as a putative endophenotype for autism spectrum disorders. Front Integr Neurosci 2013; 7(Pt 43): 1-5. CW, Dierker D, Mostafavi I, Schumann CM, Rivera SM, Amaral DG, et al. Cortical folding abnormalities in autism revealed by surface-based morphometry. J Neurosci 2007; 27(Pt 43): 11725 – 11735. S, O’Briende LM, Filipekf PA, Rapina I, Herbertgh MR. Motor stereotypies and volumetric brain alterations in children with autistic disorder. Res Autism Spect Disor 2013; 7(Pt 1): 82-92. BA, Conroy MA, Richmond Mancil G, Nakao T, Alter P. Effects of circumscribed interests on the social behaviors of children with autism spectrum disorders. J Autism Dev Disord 2007; 37: 1555 – 1561Richler J, Bishop SL, Kleinke J, Lord C. Restricted and repetitive behaviors in young children with autism spectrum disorders. J Autism Dev Disord 2007; 37: 73-85.Estes A, Shaw DWW, Sparks BF, Friedman S, Giedd JN, Dawson G, et al. Basal ganglia morphometry and repetitive behavior in young children with autism spectrum disorder. Autism Res 2011; 4(Pt 3): 212-220. M, Schnack HG, Nederveen H, Bos D, Lahuis BE, de Jonge MV, et al. Changes in the developmental trajectories of striatum in autism. Biol Psychiatry 2009; 66: 327-333. M, Durston S, Staal WG, Palmen SJMC, vanEngeland H. Caudate nucleus is enlarged in high-functioning medication-na?ve subjects with autism. Biol Psychiatry 2007; 62(Pt 3): 262 – 266. MR, Ziegler DA, Deutsch CK, O’Brien LM, Lange N, Bakardjiev A, et al. Dissociations of cerebral cortex, subcortical and white matter volumes in autistic boys. Brain 2003; 126(Pt 5): 1182 – 1192. G, Bates M, Buckman J, Pandina G, Hendren R. Caudate nucleus volume and cognitive performance: Are they related in childhood psychopathology? Biol Psychiatry 2006; 60(Pt 9): 942-950. M, Ozonoff S, McMahon WM. Repetitive behavior profiles in Asperger syndrome and high functioning autism. J Autism Dev Disord 2005; 35(Pt 2): 145- 158. Kringelbach ML, Rolls ET. The functional neuroanatomy of the human orbitofrontal cortex: evidence from neuroimaging and neuropsychology. Prog Neurobiol 2004; 72(Pt 5): 341-372. DW, Lewis MD, Lobust E. The role of the orbitofrontal cortex in normally developing compulsive-like behaviors and obsessive compulsive disorder. Brain Cogn 2004; 55(Pt 1): 220-234.(03)00274-4Vink M, Kahn RS, Raemaekers M, van den Heuvel M, Boersma M, Ramsey NF. Function of striatum beyond inhibition and execution of motor responses. Hum Brain Mapp 2005; 25: 336 – 344. BB, Vink M. On the role of the striatum in response inhibition. PLoS ONE 2010; 5(Pt 11): 1384. BW, Delgado MR, Hikosaka, O. The role of the dorsal striatum in reward and decision-making. J Neurosci 2007; 27(Pt 31): 8161-8165. JR. Why do we have a caudate nucleus? Acta Neurobiol Exp (WARS) 2010; 70: 95-105.Hagmann P, Cammoun L, Gigandet X, Meuli R, Honey CJ, Wedeen Van J et al. Mapping the structural core of human cerebral cortex. PLoS Biol 2008; 6(Pt 7): e159. X, Rao H. Progress in functional connectivity analysis. Progress in Biochemistry and Biophysics 2007; 1: 34-35.Agam Y, Joseph RM, Barton JJS, Manoach DS. Reduced cognitive control of response inhibition by the anterior cingulate cortex in autism spectrum disorders. NeuroImage 2010; 52: 336-347. RK, Keller TA, Minsher NJ, Just MA. Inhibitory control in high-functioning autism: Decreased activation and underconnectivity in inhibition networks. Biol Psychiatry 2007; 62(Pt 3): 198 – 206. N, Rubia K, Daly E, Smith A, Williams S, Murphy DGM. Neural correlates of executive function in autistic spectrum disorders. Biol Psychiatry 2006; 59: 7-16. A, Villalobosa ME, Daviesa MM, Dahla BC, Muller R. Partially enhanced thalamocortical functional connectivity in autism. Brain Res 2006; 1104: 160-174. KC, Frost L, Linsencardt D, McIlroy JR, Muller R. Atypical diffuse functional connectivity between caudate nuclei and cerebral cortex in autism. Behav Brain Funct 2006; 2(Pt 1): 34. BA, McBee M, Holtzclaw T, Baranek GT, Bodfish JW. Relationships among repetitive behaviors, sensory features, and executive functions in high functioning autism. Res Autism Spect Disord 2009; 3(Pt 4): 959-966. Bolling DZ, Pitskel NB, Deen B, Crowley MJ, McPartland JC, Kaiser MD, et al. Enhanced neural responses to rule violation in children with autism: A comparison to social exclusion. Dev Cogn Neurosci 2011; 1(Pt 3): 280-294. M, Ozonoff SJ, Ursu S, Ravizza S, Cummings N, Ly S, Carter CS. The neural substrates of cognitive control deficits in autism spectrum disorder. Neuropsychologia 2009; 47: 2515-2526. LQ, Supekar K, Lynch CJ, Khouzam A, Phillips J, Feinstein C, et al. Salience network-based classification and prediction of symptom severity in children with autism. JAMA Psychiatry 2013; 1 – 11. MP, Stein MB. An insular view of anxiety. Biol Psychiatry 2006; 60: 383 – 387. R. The frontoparietal attention network of the human brain: action, saliency, and a priority map of the environment. Neuroscientist 2012; 18(Pt 5): 502 – 515. J. Attentional network deficits in autism spectrum disorders. In: Buxbaum, JD and Hof PR, editors. The Neuroscience of Autism Spectrum Disorders. Elsevier Inc 2013, p. 281-288. Clery H, Andersson F, Bonnet-Brilhault F, Philippe A, Wicker B, Gornot M. fMRI investigation of visual change detection in adults with autism. Neuroimage Clin 2013; 2: 303-312. KM, Dichter GS, Baranek GT, Belger A. The neural circuitry mediating shifts in behavioral response and cognitive set in autism. Biol Psychiatry 2008; 63(Pt 10): 974-980. JC, Foss-Feig JH, Heacock J, Schauder KB, Loring WA, Rogers BP, et al. Affective neural response to restricted interests in autism spectrum disorders. J Child Psychol Psychiatry 2013; 55(Pt 2): 162-171. GS, Felder JN, Green SR, Rittenberg AM, Sasson NJ, Bodfish JW. Reward circuitry function in autism spectrum disorders. SCAN 2012; 7: 160 – 172. A, Rittenberg A, Sasson NJ, Turner-Brown L, Bodfish JW, Dichter GS. Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord 2013; 43(Pt 12): 2903 – 2913. E, Townsend J, Akshoomoff NA, Saitoh O, Yeung-Courchesne R, Lincoln AJ et al. Impairment in shifting attention in autistic and cerebellar patients. Behav Neurosci 1994; 108: 848 – 856. S, Gallagher L, O’Hanlon E, McGrath J, Balsters JH. Functional and structural connectivity of frontostriatal circuitry in autism spectrum disorder. Front Hum Neurosci 2013; 7 (Pt 430): 1-14. MJ, Jung TP, Makeig S, Brown G, Kindermann SS, Lee T, Sejnowski TJ. Spatially independent activity patterns in functional MRI data during the Stroop color- naming task. Proc Natl Acad Sci U S A 1998; 95: 803-810. Boyd BA, McDonough SG, Bodfish JW. Evidence-based behavioral interventions for repetitive behaviors in autism. J Autism Dev Disord 2012; 42: 1236 – 1248. BB, Mennes M, Zuo X, Gohel S, Kelly C, Smith SM, et al. Toward discovery science of human brain function. PNAS 2010; 107(Pt 10): 4734 – 4739. CS, Peltier SJ, Wiggins JL, Weng S, Carrasco M, Risi S, Lord C. Abnormalities of intrinsic functional connectivity in autism spectrum disorders. Neuroimage 2009; 47: 764-772. SJ, Wiggins JL, Peltier SJ, Carrasco M, Risi S, Lord C. Alterations of resting state functional connectivity in the default network in adolescents with autism spectrum disorders. Brain Res 2010; 1313: 202 – 214. B, Kaufmann WE, van Zijl PC, Fredericksen K, Pearlson GD, Solaiyappan M et al. Diffusion tensor imaging and axonal tracking in the human brainstem. Neuroimage 2001; 14: 723 – 735. C, Humphreys K, Jung K, Minshew N, Behrmann M. The anatomy of the callosal and visual association pathways in high-functioning autism: a DTI tractography study. Cortex 2011; 47(Pt 7): 863-873. KN, Polli FE, Joseph R, Tuch D, Hadjikhani N, Barton J, et al. Response monitoring, repetitive behavior and anterior cingulate abnormalities in autism spectrum disorders. Brain 2008; I3I: 2464 – 2478. B, Barbas H. Changes in prefrontal axons may disrupt the network in autism. J Neurosci 2010; 30(Pt 44): 14595 – 14609. M, Leemans A, Johnston P, Ecker C, Daly E, Murphy CM, et al. Frontal striatal circuitry and inhibitory control in autism: Findings from diffusion tensor imaging tractography. Cortex 2012; 48: 183 – 193. 3Eye Tracking Effort Expenditure and Autonomic Arousal to Social and Circumscribed Interest Stimuli in Autism Spectrum DisorderJenna M. Traynor 1, Alex Gough 1, Eric Duku 2, David I. Shore 3, and Geoffrey B.C. Hall 31 PhD Candidate, Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, Ontario, Canada2 Assistant Professor, Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada3 Associate Professor, Department of Psychology, Neuroscience, & Behaviour, McMaster University, Hamilton, Ontario, CanadaTraynor, J., Gough, A., Duku, E., Shore, D.I., & Hall, G.B.C. Eye tracking effort expenditure and autonomic arousal to social and circumscribed interest stimuli in autism spectrum disorder. Revisions submitted to Journal of Autism and Developmental Disorders, July 2018.Abstract and Key WordsThe social communicative deficits and repetitive behaviours seen in Autism Spectrum Disorder (ASD) may be affected by altered stimulus salience and reward attribution. The present study used eye tracking and a behavioural measure to index effort expenditure, arousal, and attention, during viewing of images depicting social scenes and subject-specific circumscribed interests in a group of 10 adults with ASD (mean age 25.4 years) and 19 typically-developing controls (mean age 20.7 years) Split-plot and one-way repeated measures ANOVAs were used to explore results. A significant difference between the ASD and control group was found in the amount of effort expended to view social and circumscribed images. The ASD group also displayed significant differences in pupillary response to social and circumscribed images, indicative of changes in autonomic arousal. Overall, the results support the social motivation hypothesis in ASD (Chevallier et al., 2012) and suggest a role for autonomic arousal in the ASD symptom dyad.KEYWORDS: autism spectrum disorder, repetitive behaviours, circumscribed interests, social communication deficits, eye tracking, pupillometryIntroductionThe Autism Spectrum Disorder (ASD) symptom dyad includes deficits in social communication and repetitive behaviours (American Psychiatric Association, 2013). Recently, a parallel between the etiology of social deficits and repetitive behaviour has been identified (Benning et al., 2016; Cascio et al., 2014; Dichter et al., 2012; Foss-Feig et al., 2016; Pierce et al., 2015; Sabatino et al., 2013; Sasson et al., 2008; 2011; Watson et al., 2015). Specifically, altered salience and reward processing have been implicated in the etiology of both symptoms, but research findings remain varied and difficult to integrate.Social communication symptoms in ASD include deficits in social responding, orientating, attention, and engagement that are often present by the first year of life (Zwaigenbaum et al., 2005). The social motivation hypothesis (Chevallier et al., 2012) suggests that these deficits result from an early impairment in social motivation, which is understood to have a top-down effect on behaviour throughout development, and is hypothesized to originate in neural circuits that subserve salience and reward processing; namely, limbic and frontostriatal circuits. While neurotypicals display well-established and heightened activity in these circuits in response to the presentation of social stimuli (Kampe et al., 2001; O’Doherty et al., 2003; Phillips et al., 1998), early functional abnormalities in these areas are thought to result in reduced motivation to socially engage in ASD. The broad range of social deficits that result can lower quality of life and preclude the development of meaningful relationships (Kasari & Patterson, 2012).Repetitive behaviour in ASD is commonly divided into two broad categories: “lower-order” behaviour, which includes more rudimentary behaviour such as repetitive motor behaviour, vocal stereotypies, and preoccupations with parts of objects, and “higher-order” behaviour which is more cognitively-oriented, and includes insistence on sameness in routine, and circumscribed interests (CIs) (Turner, 1999). Animal and human studies have presented strong evidence to support the role of altered salience and reward-related, frontostriatal function in the etiology of repetitive behaviour, specifically in the orbitofrontal cortex, anterior cingulate cortex, caudate nuclei, putamen, and globus pallidus (Hardan et al., 2006; Hollander et al., 2005; Langen et al., 2011; Lewis et al., 2007; Sears et al., 1999). Functional magnetic resonance imaging (fMRI) studies have also identified significant correlations between repetitive behaviour and activation in these areas (Uddin et al., 2013; Shafritz et al., 2008). Recently, fMRI and eye tracking studies have investigated salience and reward processing in the ASD symptom dyad through an examination of CIs. CIs encompass a broad range of heterogeneous and intense interests that are prevalent across the autism spectrum, are non-social in nature (i.e., mechanical objects), and due to their intensity, can interfere with important areas of functioning, such as socialization (Turner-Brown et al., 2011). However, it should be noted that CIs in ASD are not always conceptualized as functionally interfering behaviour (Mercier et al., 2000), as they can also result in significantly increased knowledge and expertise in a certain area (e.g., vehicle mechanics), and be considered an area of strength.To date, studies have compared neural and behavioural responses to the presentation of both social stimuli and high autism interest (HAI) stimuli. HAI stimuli include pictures of planes, trains, fans, and other objects commonly identified by individuals with ASD as CIs (Turner-Brown et al., 2011). However, given the heterogeneity in CIs across the autism spectrum, it has been suggested that not all ASD subjects in any given sample are strongly interested in all stimuli within a given HAI stimuli set, thereby decreasing the validity of these paradigms (Parsons et al., 2016). As such, paradigms that use individual, subject-specific CI stimuli have been pared to social stimuli, the presentation of HAI stimuli sets elicit greater affective-, salience-, and reward-related brain activation in individuals with ASD (Dichter et al., 2012; Kohls et al., 2018; Sabatino et al., 2013), as well as greater motivation-related, event-related potential (Benning et al., 2016), compared to controls. Relatedly, it has been shown that individuals with ASD are willing to receive less of a financial reward in order to view images related to restricted interests, compared to control subjects, indicating a higher reward value of interest images in ASD (Watson et al., 2015). Additionally, results from individualized paradigms have shown increased activation to subject-specific interest images in ASD subjects compared to controls, in affective regions (Cascio et al., 2014), and in the fusiform gyrus, an expertise-related area of the brain commonly activated to faces in neurotypicals (Foss-Feig et al., 2016). Altered attention and arousal may also play a role in the ASD symptom dyad. In particular, with reference to attention, differences in attentional gaze to social and HAI stimuli have been clearly identified, with most studies showing that ASD subjects demonstrate preferential attention to non-social and HAI stimuli across development (Pierce et al., 2015; Sasson et al., 2008; 2011; Unruh et al., 2016). However, counter to this hypothesis, the presentation of subject-specific CI stimuli did not interfere with performance on a selective attention task in ASD subjects (Parsons et al., 2016). As such, more studies using subject-specific interest stimuli are needed to understand the role that attention plays in the ASD symptom dyad. Additionally, many eye tracking studies measuring pupillary response, a well-known marker of autonomic arousal (Bradley et al., 2008; Hess & Polt, 1960) and reward (Bijleveld et al., 2009), have demonstrated decreased arousal and reward to social stimuli in ASD. For example, when gazing at happy faces, which are intrinsically rewarding for neurotypicals (Kampe et al., 2001; O’Doherty et al., 2003; Phillips et al., 1998), ASD subjects do not display increased pupillary size as do their control counterparts (Sepeta et al., 2014). In fact, ASD subjects can show pupillary restriction to faces (Anderson et al., 2006). However, other studies have found no difference between ASD and control subjects in pupillary response to faces (Falck-Ytter, 2008; Sabatino DeCriscio et al., 2016), but instead, have found increased pupil size to inverted faces in ASD only (Falck-Ytter, 2008), or direct attentional gaze-mediated pupillary response in control subjects only (Sabatino DeCriscio et al., 2016). Given the established association between motivation and arousal (Eysenck, 1981) and motivation and reward (Wise, 2004), the majority of these findings support the social motivation hypothesis, although to date, pupillary response has only been investigated using social, but not CI stimuli, precluding an understanding of whether the latter may be overvalued in relation to increased arousal. More indirect investigations of the role that arousal plays in CI behaviour have yielded different results. For example, one study found that the presence of CI behaviour in ASD significantly predicted increased willingness to expend effort in order to receive a reward on an effort expenditure task (Damiano et al., 2012). However, another found that although ASD persons rate HAI stimuli as more subjectively pleasing, they do not rate HAI as more arousing than social stimuli (Sasson et al., 2012). Moreover, a third study found that individuals belonging to the broader autism phenotype did rate HAI as more subjectively arousing, compared to a phenotype negative group (Morrison et al., 2018). Thus, a more direct and objective measure of arousal in CI is needed.As such, the primary purpose of the present study was to objectively quantify autonomic arousal to both social and CI stimuli, using pupillary response. A measure of blink rate was also collected to index attention (Shultz et al., 2011). As previous work has identified a positive association between effort expenditure and CIs, this study also included an objective measure of effort. Finally, following recent imaging and eye tracking studies (Cascio et al., 2014; Parsons et al., 2016), this study used subject-specific CI stimuli. This work is important for several reasons. First, and to our knowledge, this is the first study to use pupillometry to simultaneously measure responses to the presentation of subject-specific CI and social stimuli, which will allow for a more objective and thorough examination of the role of arousal and reward in the ASD symptom dyad. Second, the inclusion of multiple behavioural measures, including attention, arousal, and effort, will permit a more comprehensive analysis, which may help to interpret the discrepant results found to date, and provide evidence to support or challenge the social motivation hypothesis (Chevallier et al., 2012). Third, as strong interests are common in neurotypicals, this study included a control group with strong interests, in order to identify differences between interests in neurotypical and ASD persons, a sparsely investigated area to date (Gutermuth Anthony et al., 2013).MethodsParticipantsTable 1 depicts demographic information for all subjects (10 ASD and 19 typically-developing controls). All participants were between 16 and 35 years of age with normal-to-corrected vision. Exclusion criteria included previous head-injury causing loss of consciousness, use of contact lenses, obsessive-compulsive disorder (OCD), or a score above the clinical cut-off of 53 on the Padua Inventory OCD self-report (Sanavio, 1988). Additionally, control subjects were excluded based on the presence of any psychiatric disorder, a first-degree relative with an ASD diagnosis, or a score above 32 on the Autism Quotient (AQ; Baron-Cohen, 2001). Psychiatric comorbidities were permitted in the ASD group due to their high prevalence in the ASD population (Mazzone et al., 2012); one participant had generalized anxiety disorder and one had attention-deficit-hyperactivity disorder, and these participants took a prescribed antidepressant and stimulant (lisdexamfetamine dimesylate), respectively. ASD participants were recruited from a local center for adults with ASD, had a pre-existing diagnosis given by a licensed practitioner, and provided assessment paperwork prior to participating. A personal support worker or parent/guardian of ASD subjects confirmed the circumscribed nature of their interest. Control participants were recruited from an undergraduate psychology program and had an interest that met a frequency and intensity cut-off score on an author-designed, likert-type scale used to quantify interests in this group (Table 2). All participants provided written consent to participate and a McMaster University research ethics board approved the study. A total of 13 young adults with ASD and 20 control subjects were recruited. Recruitment of ASD subjects posed a challenge, as participants were required to be verbal and have a CI that could be displayed using images; participants with CIs pertaining to factual or statistical information could not participate. One ASD participant was disqualified due to the non-circumscribed nature of their interest, and two others due to incomplete calibration of the eye tracking system. One control subject was excluded due to an elevated Padua Inventory score, which left 19 control and 10 ASD subjects in the final analysis. All interests are listed in Table 3.MeasuresInterest ScaleControl participants filled out an author-designed questionnaire in order to quantify the frequency and intensity of their interest. CIs in ASD are more intense and functionally interfering than non-circumscribed interests (Turner-Brown et al., 2011). The scale was used to create a control group with highly intense, and somewhat interfering interests; subjects were asked about the number of hours they spend per week engaged in their interest and about the level of functional interference introduced by the interest (rated a scale from 0 to 10). Additionally, subjects rated their level of interest and level of excitement about the interest on a scale from 0 to 10. Only subjects with ratings of 7 out of 10 or higher for both interest and excitement were included in the study (Table 2). Wechsler Adult Scale of Intelligence – 2nd Edition (WASI-II)To quantify verbal and non-verbal abilities, all subjects completed the vocabulary and matrix reasoning subtests of the WASI-II (Table 1), a brief measure of intellectual functioning that has been validated in individuals from 6 to 90 years of age and includes individuals with ASD in the normative sample (Wechsler, 2011). The WASI-II provides an estimate of total IQ from these two subtests. Padua Inventory (PI)In order to ensure that interests were not associated with clinically significant obsessive or compulsive behaviours, all subjects completed the PI, a 60-item questionnaire providing a comprehensive, dimensional measure of obsessive and compulsive thoughts and behaviours (Sanavio, 1988). Items are answered on a scale from 0 (not at all) to 4 (very much). The PI has demonstrated good internal consistency, with Chronbach’s alpha for total scores ranging from 0.90 – 0.94 (Sanavio, 1998; Van Oppen, 1992), and good convergent validity with the Symptom Checklist 90-Revised obsessive-compulsive dimension (r = 0.72) and Maudsley Obsessive-Compulsive Inventory (r = 0.74) (Van Oppen, 1992). All subjects included in our sample scored well below the clinical cut-off score of 53, with mean scores of 22.4 (ASD) and 26.4 (TD).Autism Quotient (AQ)To screen for ASD in the control group, participants completed the AQ, a 50-item self-report measure of autistic traits for individuals with at least average intelligence (Baron-Cohen et al., 2001). The majority of nonclinical samples obtain an average score of 17, with a clinical cut-off score of 26 indicating need for further assessment (Ruzich et al., 2015). Our control group obtained a mean score of 16.5, and no control subjects met the clinical cut-off. Our ASD group obtained a mean score of 24.6, which was significantly higher than controls (p= 0.002). Some ASD subjects had below-average IQ, thereby decreasing the validity of their AQ self-report. Eyetracking ParadigmAn Eyelink II head-mounted eye tracking system (SR Research Ltd.) was used to record participants' monocular gaze position and pupil size at a sampling rate of 500 Hz. Participants were seated 60cm from a 17” inch display screen. Calibration took place before the experiment began and during a mid-way break in the experiment, using a five-point calibration method. Two five-minute baseline-recording periods occurred, prior to and immediately after the experiment, which collected baseline measures of blink rate (per second) and pupil size (arbitrary units; SR Research, 2009). These measures were averaged in order to account for eye fatigue across the experiment, producing one mean baseline measure of blink rate and pupil size for each participant, which were used as covariates in the main analyses. During baseline recording, a fixation cross remained on the screen, and participants were instructed to keep their eyes on the screen. Participants were told that they would see images of people and objects on the screen, and that some stimuli would depict their interest and some would not. They were instructed to remain still while viewing the stimuli. The paradigm was built using Experiment Builder (SR Research Ltd). Stimuli were randomly presented over two blocks, and consisted of a total of 40 unsaturated, black-and-white images of each participant’s self-identified interest, 40 social images, and 40 images of neutral objects (e.g., broom, telephone, chair). None of the neutral objects corresponded to any subject’s self-identified interest. Social stimuli included naturalistic scenes depicting social interaction. Example stimuli are depicted in Figure 1. Stimuli presentation was preceded by a fixation cross. All pictures were presented following a 500 ms shape that appeared in the same location as the fixation cross; interest pictures were preceded by a square, social images by a circle, and random images by a diamond, and these shapes served to prime the subject as to the type of image to be subsequently presented. The experiment consisted of a passive and an active block. During the passive block, participants passively viewed the stimuli. This block consisted of 60 trials (20 of each image type), each presented for 5 seconds. At the mid-way point, participants were told that they would again see similar pictures. However, this time the pictures would be masked by a dark gray filter, which would slowly reveal one fifth of the image, and then pause. After this pause, if participants wished to view more of the image, they would need to press the space bar five times as quickly as possible, which would cue the computer to continue revealing the image. This pause-reveal sequence continued until either the full picture was visible after a total of four pauses, or until the participant chose to move on to the next picture by abstaining from pressing the space bar. Participants were told that it was entirely their decision as to how much of any picture they chose to see more of, if any. They were each given three practice trials in order to familiarize themselves with how quickly the space bar needed to be pressed. All participants were able to press the space bar quickly enough to reveal the images during practice trials. Space bar button presses (per trial) served as a measure of effort. The active block consisted of 60 trials (20 of each image type). Active block trials lasted various durations depending on the number of times the participant chose to view additional parts of images by pressing the space bar. AnalysisData were extracted from the Eyelink II system using Data Viewer software (SR Research Ltd.), and analyzed using R64 programming software (R Core Team, 2013). The passive and active blocks were analyzed separately in order to control for potential button by dependent variable interaction effects during the active block.A split plot analysis of variance (ANOVA) was used to analyze the button press data to examine within- and between-group differences across the three conditions (social, neutral, interest). Pupil size (arbitrary camera sensor units; SR Research, 2009) and blink rate data (blink events as defined by the Eyelink II parser, per second) were examined using within-group analyses only. Separate, one-way repeated measures ANOVA models were used for the ASD and control group because baseline pupil size (p = 0.03) and baseline blink rate (p = 0.01) differed significantly between the two groups, and this difference violated the assumption of no significant interaction between the independent variable and the covariate (Maxwell & Delaney, 1990), which precluded a between-group analysis. Due to individual differences in baseline pupil size and baseline blink rate, these were included as variables of no interest in these models. The mean percent change between baseline pupil size and pupil size measured during each trial, averaged across all trials in that condition, was used in the analysis. To account for the between-subjects button press analysis in the active and passive blocks (2 comparisons), and the separate, within-subjects comparisons of pupil size and blink rate data during the active and passive blocks for each group (6 comparisons), a Bonferroni correction was applied to all analyses by dividing alpha = 0.05/8. This conservative correction resulted in an adjusted alpha level of p = 0.006. Violations of the Mauchly Sphericity assumption (p < 0.05) were accounted for with a conservative Greenhouse-Geisser (GG) correction (Girden, 1992). Examination of simple main effects was implemented where appropriate using Tukey’s Honestly Significant Difference (HSD) test, which examined pair-wise comparisons while appropriately controlling for the additional number of comparisons introduced by post-hoc testing (Tukey, 1949). Significantly different patterns of results in the ASD and control group for each dependent variable were expected. Reflecting the overvaluation of non-social stimuli, and especially of interest stimuli in the ASD group, we predicted that this group would display: significantly more button presses, increased pupil size, and decreased blink rate in the interest condition, compared to both the social and neutral conditions, indicative of increased willingness to expend effort (Treadway et al., 2009), arousal (Bradley et al., 2008; Hess & Polt, 1960), and attention (Kaufman & Alm, 2003; Shultz et al., 2011), respectively. For control subjects, an opposite trend was predicted: a higher valuation of social stimuli, compared to both the interest and neutral stimuli, reflected by these same patterns in the social condition, compared to both the interest and neutral conditions. ResultsCorrelationsThere were no significant correlations found between any of the blink rate or pupil size variables (p > 0.05), aside from expected correlations for the same variables across the two blocks (e.g., passive and active blink rate). During the active block, button presses were not significantly correlated with any of the dependent variables (p > 0.05). Although there was a small significant difference in age between the two groups (p = 0.03), there were no significant correlations between age and any of the dependent variables (p > 0.05). There were no significant correlations between IQ and any of the dependent variables, for either the ASD group (p > 0.05), or control group (p > 0.05). Testing for Violations of the Normality and Homogeneity of Variance AssumptionsViolations of the assumptions of normality and homogeneity of variance of the model’s residuals were tested in both the passive and active blocks for each dependent variable using the Shapiro-Wilk normality test (Shapiro & Wilk, 1965) and the Levene’s Test for Homogeneity of Variance (Levene, 1960) (Table 1). The button press data did not violate either of these assumptions (p > 0.05). The blink data did not violate either of these assumptions in the active block (p > 0.05). In the passive block, the blink data violated the assumption of normality of residuals only (p = 0.003), which is not of imminent concern as ANOVA models are generally robust to violations of normality (Schmider et al., 2010). The mean percent change in pupil size from baseline data violated both of these assumptions in the passive and active blocks; the implications of these violations are discussed and a supplementary model is proposed (see Discussion).Button PressesThe number of button presses made in each condition depended on whether the participant belonged to the ASD or control group (F (2,54) = 13.054, p = 0.000, GG corrected for violation of the Sphericity assumption). Subjects with ASD pushed buttons to their interest pictures significantly more than control subjects did (F (1, 27) = 9.209, p = 0.005). Additionally, within the ASD group, significantly more button presses were made during the interest vs. the social condition (z = 8.207, p < 0.001), and during the interest vs. the neutral condition (z = 7.745, p < 0.001). Within the control group, significantly more button presses were made during the interest vs. social condition (z = 4.013, p < 0.001), the interest vs. the neutral condition (z = 7.681, p < 0.001), and during the social vs. neutral condition (z = -3.668, p < 0.001). These data are displayed in Figure 2.Pupil SizeWithin the ASD group, the effect of condition on pupil size during the passive block was not significant after correction for multiple comparisons (F (2, 18) = 5.62, p = 0.013). ASD subjects displayed the largest pupil size during the interest condition, followed by the social condition, and the smallest pupil size during the neutral condition (Figure 3a). During the active block for ASD subjects, a significant effect of condition on pupil size was found (F (2, 15) = 9.823, p = 0.002), and there was no significant interaction between button presses and condition (F (2, 15) = 2.346, p = 0.114). Significantly larger pupil size during the interest vs. the neutral condition (z = 3.967, p < 0.001), and in the interest vs. the social condition (z = 2.463, p = 0.036) was found. There was no significant difference in pupil size between the social and neutral condition (z = -1.504, p = 0.289, Figure 3b). Within the control group, the effect of condition on pupil size during the passive block (F (2, 36) = 5.014, p = 0.012) did not survive correction for multiple comparisons. Control subjects displayed the largest pupil size to social images, followed by interest images, and the smallest pupil size to neutral images (Figure 3c).During the active block for control subjects, there was a significant main effect of condition on pupil size (F (2, 33) = 17.74, p = 0.000), and no significant interaction between button presses and condition (F (2, 33) = 0.698, p = 0.505). Significantly larger pupil size in the social versus the neutral condition (z = -2.976, p = 0.008), and in the interest versus the neutral condition (z = 2.424, p = 0.041) was found. There was no significant difference in pupil size between the interest and social condition (z = -0.552, p = 0.845) for control subjects (Figure 3d).As these results due not explore between-subjects effects (see Analysis for rationale), a supplementary between-subjects model is proposed (see Discussion, Statistical Considerations).Blink RateWithin the ASD group, there was no significant difference in blink rate between the three conditions, in either the passive block (F (2, 18) = 0.212, p = 0.811), or the active block (F (2, 15) = 0.239, p = 0.790). During the passive block, ASD subjects demonstrated the highest blink rate during the neutral condition, and a lower, similar blink rate during the interest and social conditions (Figure 4a). During the active block, ASD subjects displayed a similar blink rate during all three conditions (Figure 4b).Within the control group, the effect of condition on blink rate in the passive block (F (2, 36) = 3.349, p = 0.045) did not survive correction for multiple comparisons. In the active block, a trend toward an effect of condition on blink rate was found after correcting for multiple comparisons (F (2, 33) = 6.004, p = 0.006). Post hoc testing revealed a significantly lower blink rate in the interest vs. the neutral condition only (z= -3.184, p = 0.000). During both blocks, control subjects demonstrated the lowest blink rate during the interest conditions, but their blink rate during the social and neutral conditions varied slightly across the two blocks (Figure 4c and d). As these results do not explore between-subjects effects (see Analysis for rationale) a supplementary between-subjects analysis is proposed (see Discussion, Statistical Considerations).Discussion The current study found significant between-subject differences in the amount of times subjects chose to view more of social and interest pictures, and in the amount of effort expended to view these images. Significant within-subject effects of condition on pupil size while making these choices were found, indicative of changes in autonomic arousal. Importantly, these results differed depending on whether participants belonged to the ASD or control group. Specifically, compared to control subjects, ASD subjects chose to view more of, and expend more effort to view, their subject-specific CI stimuli. They also chose to view more CI stimuli over both social and neutral pictures. Importantly, no differences in the amount of times they chose to view social versus neutral pictures were found. Further, when making these choices, ASD subjects demonstrated larger pupil size to CI images, compared to both social and neutral images, and importantly, no difference in pupil size when viewing social or neutral images. Overall, these results provide support for the social motivation hypothesis (Chevallier et al., 2012), and are in line with our a priori hypotheses indicating increased autonomic arousal and reward in CI in ASD. Importantly, although both groups chose to view more of, and expend more effort to view, their interest images over social images, the control group demonstrated no differences in pupil size when viewing interest and social stimuli; unlike the ASD group, this likely reflects the additional salience and reward attribution to social stimuli in neurotypicals. Additionally, unlike the ASD group, control subjects demonstrated a significant preference for social over neutral images, and this preference was reflected by increased pupil size during viewing of social over neutral images, supporting our a priori hypothesis that social stimuli are highly valued in typically developing individuals.This study also tracked blink rate as an indirect measure of attention and found no differences in blink rate between any of the conditions for ASD subjects. In the control group, trends were identified that suggest greater attention paid to interest images, as measured by overall lower blink rates during interest trials. These data suggest that, counter to our a priori hypothesis, ASD subjects did not preferentially allocate attention to interest images. It is possible that our small ASD sample may have been insufficient to detect differences. It is also possible that blink rate in the current study was confounded by level of arousal; although decreased blink rate to visual stimuli is indicative of increased attention (Kaufman & Alm, 2003), there is work showing that increased blink rate is correlated with higher levels of arousal (De Jong & Merckelbach, 1990). It is therefore possible that if ASD subjects experienced a higher level of arousal to interest images compared to control subjects, as our pupil data indicate, that the effect of attention on blink rate may have been confounded, resulting in no difference in blink rate between the arousing interest condition and the non-arousing social and neutral conditions. However, this is unlikely, as if this were true, a similar confounding effect would have likely been observed in the control group, which was not the case. Additionally, recruitment of a control group with higher than average, intense interests may have influenced the results; it is possible that a high level of attentiveness to interest images in both the ASD and control group minimized any between-group differences in attention to stimuli.Statistical ConsiderationsNotably, after correcting for multiple comparisons, differences in pupil size were only significant when subjects were choosing to view images and expending effort in the active block, although statistical trends paralleling these patterns were identified in the passive block. Although there were no interaction effects found between pupil size and button presses, increased effort expenditure while button pressing likely contributed to increased arousal in participants during the active block. However, if the majority of the variance in the effect of increased pupil size were accounted for by effort expenditure alone, we would have expected to find no differences in pupil size between ASD and control subjects across the conditions (as both groups demonstrated a similar pattern of effort expenditure). This was not the case; although both groups chose to expend significantly more effort to view interest over social images, increased pupil size to interest over social images was only found in the ASD group. Therefore, the effect of increased pupil size may reflect the combination of increased autonomic arousal due to effort expenditure and increased reward attribution to CI stimuli over social stimuli in the ASD group. Further, the implications of the significant within-subjects effects found for the pupil data should be explored and validated using a between-subjects model, but due to significant group differences in the baseline pupil size covariate, which violated the assumptions of our ANOVA model (Maxwell & Delaney, 1990), a statistically sound between-subjects analysis was precluded. This is however, an important area to address as the current results do not provide information about between-subjects effects. As such, an exploratory, split plot between-subjects ANOVA is presented (Supplementary Material 1.0), which was analyzed without including the baseline pupil size covariate. We explore this model in a supplementary section as baseline pupil size is a meaningful covariate, particularly due to the existing differences in baseline pupil size between groups, and excluding this covariate removes intrinsic information about our ASD group. For example, previous work has also found significantly increased baseline pupil size in ASD, relative to neurotypicals (Anderson & Colombo, 2009), which is thought to reflect overall increased arousal of the autonomic nervous system in ASD, relative to controls (Hirstein et al., 2001; Ming et al., 2005; Zahn et al., 1987). Nevertheless, our supplementary model exclusive of this covariate offers a preliminary investigation of between-subjects effects, and returned a statistical trend similar to the pattern of results in the main analysis. Specifically, relative to controls, ASD subjects displayed larger pupil size during the interest condition, compared to the social condition (p = 0.081). It is also pertinent to note that compared to the within-subjects analysis in the main body of the paper, this between-subjects analysis was underpowered (Maxwell & Delaney, 1990), and especially due to our small sample size. Note that a between-subjects analysis of blink rate is also included in Supplementary Material 1.0. Similarly, to satisfy the assumptions of the ANOVA, this model did not include the baseline blink rate covariate. This model returned essentially null results, similar to our within-subjects analysis.It is also worth considering that the repeated measures ANOVA model used to analyze the pupil data in the current study assumes a normal distribution and homogeneity of variance of the model’s residuals. These assumptions should be taken into account, as the dependent variable of mean percent change in pupil size from baseline in the current study violated both of these assumptions, but was used as per common practice of using a percent change from baseline measure in pupillometry, and importantly, to adequately account for the arbitrary nature of the pupil size measurements collected by a video-based eye tracker (i.e. that neither actual physical pupil size nor the physical change in pupil size can be determined from the pupil size measurement returned by the system alone). As such, it is worth considering how violations of these assumptions may have impacted the experimental results. ANOVA models are generally robust to violations of the normality assumption (Schmider et al., 2010). However, violations of the assumption of homogeneity of variance are of greater concern, especially when such violations are observed in models with unequal sample sizes, as is observed the current study (10 ASD, 19 TD). In this case, there is rationale to suspect that the risk of making a Type I error is inflated (Maxwell & Delaney, 1990). To investigate this further, a repeated measures ANOVA model using raw, observed pupil size averaged across all trials in each condition was created (Supplementary Material 2.0). Compared to the percent change pupil data used in the main analysis, these observed pupil size data more adequately satisfied the assumptions of the ANOVA model; violations of normality were significantly smaller and the more meaningful assumption of homogeneity of variance was not violated (Table 4, Supplementary Material 2.0). This model did not account for the arbitrary nature of the pupil size measurements by using a percent change from baseline calculation, but instead, only included baseline pupil size in the model as a covariate of no interest. In other words, this model adequately controlled for the effect of individual differences in baseline pupil size on the dependent variable, but did not completely account for the arbitrary nature of the pupil data (e.g. the particular pupil size measurements obtained could have been influenced by factors irrelevant to both the participants and experimental conditions of interest to our analyses, such as eye-camera distance, camera-eye angle, lighting conditions, etc.). The pattern of results from this analysis was virtually identical to the results of the model that used percent change in pupil size, except the supplementary ANOVA model returned p-values of slightly less significance (albeit still meeting the alpha threshold of significance), possibly reflecting a small inflation of significance in the main pupil analysis due to the aforementioned violations (Supplementary Figure 5). As an identical pattern of results was observed, this supplementary analysis supports the validity of our main results.LimitationsFirst and foremost, our individualized study design limited the number of eligible participants included in the final analysis (10 ASD, 19 controls), which limits the stability of the discovered effects. Additionally, due to statistical limitations, we explored the pupil and blink data in our main analysis using within-subjects ANOVAs only. Although the preliminary, between-subjects model offered in our supplementary section demonstrated a trend toward a significant pattern of results, it should be noted that these results did not reach significance. Although it is likely that this analysis was significantly underpowered (Maxwell & Delaney, 1990), it is also possible that a true between-subjects effect does not exist, and it would be helpful to replicate these results in the future using a statistically appropriate between-subjects model and a larger sample size.Additionally, although some estimates suggest that up to 88% of children with ASD have a CI (Mercier et al., 2000), our study used an adult sample of high-functioning individuals with intense CIs. The conclusions pertaining to reward processing in social and repetitive behaviour that are drawn from the current study are only generalizable to high-functioning individuals with CIs, particularly due to the heterogeneity in repetitive behaviour across the autism spectrum (Turner-Brown et al., 2011). Future work should focus on investigating this relationship in lower-functioning individuals.Finally, the current study used an author-designed, self-report measure of control participants’ interests, which is a less valid measure of restricted interests compared to existing, well-validated interest measures (e.g., Bodfish, 2003; South et al., 1999). However, an author-designed scale was chosen over existing measures, as the latter are used to describe common CIs in ASD or require a caregiver report of childhood behaviour, and were not appropriate for the quantification of interests in a neurotypical, adult sample. As we used a subjective measure of participant interests, the risk of over- or underreporting should be acknowledged, although this risk also exists in many validated self-report adult assessments (e.g., The Ritvo Autism Asperger Diagnostic Scale – Revised; Ritvo et al., 2011, and the AQ; Baron-Cohen et al., 2001). Further, it is perhaps more likely that participants would have underreported any functional interference of their interest, due to a social desirability bias (Edwards, 1953), and in this case, the effects in the current study would be a conservative estimate of existing differences between individuals with ASD and the general population. Conclusions and Future DirectionsThe current study posits that abnormal reward processing may contribute to the development of both social amotivation and restricted interests in ASD, and provides objective evidence to suggest that altered arousal and effort expenditure may be mechanisms by which these abnormal processes are supported and/or maintained. A strength of the current study is that it included idiosyncratic interest stimuli, in addition to social stimuli and neutral stimuli depicting images of random, non-social objects (i.e., umbrella, table). Overall, we found decreased arousal to social images in ASD, relative to images of restricted interests, but not relative to the neutral, non-social images (which produced virtually the same level of arousal in our ASD group as the social stimuli did). Therefore, our findings demonstrate the specificity of altered arousal and effort expenditure mechanisms to circumscribed interests specifically, and to social deficits more broadly. These findings are in line with previous studies that have found increased attention (Pierce et al., 2015; Sasson et al., 2008; 2011) and reward- and motivation-related neural function (Benning et al., 2016; Cascio et al., 2014; Dichter et al., 2012; Sabatino et al., 2013) to mechanical, non-social stimuli and HAI stimuli, relative to social stimuli.Moving forward, given evidence of increased autonomic arousal in ASD at rest (Hirstein et al., 2001; Ming et al., 2005; Zahn et al., 1987), the developmental trajectory of the current effects should be explored; a primary social motivation deficit and increased reward value of restricted interests may interact over time, and without intervention, result in more prominent non-social behaviour across development in ASD. Future study of the developmental trajectory of arousal in the ASD symptom dyad may allow for the identification of critical developmental periods in which intervention may be most effective to support pro-social behaviour. At present, the current findings support social skills interventions for children with ASD that focus on incorporating CIs into treatment in order to reinforce social communicative behaviour (Baker et al., 1998; 2000; Boyd et al., 2007). Overall, findings from our button press and pupil size data support the social motivation hypothesis. As this is the first study to provide a simultaneous, objective measure of autonomic arousal to the presentation of social and subject-specific interest stimuli, our data provide incremental evidence by suggesting that disproportionate effort expenditure and autonomic arousal play a role in social and non-social, repetitive behaviour in ASD. Additionally, as we recruited a control group with higher than average, intense interests, the effects found in the current study may be a conservative estimate of the true differences that exist between individuals with ASD and the general population. Tables and FiguresTable 1. Demographic information and descriptive statistics for variables of no interest (top) and variables of interest (bottom). VARIABLE OF NO INTERESTASD mean, (sd), rangeCONTROLmean, (sd), rangeT-value, p-valueIQ score (raw)76.5 (15.3), 39 - 9486.6 (10.1), 64 - 100T = 1.86, p = 0.09Age25.4 (5.25), 16 - 3420.7 (2.42), 18 - 28T= 2.53, p = 0.03Sex (m, f)8, 2 10, 9Fisher’s Exact p = 0.23Baseline Blink 0.30 (0.26), 0.027 – 0.8830.50 (0.175), 0.113 – 0.843T = -2.37, p = 0.01Baseline Pupil 882.319 (420.627), 371.660 – 1594.350631.835 (240.447), 337.340 – 1277.685T = 1.97, p = 0.03 VARIABLE OF INTEREST ASD mean (sd) conditionCONTROL mean (sd) conditionShapiro-Wilk Normality Test (W, p-value)Levene’s Test of Homogeneity of Variance F (1, 27), p-valueButton presses: active2.03 (0.83) social2.16 (0.58) neutral4.26 (0.62) interest2.55 (0.92) social1.93 (0.78) neutral3.22(0.94) interest W = 0.990, p = 0.731F = 0.171, p = 0.68 (social)F = 0.516, p = 0.47 (neutral)F = 1.369, p = 0.25 (interest)Blink rate per sec: passive0.35 (0.22) social0.38 (0.24) neutral0.35 (0.17) interest0.71 (0.37) social0.70 (0.37) neutral0.66 (0.38) interestW = 0.952,p = 0.003F = 1.927, p = 0.18 (social)F= 2.101, p = 0.16 (neutral)F= 1.979, p = 0.17 (interest)Blink rate per sec: active 0.21 (0.08) social0.21 (0.08) neutral0.20 (0.09) interest0.42 (0.18) social0.44 (0.16) neutral0.39 (0.16) interestW = 0.978, p = 0.150F = 4.208, p = 0.05 (social)F = 1.190, p = 0.29 (neutral)F= 3.556, p = 0.07 (interest)Percent change in pupil size: passive 0.20 (0.34) social 0.16 (0.31) neutral 0.22 (0.33) interest 0.09 (0.16) social 0.05(0.14) neutral 0.08 (0.15) interestW = 0.917,p = 0.002F = 4.528, p = 0.044 (social)F = 3.942, p = 0.057 (neutral)F = 5.307, p = 0.029 (interest)Percent change in pupil size: active 0.37 (0.39) social 0.34 (0.38) neutral 0.43 (0.41) interest 0.13 (0.16) social 0.05 (0.17) neutral 0.14 (0.17) interestW = 0.933,p = 0.0002F = 4.619, p = 0.041 (social)F = 5.060, p = 0.032 (neutral)F = 6.965, p = 0.014 (interest)Table 2. Descriptive statistics for control participants’ self-identified interestsInterest (/10)Excitement (/10)Length of interest (yrs)Time spent on interest (hrs/week)Level of functional interference (/10)Mean (SD) range8.94 (0.94), 7 - 108.88 (1.07),7 - 107.71 (2.28),4 - 105.79 (2.43),1 - 103.88 (2.55),0 - 8Table 3. List of participant interests.ASD InterestsTD Interestsgrand theft auto video gamerome total war video gamecamping trailerscosmeticsCanadian landscapesbirdscosplay characters Evangelion and Madoka Magicanime characters yveltal and natural harmonia metal armor, plate mail, and shields for larpingallied forces war planes league of legends video gameworld of warcraft video gameNHL video games bollywood dancepianodrumsweight liftingdancecheerleadingsoftballbasketballlacrossehockeyfencingknittingbakingcookingFigure 1. Example of stimuli in the three different conditions.Figure 2. Results of a split plot ANOVA investigating effort expenditure to social, neutral, and interest stimuli in ASD and control subjects.Figure 3. Results of within-subjects, repeated measures ANOVAs investigating percent change in pupil size to social, neutral, and interest stimuli in ASD (top) and control (bottom) subjects.Figure 4. Results of within-subjects, repeated measures ANOVAs investigating changes in blink rate to social, neutral, and interest stimuli in ASD (top) and control (bottom) subjects.ReferencesAmerican Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. Arlington, VA, American Psychiatric Association, 2013.Anderson, C.J., Colombo, J. (2009). Larger tonic pupil size in young children with autism spectrum disorder. Dev Psychobiol, 51(2), 207 – 211.Anderson, C.J., Colombo, J., Shaddy, J.D. (2006). Visual scanning and pupillary responses in young children with Autism Spectrum Disorder. J Clin Exp Neuropsychol, 28, 1238–1256. Baker, M.J. (2000). Incorporating the thematic ritualistic behaviors of children with autism into games: increasing social play interactions with siblings. Journal of Positive Behavior Interventions, 2(2), 66 – 84.Baker, M.J., Koegel, R.L., Koegel, L.K. (1998). Increasing the social behavior of young children with autism using their obsessive behaviors. Journal of the Association for Persons with Severe Handicaps, 23, 300 – 308.Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., Clubley, E. (2001). The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians.?J Autism Dev Disord, 31, 5–17.Benning, S.D., Kovac, M., Campbell, A., Miller, S., Hanna, E.K., Damiano, C.R., Sabatino-DiCriscio, A., Turner-Brown, L., Sasson, N.J., Aaron, R.V., Kinard, J., Dichter, G.S. (2016). Late positive potential ERP responses to social and nonsocial stimuli in youth with autism spectrum disorder. J Autism Dev Disord, 46(9), 3068 – 3077.Bijleveld, E., Custers, R., Aarts, H. (2009). The unconscious eye opener: pupil dilation reveals strategic recruitment of resources upon presentation of subliminal reward cues. Psychol Sci, 20,1313–1315. ?Bodfish, J.W. (2003).?Interests Scale.?Chapel Hill, NC.Bodfish, J. W., and Lewis, M. H. (2002). Repetitive Behavior in Autism. Paper presented at the International Meeting for Autism Research. (IMFAR), Orlando, FL.Bodfish, J.W., Symons, F.J., Parker, D.E., Lewis, M.H. (2000). Varieties of repetitive behavior in autism: Comparisons to mental retardation. J Autism Dev Disord, 30, 237–243.Boyd, B.A., Conroy, M.A., Richmond Mancil, G., Nakao, T., Alter, P.J. (2007). Effects of circumscribed interests on the social behaviors of children with autism spectrum disorders. J Autism Dev Disord, 37, 1550 – 1561.Bradley, M.M., Miccoli, L., Escrig, M.A., Lang, P.J. (2008). The pupil as a measure of emotional arousal and autonomic activation. Psychophysiology, 45, 602–607. ?Cascio, C.J., Foss-Feig, J.H., Heacock, J., Schauder, K.B., Loring, W.A., Rogers, B.P., Pryweller, J.R., Newsom, C.R., Cockhren, J., Cao, A., Bolton, S. (2014). Affective neural response to restricted interests in autism spectrum disorders. Journal of Child Psychology and Psychiatry, 55(2), 162 – 171.Chevallier, C., Kohls, G., Troiani, V., Brodkin, E.S., Schultz, R.T. (2012). The social motivation theory of autism. Trends Cogn Sci, 16(4), 231–239. Constantino J. N., Gruber, C.P. (2005).?Social Responsiveness Scale (SRS)?Los Angeles, CA: Western Psychological Services.Damiano, C.R., Aloi, J., Treadway, M., Bodfish, J.W., Dichter, G.S. (2012). Adults with autism spectrum disorders exhibit decreased sensitivity to reward parameters when making effort-based decisions. Journal of Neurodevelopmental Disorders, 4, 13.Dichter, G.S., Felder, J.N., Green, S.R., Rittenberg, A.M., Sasson, N.J., Bodfish, J.W. (2012). Reward circuitry function in autism spectrum disorders. SCAN, 7, 160 – 172.De Jong, P., Merckelbach, H. (1990). Eyeblink frequency, rehearsal activity, and sympathetic arousal. International Journal of Neuroscience, 51 (1-2), 89–94.Edwards, A. (1953). The relationship between the judged desirability of a trait and the probability that the trait will be endorsed. Journal of Applied Psychology, 37, 90 – 93.Eysenck, H.J. (1981) Arousal, Intrinsic Motivation, and Personality. In: Day H.I. (eds) Advances in Intrinsic Motivation and Aesthetics. Springer, Boston, MAFalck-Ytter, T. (2008). Face inversion effects in autism: a combined looking time and pupillometric study. Autism Res, 1, 297–306. ?Foss-Feig, J.H., McGugin, R.W., Gauthier, I., Mash, L.E., Ventola, P., Cascio, C.J. (2016). A functional neuroimaging study of fusiform response to restricted interests in children and adolescents with autism spectrum disorder. Journal of Neurodevelopmental Disorders, 8, 15.Girden, E. (1992).?ANOVA: Repeated measures. Newbury Park, CA: Sage.Goodman, R. (1997). The strengths and difficulties questionnaire: a research note.?Journal of Child Psychology and Psychiatry, 38, 581.Gutermuth Anthony, L., Kenworthy, L., Yerys, B.E., Jankowski, K.F., James, J.D., Harms, M.B., Martin, A., Wallace, G.L. (2013). Interests in high-functioning autism are more intense, interfering, and idiosyncratic, but not more circumscribed, than those in neurotypical development. Dev Psychopathol, 25(3), 643-652.Hardan, A.Y., Girgis, R.R., Lacerda, A.L.T., Yorbik, O., Kilpatrick, M., Keshavan, M.S., Minshew, N.J. (2006). Magnetic resonance imaging study of the orbitofrontal cortex in autism. J Child Neurol. 21, 866. Hess, E.H., Polt, J.M. (1960). Pupil size as related to interest value of visual stimuli. Science, 132, 349–350. Hirstein, W., Iversen, P., Ramachandran, V.S. (2001). Autonomic responses of autistic children to people and objects. Proceedings of Biological Sciences, 268, 1883 – 1888.Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M., Licalzi, E., Wasserman, S., Soorya, L., Buchsbaum, M. (2005). Striatal volume on magnetic resonance imaging and repetitive behaviors in autism. Biol Psychiatry. 58(3), 226–232.Hus, V., Bishop, S., Gotham, K., Huerta, M., Lord, C. (2013). Factors influencing scores on the social responsiveness scale. J Child Psychol Psychiatry, 54(2), 216–224.Kampe, K.K., Frith, C.D., Dolan, R.J., Frith, U. (2001). Reward value of attractiveness and gaze. Nature, 413, 589-610.Kasari, C., and Patterson, S. (2012). Interventions addressing social impairment in autism. Curr Psychiatry Rep, 14(6), 713–725.Kaufman,?P.,?Alm,?A. (2003).?Adler's Physiology of the Eye: Clinical Applications, Elsevier PhilidalphiaKohls, G., Antezana, L., Mosner, M.G., Schultz, R.T., Yerys, B.E. (2018). Altered reward system reactivity for personalized circumscribed interests in autism. Molecular Autism, 9, 9.Langen, M., Durston, S., Kas, M.J., van Engeland, H., Staal, W.G. (2011). The neurobiology of repetitive behavior:…and men. Neurosci Biobehav Rev. 35(3), 356–265. Leekam, S., Tandos, J., McConachie, H., Meins, E., Parkinson, K., Wright, C., ... & Couteur, A. L. (2007). Repetitive behaviours in typically developing 2‐year‐olds.?Journal of Child Psychology and Psychiatry,?48(11), 1131–1138.Levene, H. (1960). Robust tests for equality of variances. In: Olkin, I., Hotelling, H., et al.?Contributions to Probability and Statistics: Essays in Honor of Harold Hotelling. Stanford University Press, 278–292.Lewis, M.H., Tanimura, Y., Lee, L.W., Bodfish, J.W. (2007). Animal models of restricted repetitive behavior in autism. Beh Brain Res. 176, 66–74. Maxwell, S.E., Delaney, H.D. (1990).?Designing experiments and analyzing data: A model comparison perspective. Belmont: Wadsworth.Mazzone, L., Ruta, L., Reale, L. (2012). Psychiatric comorbidities in asperger syndrome and high functioning autism: diagnostic challenges. Annals of General Psychiatry, 11, 16.Mercier, C., Mottron, L., Belleville, S. (2000). A psychosocial study on restricted interests in high-functioning persons with pervasive developmental disorders. Autism, 4(4), 406 – 425.Ming, X., Julu, P.O.O., Brimacombe, M., Connor, S., Daniels, M.L. (2005). Reduced cardiac parasympathetic activity in children with autism. Brain and Development, 27, 509 – 516.Morrison, K.E., Chambers, L.K., Faso, D.J., Sasson, N.J. (2018). The content and function of interests in the broad autism phenotype. Research in Autism Spectrum Disorders, 49, 25 – 33.O’Doherty, J., Winston, J., Critchley, H., Perrett, D., Burt, D.M., Dolan, R.J. (2003). Beauty in a smile: the role of medial orbitofrontal cortex in facial attractiveness. Neuropsychologia, 41, 147–155.Parsons, O.E., Bayliss, A., Remington, A. (2016). A few of my favourite things: circumscribed interests in autism are not accompanied by increased attentional salience on a personalized selective attention task. Molecular Autism, 8, 20. Phillips, M.L., Bullmore, E.T., Howard, R., Woodruff, P.W., Wright, I.C., Williams, S.C., Simmons, A., Andrew, C., Brammer, M., David, A.S. (1998). Investigation of facial recognition memory and happy and sad facial expression perception: an fMRI study. Psychiatry Res, 83, 127–138. Pierce, K., Marinero, S., Hazin, R., McKenna, B., Carter Barnes, C., Malige, A. (2015). Eye tracking reveals abnormal visual preference for geometric images as an early biomarker of an autism spectrum disorder subtype associated with increased symptom severity. Biol Psychiatry, 79, 657 – 666.R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: , R.A., Ritvo, E.R., Guthrie, D., Ritvo, M.J., Hufnagel, D.H., McMahon, W., Tonge, B., Mataix-Cols, D., Jassi, A., Attwood, T., Eloff, J. (2011). The Ritvo Autism Asperger Diagnostic Scale- Revised (RAADS-R): A scale to assist the diagnosis of autism spectrum disorder in adults: An international validation study. J Autism Dev Disord, 41(8), 1076 – 1089.Ruzich, E., Allison, C., Smith, P., Watson, P., Auyeung, B., Ring, H., Baron-Cohen, S. (2015). Measuring autistic traits in the general population: a systematic review of the Autism-Spectrum Quotient (AQ) in a nonclinical population sample of 6,900 typical adult males and females. Molecular Autism, 6, 2. Sabatino, A., Rittenberg, A., Sasson, N.J., Turner-Brown, L., Bodfish, J.W., Dichter, G.S. (2013). Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord, 43(12), 2903 – 2913.Sabatino DiCriscio, A., Miller, S.J., Hanna, E.K., Kovac, M., Turner-Brown, L., Sasson, N.J., Sapyta, J., Troiani, V., Dichter, G.S. (2016). Brief report: cognitive control of social and non-social visual attention in autism. J Autism Dev Disord, 46, 2797 – 2805.Sanavio, E. (1988). Obsessions and compulsions: The Padua Inventory. Behav Res and Therapy, 26, 167–177.Sasson, N.J., Turner-Brown, L.M., Holtzclaw, T.N., Lam, K.S.L., Bodfish, J.W. (2008). Children with autism demonstrate circumscribed attention during passive viewing of complex social and non-social picture arrays. Autism Res, 1: 1.Sasson, N.J., Elison, J.T., Turner-Brown, L.M., Dichter, G.S., Bodfish, J.W. (2011). Brief report: circumscribed attention in young children with autism. J Autism Dev Disord, 41 (2), 242 – 247.Sasson, N.J., Dichter, G.S., Bodfish, J.W. (2012). Affective responses by adults with autism are reduced to social images but elevated to images related to circumscribed interests.?PLoS One, 7(8), e42457.Schmider, E., Ziegler, M., Danay, E., Beyer, L., Buhner, M. (2010). Is it really robust? Reinvestigating the robustness of ANOVA against violations of the normal distribution assumption. Methodology, 6(4), 147-151.Sears, L.L., Vest, C., Mohammed, S., Bailey, J., Rason, B.J., Piven, J. (1999). An MRI study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol Psychiatry, 23(4), 613–624. Sepeta, L., Tsuchiya, N., Davies, M.S., Sigman, M., Bookheimer, S.Y., and Dapretto, M. (2014). Abnormal social reward processing in autism as indexed by pupillary responses to happy faces. Journal of Neurodevelopmental Disorders, 4:17.Shafritz, K.M., Dichter, G.S., Baranek, G.T., Belger, A. (2008). The neural circuitry mediating shifts in behavioral response and cognitive set in autism. Biol Psychiatry. 63(10), 974- 980. Shapiro, S.S.,?Wilk, M.B. (1965). An analysis of variance test for normality (complete samples).?Biometrika,?52?(3–4), 591–611.Shultz, S., Klin, A., Jones, W. (2011). Inhibition of eye blinking reveals subjective perceptions of stimulus salience. Proc Natl Acad Sci USA, 108(52), 21270–21275. South, M., Klin, A. & Ozonoff, S. (1999). ‘The Yale Special Interests Interview’, unpublished instrument, available from the authors.SR Research Ltd.?(2009). Eyelink II users manual, version 2.133.?Mississauga, ON: SR Research Ltd.Treadway, M.T., Buckholtz, J.W., Schwartzman, A.N., Lambert, W.E., Zald, D.H. (2009). Worth the ‘EEfRT’? The effort expenditure for rewards task as an objective measure of motivation and anhedonia. PLoS ONE, 4(8), e6598.Tukey, J, (1949). "Comparing Individual Means in the Analysis of Variance".?Biometrics.?5?(2): 99–114. Turner, M.A. (1999). Annotation: repetitive behaviour in autism: a review of psychological research. J Child Psychol Psychiatry, 40(6), 839 – 849.Turner-Brown, L.M., K.S.L., Holtzclaw, T.N., Dichter, G.S., Bodfish, J.W. (2011). Phenomenology and measurement of circumscribed interests in autism spectrum disorders. Autism, 15(4), 437–456.Uddin, L.Q., Supekar, K., Lynch, C.J., Khouzam, A., Phillips, J., Feinstein, C., Ryali, S., Menon, V. (2013). Salience network-based classification and prediction of symptom severity in children with autism. JAMA Psychiatry, 70(8), 869–879. Unruh, K.E., Sasson, N.J., Shafer, R.L., Whitten, A., Miller, S.J., Turner-Brown, L., Bodfish, J.W. (2016). Social orientating and attention is influenced by the presence of competing nonsocial information in adolescents with autism. Front Neurosci, 10, 586.Van Oppen, P. (1992). Obsessions and compulsions: Dimensional structure reliability, convergent, and divergent validity. Behav Res Ther, 20(6), 631–637.Watson, K.K., Miller, S., Hannah, E., Kovac, M., Damiano, C.R., Sabatino-DiCrisco, A., Turner-Brown, L., Sasson, N.J., Platt, M.L., Dichter, G.S. (2015). Increased reward value of non-social stimuli in children and adolescents with autism. Frontiers in Psychology, 6, 1026.Wechsler, D. (2011). Wechsler Abbreviated Scale of Intelligence–Second Edition (WASI-II).?San Antonio, TX:?NCS Pearson.Wise, R.A. (2004). Dopamine, learning, and motivation. Nature Reviews Neuroscience. 5, 483–494.Zahn, T.P., Rumsey, J.M., Van Kemmen, D.P. (1987). Autonomic nervous system activity in autistic, schizophrenic, and normal men: effects of stimulus significance. Journal of Abnormal Psychology, 96(2), 135 – 144. Zwaigenbaum, L., Bryson, S., Rogers, T., Roberts, W., Brian, J., and Szatmari, P. (2005). Behavioral manifestations of autism in the first year of life. Int J Devl Neurosci, 23, 143–152.Supplementary Material Between-Subjects EffectsResults from split plot repeated measures ANOVAs that excluded the baseline pupil size and baseline blink rate covariates (so as to satisfy the assumptions of the ANOVA model) are presented below. Descriptive information for variables included in these models is displayed in Table 1.Pupil SizeDuring the passive block, there was no main effect of group on pupil size (F (1, 27) = 1.788, p = 0.192). There was a significant main effect of condition on pupil size (F (2, 54) = 11.039, p = 0.000). There was no group by condition interaction effect (F (2, 54) = 1.292, p = 0.283). Post hoc analysis revealed that both groups demonstrated significantly larger pupil size during the interest vs. the neutral condition (F (1, 27) = 14.731, p = 0.001).During the active block, there was a main effect of group on pupil size that did not survive correction for multiple comparisons (F (1, 27) = 6.601, p = 0.016). There was a significant main effect of condition on pupil size (F (2, 54) = 21.070, p = 0.000). The group by condition interaction effect did not reach significance (F (2, 54) = 2.025, p = 0.142). Post hoc analyses revealed that both groups demonstrated significantly larger pupil size during the interest vs. the neutral condition (F(1, 27) = 40.451, p = 0.000), and in the interest versus the social condition (F (1, 27) = 5.082, p = 0.032). There was a trend toward a group by condition interaction effect, whereby the ASD group displayed larger pupil size during the interest versus the social condition (F (1, 27) = 3.292, p = 0.081), relative to controls. Blink RateDuring the passive block, there was a main effect of group on blink rate that did not survive correction for multiple comparisons (F (1, 27) = 6.461, p = 0.017), with the ASD group displaying decreased blink rate, relative to controls. There was no main effect of condition on blink rate (F (2, 54) = 1.677, p = 0.206). There was no group by condition interaction effect (F (2, 54) = 1.302, p = 0.264).During the active block, there was a significant main effect of group on blink rate (F (1, 27) = 13.859, p = 0.001), with the ASD group displaying significantly decreased blink rate, relative to controls. There was a trend toward a main effect of condition on blink rate (F (2, 54) = 3.098, p = 0.053). There was no group by condition interaction effect (F (2, 54) = 1.532, p = 0.225). Post hoc analysis revealed that both groups displayed significantly reduced blink rate during the interest vs. the neutral condition (F (1, 27) = 5.512, p = 0.026).2.0. Within-Subjects Effects, Raw Observed Pupil Size The results of a within-subjects, repeated measures ANOVA model that used pupil size as observed (arbitrary units), rather than calculating mean percent change in pupil size from baseline (arbitrary units; Table 4), so as to satisfy the assumptions of the ANOVA model, are presented below. This model included the baseline pupil size covariate.Within the ASD group, the effect of condition on pupil size during the passive block was not significant (F (2, 18) = 2.66, p = 0.097). ASD subjects displayed the largest pupil size during the interest condition, followed by the social condition, and the smallest pupil size during the neutral condition (Figure 5a). During the active block for ASD subjects, a significant effect of condition on pupil size was found: F (2, 15) = 8.022, p = 0.0043. Significantly larger pupil size during the interest vs. the neutral condition (z = 3.623, p < 0.001) was found, with a trend toward significantly larger pupil size in the interest vs. the social condition (z = 2.258, p = 0.0618), and no significant difference in pupil size between the social and neutral condition (z = -1.365, p = 0.3597), Figure 5b. Within the control group, the effect of condition on pupil size during the passive block: F (2, 36) = 3.35, p = 0.046 did not survive correction for multiple comparisons. Control subjects displayed the largest pupil size to social images, followed by interest images, and the smallest pupil size to neutral images (Figure 5c).During the active block for control subjects, there was a significant main effect of condition on pupil size: F (2, 33) = 15.580, p = 0.000. Significantly larger pupil size in the social versus the neutral condition (z = -4.755, p = 1 ^e-05), and in the interest versus the neutral condition (z = 5.104, p = 1^e-05) was found. There was no significant difference in pupil size between the interest and social condition (z = 0.349, p = 0.935) for control subjects (Figure 5d).SUPPLEMENTARY VARIABLE OF INTERESTASD mean (sd) conditionCONTROL mean (sd) conditionShapiro-Wilk Normality Test (W, p-value)Levene’s Test of Homogeneity of Variance(F-value, p-value)Observed pupil size (arbitrary units): passive 1000.50 (393.53) social971.71 (391.48) neutral1009.87 (385.94)interest686.34 (275.01) social663.38 (257.66) neutral683.80 (278.97) interestW = 0.967, p = 0.03F = 2.555, p = 0.123 (social)F = 3.467, p = 0.074 (neutral)F = 2.485, p = 0.127 (interest)Observed pupil size (arbitrary units): active 1108.85 (477.89) social1067.70 (494.18) neutral1136.73 (482.97) interest713.68 (280.48) social658.32 (254.62) neutral717.74 (285.72) interestW = 0.971,p = 0.04F = 2.755, p = 0.109 (social)F= 3.289, p = 0.081 (neutral)F = 2.203, p = 0.149 (interest)Table 4. Descriptive statistics for observed pupil size data used in a supplementary repeated measures ANOVA modelFigure 5. Results of within-subjects, repeated measures ANOVAs investigating changes in raw, observed pupil size to social, neutral, and interest stimuli in ASD (top) and control (bottom) subjects.Chapter 4Indices of Repetitive Behaviour are Correlated with Patterns of Intrinsic Functional Connectivity in Youth with Autism Spectrum DisorderJenna M. Traynor 1., Krissy A.R. Doyle-Thomas 2., Lindsay C. Hanford 1, Nicholas E. Foster 3,4, Ana Tryfon 3,4, Krista L. Hyde 3,4, Evdokia Anagnostou 2,5., Alan C. Evans 6, Lonnie Zwaigenbaum 7., and Geoffrey B.C. Hall 1; NeuroDevNet ASD Imaging Group81 McMaster University, Department of Psychology, Neuroscience & Behaviour, Hamilton, Ontario, Canada2 Bloorview Research Institute, Holland Bloorview Kids Rehabilitation Hospital, University of Toronto, Toronto, Ontario, Canada3 International Laboratory for Brain Music and Sound (BRAMS), University of Montreal, Montreal, Quebec, Canada4 Faculty of Medicine, McGill University, Montreal, Quebec, Canada5 Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada6 Montreal Neurological Institute, Montreal, Quebec, Canada7 Department of Pediatrics, University of Alberta, Edmonton, Alberta, Canada8 , Vancouver, British Columbia, CanadaTraynor, J.M., Doyle-Thomas, K.A.R., Hanford, L.C., Foster, N.E., Tryfon, A., Hyde, K.L., Anagnostou, E., Evans, A.C., Zwaigenbaum, L., Hall, G.B.C., NeuroDevNet ASD Imaging Group (2018). Indices of repetitive behaviour are correlated with patterns of intrinsic functional connectivity in youth with autism spectrum disorder. Brain Research, 1685, 79 – 90.Abstract and Key WordsThe purpose of the current study was to examine how repetitive behaviour in Autism Spectrum Disorder (ASD) is related to intrinsic functional connectivity patterns in a number of large-scale, neural networks. Resting-state fMRI scans from thirty subjects with ASD and thirty-two age-matched, typically developing control subjects were analysed. Seed-to-voxel and ROI-to-ROI functional connectivity analyses were used to examine resting-state connectivity in a number of cortical and subcortical neural networks. Bivariate correlation analysis was performed to examine the relationship between repetitive behaviour scores from the Repetitive Behaviour Scale – Revised (RBS-R) and intrinsic functional connectivity in ASD subjects. Compared to control subjects, ASD subjects displayed marked over-connectivity of the thalamus with several cortical sensory processing areas, as well as over-connectivity of the basal ganglia with somatosensory and motor cortices. Within the ASD group, significant correlations were found between functional connectivity patterns and total RBS-R scores as well as one principal component analysis-derived score from the RBS-R. These results suggest that thalamocortical resting-state connectivity is altered in individuals with ASD, and that resting-state functional connectivity is associated with ASD symptomatology.KEYWORDS: autism spectrum disorder, repetitive behaviour, resting-state functional magnetic resonance imaging, intrinsic functional connectivityIntroduction Restricted and repetitive behaviour is a prominent symptom of Autism Spectrum Disorder (ASD), and is associated with marked distress, interference with educational achievement, attention, and learning, and difficulty maintaining meaningful interpersonal relationships (Leekam, Uljarevic, & Prior, 2011). In the past, neuroimaging studies have examined the neural underpinnings of restricted and repetitive behaviour by correlating total, overall repetitive behaviour scores from diagnostic measures with patterns of neural structure and function. Most studies have used repetitive behaviour scores from the Autism Diagnostic Interview – Revised (ADI-R; Lord, Rutter, & Le Couteur, 1994); in both children and adults with ASD, total ADI-R repetitive behaviour scores have been correlated with a number of neural abnormalities in the frontal cortex (Agam, Joseph, Barton, & Manoach, 2010; Rojas et al., 2006; Shafritz, Dichter, Baranek, & Belger, 2008; Thakkar et al., 2008; Thomas, Humphreys, Jung, Minshew, & Behrmann, 2011), post central gyri (Rojas et al., 2006), parietal cortex (Nordahl et al., 2007; Shafritz et al., 2008), superior temporal gyri (Rojas et al., 2006), amygdala (Dziobek, Fleck, Rogers, Wolf, & Convit, 2006; Rojas et al., 2006) and caudate nuclei (Rojas et al., 2006). However, repetitive behaviour encompasses a broad array of atypical motor and cognitive behaviour and presents heterogeneously across the autism spectrum. For example, higher-functioning individuals with ASD tend to exhibit more cognitively based repetitive behaviour, such as cognitive rigidity and ritualistic mental acts (Szatmari et al., 2006; Turner, 1999). On the other hand, lower-functioning individuals tend to exhibit more motor-based stereotypies (Szatmari et al., 2006; Turner, 1999). Some individuals with ASD also display repetitive behaviors in both cognitive and motor domains. Given this complexity, repetitive behaviour has been divided into subtypes, and associations between subtypes and various neurobiological correlates have been examined. For example, the Repetitive Behaviour Scale – Revised (RBS-R; Bodfish, Symons, Parker, & Lewis, 2000), stratifies repetitive behaviour into six subscales; stereotyped; self-injurious; compulsive; ritualistic; sameness; and restricted behaviour. The ADI-R divides repetitive behaviour into only four subtypes; motor stereotypies, insistence on sameness, abnormal preoccupations or circumscribed interests, and preoccupations with non-functional parts of objects (Langen, Durston, Kas, van Engeland, & Staal, 2011; Lord et al., 1994). Studies using principal component and factor analyses have supported the division of repetitive behaviour into subtypes, although, different studies have supported various factor solutions, depending on the symptom measure examined and the analysis method used. For example, Lam and Aman (2007) used factor analysis of the RBS-R to identify a five-factor solution consisting of ritualistic/sameness behaviour, stereotypic behaviour, self-injurious behaviour, compulsive behaviour, and restricted interests, whereas analyses of the ADI-R have yielded both a three-factor model consisting of repetitive motor behaviour, sameness behaviour, and circumscribed patterns of interest (Lam, Bodfish, & Piven, 2008), and a two-factor model consisting of repetitive motor behaviour and insistence on sameness (Cuccaro et al., 2003; Szatmari et al., 2006). Structural MRI and task-based fMRI studies have identified associations between repetitive behaviour subtypes and an intricate array of neural abnormalities. Briefly, it has been suggested that repetitive motor behaviour in ASD may be associated with atypical connectivity in thalamocortical (Mizuno, Villalobos, Davies, Dahl, & Muller, 2006) and cortico-basal ganglia (Turner, Frost, Linsenbardt, McIlroy, & Muller, 2006) networks. On the other hand, ADI and ADI-R scores of insistence on sameness in daily routine have been specifically associated with striatal abnormalities in both children (Langen et al., 2013; Qiu et al., 2016) and adults (Hollander et al., 2005; Sears et al., 1999). Striatal abnormalities have also been found to subserve circumscribed interest behaviour, and this has been demonstrated in both structural (Hollander et al., 2005; Sears et al., 1999) and functional, task-based connectivity studies (Cascio et al., 2014; Dichter et al., 2012; Sabatino et al., 2013), with a specific focus on the role of corticostriatal networks. ADI-R scores for circumscribed interests have also shown an association with volume of the orbitofrontal cortex (Hardan et al., 2006). Finally, an association between preoccupations with non-functional parts of objects and frontal and cerebellar volume has been demonstrated (Pierce & Courchesne, 2001), although more work is needed to replicate these findings. Although most studies have used the ADI-R to examine repetitive behaviour, two recent studies have examined associations between RBS-R scores and neural correlates; Eisenberg, Wallace, Kenworthy, Gotts, and Martin (2015) found an association between RBS-R sameness behaviour scores and covariance of striatal and limbic gray matter structures, and one task-based fMRI study using an inhibitory control paradigm found no correlation between total RBS-R scores and functional connectivity within the frontoparietal attention network in children with ASD (Ambrosino et al., 2014). Recently, resting-state fMRI has also been employed to investigate the relationship between spontaneous, low frequency oscillations in neural activity and repetitive behaviour scores, again, primarily from the ADI-R. Whereas task-based fMRI activation is task-dependent, patterns of activation revealed during resting-state fMRI can be reliably identified in the absence of task performance. Resting-state studies have revealed patterns of intrinsic connectivity that are strongly associated with behavioural symptomatology in ASD. For example, resting-state connectivity patterns in the salience network (Uddin et al., 2013), default mode network (Monk et al., 2009; Weng et al., 2010) and ventro-temporal-limbic network (Glerean et al., 2016) have been correlated with total, overall ADI-R repetitive behaviour scores. Additionally, one study used RBS-R scores in conjunction with measures of social behaviour, language, and motor development, to demonstrate that intrinsic connectivity patterns in 6-month old infants can reliably predict an ASD diagnosis at 24 months of age (Emerson et al., 2017). However, given the diverse typology of repetitive behaviour across the autism spectrum, it is likely that distinct patterns of intrinsic connectivity are also associated with specific repetitive behaviour subtypes. Indeed, a recent review by Traynor and Hall (2015) indicated that several subtypes of repetitive behaviour may be associated with neural abnormalities in networks that subserve salience attribution, reward processing, motor control, cognitive control, and attention. However, the primary use of the ADI-R as an indicator of the association between neural abnormality and behavioural symptomatology is problematic. This is because the ADI-R is a categorical diagnostic measure of behaviour that was designed to support clinicians in making binary (yes/no) decisions, rather than a dimensional measure of repetitive behaviour symptoms. Although this has been common practice, use of the ADI-R in this manner may preclude a sound analysis of the correlation between subscores and neural anomalies, especially due to the restricted range of scores on some ADI-R subscales. A more statistically appropriate measure for correlation analyses would be the RBS-R (Bodfish et al., 2000), which was designed to be used as a dimensional measure of repetitive behavior in ASD and contains a wider range of repetitive behaviour subscores.Therefore, the purpose of the current study was to use the RBS-R to examine the association between repetitive behaviour in ASD and intrinsic functional connectivity patterns. Specifically, seed-to-voxel and ROI-to-ROI connectivity analyses were used to examine differences in resting-state functional connectivity between subjects with ASD and typically developing control subjects. Then, within-subject correlation analyses were conducted to investigate the relationship between intrinsic functional connectivity patterns in ASD subjects and a) RBS-R total scores, and b) two principal components (PC) derived from the RBS-R subscores via principle component analysis (PCA). Studies that examine repetitive behaviour subcategories are important for two reasons. First, they allow for more precise delineation of the neural circuitry subserving qualitatively different repetitive behaviours in ASD, and reflect the level of precision found in other ASD fields such as genetics, where the association between genotypes has been examined in relation to repetitive behaviour subtypes (Chao et al., 2010; Garner, Meehan, & Mench, 2003; Lewis, Tanimura, Lee, & Bodfish, 2007). Second, these findings have important implications for treatment interventions, which can be tailored to individuals displaying idiosyncratic profiles of repetitive behaviour. For example, specific neural signatures may indicate an increased likelihood of the presence of specific repetitive behaviours that can be targeted behaviourally, and by targeting these behaviours, neuroplastic changes in connectivity may arise.Results2.1 Between–subject, seed-to-voxel analysisBetween-subject seed-to-voxel effects are displayed in Table 1 and Figure 1. These results represent differences between the ASD and control group in the bivariate temporal correlation of each seed with every other voxel in the brain (Whitfield-Gabrieli & Nieto-Castanon, 2012). Peak- voxel threshold was set at p < 0.001 and cluster threshold was set at p <0.05. As examining connectivity patterns between each seed with every other voxel in the brain greatly increased the number of comparisons drawn, a more conservative Family Wise Error (FWE) correction was implemented (Benjamini & Hochberg, 1995). Clusters were considered significant if they survived correction for multiple comparisons at a threshold of p <0.05 (FWE). Compared to control subjects, the ASD group displayed a broader resting state network, as well as negative connectivity of the posterior cingulate cortex (PCC) with the angular gyrus, and positive connectivity of the PCC with the superior temporal gyrus. Additionally, the ASD group displayed over- connectivity of the hippocampus with the associative visual cortex, marked over-connectivity of the thalamus with several sensory processing areas of the cortex including the primary somatosensory cortex, auditory cortex, premotor cortex, superior temporal gyrus, and the insular cortex, and over-connectivity of basal ganglia structures (putamen and globus pallidus) with somatosensory and motor cortices, as well as with the fusiform gyrus. 2.2 Between-subject, ROI-to-ROI analysisBetween-subject ROI-to-ROI effects are displayed in Table 2 and Figure 2. This analysis displayed the bivariate temporal correlation of each predefined source ROI with every other target ROI in the brain (Whitfield-Gabrieli & Nieto-Castanon, 2012). Results were largely consistent with the seed-to-voxel analysis and were considered significant if they survived correction for multiple comparisons at a threshold of p < 0.05 False Discovery Rate (FDR) (Benjamini & Hochberg, 1995). This less conservative threshold was implemented as a result of examining ROI-to-ROI connectivity patterns in the brain. Compared to the seed-to-voxel analysis, this analysis greatly reduced the number of comparisons drawn. Beta values correspond to the size of the effect. Similar to the seed-to-voxel analysis, the ROI-to-ROI analysis revealed marked over-connectivity in ASD subjects compared to controls, of the thalamus with several cortical sensory processing areas, including the primary somatosensory, auditory and motor cortices, and the insular cortices. In addition, compared to control subjects, the ASD group again displayed over-connectivity of basal ganglia structures (putamen and globus pallidus) with somatosensory and motor cortices. The ROI-to-ROI analysis also revealed under-connectivity of the left hippocampus with the right perirhinal cortex (ASD vs. TD); this result may be attributed to the less conservative correction method used in this analysis.2.3 Brain-behaviour relationships in ASD: Within-subject ROI-to-ROI bivariate correlation analysisFirst, a ROI-to-ROI bivariate correlation analysis using overall, total RBS-R scores revealed a significant positive association between total RBS-R scores and connectivity between the left primary visual cortex (V1)/BA.17 and the right inferior frontal gyrus (IFG), pars orbitalis/BA.47, T (21) = 4.13, pFDR = 0.0488.Second, the results of using two PC’s from the RBS-R subscores as variables in a ROI-to-ROI bivariate correlation analysis are displayed in Table 3. First, a multivariate F test was implemented to examine whether there was any effect among either of the two components on ROI-to-ROI functional connectivity. The F test revealed a significant association between these two components and positive connectivity between the right inferior parietal lobe (IPL) and the right IFG, pars triangularis (Table 3, left). To examine the origin of this effect, post hoc testing was implemented, and revealed the simple main effect of each individual RBS-R component on this pattern of functional connectivity. A Bonferroni correction for two additional comparisons (i.e., each of the RBS-R components) resulted in a significance threshold of alpha = 0.05/2 = p < 0.025 for these post hoc tests (Table 3, right). This testing revealed a significant simple main effect of PC 2 only. Thus, PC 2 scores were positively correlated with connectivity between the right IPL and the right IFG in our sample. The simple main effect of PC 1 was not significant.DiscussionThis study examined the relationship between intrinsic connectivity patterns and repetitive behaviour in ASD. Compared to control subjects, significant differences in resting-state connectivity patterns in youth with ASD were found. Additionally, a correlation analysis revealed significant relationships between specific repetitive behaviour scores and resting-state connectivity, within the ASD group. 3.1 Thalamocortical and Cortico-Basal Ganglia Connectivity in ASDPerhaps the most striking between-subject effect was the marked thalamocortical over-connectivity observed in the ASD group, in both the seed-to-voxel and ROI-to-ROI between-subject analyses. Compared to control subjects, the ASD group displayed over-connectivity of the left thalamus with several peripheral sensory processing areas including the left premotor and auditory cortices and the bilateral somatosensory cortices. Contralateral thalamic-insular hyper-connectivity was also displayed in the ASD group, compared to controls. Although thalamocortical organization is predominantly ipsilateral, animal work has revealed contralateral organization of cortico-subthalamic projections, specifically in the mediodorsal thalamic nuclei (Negyessy, Hamori, & Bentivoglio, 1998) and in pallidothalamic areas (Hazrati & Parent, 1991). As the current study did not use sub-sectioned thalamic seeds, and is concerned only with the temporal correlation between two brain regions, contralateral thalamocortical projections are considered. The thalamus is commonly referred to as a sensory gateway structure, relaying input from subcortical structures such as the basal ganglia and cerebellum to the cortex. It is also known to modulate higher-order communication between cortices, via cortico-thalamic-cortical feedback loops (Sherman, 2007). Therefore, aberrant thalamic-sensory-cortical connectivity in ASD may contribute to abnormal processing of auditory and tactile stimuli; a hallmark characteristic of ASD. Indeed, measures of sensory abnormalities have been linked with thalamic alterations in children with ASD (Hardan et al., 2008). Importantly, the findings in the current study are in line with results from a recent study by Cerliani et al. (2015) that analyzed resting-state data from a large sample of male youth (166 ASD, 193 TD, M age: 16.2 years) from the Autism Brain Imaging Data Exchange. This study also found over-connectivity between thalamic and cortical sensory areas in ASD subjects, compared to controls. Previous work using task-based fMRI has also found over-connectivity of thalamocortical regions in adult males with ASD (Mizuno et al., 2006). However, in another resting-state fMRI study using a sample of ASD youth and typically developing controls, Nair, Treiber, Shukla, Shih, and Muller (2013) found thalamocortical under-connectivity. In sum, although at present, most studies point to over-connectivity, patterns of thalamocortical resting state connectivity may be dependent on a number of moderating factors such as the age of the research sample, level of individual functioning, symptom presentation and imaging methodology. Further, the current findings of over-connectivity are specific to the left thalamus. Consistent with this finding, previous work has demonstrated abnormal thalamic laterality in ASD. For example, reduced metabolic concentrations of choline (Hardan et al., 2008) and creatine (Friedman et al., 2003; Hardan et al., 2008) in the left thalamus of individuals with ASD have been found. Say et al. (2014) also found greater left-minus right laterality of the thalamus in a sample of children and adolescents with Asperger Syndrome. Finally, both the seed-to-voxel and ROI-to-ROI analyses revealed over-connectivity in cortico-basal ganglia circuits, in ASD subjects compared to controls. Similar to the findings for the thalamus, the putamen and the globus pallidus seeds displayed over-connectivity with somatosensory and motor cortices. Importantly, these findings are consistent with several previous ASD studies that have found cortico-basal ganglia over-connectivity in ASD subjects relative to controls. For example, a recent resting-state study also found over-connectivity of the basal ganglia (caudate and putamen) with the cerebral cortex in a group of youth with ASD (age 7 – 13 years; Di Martino et al., 2011). In another study, Turner et al. (2006) found primarily over-connected and diffuse caudate-cortical circuitry during a visuomotor coordination fMRI task. Finally, the Autism Brain Imaging Exchange study by Cerliani et al. (2015) also found resting-state over-connectivity of basal ganglia with cortical-sensory regions in ASD youth. Basal ganglia abnormalities in ASD may be associated with atypical gait (Rhinehart et al., 2006), social-communicative and motor function (Qiu et al., 2010).Of note, in contrast to the majority of our findings pertaining to over-connectivity, the current results showed under-connectivity of the left hippocampus with the right perirhinal cortex in ASD subjects, compared to controls. Pertaining to this finding, the under-connectivity theory of ASD (Just et al., 2004) posits that ASD is characterized by less synchronization between frontal and posterior areas during task performance, and more generally, that under-connectivity may contribute to information processing deficits in ASD (Just et al., 2012). As the hippocampus and perirhinal cortex are both a part of the medial temporal lobe, and as our study examined resting-state BOLD signal only, our results do not necessarily provide support for or against the under-connectivity theory of ASD; to date, resting-state functional connectivity studies in ASD have found both over and under frontal-posterior connectivity (Jones et al., 2010; Noonan, Haist, & Muller, 2009; Villalobos et al., 2005). With this said, our results are in line with the hypothesis that under-connectivity may result in information processing deficits (Just et al., 2012). Specifically, a recent study found reduced retrieval-related, hippocampal-middle temporal gyrus connectivity in ASD subjects during an episodic memory task (Cooper et al., 2017), which supports an emerging trend suggesting that ASD is characterized by selective episodic memory deficits (see Boucher, Mayes, & Bigham, 2012 for a review) that may impact information processing. Our finding of reduced hippocampal-perirhinal cortex connectivity provides additional support that may extend the hypothesis of selective memory deficits in ASD beyond task-related under-connectivity, to potentially include intrinsic hippocampal under-connectivity, and this may subserve information processing deficits. Future work is needed to investigate this hypothesis.3.2 Repetitive Behaviour and Connectivity in ASDThis work identified functional connectivity patterns within the ASD group that were significantly associated with scores from the RBS-R. Patterns of functional connectivity showing an association with RBS-R scores in the ASD group did not overlap with the between-subject differences in connectivity found during the between-subject ROI-to-ROI analysis. This finding implies that similar connectivity patterns in both individuals with ASD and typically developing controls may be associated with very different behavioural presentations between these two groups. This demonstrates the principle of multifinality, whereby depending on the system in which they operate, similar factors or components may function differently and lead to different outcomes (Wilden, 1980). In this case, patterns of functional connectivity common to both subjects with ASD and typically developing controls (similar factors) uniquely function to subserve repetitive behaviour in the ASD group (the ‘system’ in which they operate), but not in the TD group. 3.2.1 Total RBR-S Scores and Functional ConnectivityThis study found a significant positive correlation between repetitive behaviour in ASD as measured by total, overall RBS-R scores and connectivity between the left primary visual cortex (V1) and the right IFG, pars orbitalis. That is, individuals with greater amounts of repetitive behaviour display stronger connectivity between these two areas. Although it is commonly accepted that V1 is strongly implicated in the processing of visual information, Kaster, Pinsk, De Weerd, Desimone, and Ungerleider, (1999) have also demonstrated attentional modulation of the primary visual cortex, such that attended-to visual stimuli elicit greater BOLD activation than unattended-to, distractor stimuli. Similarly, although the right IFG has demonstrated strong involvement in response inhibition (Aron, Fletcher, Bullmore, Sahakian, Robbins, 2003; Menon, Adleman, White, Glover, & Reiss, 2001; Rubia, Smith, Brammer, & Taylor, 2003), this area has also been recently implicated in directing attention to behaviourally salient stimuli (Corbetta & Shulman, 2002). For example, work that has separated attentional aspects of response inhibition tasks (i.e., stop cue detection) from the inhibitory response period has shown that the right IFG may be more involved in general attentional control, activating in temporal synchrony with the IPL (Hampshire, Chamberlain, Monti, Duncan, & Owen, 2010). Thus, the right IFG is not only implicated in response inhibition, but also functions in conjunction with a task-on, attentional network (i.e., frontoparietal network), supporting an individual’s capacity to select cues from the environment that are salient to the task at hand (i.e., inhibitory demands). Although the current finding associates stronger V1- right IFG connectivity with repetitive behaviour, this relationship is likely quite complex, especially due to right IFG involvement in both inhibition and attentional processes. It is therefore possible that stronger connectivity may be the result of behaviourally-reinforced neural plasticity in a number of ways. For example, in the context of inhibitory control, repetitive behaviours may be understood as overlearned responses to environmental stimuli, whereby individuals with greater frequencies of repetitive behaviour are also required to inhibit these repetitive tendencies more often than individuals with less frequency of such behaviour, thereby strengthening V1 – right IFG connectivity. On the other hand, in the context of attention, stronger connectivity may indicate increased attention to cues salient to either the initiation or termination of sameness behaviour, also strengthening connectivity. It is also possible that a linear conceptualization may not be sufficient to explain this relationship, which may involve more dynamic functional processes that could be elucidated by investigating dynamic, network-based connectivity patterns in relation to behaviour (i.e., physiophysiological interaction analysis). Related to our finding, a previous task-based fMRI study found that impaired visual motor learning (i.e., action planning) was associated with reduced functional connectivity between the left primary visual cortex (V1) and the right IFG in individuals with ASD (Villalobos, Mizuno, Branelle, Kemmotsu, & Muller, 2005). Despite the directional difference between this finding (i.e., less connectivity associated with greater impairment) and ours, it does provide support for involvement of V1 – right IFG connectivity in repetitive behaviour. This directional difference does not necessarily contrast with our findings, as task-based fMRI is sensitive to BOLD signal at higher frequencies than the low-frequency oscillations that resting-state studies are designed to detect.3.2.2 Repetitive Behaviour Subtypes and Functional ConnectivityThis study found a significant positive association between RBS-R-derived PC 2 and connectivity between the right IPL and the right IFG, pars triangularis. Behaviorally, subjects who score high on PC 2 can be described as individuals with higher scores for insistence on sameness behaviour, and lower scores for self-injurious and restricted behaviour. Broadly, this finding therefore shows a general positive association between sameness behaviour and connectivity within the frontoparietal attention network, and it is probable that sameness behaviour may be modulated by altered neural functioning that subserves directed attention to environmental cues signaling a need for behavioural change. This finding also supports previous work that has identified insistence on sameness as distinct statistically, from other restrictive repetitive behaviour (Cuccaro et al., 2003; Lam et al., 2008; Szatmari et al., 2006).Previous findings supporting the current results have found reduced integration of the IFG-IPL resting-state network in ASD subjects (Bos et al., 2014), and have also identified that characteristics of the IFG, pars triangularis are able to differentiate individuals with ASD from typically-developing control subjects. For example, Jiao and colleagues (2010) demonstrated that decreased cortical thickness in the bilateral IFG, pars triangularis was predictive of membership in an ASD diagnostic group in a sample of youth (M age = 9.2) with and without ASD. Finally, much previous work has been conducted on the insistence on sameness subtype of repetitive behaviour. In addition to attention, the IPL also plays a role in the integration of visual and sensory stimuli (Anderson, 2011), and the current finding is supported by work that has demonstrated a positive association between abnormal sensory processing and insistence on sameness (Wigham, Rodgers, South, McConachie, & Freeston, 2015). Supporting involvement of the IFG in sameness behaviour, previous functional studies using task-based fMRI paradigms to indirectly examine the association between sameness behaviour and neural function have revealed reduced activation of the right IFG to novel auditory events during an auditory change detection task (Gomot et al., 2006) as well as altered activation of frontal areas during a visual change detection task (Clery et al., 2013). Structurally, sameness behaviour has been most strongly associated with aberrant growth rate (Langen et al., 2013) and morphometry (Hollander et al., 2005; Sears et al., 1999) of the striatum.Of note, post hoc testing did not reveal a significant association between PC 1 and functional connectivity of right IPL and right IFG, pars triangularis (see Table 3). This may be understood via an examination of the direction and magnitude of strength (i.e., correlation) between each of the original six RBS-R subscale scores and PC 1, which indicates that individuals who scored high on this component displayed relatively low amounts of all six types of repetitive behaviour; there is a sizeable and similar negative association between each RBS-R subscore and PC 1 (see Table 5). Lower amounts of repetitive behaviour in general were therefore not associated with specific patterns of attention-related connectivity of the IPL and IFG, pars triangularis.The current findings should be interpreted within the context of some limitations. First, due to the small sample size used in our analysis, these findings should be replicated and validated using a larger sample. Second, our sample of youth with ASD contained a wide age range (i.e., 10 – 21 years), but our relatively smaller within-subject sample size (n = 30) precluded an exploratory age analysis. Examination of the developmental trajectory of these neural patterns is an important future direction, as adolescence is a critical developmental time period. Future work examining the developmental trajectory of these behaviours in the context of neural function would increase the ecological validity of findings. With this said, the PCA decomposition in the current study primarily highlighted the neural circuitry associated with insistence on sameness behaviour, and sameness behaviour is one repetitive behaviour subtype that has demonstrated consistency across the autism spectrum, and is thought to be relatively independent of both intelligence and age (Bishop, Richler, & Lord, 2006; Hus, Pickles, Cook, Risi, & Lord, 2007). Therefore, despite our wider age range, the current results may be robust to the effects of age on identified brain-behaviour relationships. Finally, our sample included an uneven distribution of males and females in the ASD and TD groups. This sex difference was controlled for in the main analysis and was further addressed in a supplementary analysis (5. Supplementary Material). Although this difference did not have an effect on the results, future work may employ a sex-balanced sample of participants so as to entirely preclude the possibility of sex differences affecting the findings.ConclusionsThis study identified significant differences in patterns of resting-state functional connectivity between youth with ASD and a group of age-matched, typically developing control subjects. Additionally, we found that individuals with ASD display intrinsic connectivity patterns that are uniquely associated with different indices of repetitive behaviour from the RBS-R. Two major themes present from these findings, and are supported by previous work in ASD. First, compared to typically developing subjects, the left thalamus in ASD subjects displayed broad over-connectivity with a number of sensory-cortical areas. Second, both attention to salient cues and inhibitory control processes are likely involved in repetitive behaviour in ASD, and these process are associated with connectivity of the right IFG with other cortical areas involved in visual (V1) and sensory processing (IPL). The frontal cortex, and the IFG specifically, contain high concentrations of inhibitory gamma-aminobutyric acid (GABA) neurons (Kadosh, Krause, King, Near, & Kadosh, 2015) and animal work has pointed to abnormalities in GABA function as a likely factor in the maintenance of repetitive sensory and motor behaviour in ASD (Han, Tai, Jones, Scheuer, Catterall, 2014). Two recent human studies have also initiated the investigation of GABAergic function in vivo in the ASD cortex (Harada et al., 2011; Rojas et al., 2014), and have found decreased GABA concentration in left frontal lobe and left perisylvian region, respectively. As the current study demonstrated an association between right IFG activity and both total RBS-R scores and sameness behaviour, further inquiry into the role of inhibitory GABA function in repetitive behaviours is warranted.Our study found also vast between-subject differences in thalamo-sensory-cortical connectivity, but our within subject analyses did not show thalamic associations with repetitive behaviour. However, the structural connectivity of the thalamus with the primary visual cortices as well as its role as a sensory gateway structure indicates it may also play an indirect role in repetitive behaviour. Future work may more efficiently examine thalamo-cortical connectivity in relation to repetitive behaviour with the use of dynamic and complex patterns of connectivity analyses, which involve more than two regions of the brain (i.e., physiophysiological interaction analysis or network-based connectivity analysis).Finally, future work may also focus on a more comprehensive examination of repetitive behaviour by including scores from the ADI-R in analyses of resting-state connectivity patterns. Although the ADI-R was originally designed for diagnostic classification purposes, work has suggested that it can be used as a measure of symptomatology in biological research studies (Lord et al., 2001). Indeed, use of ADI-R repetitive behaviour subscores has greatly aided in the discovery of important findings in the field of neuroimaging in ASD (Hardan et al., 2006; Hollander et al., 2005; Langen et al., 2013; Mosconi et al., 2009; Qui et al., 2016; Sears et al., 1999), and has been the most commonly used measure of repetitive behaviour in the ASD literature. Despite the statistical limitations that would have precluded the use of ADI-R scores in the current analysis, work by Tadevosyan–Leyfer et al. (2003) has shown that individual repetitive behaviour items on the ADI-R may be used as proxies to continuously distributed measures of repetitive behaviour. As such, future work may compare associations found between resting-state functional connectivity and scores on both the ADI-R and RBS-R. Given the complex presentation and etiology of repetitive behaviours in ASD, many approaches to the neurobiological profiling of symptoms are necessary. In this regard, both categorical and dimensional classification of symptoms and their relation to neural function in ASD will help to better conceptualize the symptoms of this disorder. Taken together, these findings imply that repetitive behaviours are associated with unique neural underpinnings that are characterized by connectivity patterns associated with inhibitory and attentional processes. The current results also assert that in addition to being a distinct statistical factor, sameness behaviour may be delineated from other types of repetitive behaviour on a neurobiological level, and parallel previous work that has supported this hypothesis (Hollander et al., 2005; Langen et al., 2009; 2013; Sears et al., 1999). These findings may be clinically useful for the development of targeted interventions for individuals displaying idiosyncratic profiles of repetitive behaviour. Experimental Procedure4.1 Materials and MethodsData from 36 individuals with an ASD and 35 typically developing (TD) control subjects was included in initial preprocessing. Data from 9 subjects (4 ASD, 5 TD) was removed due to excessive head motion (greater than 2mm in any direction), leaving 30 ASD (M age = 16 +/- 3.0) and 32 TD subjects (M age = 15+/- 3.0) in the final analysis (Table 4). Subjects were recruited as a part of the NeuroDevNet Autism Demonstration Project, a multisite study to examine brain structural and behavioural development in children with ASD (Zwaigenbaum et al., 2011). All subjects gave written informed consent/assent where applicable. ASD participants had a pre-existing diagnosis, which was assigned by a licensed clinician prior to participating in the study. ASD diagnoses were supported according to DSM-IV-TR criteria, using the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2001) and the ADI-R (Lord et al., 1994); these assessments were completed with all participants upon entering the study. Completion of these instruments yielded a score at or above an algorithm threshold, which confirmed the presence of an ASD. Participants with a primary psychiatric diagnosis (other than ASD), previous head injury, epilepsy, an ASD-associated genetic disorder or a history of neuromotor impairment were excluded. ASD participants were medication free at the time of scanning. Control participants with a first-degree relative with ASD were also excluded. Intelligence was measured by either the Wechsler Abbreviated Scale of Intelligence (WASI) (Wechsler, 1999) or the WASI-Second Edition (Wechsler, 2011). Repetitive behaviour scores were derived from the RBS-R (Bodfish et al., 2000) and included subscales for: stereotyped behaviour; self-injurious behaviour; compulsive behaviour; ritualistic behaviour; sameness behaviour; and restricted behaviour. Each subscale contained between 4 and 11 items and items were scored using a range from 0 (behaviour does not occur) to 3 (behaviour occurs and is a severe problem). Thus, the full possible range for each subscale score differed depending on the number of items it contained (see Table 4). Bodfish and Lewis (2002) have demonstrated good subscale inter-rater reliability (ranging from 0.55 to 0.78) and subscale test-retest reliability (ranging from 0.52 to 0.96) of the RBS-R. Additionally, the scale has been validated using factor analyses yielding satisfactory Cronbach’s alpha subscale scores, ranging from 0.78 to 0.91 (Lam & Aman, 2007). It has also been validated for use across development, including samples of young children with autism (Mirenda et al., 2010).4.2 Imaging Data AcquisitionMR data was obtained from the Hospital for Sick Children and the Montreal Neurologic Institute using Siemens TimTrio 3 Telsa scanners and identical imaging sequences. All subjects underwent a 4:43 min resting-state MRI scan using an axial 2D EPI sequence, TR/TE = 2340/30 ms; 40 interleaved slices; thickness = 3.5 mm; voxel size=3.5×3.5×3.5mm; flip angle = 70?; FOV = 224 x 224mm; matrix = 64 × 64 x 40mm; 120 acquisitions. A high-resolution structural T1 scan was acquired using a sagittal three-dimensional magnetization-prepared rapid acquisition gradient echo (3D-MPRAGE). Subjects were instructed to relax and to keep their eyes open and gaze centered on a fixation cross for the duration of the scan.4.3 Data PreprocessingData were preprocessed using custom scripts created in MATLAB and run through SPM8 (). Preprocessing involved discarding the first 3 volumes of each subjects fMRI scan to allow for T2 equilibration effects, motion correction via realignment of the time series based on a least squares approach and a six parameter, rigid-body transformation (Ashburner et al., 2013), unwarping of the time series to further reduce the susceptibility-distortion-by-movement interaction (Andersson, Hutton, Ashburner, Turner, & Friston, 2001), coregistration of functional images to each subject’s T1-weighted anatomical image, segmentation of each subject’s T1-weighted anatomical image into gray matter, white matter and cerebral spinal fluid tissue priors, normalization of all images to standard Montreal Neurologic Institute (MNI) space, resampling into 2mm3 isotropic voxels, and smoothing with a 6mm 3D full width half maximum kernel. Data was then further preprocessed to remove possible confounding variance from white matter and cerebral spinal fluid BOLD signal, as well as from head motion artifacts. This noise was removed using the default preprocessing steps in the CONN fMRI Functional Connectivity Toolbox () and using the principle component based noise correction method “CompCor” (Behzadi, Restom, Liau, & Liu, 2007), which has been shown to more effectively remove motion artifacts compared to traditional mean signal methods (Muschelli et al., 2014). Motion scrubbing was also incorporated into preprocessing and removed additional scans identified as motion outliers (Power, Barnes, Snyder, Schlaggar, & Petersen, 2012). A t test revealed no significant differences in head motion between the ASD and control group, T (60) = 1.18, p (two-sided) = 0.24. All preprocessed scans were then band pass filtered at 0.008 – 0.09 Hz. 4.4 Region of Interest DefinitionA priori regions of interest (ROIs) were set in several cortical and subcortical areas in order to examine connectivity patterns in large-scale neural networks that have been previously implicated in repetitive behaviours, including the default mode network (Monk et al., 2009; Weng et al., 2010), salience network (Uddin et al., 2013), limbic (Dziobek et al., 2006; Glerean et al., 2016; Rojas et al., 2006) and frontostriatal networks (Cascio et al., 2014; Dichter et al., 2012, Eisenberg et al., 2015; Hollander et al., 2005; Langen et al., 2013; Qiu et al., 2016; Sears et al., 1999; Turner et al., 2006), and thalamocortical (Mizuno et al., 2006), motor (Villalobos et al., 2005) and cognitive control networks (Sabatino et al., 2013) Accordingly, predefined 10 mm sphere ROIs from the CONN toolbox were centered at MNI coordinates: left medial prefrontal cortex (MPFC; -1, 49, -5), left posterior cingulate cortex (PCC; -6, -52, 40), the bilateral (left/right) inferior parietal lobes (IPL; -46, -70, 36; 46, -70, 36) (Fox et al., 2005) and bilaterally (left/right) in the insular cortex/Brodmann area (BA) 13 (-40, 14, -8; 40, 14, -8) and primary visual cortex/BA 17 (-8, -96, 2; 16, -96, 4). Subcortical ROIs were manually created using probability maps from the Harvard-Oxford Subcortical Structural Atlas in FSL (Desikan et al., 2006; Frazier et al., 2005; Goldstein et al., 2007; Makris et al., 2006) that were binarised at a 10% threshold. These ROIs were placed bilaterally and centered at MNI coordinates (left/right): amygdala (-24, -4, -18; 26, 0, -22), hippocampus (-28, -18, -18; 28, -12, -22), nucleus accumbens (-10, 14, -6; 10, 12, -8), thalamus (-12, -24, 6; 14, -24, 6), globus pallidus (-18, -2, -2; 20, -4, -2), putamen (-22, 8, -4; 22, 10, -4) and caudate nucleus (-12, 16, 4; 12, 18, 2). 4.5 Functional Connectivity AnalysisTo identify differences in intrinsic functional connectivity patterns between ASD and control subjects, seed-to-voxel and ROI-to-ROI based analyses were implemented using the CONN Toolbox (). Each ROI’s time-series was defined as the mean BOLD activation within the ROI voxels (Whitfield-Gabrieli & Nieto-Castanon, 2012). Subject-specific connectivity maps were produced in a first level analysis, with motion parameters for each individual subject included as covariates of no-interest. Second-level seed-to-voxel and ROI-to-ROI analyses were then conducted to examine group differences in connectivity patterns, while controlling for significant differences in gender and IQ scores between the ASD and TD group (Tables 1 and 2). Significant differences in gender and IQ scores between the two groups were controlled for by dummy coding the gender variable with arbitrary 0 and 1 values, and then entering both gender and IQ scores into the general linear model as covariates of no interest. Additionally, as the number of females in the TD group (n= 16) was significantly greater than the number of females in the ASD group (n=3), two supplementary analyses were performed in addition to controlling for this gender difference in the main analyses. These supplementary analyses were completed to provide additional evidence that the current results were not influenced by the prominent gender difference between the ASD and TD group (5. Supplementary Material). Finally, to examine the relationship between RBS-R scores and functional connectivity within the ASD group, ROI-to-ROI analyses were performed, using bivariate correlation analysis within the ASD group only. First, the correlation between total, overall RBS-R scores and functional connectivity within the ASD group was examined. Then, following the dimensional structure of the RBS-R, an analysis was performed in order to examine more specific relationships between RBS-R subscale scores and functional connectivity within the ASD group. This subscore analysis was done by first conducting a PCA decomposition of the RBS-R subscale scores. Subsequent to the PCA decomposition, a correlation analysis was performed using the CONN toolbox; this correlation analysis was first implemented as a multivariate F test to examine for an effect among either of PC 1 or PC 2 (i.e., the PC’s accounting for the largest amount of variance in the RBS-R data). Post hoc testing included examination of the simple main effect of each of the RBS-R components on the connectivity patterns that were significant in the multivariate F test. A Bonferroni correction was applied to the post hoc data by dividing alpha = 0.05 by two (for each of the two RBS-R components identified during the PCA decomposition), resulting in a significance value of p = 0.025. 4.6 Principal Component Decomposition of RBS-R subscoresGiven the number of a priori selected bilateral ROIs (22) and RBS-R subscales (6) in our analysis, and the comparatively small number of subjects in the ASD group (n = 30), the analysis that examined correlations between functional connectivity and RBS-R subscale scores was implemented after reducing the dimensionality of the RBS-R data using a PCA decomposition of the 6 subscale scores. This decomposition was implemented using a custom script created in MATLAB and served to increase the power of the analysis by reducing the number of covariates entered into the model; rather than entering all 6 RBS-R subscale scores into the model only the first two largest PCA-derived component scores were entered. Cumulatively, these two components accounted for 79.1 percent of the total variance in the RBS-R data, with the first and second components accounting for 64.2 and 14.9 percent of the variance, respectively. The first two components were retained as per the common criteria of component selection that retains the number of components that account for at least 70 to 80 percent of the cumulative variance in the data (O’Rourke & Hatcher, 2013).4.7 Interpretation of the Principal Component Decomposition Table 5 contains the normalized vectors for each PCA-derived component as well as the cumulative percentage of variance in the RBS-R data accounted for per PC. The column for each component contains a normalized vector of six values, which characterize the relative weights associated with that component. These weights can be interpreted as the direction and magnitude of strength (i.e., correlation) between each of the original six RBS-R subscale scores and that component. For example, unlike PC 2 in the second column, the values in the first column of Table 5 indicate that each of the six original RBS-R subscores explain a relatively similar amount of variation in PC 1. On the other hand, PC 2 (Table 5, column 2) contains values indicating that a) self-injurious behaviour, sameness behaviour, and restricted behaviour play a sizable role in explaining the variation in this component; 2) stereotyped, compulsive, and ritualistic behaviour play a significantly smaller role in explaining the variance in this component; 3) there is a considerable positive association between sameness behaviour and this component; 4) there is a strong, inverse relationship between self-injurious behaviour and this component and 5) there is considerable inverse relationship between restricted behaviour and this component. In sum, an examination of the directionality of the strongest values for PC 2 indicates that ASD subjects with positive component 2 scores would tend to score highest on the sameness behaviour subscale, and lowest on the self-injurious and restricted behaviour subscales. Likewise, individuals with SeedBrain Region/Brodmann AreaDirection of Connectivityk-voxelsPeak cluster location (x, y, z)cluster-level p (unc.) cluster-levelp (FWE)PCC (L)Angular Gyrus (R) / BA.39Superior Temporal Gyrus (L) / BA.22NegativePositive197138( +58 -64 +32 )( -64 -40 +22 )0.0004390.0021180.0082780.039319Hippo(R)Associative Visual Cortex (L) / BA.19Positive173( -48 -64 +16 )0.0008020.015117Thalamus (L)Insular Cortex (R) / BA.13Premotor Cortex (L) /BA.6Primary Somatosensory Cortex (R) / BA.2Superior Temporal Gyrus (L) / BA.22Primary Somatosensory Cortex (R) /BA.3Primary Auditory Cortex (L) / BA.42Primary Somatosensory Cortex (L) / BA.3Brodmann Area Not LabelledPositivePositivePositivePositivePositivePositivePositiveNegative372356225191177162158140( +34 -22 +10 )( -26 -34 +72 ) ( +40 -36 +60 ) ( -54 +02 +00 ) ( +46 -24 +36 )( -60 -28 +16 ) ( -44 -20 +50 ) ( +04 +10 +14 ) 0.0000100.0000140.0002290.0005240.0007490.0011130.0012400.0020440.0001930.0002620.0042980.0098170.0140140.0207510.0230910.037766 Putamen (L)Fusiform Gyrus (R) / BA.37Primary Somatosensory Cortex (L) / BA.2PositivePositive235153( +48 -50 -14 )( -50 -24 +54 )0.0001510.0012390.0029410.023819Globus Pallidus (R)Primary Motor Cortex (L) / BA.4Primary Motor Cortex (L) / BA.4PositivePositive500130( -28 -30 +68 )( -36 -18 +44 ) 0.0000010.0026450.0000200.048980negative PC 2 scores would tend to score highest on the self-injurious and restricted behaviour subscales and lowest on the sameness behaviour subscale.Tables and FiguresTable 1. Significant differences in seed-to-voxel resting-state functional connectivity (ASD > TD, 2-sided contrast).PCC – Posterior Cingulate Cortex; Hippo - Hippocampusp (unc.) – statistically uncorrected p-valuesp FWE – statistically corrected using the Family Wise Error Rate (Benjamini & Hochberg, 1995)*Positive and negative direction of connectivity refers to regions where ASD subjects demonstrated over or under-connectivity compared to TD subjects, respectively.Table 2. Significant differences in ROI-to-ROI resting-state functional connectivity (ASD >TD, 2-sided contrast).SeedBrain Region/Brodmann AreaDirection of ConnectivitybetaT(58)p (unc.) p (FDR)Hippo(L)Perirhinal Cortex (R)/BA.35Negative-0.26-3.890.0002610.026574Thalamus (L)Primary Somatosensory Cortex (R)/BA.3Primary Somatosensory Cortex (R)/BA.2Primary Auditory Cortex (L)/BA.41Premotor Cortex (L)/BA.6Primary Somatosensory Cortex (L)/BA.3Primary Motor Cortex (R)/BA.4Primary Auditory Cortex (L)/BA.42Primary Somatosensory Cortex (R)/BA.1Premotor Cortex (R)/BA.6Primary Auditory Cortex (R)/BA.41Primary Motor Cortex (L)/BA.4Insular Cortex (R)/BA.13Primary Somatosensory Cortex (L)/BA.1PositivePositivePositivePositivePositivePositivePositivePositivePositivePositivePositivePositivePositive0.28 0.24 0.21 0.22 0.27 0.26 0.21 0.24 0.23 0.22 0.24 0.21 0.204.153.683.643.613.563.513.463.313.243.233.183.082.870.0001110.0005070.0005900.0006300.0007490.0008800.0010110.0015930.0020050.0020600.0023920.0031360.0057510.0113610.0147380.0147380.0147380.0147380.0147380.0147380.0203090.0210130.0210130.0221790.0266570.045125Putamen (L)Primary Motor Cortex (L)/BA.4Primary Somatosensory Cortex (L)/BA.2Primary Somatosensory Cortex (L)/BA.1Primary Motor Cortex (R)/BA.4PositivePositivePositivePositive0.220.220.180.193.623.563.453.410.0006250.0007520.0010520.0011890.0303230.0303230.0303230.030323Globus Pallidus (R)Primary Motor Cortex (L)/BA.4Primary Somatosensory Cortex (L)/BA.3PositivePositive0.240.204.303.510.0000670.0008830.0068160.045035Hippo – Hippocampusp (unc.) – statistically uncorrected p-valuesp (FDR) – statistically corrected using the False Discovery Rate (Benjamini & Hochberg, 1995)*Positive and negative direction of connectivity refers to regions where ASD subjects demonstrated over or under-connectivity compared to TD subjects, respectively.Table 3. Significant results from a bivariate correlation analysis demonstrating the relationship between two PCA-derived factors from the RBS-R and connectivity within the ASD group. Multivariate F Test for any effect among RBS-R components 1 and 2 Post Hoc Test of Simple Main EffectsSeedBrain Region/Brodmann Areabeta (effect size)F(2,27)p < 0.01, FDRCorrelation with RBS-R ComponentR2beta (effect size)T (28)p < 0.025 (Bonferroni correction)RLPIFG, pars triangularis (R)/BA.450.8711.640.02305Component 10.030.210.890.38319Component 20.440.854.650.00007*RPL – Right Inferior Parietal Lobe; IFG – Inferior Frontal GyrusTable 4. Demographic Data.Autism Group(n = 30)Control Group(n=32)T(df) = t value/ X2, p valueGender: Male/Female27/316/16X2 = 9.85, p = 0.0017*Age: Mean (SD), Range16 (3.0), 10 – 21 yrs15 (3.0), 10 – 20 yrsT (59.7) = 1.3, p= 0.171WASI IQ: Mean (SD)102.88 (17.2)114 (9.3)T (45.8) = - 3.1, p= 0.003 *RBS-R Overall total scoreMean (SD), Range21.22 (17.32), 0 - 129RBS-R Stereotyped behaviour Mean (SD), Range2.83 (2.29), 0 - 18RBS-R Self-injurious behaviour Mean (SD), Range1.74 (1.69), 0 - 24RBS-R Compulsive behaviourMean (SD), Range3.09 (3.95), 0 - 24RBS-R Ritualistic behaviourMean (SD), Range4.00 (3.73), 0 - 18RBS-R Sameness behaviourMean (SD), Range7.04 (6.56), 0 - 33RBS-R Restricted behaviourMean (SD), Range2.52 (2.50), 0 - 12Table 5. Normalized vectors for each PCA-derived component.RBS-R SubscaleComponent 1 (64.15 cum %)Component 2(79.09 cum %)Component 3(86.9 cum %)Component 4(92.48 cum %)Component 5(96.88 cum %)Component 6(100 cum %)I. Stereotyped behaviour-0.43700.18270.3836-0.13720.7715-0.1206II. Self-injurious behaviour-0.3209-0.7129-0.4244-0.40220.0963-0.1939III. Compulsive behaviour-0.45280.14560.1567-0.4573-0.34910.6467IV. Ritualistic behaviour-0.42400.1231-0.56200.59700.16690.3239V. Sameness behaviour-0.41320.4828-0.1840-0.1147-0.3770-0.6380VI. Restricted behaviour-0.3880-0.43460.54630.4906-0.3220-0.1331Figure 1. Clusters from a seed to voxel analysis (ASD>TD, 2-sided contrast, p <0.05, FWE) showing significant connectivity with each ROI. Warm colours indicate clusters demonstrating increased connectivity with that ROI in ASD subjects compared to controls. Cool colours indicate clusters demonstrating decreased connectivity with that ROI in ASD subjects compared to controls. Figure 2. Significant patterns of ROI-to-ROI functional connectivity (ASD > TD, 2-sided contrast, p < 0.05 FDR). ROIs are indicated by black circles. Areas of the brain demonstrating significantly greater ROI-to-ROI connectivity in ASD subjects compared to controls are displayed using red connectivity lines. Areas of the brain demonstrating significantly less ROI-to-ROI connectivity in ASD subjects compared to controls are displayed using blue connectivity lines. Top left: Regions demonstrating under-connectivity with left hippocampus. Top right: Regions demonstrating over-connectivity with left putamen. Bottom left: Regions demonstrating over-connectivity with left thalamus. Bottom right: Regions demonstrating over-connectivity with right globus pallidus.AcknowledgmentsThis research was supported by Kids Brain Health Network (formerly NeuroDevNet) and the Women and Children’s Health Research Institute. Author LZ is supported by the Stollery Children’s Hospital Foundation Chair in Autism. We are very grateful to the following for providing the segmentations used to create the Harvard Cortical and Subcortical Structural Atlases: David Kennedy and Christian Haselgrove, Centre for Morphometric Analysis, Harvard; Bruce Fischl, the Martinos Center for Biomedical Imaging, MGH (NIH grants P41-RR14075, R01 RR16594-01A1, R01 NS052585-01); Janis Breeze and Jean Frazier from the Child and Adolescent Neuropsychiatric Research Program, Cambridge Health Alliance (NIH grants K08 MH01573, K01 MH01798); Larry Seidman and Jill Goldstein from the Department of Psychiatry of Harvard Medical School.ReferencesAgam, Y., Joseph, R.M., Barton, J.J.S, Manoach, D.S., 2010. Reduced cognitive control of response inhibition by the anterior cingulate cortex in autism spectrum disorders. NeuroImage. 52, 336-347. doi:10.1016/j.neuroimage.2010.04.010Ambrosino, S., Bos, D.J., van Raalten, T.R., Kobussen, N.A., van Belle, J., Orange, B., Durston, S., 2014. Functional connectivity during cognitive control in children with autism spectrum disorder: an independent component analysis. J Neural Transm. 121, 1145 – 1155. doi:10.1007/s00702-014-1237-8Anderson, R.A., 2011. Inferior parietal lobule function in spatial perception and visuomotor integreation. Comprehensive Physiology. 438 – 518. doi: 10.1002/cphy.cp010512Andersson, J.L.R., Hutton, C., Ashburner, J., Turner, R., Friston, K., 2001. Modelling geometric deformations in EPI time series. Neuroimage. 13, 90 -919. doi:10.1006/nimg.2001.0746Aron A.R., Fletcher P.C., Bullmore E.T., Sahakian B.J., Robbins T.W., 2003. Stop-signal inhibition disrupted by damage to right inferior frontal gyrus in humans (vol 6, pg 115, 2003)?Nat. Neurosci. 6:1329.Ashburner, J., Barnes, G., Chen, C., Daunizeau, J., Flandin, G., Friston, K., et al., 2013. SPM8 Manual, Trust Centre for Neuroimaging, London, UK. , Y., Restom, K., Liau, J., Liu, T.T., 2007. A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. NeuroImage. 37(1), 90-101. doi:10.1016/j.neuroimage.2007.04.042Benjamini, Y. and Hochberg, Y., 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 57(1), 289–300. Stable URL , S.L., Richler, J., Lord, C., 2006. Association between restricted and repetitive behaviors and nonverbal IQ in children with autism spectrum disorders. Child Neuropsychol. 12(4–5), 247–67.Bodfish, J. W., and Lewis, M. H., 2002. Repetitive Behavior in Autism. Paper presented at the International Meeting for Autism Research. (IMFAR), Orlando, FL.Bodfish, J. W., Symons, F. J., Parker, D. E., Lewis, M. H., 2000. Varieties of repetitive behavior in autism: Comparisons to mental retardation. Journal of Autism and Developmental Disorders. 30, 237–243.Bos, D., van Raalten, T., Oranje, B., Smits, A.R., Kobussen, N.A., van Belle, J., Rombouts, S.A.R.B., Durston, S., 2014. Developmental differences in higher-order resting-state networks in autism spectrum disorder. Neuroimage Clin. 820 – 827.Boucher, J., Mayes, A., Bigham, S., 2012. Memory in autistic spectrum disorder. Psychol Bull, 138, 458 – 496.Cascio, C.J., Foss-Feig, J.H., Heacock, J., Schauder, K.B., Loring, W.A., Rogers, B.P., Pryweller, J.R., Newsom, C.R., Cockhren, J., Cao, A., Bolton, S., 2014. Affective neural response to restricted interests in autism spectrum disorders. J Child Psychol Psychiatry. 55(2), 162-171. doi: 10.1111/jcpp.12147 Cerliani, L., Mennes, M., Thomas, R.M., Di Martino, A., Thioux, M., Keysers, C., 2015. Increased functional connectivity between subcortical and cortical resting-state networks in autism spectrum disorder. JAMA Psychiatry. 72(8), 767-777Chao, H., Chen, H., Samaco, R.C., Xue, M., Chahrour, M., Yoo, J., Neul, J.L., Gong, S., Lu, H., Heintz, N., Ekker, M., Rubenstein, J.L.R., Noebels, J.L., Rosenmund, C., Zoghbi, H.Y., 2010. GABAergic dysfunction mediates autism-like stereotypies and Rett syndrome phenotypes. Nature. 468(7321), 263 – 269. doi: 10.1038/nature09582Clery, H., Andersson, F., Bonnet-Brilhault, F., Philippe, A., Wicker, B., Gomot, M., 2013. fMRI investigation of visual change detection in adults with autism. Neuroimage Clin. 2, 303-312.Cooper, R.A., Richter, F.R., Bays, P.M., Plaisted-Grant, K.C., Baron-Cohen, S., Simons, J.S., 2017. Reduced hippocampal functional connectivity during episodic memory retrieval in autism. Cerebral Cortex, 27 (2), 888 – 902.Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neuroscience, 3(3), 201-215.Cuccaro, M.L., Shao, Y., Grubber, J., Slifer, M., Wolpert, C.M., Donnelly, S.L., Abramson, R.K., Ravan, S.A., Wright, H.H., DeLong, G.R., Pericak-Vance, M.A., 2003. Factor analysis of restricted and repetitive behaviors in autism using the Autism Diagnostic Interview –R. Child Psychiatry Hum Dev. 34(1), 3 – 17. doi: 10.1023/A:1025321707947Desikan, R.S., Ségonne, F., Fischl, B., Quinn, B.T., Dickerson, B.C., Blacker, D., Buckner, R.L., Dale, A.M., Maguire, R.P., Hyman, B.T., Albert, M.S., Killiany, R.J., 2006. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 31(3), 968-980. doi:10.1016/j.neuroimage.2006.01.021Dichter, G.S., Felder, J.N., Green, S.R., Rittenberg, A.M., Sasson, N.J., Bodfish, J.W., 2012. Reward circuitry function in autism spectrum disorders. SCAN. 7, 160-172. doi: 10.1093/scan/nsq095Di Martino, A., Kelly, C., Grzadzinski, R., Zio, X.N., Mennes, M., Mairena, M.A., Lord, C., Castellanos, F.X., Millham, M.P., 2011. Aberrant striatal functional connectivity in children with autism. Biol Psych. 69(9), 847 – 856. doi:10.1016/j.biopsych.2010.10.029Dziobek, I., Fleck, S., Rogers, K., Wolf, O.T., Convit, A., 2006. The ‘amygdala theory of autism’ revisited: linking structure to behaviour. Neuropsychologia. 44 (10), 1891 – 1899. doi:10.1016/j.neuropsychologia.2006.02.005Eisenberg, I.W., Wallace, G.L., Kenworthy, L., Gotts, S.J., Martin, A., 2015. Insistence on sameness relates to increased covariance of gray matter structure in autism spectrum disorder. Molecular Autism. 6, 54. doi:10.1186/s13229-015-0047-7Emerson, R.W., Adams, C., Nishino, T., Hazlett, H.C., Wolff, J.J., Zwaigenbaum, L., Constantino, J.N., Shen, M.D., Swanson, M.R., Elison, J.T., Kandala, S., Estes, A.M., Botteron, K.N., Collins, L., Dager, S.R., Evans, A.C., Gerig, G., Gu, H., McKinstry, R.C., Paterson, S., Schultz, R.T., Styner, M., IBIS Network, Schlaggar, B.L., Pruett Jr, J.R., Piven, J., 2017. Functional neuroimaging of high-risk 6-month-old infants predicts a diagnosis of autism at 24 months of age. Sci Transl Med. 9, eaag2882.Fox, M.D., Snyder, A.Z., Vincent, J.L., Corbetta, M., Van Essen, D.C., Raichle, M.E., 2005. The human brain is intrinsically organized into dynamic, anticorrelated functional networks. PNAS. 102(27), 9673-9678. doi:10.1073/pnas.0504136102Frazier, J.A., Chiu, S., Breeze, J.L., Makris, N., Lange, N., Kennedy, D.N., Herbert, M.R., Bent, E.K., Koneru, V.K., Dieterich, M.E., Hodge, S.M., Rauch, S.L., Grant, P.E., Cohen, B.M., Seidman, L.J., Caviness, V.S., Biederman, J., 2005. Structural brain magnetic resonance imaging of limbic and thalamic volumes in pediatric bipolar disorder. Am J Psychiatry. 162(7), 1256-1265. doi: 10.1176/appi.ajp.162.7.1256Friedman, S.D., Shaw, D.W., Artru, A.A., Richards, T.L., Gardner, J., Dawson, G., Posse, S., Dager, S.R., 2003. Regional brain chemical alterations in young children with autism spectrum disorder. Neurology. 60(1), 100–107. doi: 10.?1212/?WNL.?60.?1.?100Garner, J.P., Meehan, C.L., Mench, J.A., 2003. Stereotypies in caged parrots, schizophrenia and autism: evidence for a common mechanism. Behav Brain Res. 145(1/2), 125 – 134.Glerean, E., Pan, R.K., Salmi, J., Kujala, R., Lahnakoski, J.M., Roine, U., Nummenmaa, L., Leppamaki, S., Nieminen-von Wendt, T., Tani, P., Saramaki, J., Sams, M., Jaaskelainen, I.P., 2016. Reorganization of functionally connected brain subnetworks in high-functioning autism. Hum Brain Mapp. 37(3), 1066 – 1079.Goldstein, J.M., Seidman, L.J., Makris, N., Ahern, T., O'Brien, L.M., Caviness, V.S. Jr., Kennedy, D.N., Faraone, S.V., Tsuang, M.T., 2007. Hypothalamic abnormalities in schizophrenia: sex effects and genetic vulnerability. Biol Psychiatry. 61(8), 935-945. doi:10.1016/j.biopsych.2006.06.027Gomot, M., Bernard, F.A., Davis, M.H., Belmonte, M.K., Ashwin, C., Bullmore, E.T., Baron-Cohen, S., 2006. Change detection in children with autism: an auditory event-related fMRI study. Neuroimage. 29(2), 475 – 484. doi: 10.1016/j.neuroimage.2005.07.027Hampshire, A., Chamberlain, S.R., Monti, M.M., Duncan, J., Owen, A.M., 2010. The role of the right inferior frontal gyrus: inhibition and attentional control. Neuroimage. 50(3-3), 1313-1319.Han, S., Tai, C., Jones, C.J., Scheuer, T., Catterall, W.A., 2014. Enhancement of inhibitory neurotransmission by GABAA receptors having A2,3- subunits ameliorates behavioural deficits in a mouse mode of autism. Neuron, 81(6), 1282 – 1289.Harada M., Taki M. M., Nose A., Kubo H., Mori K., Nishitani H., Matsuda, T., 2011.?Non-invasive evaluation of the GABAergic/glutamatergic system in autistic patients observed by MEGA-editing proton MR spectroscopy using a clinical 3 tesla instrument.?J. Autism Dev. Disord.?41, 447–454. doi:10.1007/s10803-010-1065-0Hardan, A.Y., Girgis, R.R., Lacerda, A.L.T., Yorbik, O., Kilpatrick, M., Keshavan, M.S., Minshew, N.J., 2006. Magnetic resonance imaging study of the orbitofrontal cortex in autism. J Child Neurol. 21, 866. doi:10.1177/08830738060210100710 Hardan, A.Y., Minshew, N.J., Melhem, N.M., Srihari, S., Jo, B., Bansal, R., Kevshavan, M.S., Stanley, J.A., 2008. An MRI and proton spectroscopy study of the thalamus in children with autism. Psychiatry Res. 163, 97–105. doi: 10.1016/j.pscychresns.2007.12.002Hazrati, L.N., and Parent, A., 1991. Contralateral pallidothalamic and pallidotegmental projections in primates: an anterograde and retrograde labeling study. Brain Res, 576(2), 212 – 223. doi:10.1016/0006-8993(91)90798-ZHollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M., Licalzi, E., Wasserman, S., Soorya, L., Buchsbaum, M., 2005. Striatal volume on magnetic resonance imaging and repetitive behaviors in autism. Biol Psychiatry. 58(3), 226-232. , V., Pickles, A., Cook, E.H., Risi, S., Lord, C., 2007. Using the autism diagnostic interview–revised to increase phenotypic homogeneity in genetic studies of autism. Biol Psychiatry. 61(4),438–48.Jiao, Y., Chen, R., Ke, X., Chu, K., Lu, Z., Herskovits, E.H., 2010. Predictive models of autism spectrum disorder based on brain regional cortical thickness. Neuroimage, 50 (20), 589 – 599.Jones, T.B., Bandettini, P.A., Kenworthy, L., Case, L.K., Milleville, S.C., Martin, A., Birn, R.M., 2010. Sources of group differences in functional connectivity: An investigation applied to autism spectrum disorder. Neuroimage, 49, 401 – 414.Just, M.A., Cherkassy, V.L., Keller, T.A., Kana, R.K., Minshew, N.J., 2004. Cortical activation and synchronization during sentence comprehension in high functioning autism: Evidence of underconnectivity. Brain, 127, 1811-1821. Just, M.A., Keller, T.A., Malave, V.L., Kana, R.K., Varma, S., 2012. Autism as a neural systems disorder: A theory of frontal-posterior underconnectivity. Neurosci Biobehav Rev, 36, 1292 – 1313.Kadosh, C.K., Krause, B., King, A.J., Near, J., Kadosh, R.C., 2015. Linking GABA and glutamate levels to cognitive skill acquisition during development. Hum Brain Mapp. 36(11), 4334-4345. Kaster, S., Pinsk, M.A., De Weerd, P., Desimone, R., Ungerleider, L.G., 1999. Increased activity in human visual cortex during directed attention in the absence of visual stimulation. Neuron, 22(4), 751-761.Lam, K.S., and Aman, M.G., 2007. The repetitive behavior scale – revised: Independent validation in individuals with autism spectrum disorder. J Aut Dev Disord. 37, 855 – 866.Lam, K.S., Bodfish, J.W., Piven, J., 2008. Evidence for three subtypes of repetitive behavior in autism that differ in familiality and association with other symptoms. J Child Psychol Psychiatry. 49(11), 1193–1200. doi: 10.1111/j.1469-7610.2008.01944.xLangen, M., Schnack, H.G., Nederveen, H., Bos, D., Lahuis, B.E., de Jonge, M.V., van Engeland, H., Durston, S., 2009. Changes in the developmental trajectories of striatum in autism. Biol Psychiat, 66(4), 327 – 333.Langen, M., Durston, S., Kas, M.J., van Engeland, H., Staal, W.G., 2011. The neurobiology of repetitive behavior:…and men. Neurosci Biobehav Rev. 35(3), 356 – 265. doi: 10.1016/j.neubiorev.2010.02.005Langen, M., Bos, D., Noordermeer, S.D., Nederveen, H., van Engeland, H., Durston, S., 2013. Changes in the development of the striatum are involved in repetitive behaviors in autism. Biol Psychiatry. 75(5), 405-411. doi:10.1016/j.biopsych.2013.08.013Leekam, S.R., Uljarevic, M., Prior, M.R., 2011. Restricted and repetitive behaviours in autism spectrum disorders: a review of research in the last decade. Psychol Bull. 137 (4), 562 – 593. doi: 10.1037/a0023341Lewis, M.H., Tanimura, Y., Lee, L.W., Bodfish, J.W., 2007. Animal models of restricted repetitive behavior is autism. Beh Brain Res. 176, 66 – 74. Lord, C., Risi, S., Lambrecht, L., Cook, E.H. Jr., Leventhal, B.L., DiLavore, P.C., Pickles, A., Rutter, M., 2001. The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord. 30, 205–223. doi:10.1023/A:1005592401947Lord, C., Rutter, M., Le Couteur, A., 1994. Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord. 24, 659–685. doi:10.1007/BF02172145Makris, N., Goldstein, J.M., Kennedy, D., Hodge, S.M., Caviness, V.S., Faraone, S.V., Tsuang, M.T., Seidman, L.J., 2006. Decreased volume of left and total anterior insular lobule in schizophrenia. Schizophr Res. 83(2-3), 155-171. doi:10.1016/j.schres.2005.11.020Menon V., Adleman N.E., White C.D., Glover G.H., Reiss A.L., 2001. Error-related brain activation during a Go/NoGo response inhibition task.?Hum. Brain Mapp.12:131–143.Mirenda, P., Smith, I.M., Vaillancourt, T., Georgiades, S., Duku, E., Szatmari, P., Bryson, S., Fombonne, E., Roberts, W., Volden, J., Waddell, C., Zwaigenbaum, L., The Pathways in ASD Study Team., 2010. Validating the repetitive behavior scale –revised in young children with autism spectrum disorder. J Autism Dev Disord. 40: 1521 – 1530. doi: 10.1007/s10803-010-1012-0?Mizuno, A., Villalobos, M.E., Davies, M.M., Dahl, B.C., Muller, R., 2006. Partially enhanced thalamocortical functional connectivity in autism. Brain Res. 1104, 160 – 174. doi:10.1016/j.brainres.2006.05.064Mosconi, M.W., Kay, M., D’Cruz, A.-M., Seidenfeld, A., Guter, S., Stanford, L.D., Sweeney, J.A., 2009. Impaired inhibitory control is associated with higher-order repetitive behaviours in autism spectrum disorders. Psychol Med. 39(9), 1559-1566/ doi:10.1017/S0033291708004984Monk, C.S., Peltier, S.J., Wiggins, J.L., Weng, S., Carrasco, M., Risi, S., Lord, C., 2009. Abnormalities of intrinsic functional connectivity in autism spectrum disorders. Neuroimage. 47, 764-772. doi:10.1016/j.neuroimage.2009.04.069Muschelli, J., Nebel, M.B., Caffo, B.S., Barber, A.D., Pekar, J.J., Mostofsky, S.H., 2014. Reduction of motion-related artifacts in resting state fMRI using aCompCor. Neuroimage. 96, 22 – 35. doi:10.1016/j.neuroimage.2014.03.028 Nair, A., Treiber, J.M., Shukla, D.K., Shih, P., Muller, R., 2013. Impaired thalamocortical connectivity in autism spectrum disorder: a study of functional and anatomical connectivity. Brain. 136, 1942-1955.Negyessy, L., Hamori, J., Bentivoglio, M., 1998. Contralateral cortical projection to the medialdorsal thalamic nucleus: origin and synaptic organization in the rat. Neuroscience. 84(3), 741 – 753. doi:10.1016/S0306-4522(97)00559-9Noonan, S.K., Haist, F., Muller, R.A., 2009. Aberrant functional connectivity in autism: evidence from low-frequency BOLD signal fluctuations. Brain Res, 1262, 48 – 63.Nordahl, C.W., Dierker, D., Mostafavi, I., Schumann, C.M., Rivera, S.M., Amaral, D.G., Van Essen, D.C., 2007. Cortical folding abnormalities in autism revealed by surface-based morphometry. J Neurosci. 27(43), 11725-11735. doi:10.1523/JNEUROSCI.0777-07.2007O’Rourke, N. and Hatcher, L., 2013. Principal Component Analysis in: A Step-by-Step Approach to Using SAS for Factor Analysis and Structural Equation Modeling, Second Edition, 2nd Edition: SAS Institute Inc., Cary, North Carolina, USAPierce, K., and Courchesne, E., 2001. Evidence for a cerebellar role in reduced exploration and stereotyped behavior in autism. Biol Psychiatry. 49(8), 655-664. doi: 10.1016/S0006-3223(00)01008-8.Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E., 2012. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuro Image. 59, 2142– 2154.Qiu, A., Adler, M., Crocetti, D., Miller, M.I., Mostofsky, S.H., 2010. Basal ganglia shapes predict social, communication and motor dysfunction in boys with autism spectrum disorder. J Am Acad of Child Adolesc Psychiatry. 49(6), 539 – 551. doi: 10.1016/j.jaac.2010.02.012Qui, T., Chang, C., Li, Y., Qian, L., Xiao, C.Y., Xiao, T., Xiao, X., Xiao, Y.H., Chu, K.K., Lewis, M.H., Ke, X., 2016. Two years changes in the development of caudate nucleus are involved in restricted and repetitive behaviors in 2 – 5 year old children with autism spectrum disorder. Developmental Cognitive Neuroscience. 19, 137 – 143. doi:10.1016/j.dcn.2016.02.010Rhinehart, N.J., Tonge, B.J., Iansek, R., McGinley, J., Brereton, A.V., Enticott, P.G., Bradshaw, J.L., 2006. Gait function in newly diagnosed children with autism: cerebellar and basal ganglia related motor disorder. Dev Med Child Neurol. 10, 819 – 824. doi: 10.1111/j.1469-8749.2006.tb01229.xRojas, D.C., Peterson, E., Winterrowd, E., Reite, M.L., Rogers, S.J., Tregallas, J.R., 2006. Regional gray matter volumetric changes in autism associated with social and repetitive behavior symptoms. BMC Psychiatry. 6, 56. doi:10.1186/I47I-244X-6-56Rojas, D.C., Becker, K.M., Wilson, L.B. 2014. Magnetic resonance spectroscopy studies of glutamate and GABA in Autism: Implications for Excitation-Inhibition Imbalance Theory. Curr Dev Disord Report. 2 (1), 46-57.Rubia K., Smith A.B., Brammer M.J., Taylor E., 2003. Right inferior prefrontal cortex mediates response inhibition while mesial prefrontal cortex is responsible for error detection.?Neuroimage.?20:351–358.Sabatino, A., Rittenberg, A., Sasson, N.J., Turner-Brown, L., Bodfish, J.W., Dichter, G.S., 2013. Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord. 43(12), 2903-2913. doi:10.1007/s10803-013-1837-4 Say, G.N., Sahin, B., Aslan, K., Akbas, S., Incesu, L., Ceyhan, M., 2014. Increased laterality of the thalamus in children and adolescents with Asperger’s Disorder: an MRI and proton spectroscopy study. Psychiatry Investig. 11(3), 237 – 242. doi: 10.4306/pi.2014.11.3.237Sears, L.L., Vest, C., Mohammed, S., Bailey, J., Rason, B.J., Piven, J., 1999. An MRI study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol Psychiatry. 23(4), 613- 624. doi:10.1016/S0278-5846(99)00020-2Shafritz, K.M., Dichter, G.S., Baranek, G.T., Belger, A., 2008. The neural circuitry mediating shifts in behavioral response and cognitive set in autism. Biol Psychiatry. 63(10), 974- 980. ?doi:10.1016/j.biopsych.2007.06.028 Sherman, S.M., 2007. The thalamus is more than just a relay. Curr Opin Neurobiol. 17(4), 417 – 422. doi:10.1016/j.conb.2007.07.003Stevens, J.P., 2002. Applied multivariate statistics for the social sciences (4th ed.). Mahwah, NJ: LEA.Szatmari, P., Georgiades, S., Bryson, S., Zwaigenbaum, L., Roberts, W., Mahoney, W., Goldberg, J., Tuff, L., 2006. Investigating the structure of the restricted and repetitive behaviors and interests domain of autism. J Child Psychol Psychiatry. 47(6), 582 – 591. doi:?10.1111/j.1469-7610.2005.01537.xTadevosyan-Leyfer, O., Dowd, M., Mankoski, R., Winklowsky, B., Putnam, S., McGrath, L., Tager-Flusberg, Folstein, S.E., 2003. A principal components analysis of the Autism Diagnostic Interview-Revised. J Am Acad Child Adolesc Psychiatry. 42(7), 864 – 872.Thakkar, K.N., Polli, F.E., Joseph, R., Tuch, D., Hadjikhani, N., Barton, J., Manoach, D.S., 2008. Response monitoring, repetitive behavior and anterior cingulate abnormalities in autism spectrum disorders. Brain. I3I: 2464-2478. , C., Humphreys, K., Jung, K., Minshew, N., Behrmann, M., 2011. The anatomy of the callosal and visual association pathways in high-functioning autism: a DTI tractography study. Cortex. 47(7): 863-873. , J.M., Hall, G,B.C, 2015. Structural and functional neuroimaging of restricted and repetitive behaviour in autism spectrum disorder. J Intellectual Disabl – Diagnosis and Treatment. 3, 21 -34. doi:10.6000/2292-2598.2015.03.01.4Turner M., 1999. Annotation: Repetitive behaviour in autism: a review of psychological ?research. J Child Psychol Psychiatry. 40, 839–849. Turner, K.C., Frost, L., Linsenbardt, D., McIlroy, J.R., Muller, R., 2006. Atypically diffuse functional connectivity between caudate nuclei and cerebral cortex in autism. Behavioural and Brain Functions. 2, 34. doi:10.1186/1744-9081-2-34Uddin, L.Q., Supekar, K., Lynch, C.J., Khouzam, A., Phillips, J., Feinstein, C., Ryali, S., Menon, V., 2013. Salience network-based classification and prediction of symptom severity in children with autism. JAMA Psychiatry. 70(8), 869 – 879. doi:10.1001/jamapsychiatry.2013.104 Villalobos, M.E., Mizuno, A., Branelle, C.D., Kemmotsu, N., Muller, R.A. (2005). Reduced functional connectivity between V1 and inferior frontal cortex associated with visuomotor performance in autism. Neuroimage. 25(3), 916-925. doi: 10.1016/j.neuroimage.2004.12.022Wechsler,D., 1999. Wechsler Abbreviated Scale of Intelligence. Pearson, Minnesota.Wechsler, D., 2011. Wechsler Abbreviated Scale of Intelligence, second ed. Pearson, Minnesota.Weng ,S.J., Wiggins, J.L., Peltier, S.J., Carrasco, M., Risi, S., Lord, C., 2010. Alterations of resting state functional connectivity in the default network in adolescents with autism spectrum disorders. Brain Res. 1313, 202-214. doi:10.1016/j.brainres.2009.11.057Whitfield-Gabrieli, S., Nieto-Castanon, A., 2012. Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connect. 2(3), 125 – 141. doi: 10.1089/brain.2012.0073Wigham, S., Rodgers, J., South, M., McConachie, H., Freeston, M., 2015. The interplay between sensory processing abnormalities, intolerance of uncertainty, anxiety and restricted and repetitive behaviours in autism spectrum disorder. J Autism Dev Disord. 45(4), 943-952. doi: 10.1007/s10803-014-2248-xWilden, A., 1980. System and Structure: Essays in Communication and Exchange, second ed. Tavistock Publications Limited, Great Britain.Zwaigenbaum, L., Scherer, S., Szatmari, P., Fombonne, E., Bryson, S.E., Hyde, K., Anagnostou, E., Brian, J., Evans, A., Hall, G., Nicholas, D., Roberts, W., Smith, I., Vaillancourt, T., Volden, J., 2011. The neurodevnet autism spectrum disorders demonstration project, Semin Pediatr Neurol. 18(1), 40 – 48. doi: 10.1016/j.spen.2011.02.0035. Supplementary Material5.1 Supplementary Analyses of Sex Differences Table 6. Within-subject differences in ROI-to-ROI resting-state functional connectivity (TD male > TD female, 2-sided contrast).SeedBrain Region/Brodmann AreaDirection of ConnectivitybetaT(30)p (unc.) p (FDR)Hippo(L)Perirhinal Cortex (R)/BA.35Negative-0.07-0.970.3383920.990629Thalamus (L)Primary Somatosensory Cortex (R)/BA.3Primary Somatosensory Cortex (R)/BA.2Primary Auditory Cortex (L)/BA.41Premotor Cortex (L)/BA.6Primary Somatosensory Cortex (L)/BA.3Primary Motor Cortex (R)/BA.4Primary Auditory Cortex (L)/BA.42Primary Somatosensory Cortex (R)/BA.1Premotor Cortex (R)/BA.6Primary Auditory Cortex (R)/BA.41Primary Motor Cortex (L)/BA.4Insular Cortex (R)/BA.13Primary Somatosensory Cortex (L)/BA.1PositiveNegativeNegativePositiveNegativeNegativeNegativePositivePositiveNegativeNegativePositiveNegative0.02 -0.01 -0.09 0.01 -0.05 -0.05-0.09 0.02 0.03 -0.04 -0.09 0.01 -0.010.17-0.08-1.150.20-0.52-0.51-1.170.240.38-0.42-0.880.09-0.130.8672830.9332760.2579930.8390820.6057380.6144220.2510040.8094560.7090780.6772880.3859290.9311510.8980820.9713690.9713690.9713690.9713690.9713690.9713690.9713690.9713690.9713690.9713690.9713690.9713690.971369Putamen (L)Primary Motor Cortex (L)/BA.4Primary Somatosensory Cortex (L)/BA.2Primary Somatosensory Cortex (L)/BA.1Primary Motor Cortex (R)/BA.4NegativeNegativeNegativeNegative-0.04-0.12-0.08-0.05-0.52-1.69-1.23-0.720.6042040.1009430.2273220.4782840.9297110.5714520.7786210.929711Globus Pallidus (R)Primary Motor Cortex (L)/BA.4Primary Somatosensory Cortex (L)/BA.3PositiveNegative0.00-0.070.01-1.250.9920320.2208110.9920320.594698As there was a prominent gender difference between the ASD and TD groups, the following analyses were completed in addition to controlling for gender differences in the main analyses. First, it was determined that there were no significant within-subject differences between male and female subjects in the TD group. In other words, there were no differences between male and female TD subjects in the bivariate temporal correlation of each predefined source ROI with every other target ROI in the brain (Whitfield-Gabrieli and Nieto-Castanon, 2012). Results did not survive correction for multiple comparisons at threshold of p(FDR) < 0.05 in any of the 8 cortical or 14 subcortical predefined ROIs. Thus, combining both male and female TD subjects into one group was not confounded by pre-existing differences accounted for by gender. To exemplify this result, connectivity values for this within-subject analysis are displayed in Table 6 for the source and target ROIs found to be significant in the main ROI-to-ROI analysis.Additionally, two analyses that compared ROI-to-ROI connectivity in the ASD group to i) the TD males only and ii) the TD females only, were conducted in order to examine whether the significant findings in the main ROI-to-ROI analysis were preferentially accounted for by either TD gender alone. Both analyses controlled for significant differences in IQ scores between the ASD group and the male and female TD groups. First, after TD females were removed from the analyses, no significant differences in ROI-to-ROI connectivity remained in any of the 8 cortical or 14 subcortical predefined ROIs (pFDR > 0.05 for all ROIs). Similarly, after removing TD males from the analysis and comparing the ASD group to TD females alone, only one significant difference in positive connectivity of the left hippocampus and right primary auditory cortex/ BA.42 (beta = 2.80, T(42) = 3.99, p(FDR) = 0.026) remained; a result that did not survive correction for gender in the main ROI-to-ROI analysis. It therefore appears that the variable of TD gender did not account for the results in the main ROI-to-ROI analyses. Rather, it is likely that combining the TD male and females into one group increased the overall power of the sample and allowed for the detection of true effects.Chapter 5Frontostriatal functional connectivity during inhibitory control task performance is atypically correlated with cortical GABA concentration in autism spectrum disorder: A pilot studyJenna M. Traynor 1, Norm Konyer 2, Michael D. Noseworthy 3, and Geoffrey B.C. Hall 41 PhD Candidate, Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, Ontario, Canada2 Senior Scientific Research Officer, St. Joseph’s Healthcare Hamilton, Imaging Research Centre, Hamilton, Ontario, Canada3 Associate Professor, Electrical and Computer Engineering, McMaster University, Hamilton, Ontario Canada4 Associate Professor, Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, Ontario CanadaTraynor, J.M., Konyer, N., Noseworthy, M.D., & Hall, G.B.C. Frontostriatal functional connectivity during inhibitory control task performance is atypically correlated with cortical GABA concentration in autism spectrum disorder. Submitted to Brain Research, Sept 2018.Abstract and KeywordsDeficits in inhibitory control are correlated with repetitive behaviour in Autism Spectrum Disorder (ASD). In vivo quantification of cortical gamma-aminobutyric acid (GABA) via single proton magnetic resonance spectroscopy (1H MRS) is a powerful tool that can be used to examine GABA in the ASD brain. This pilot study combined functional magnetic resonance imaging (fMRI) with 1H MRS to examine the relationship between functional connectivity during inhibitory control task performance and cortical GABA concentration, in a group of 8 adults with ASD and 7 age- and sex-matched neurotypicals. Intrastriatal under-connectivity during behavioural inhibition was found in ASD subjects, relative to controls. Additionally, an inverse correlation between frontostriatal functional connectivity during behavioural inhibition and cortical GABA was identified in the ASD group, relative to controls. There was no difference in the concentration of cortical GABA between ASD and control subjects. These findings suggest that GABA may have a functional impact on response inhibition and repetitive behaviour in ASD.KEYWORDS: gamma-aminobutyric acid, functional connectivity, functional magnetic resonance imaging, magnetic resonance spectroscopy, autism spectrum disorder, repetitive behavior, inhibitory controlIntroductionRepetitive behaviour (RB) in ASD is characterized by a broad range of behaviours, including unusual sensory behaviour, motor and vocal stereotypies, insistence on sameness in daily routine, unusual preoccupations, and circumscribed interests (Turner, 1999). Deficits in inhibitory control, defined as ‘the suppression of actions and resistance to interference from irrelevant stimuli’ (Bjorklund & Harnishfegar, 1995), have long been hypothesized to contribute to motor and cognitive RB in ASD (Ozonoff, 1991; Schmitz et al., 2006). Indeed, research has demonstrated that deficits in inhibitory control task performance in ASD (Kleinhans et al., 2005; Luna et al., 2006; Ozonoff et al., 1994) are significantly correlated with several indices of motor and cognitive RB (Lopez et al., 2005; Mosconi et al., 2009; South et al., 2007).Given this association, it follows that research examining the neurobiological basis of RB has investigated neural activation during inhibitory control task performance. In fact, tasks that require the inhibition of a prepotent response have been labeled as “repetitive behaviour proxies” (Anagnostou and Taylor, 2011). Most consistently, abnormalities in the frontostriatal network during inhibitory control task performance have been found in ASD subjects, relative to controls (Anagnostou, 2006; Langen et al., 2012; Shafritz et al., 2008). Furthermore, frontostriatal abnormalities have been correlated with indices of RB in ASD (Abbot et al., 2017; Cascio et al., 2013; Dichter et al., 2012; Langen et al., 2010; 2013, Sabatino et al., 2013; Thakkar et al., 2008; Traynor et al., 2018). More specifically, the right inferior frontal gyrus (rIFG; Lee et al., 2009; Traynor et al., 2018, Xiao et al., 2012) and anterior cingulate cortex (ACC; Cascio et al., 2013; Thakkar et al., 2008; Zhou et al., 2016) have been highly implicated in the pathophysiology of RB given their role in impulse control-related (Hampshire et al., 2010), and error monitoring-related (Thakkar et al., 2008) functions, respectively. Finally, the striatum has also been consistently associated with inhibitory control deficits in ASD (Cascio et al., 2013; Dichter et al., 2012; Langen et al., 2010; 2013; Sabatino et al., 2013; Traynor et al., 2018). The striatum is comprised of the caudate nuclei and putamen, and is well recognized as important in the regulation of motor movement, learning, and cognitive control (Rolls, 1994; Villablanca, 2010).The brain’s main inhibitory neurotransmitter, gamma aminobutyric acid (GABA), is intricately involved in frontostriatal brain function. GABAergic interneurons that provide inhibitory input to the striatum are located in the IFG and ACC (Akil et al., 2003; Perlman et al., 2004; Tremblay, Lee, & Rudy, 2016). Additionally, high concentrations of GABA neurons can be found in the striatum (Koos & Tepper, 1999). Therefore, it has been suggested that dysfunction in the GABAergic system may contribute to RB in ASD, and a significant amount of animal research has supported this hypothesis (e.g., Han et al., 2014; review in Kim et al., 2016). Overall, animal studies have demonstrated that functioning of the cortico-basal ganglia-thalamic motor pathway may be compromised in ASD, and mediated by atypical functioning of the GABAergic system (Kim et al., 2016). However, the specific pathways by which the GABAergic system mediates ASD behaviours in humans are complex and poorly understood. For example, different types of RB are thought to be mediated by sub-pathways within the cortico-basal ganglia-thalamic circuit, and these sub-circuits are in turn contingent on more intricate molecular processes (Haber et al., 2009). Further, the interplay between the brain’s dopaminergic and GABAergic systems (Kim et al., 2016), as well as the excitatory/inhibitory (+/-) ratio of glutamate/GABA (Presti et al., 2004) are hypothesized to be factors in the expression of RB. Mouse models have found that stereotyped behaviour can be reduced via the administration of a GABA agonist (Han et al., 2014; Silverman et al., 2015). However, the translation of animal research to human network dynamics has remained limited and there have been no randomized, double-blind, placebo-controlled trials of GABAergic medications for RB in ASD persons. To date, there are no medications that have been identified as efficacious for the treatment of RB in ASD, and off label usage of medications demonstrate limited efficacy for the treatment of RB symptoms. One recent double-blind placebo-controlled trial of bumetanide, which enhances GABAergic inhibition, significantly reduced scores on the Child Autism Rating Scale (Lemonnier et al., 2017). These findings support the utility of GABAergic medications in the treatment of general ASD symptoms, but their application to RB specifically, has not been studied. As such, more research into the neurobiological mechanisms of RB in humans is needed.Recently, in vivo quantification of cortical GABA via single proton magnetic resonance spectroscopy (1H MRS) has emerged as a growing trend in ASD research. When used in combination with 3 Tesla MRI, the MEGAPRESS spectral editing method can resolve GABA from other confounding metabolite signals, such as glutamate and glutamine (Mescher et al., 1998). To date, there have been three studies in the reported literature investigating the in vivo concentration of GABA in ASD adults (Ajram et al., 2017; Port et al., 2016; Robertson et al., 2015). Relative to controls, these studies found unchanged GABA levels in the medial prefrontal cortex/ACC (Ajram et al., 2017), left motor cortex (Robertson et al., 2015) and bilateral occipital cortex (Port et al., 2016). A handful of other investigations have used children and adolescent subjects, and have found either unchanged or decreased GABA in a number of cortical areas (Brix et al., 2015; Cochran et al., 2015; Drenthen et al., 2016; Gaetz et al., 2014; Goji et al., 2107; Harada et al., 2011; Ito et al., 2017; Port et al., 2016; Puts et al., 2016; Rojas et al., 2014). Overall, these studies indicate that relative to neurotypicals, variation in regional GABA in ASD may exist in childhood, but has not been demonstrated in adulthood.The investigations of cortical GABA variation provide no information about the functionality of GABA in ASD. In fact, only two studies to date have associated GABA concentration with ASD symptoms; correlations between GABA and tactile processing (Puts et al., 2016), and scores on the Autism Symptoms Screening Questionnaire (Brix et al., 2015) have been demonstrated in ASD children. Additionally, one study investigated the relationship between GABA and brain function in ASD, by combining MRS with magnetoencephalography (MEG) (Port et al., 2016). This study found an altered developmental trajectory of gamma-band coherence and cortical GABA coupling in ASD children, indicating that alterations in cortical GABA in ASD may have functional consequences during childhood (Port et al., 2016). Indeed, a more informative research question pertains to whether the relationship between cortical GABA and brain function in ASD differs from neurotypicals. This research question may be investigated by combining MRS and task-based functional magnetic resonance imaging (fMRI), to gather important information about the relationship between GABA and brain function in the context of behaviour. Given the established links between RB, inhibitory control, and task-based BOLD activation in GABA-rich areas, the combination of MRS and task-based fMRI is a logical next step. However, the combinations of these methods have not been carried out in an ASD population. As such, the objective of this pilot study was to examine the feasibility and utility of combining MRS and task-based fMRI to investigate the relationship between GABA and task-based fMRI BOLD activation in an ASD population. Additional research aims were to:characterize patterns of frontostriatal connectivity, during an inhibitory control task, performance between ASD and control subjects, and;identify whether cortical GABA concentration is significantly correlated with frontostriatal connectivity patterns between groups. Results from the current study could provide a basis for further investigation into the relationship between GABA and brain function. Further, given the lack of interventions for RB in ASD (Boyd et al., 2012), research into this area is important, and may ultimately provide a foundation for the development of targeted pharmacological and neurocognitive interventions. Methods2.0 Experimental Procedure2.1 Sample Size EstimationWe recruited a sample of n= 16, which is greater than calculations that indicate that a sample size of 12 is adequate to measure a 10% change in MRS quantified metabolite concentration, at an alpha level of 0.05, with 80% power (Stone et al., 2012). Additionally, sample size estimation for fMRI has indicated that approximately 20 subjects are required for high-powered fMRI studies, but that 12 subjects are sufficient for typical activations detected at an alpha threshold of 0.05 with 80% power at the single voxel level (Desmond & Glover, 2002). Due to one subject’s unusable scan, our final n was 15, which is suggested as sufficient for a pilot investigation, as the above calculations are estimates of full sample sizes for non-pilot studies. To further increase fMRI power, we used a specific a priori hypothesis and a small, targeted number of a priori regions of interest (ROIs), which were selected according to previous studies that used larger sample sizes (Cascio et al., 2013; Dichter et al., 2012; Langen et al., 2010; 2013; Lee et al., 2009; Sabatino et al., 2013; Thakkar et al., 2008; Traynor et al., 2018, Xiao et al., 2012; Zhou et al., 2016). We also used the Go/No Go task, which is a well-known and validated inhibitory control task (Uzefovsky et al., 2016). Finally, we employed a lengthy, blocked fMRI design with a large number of trials, both of which have been shown to significantly increase study power; previous studies have estimated that 25 trials are indicated for efficient fMRI paradigms, with upwards of 100 trials required for high-powered fMRI studies (Huettel & McCarthy, 2001), and our task contained 224 trials. Considering these factors in their totality, it is suggested that a final sample size of n=15 was sufficient for the purposes of a pilot investigation.2.2 ParticipantsData from 16 young adult subjects (9 ASD, 7 TD) was included in initial preprocessing. Data from 1 ASD participant was removed due to gross brain abnormality (hydrocephalus), which resulted in 8 ASD and 7 TD subjects in the final analysis (n = 15). There were no significant differences in head motion, age, sex, or IQ scores between the two groups (p < 0.05). Raw IQ scores were approximated using vocabulary and matrix reasoning subtests from the Wechsler Abbreviated Scale of Intelligence – 2nd Edition (WASI-II; Wechsler, 2011), which provides an estimate of overall IQ. Participant demographics are displayed in Table 1. The study was approved by the Hamilton Integrated Research Ethics Board (HiREB) and all subjects provided written consent to participate in the study. All subjects were right-handed, and had no history of head trauma causing loss of consciousness. Exclusion criteria for TD subjects included any psychiatric condition and/or use of psychiatric medication. ASD subjects had a pre-existing ASD DSM diagnosis confirmed by a licensed clinician prior to participating in the study, and provided a copy of diagnostic assessment paperwork prior to scanning. ASD subjects were permitted to have comorbid psychiatric conditions, due to the high prevalence of these conditions in this population. As such, four ASD participants had an anxiety disorder or depression and two subjects had attention deficit hyperactivity disorder (ADHD). These subjects were using SSRI and/or stimulant medication as prescribed. Individuals taking medications that act directly on the GABAergic system, such as benzodiazepines, were excluded from participation. 2.3 Go/No Go TaskTo measure behavioural inhibition, a blocked design using a Go/No-Go task was employed. Stimuli were presented to subjects using a computer and display projected into the bore of the MRI scanner. During the Go/No-Go task, participants selectively inhibited or executed a motor response, depending on whether they were visually presented with a Go signal (i.e., circle; execute) or a No-Go signal (square; inhibit). Subjects responded to the task using a button box in their right hand. The task contained a total of 224 trials, 65 percent of which contained an execute/Go signal (184 trials), and 35 percent of which contained an inhibit/No-Go signal (40 trials). As such, subjects were primed to inhibit a prepotent Go response. All stimuli were preceded by a center fixation and each trial was preceded by a jittered interstimulus interval ranging from 1500 to 2500 ms, so that participants would be less able to predict the onset of the Go or No-Go signal. More specifically, Go blocks contained a total of 112 Go trials, spread across 8 blocks (i.e., 14 Go trials per block). No Go blocks contained a total of 72 Go trials and 40 No-Go trials, spread across 8 blocks (i.e., 9 Go trials and 5 No-Go trials per block). Prior to completing each block, subjects received visual instructions regarding which type of block they were about to complete. Therefore, subjects were aware of which blocks would require them to selectively inhibit a prepotent response, but due to the random order of stimuli during No-Go blocks, they were unaware of when they would need to inhibit their response. The entire task was 12 minutes in length. 2.4 fMRI and MRS Data AcquisitionAll participants were trained in a mock scanner for a 10-minute single session prior to participation in the study, in order to familiarize them with the experience of MRI scanning. This practice has been shown to improve scan quality in clinical populations (Greene et al., 2016). All participants were given the opportunity to practice the task before scanning until they communicated that they understood the instructions and were comfortable performing the task. GABA-edited MRS data were collected prior to completion of the fMRI Go/No-Go task in order to reduce any potential effect of task performance on baseline GABA. All GABA-edited MRS data were collected using a MR 750 3T MRI Scanner (General Electric Healthcare, Waukesha, WI) and 32-channel head coil (MR Instruments, Minneapolis, MN). A 30x30x20mm (18mL) single voxel was placed over the cortical midline to image the bilateral medial prefrontal cortex/ACC (Figure 1). Each voxel was acquired using an editing on/off pulse sequence (MEGA PRESS) with an 8-step phase cycle and 160 total acquired averages in ~ 12 minutes (TE = 68ms, 16ms editing pulses at 7.5 ppm (edit-OFF) and 1.9 ppm (edit – ON), ordered ON-first, TR = 2000ms, 2048 data points, 2000 kHz spectral width, CHESS-based water suppression). An unsuppressed water reference was acquired, as a part of the sequence, to be used for metabolite quantification. No saturation bands were used. Prior to voxel placement, a 1 mm cubed isotropic T1-weighted imaged (3D-FSPGR) was acquired for voxel localization and segmentation (TR = 10.3 ms, TE = 3.22 ms, TI = 900ms, Flip Angle 9) for each participant. No suppression of macromolecules was performed as such an approach shown to increase the variability in GABA measures at 3T (Mikkelsen et al., 2017). Thus our measures are denoted as GABA+.Functional MR images were collected using a T2*-weighted interleaved, bottom-up echo-planar imaging (EPI) sequence: axial plane, 42 slices, flip angle 75 degrees, TE 30 ms, TR = 2250 ms, FOV = 25.6 cm, matrix 64 x 64, slice thickness = 4 cm (i.e. 4mm isotropic voxels), over 224 volumes (i.e. total scan time =8.4min). 2.5 Processing of MRS dataAll MRS spectra were processed using Gannet 3.0 (). The following steps were performed as per outlined in the Gannet 3.0 GABA Analysis Toolkit, according to Edden and colleagues’ (2013): (i) combination of phased array coil data using a single-weighted approach (Hardy et al., 1992); (ii) time-domain frequency-and-phase correction using spectral correction to maximize the quality of spectra and remove subtraction artifacts subsequent to subject movement and scanner drift (Waddell et al., 2007); (iii) exponential line broadening (3Hz) on the time domain; (iv) fast Fourier transform applied to the t2 dimensions; (v) time averaging to produce the edited difference and OFF spectra (Edden et al., 2013); (vi) pairwise rejection of outlier data with fitting parameters greater than 3 standard deviations from the mean; and (vii) subtraction of the edited ON-OFF acquisitions to generate the GABA+ signal. The MRS voxel was coregistered to T1-weighted images and images were segmented into grey matter (GM), white matter (WM), and cerebrospinal fluid (CSF) tissue priors using the SPM8 segment function (Ashburner & Friston, 2003). Final GABA+ estimates were tissue corrected for GM, WM and CSF tissue content and water relaxation of the three tissue fractions according to Harris, Puts, and Edden (2015), and were expressed in institutional units, relative to water. All Gannet spectra outputs were visually inspected to ensure identification of clear GABA+ signal. Four MRS scans were discarded due to failed fit of the Gannet model (2 ASD and 2 TD), which resulted in 11 included MRS scans (6 ASD and 5 TD). 2.6 Preprocessing of fMRI dataPreprocessing was carried out using the CONN Toolbox (). Preprocessing steps included discarding the first 2 functional volumes of each subject’s scan to allow for T2 equilibration effects. Subsequently the following were performed: (i) interleaved slice timing correction (as advised by Sladky et al., 2011); (ii) least squares time series realignment using a six parameter, rigid-body transformation (Ashburner et al., 2013); (iii) unwarping to remove movement-by-susceptibility artifacts induced by head movement greater than 1 mm in any direction, especially in the particularly susceptible frontal areas investigated in the current study (Friston et al., 1995; Wu et al., 1997); (iv) coregistration of structural and functional images; (v) normalization to MNI space; (vi) segmentation of the anatomical image into grey matter, white matter, and cerebrospinal fluid tissue priors; (vii) ART-based motion scrubbing to identify and effectively remove outliers outside of the 97th percentile of movement (i.e., 0.9mm; Power et al., 2012); and (viii) smoothing of functional images using an 8mm full-width half-maximum Gaussian kernel, which has been shown to be optimal for datasets with low N and/or high inter-subject variability (Mikl et al., 2008). Principal component analysis (PCA) based noise correction using ‘aCompCor’ (Behzadi et al., 2007) was used to reduce physiological and subject movement effects. All preprocessed data were high-pass filtered (0.008 Hz) using a temporal filter with a cut-off at 128 s. Data were further preprocessed using the default first-level analysis steps in the CONN toolbox to regress out the effect of each subject’s realignment parameters, WM and CSF signal, any outlier scans identified by motion scrubbing, and the BOLD signal associated with the presence/absence of any condition (i.e., fixation periods, GO and NO GO blocks). 2.7 fMRI region of interest definitionWe used seven frontostriatal ROIs from the FSL Harvard-Oxford Atlas, available in the CONN toolbox (), for the analysis. Accordingly, MNI coordinates (x, y, z) were centered at the right inferior frontal gyrus (rIFG) pars triangularis (52, 30, 18), rIFG pars opercularis (52, 18, 20), anterior cingulate gyrus (0, 26, 24), and bilateral (left/right) caudate nucleus (-12, 16, 4; 12, 18, 2) and putamen (-22, 8, -4; 22, 10, -4). 2.8 Functional connectivity analysisROI-to-ROI, bivariate correlation analysis was used to examine between-group differences in functional connectivity during task performance. This analysis was implemented using the CONN Toolbox () in the Matlab 2016a programming environment (; Mathworks Natick MA). ROI time series were defined as the mean BOLD activation within the ROI voxels (Whitfield-Gabrieli & Nieto-Castanon, 2012). Subject-specific functional connectivity maps were produced in first-level analysis and motion parameters were included as covariates of no interest. Second-level ROI-to-ROI analyses were then carried out to investigate i) between-group differences in functional connectivity during task performance and i) between group differences in the relationship between GABA+ and functional connectivity during task performance. In order to appropriately control for the number of comparisons made, the ROI-to-ROI between-subjects analyses were first implemented as multivariate F-tests, which identified i) between group differences in connectivity (ASD > TD, 2-sided contrast) across either of the GO and NO GO blocks and any of the ROIs and ii) between group differences in the relationship between cortical GABA+ and functional connectivity (ASD > TD, 2 sided-contrast) across either of the GO or NO GO blocks and any of the ROIs. All significant results were statistically corrected for False Discovery Rate (FDR; Benjamini & Hochberg, 1995). Post hoc testing was then carried out to examine simple main effects of each block and ROI on the connectivity patterns that were significant in the multivariate F tests. A Bonferroni correction was applied to the post hoc data by dividing alpha = 0.05 by nine (two conditions: GO and NO GO, and seven ROIs), resulting in a significance threshold of p = 0.0056.2.9 Analysis of between-subject differences in behavioural performance and cortical GABA+Two sample t-tests were implemented to examine potential between-group differences in behavioural performance during the Go/No-Go task and in the concentration of cortical GABA+ (Snedecor & Cochran, 1989). Pearson R correlation was used to investigate the relationship between GABA+ concentration and behavioural task performance. These analyses were carried out using R Statistical Programming software ().3. Results3.1 Between-subject, behavioural task performance on Go trialsTD subjects responded correctly during 93.9% of Go trials (172.9 out of 184 trials), and ASD subjects responded correctly during 89.5% of Go trials (164.6 out of 184 trials). As such, TD subjects made errors of omission on 6.1% of Go trials (11.1 out of 184 trials) and ASD subjects made errors of omission on 10.5% of Go trials (19.4 out of 184 trials). A two-sample t-test revealed no significant between-group differences in errors of omission during Go trials: t = -1.130, p = 0.279 (Figure 2, left). 3.2 Between-subject, behavioural task performance on No-Go trialsTD subjects correctly inhibited 86.1% of No-Go trials (34.4 out of 40 trials), whereas ASD subjects correctly inhibited 73.1% of No-Go trials (29.3 out of 40 trials). As such, TD subjects made errors of commission on 13.9% of No-Go trials (5.57 out of 40 trials), whereas ASD subjects made errors of commission on 26.9% of No-Go trials (10.8 out of 40 trials). A two-sample t-test revealed a significant between-group difference in errors of commission during No-Go trials: t = -2.747, p = 0.017 (Figure 2, right).3.3 Between-subject, analysis of cortical GABA+ concentrationA two-sample t-test revealed no significant differences in cortical, tissue-corrected GABA+ concentration between the ASD (mean = 2.076) and TD (mean = 2.068) group: t = -0.05, p = 0.965 (Table 1).3.4 Analysis of correlation between cortical GABA+ and task performanceThere was a moderate, non-significant Pearson R correlation between cortical GABA+ and correct inhibition trials for the ASD (r = 0.44, p = 0.28) and TD (r = 0.54, p = 0.21) group. There was no correlation between GABA+ concentration and correct Go responses in the ASD (r = -0.01, p = 0.99) or TD (r = -0.08, p = 0.86) groups. 3.5 Between-subject, ROI-to-ROI functional connectivity analysis during task performanceA multivariate F-test revealed a group by condition interaction effect in connectivity between the right caudate nucleus and the right putamen: F (2, 12) = 8.00, pFDR = 0.037. Post hoc testing revealed weaker connectivity between these regions during No-Go blocks in ASD subjects, compared to controls (p = 0.001). These results are displayed in Table 2 and Figure 3.3.6 Between-subject, correlation between functional connectivity during task performance and cortical GABA+A multivariate F-test revealed a group by condition interaction effect of cortical GABA+ on connectivity of the right inferior frontal gyrus (IFG), pars opercularis with the left putamen: F (2, 6) = 10.81, pFDR = 0.044, and a group by condition interaction effect of cortical GABA+ on connectivity between the right IFG, pars opercularis and the right putamen: F (2, 6) = 9.28, pFDR = 0.044.Post hoc testing revealed a negative association between connectivity of the rIFG, pars opercularis and left putamen during the No-Go block and cortical GABA+, in ASD subjects, compared to controls: T(7) = -4.71, p = 0.002 (Figure 3). Note this effect is bidirectional, and a positive association between connectivity of these regions during the No-Go block and cortical GABA+ was observed in the TD group. (Table 3, Figure 4).Post hoc testing also revealed a negative association between connectivity of the rIFG, pars opercularis and the right putamen during the GO block and GABA+, in ASD subjects, compared to controls, but this effect did not survive correction for multiple comparisons: T(7) = -3.70, p = 0.008. (Table 3).4.0 DiscussionIn this study we investigated the relationship between Go/No-Go task performance and cortical GABA+ in a group of young adults with ASD. A significant, atypical, negative correlation was found between frontostriatal connectivity during the No-Go block and cortical GABA+ in the ASD group, relative to controls. The ASD group also displayed ipsilateral, intrastriatal underconnectivity of the right caudate and right putamen during the No-Go block, relative to controls, that was not correlated with GABA+. To our knowledge, this is the first study to examine the functional relevance of cortical GABA+ concentration as it pertains to task-evoked brain activation in ASD. 4.1 Methodological Feasibility and UtilityFirst and foremost, we demonstrate that combining task-based fMRI with GABA+ MRS is a feasible approach in an ASD population. Furthermore, we believe that practicing the task in a mock MRI scanner environment improved data quality. Despite the ~ 35-minute length of the total scan, all participants who completed scanning generated useable data, and with the application of motion scrubbing to remove motion outliers, all of our participants’ fMRI scans were used in the analysis (except for the exclusion on one scan due to brain anatomic abnormality unrelated to scan quality). Discarded MRS scans were also unrelated to subject performance or movement and were excluded based on failed fit of the Gannet model.Second, our study demonstrates the utility of combining these two MRI acquisition methods. Most importantly, our study addresses a gap in the literature; that is, whether cortical GABA+ concentration is empirically associated with functional brain processes during inhibitory control. In line with other recent findings, our study found no significant differences in the concentration of cortical, tissue-corrected GABA+ between adult ASD and TD subjects (Ajram et al., 2017; Carvalho Pereira et al., 2017; Port et al., 2016; Robertson et al., 2015). Despite virtually identical levels of cortical GABA+ in both groups, an atypical relationship was found between task-based functional connectivity during inhibition control and cortical GABA+ concentration, in the ASD group. This finding demonstrates the principle of multifinality, whereby similar parameters (i.e., GABA+ levels) may lead to functionally different outcomes depending on the ‘system’ (i.e., ASD or control group) in which they operate (Wilden, 1980) and strongly implies that GABA+ levels, whether typical or atypical, are related to functional outcomes in ASD.4.2 Preliminary Interpretation of Findings Although an explanation of our results is likely complex and not easily interpreted from pilot data, we propose a simple and tentative interpretation, while emphasizing the necessity of future replication of our results to support such hypotheses. As the current study’s results were specific to the No-Go block, interpretations are made in light of the increased number of behavioural errors of commission made during No-Go trials in ASD subjects. That said, the weaker intra-striatal connectivity found in ASD subjects during No-Go blocks was likely a factor in the significant difficulty with behavioural inhibition displayed in the ASD group. Given the caudate nucleus has been implicated in both motor and cognitive control (Villablanca, 2010), a bottom-up effect may have emerged, whereby weaker motor regulatory-related striatal connectivity resulted in poorer behavioural performance in ASD.The atypical relationship found between cortical GABA+ and frontostriatal connectivity in ASD may be examined in the context of our control group’s pattern of results. In our control group, connectivity during No-Go block task performance was positively correlated with cortical GABA+, and we presume that the direction of this correlation supports the successful inhibition of behaviour. However, an opposite association was found in the ASD group: stronger rIFG-striatal connectivity during behavioural inhibition blocks was associated with lower cortical GABA+. The correlational nature of this finding precludes a causational explanation. However, these findings support the hypothesis that GABA+ and frontostriatal connectivity function in an atypical manner during behavioural inhibition in ASD.4.3 LimitationsThe current findings should be interpreted in the context of some limitations. First, our small sample size limits the reliability of the current findings and future investigation using larger sample size is needed. However, our final n was greater than previous sample size calculations that are sufficiently powered for differentiation of GABA from MRS spectra (Stone et al., 2012), and similar to sample size estimates that provide sufficient power for fMRI studies (Desmond & Glover, 2002). Additionally, by using a well validated, block design task with a high number of trials, the power of our study was significantly bolstered. Finally, our use of a precise number of a priori selected ROIs that have been strongly implicated in response inhibition, combined with conservative correction for multiple comparisons, would have increased the reliability of the current findings. Overall, we feel that a preliminary interpretation is possible, in light of our thoughtful study design.Some participants in our ASD group had comorbid depression, anxiety and/or ADHD and were taking prescribed SSRIs or stimulant medications. Thus, our observations were not derived from a ‘pure’ ASD sample, and comorbidities in our ASD group may have added to the variability in our results. Despite this limitation, inclusion of ASD participants with such comorbidities increased the ecological validity of our results, due to the high prevalence of psychiatric comorbidities in the ASD population. For example, up to 50% of individuals with ASD display symptoms of ADHD. Similarly, up to 66% of individuals with ADHD may manifest ASD symptoms (Davis & Kollins, 2012). Further, both ASD and ADHD populations display similar patterns of executive dysfunction, including similar response inhibition deficits (Leitner, 2014), and shared functional abnormalities of the basal ganglia (Di Martino et al., 2013). Growing evidence also suggests that adults with ASD experience significantly elevated levels of both depression and anxiety (Howlin & Moss, 2012). Therefore, had we recruited a pure ASD sample, the generalizability of our results to the large proportion of individuals with ASD who have such psychiatric comorbidities would have been low. A better study would be a larger covariate statistical design.Finally, although our blocked fMRI design significantly bolstered our study power and was necessary for a pilot investigation, this design also introduced a limitation in terms of the specificity of demonstrated effects. Specifically, it was during No-Go blocks that ASD subjects displayed both striatal underconnectivity and an atypical correlation between cortical GABA+ and functional connectivity. Since No-Go blocks contained both Go and No-Go trials, the current findings implicate that abnormalities may manifest in an environment where behavioural inhibition is required, but not necessarily during the act of behavioural inhibition in isolation. In future, an event-related design would allow for more fine-tuned analysis of patterns of brain function during inhibition trials. For example, we did not find differences in ACC-related connectivity during the No-Go block. Given the functional implication of the ACC in error monitoring, an event-related analysis of commission errors may provide information on the function of the ACC during this task (Chevrier et al., 2007). However, our findings are congruent with the implied functional roles of the rIFG and putamen, which are to support inhibition (Hampshire et al., 2010) and regulate (i.e., plan and execute) motor movement (DeLong et al., 1984), respectively.4.4 Conclusion and Future DirectionsSpeculatively, a relationship between GABA+ and functional connectivity during more rudimentary, repetitive motor behaviour in ASD may be drawn from the current results (due to the primarily motoric nature of the Go/No-Go task). Moving forward, our research group plans to examine the relationship between GABA+ and functional connectivity during a more a cognitively oriented task, which requires the inhibition of motor behaviour in the context of cognitive interference (i.e., the Spatial Stroop task). It would be valuable to investigate whether there is a different relationship between cortical GABA+ and connectivity during performance on the Spatial Stroop task in order to begin drawing associations between GABA+ and more specific profiles of RB (e.g., more cognitive, higher-order RB such as circumscribed interests). Given the vast heterogeneity of RB across the autism spectrum, treatments targeted towards reducing RB will ultimately need to be individualized according to unique RB profiles. Future work may also investigate correlations between such imaging data and dimensional measures of RB, using tools such as the Repetitive Behaviour Scale-Revised (Bodfish et al., 2000). Investigations of the current type using child participants would also increase our knowledge of the developmental trajectory of GABA+ function in ASD. Overall, the current study demonstrates that combining MRS and task-based fMRI to investigate the functional impact of GABA+ in ASD is a feasible and useful approach. Our study adds incremental evidence to the rapidly growing area of GABA+ MRS research in ASD by laying the groundwork for future investigation into the function of GABA+ in ASD. Ultimately, investigations of this type may lead to the development of urgently needed cognitive and pharmacological interventions for RB in ASD, which can improve quality of life for individuals living with this life-long disorder.Tables and FiguresTable 1. Demographic information and descriptive statistics for ASD and control subjects.VARIABLE ASD mean, (sd)CONTROLmean, (sd)T-value, p-valueIQ score (raw)80 (17.5)82.6 (8.7)T = 0.16, p = 0.87Age25.4 (6.8)23.6 (4.8)T= -0.55 p = 0.60Sex (m, f)6, 2 5, 2Fisher’s Exact p = 1.00GABA i.u. 2.076 (0.24)2.068 (0.34)T = -0.04 p = 0.96Table 2. Significant between-group differences in ROI-to-ROI functional connectivity. Results of a multivariate F-test examining differences in FC across any of the ROIs and either of the GO or NO-GO blocks are displayed on the left. Post hoc testing of simple main effects are displayed on the right. Multivariate F Test for any effect across either of the blocks and any of the ROIs (ASD > TD)Post Hoc Test of Simple Main Effects (ASD > TD)ROIROIF(2, 12)p unc.p < 0.05, FDRCorrelation with Blockbeta (effect size)T(13)p < 0.0056 (Bonferroni correction)Caudate Nucleus (R)Putamen (R)8.000.00620.0371GO -0.16-2.120.0537NO GO-0.23-4.100.0013*Table 3. Significant between-group differences in the effect of GABA+ on ROI-to-ROI functional connectivity. Results of a multivariate F-test examining between-group differences in the effect of GABA+ on FC across any of the ROIs and either of the GO or NO-GO blocks are displayed on the left. Post hoc testing of simple main effects are displayed on the right. IFG – Inferior frontal gyrus Multivariate F Test for any effect of GABA across either of the blocks and any of the ROIs (ASD > TD)Post Hoc Test of Simple Main Effects (ASD > TD)ROIROIF(2, 6)p unc.p < 0.05, FDR Blockbeta (effect size)T(7)p < 0.0056 (Bonferroni correction)IFG, pars opercularis (R)Putamen (L)10.810.01030.0438GO -0.46-1.770.1192NO GO-0.61-4.710.0022*IFG, pars opercularis (R)Putamen (R)9.280.14600.0438GO-1.08-3.700.0077NO GO-0.53-1.560.1622Figure 2. Behavioural performance on the Go/No-Go task. Non-significant between-group differences in errors of omission during Go trials are displayed on the left. Significant between-group differences in errors of commission during No-Go (inhibition) trials are displayed on the right. Figure 3. Significant ROI-to-ROI under-connectivity between the right caudate nucleus and the right putamen during No-Go blocks in ASD subjects, relative to controls (2-sided contrast, p = 0.001 FDR). ROIs are indicated by black circles. Figure 4. Significant negative correlation between connectivity of the right inferior frontal gyrus, pars opercularis and left putamen during the No-Go block and cortical GABA+, in ASD subjects, compared to controls (p = 0.002). This image depicts less connectivity between these regions, associated with a higher concentration of GABA in the ASD group.ReferencesAnagnostou, E., Sorrya, L., Fan, J., Stamper, K., & Hollander, E. (2006). fMRI for the study of response inhibition, and face and linguistic processing in autism. Neuropeadiatrics, 37.Anagnostou, E., & Taylor, M. (2011). Review of neuroimaging in autism spectrum disorders: what have we learned and where do we go from here. Molecular Autism, 2(1), 4.Ajram, L.A., Horder, J., Mendez, M.A., Galanopolous, A., Brennan, L.P., Wichers, R.H., . . . & McAlonan, G.M. (2017). Shifting brain inhibitory balance and connectivity of the prefrontal cortex of adults with autism spectrum disorder. Translational Psychiatry, 7, e1137.Akil, M., Kolachana, B.S., Rothmond, D.A., Hyde, T.M., Weinberger, D.R., & Kleinman, J.E. (2003). Catechol-O-methyltransferase genotype and dopamine regulation in the human brain. Journal of Neuroscience, 2008-2013.Ashburner, J., & Friston, K.J. (2003) Human Brain Function. New York: AcademicAshburner, J., Barnes, G., Chen, C., Daunizeau, J., Flandin, G., Friston, K., . . . & Phillips, C. (2013). SPM8 Manual, Trust Centre for Neuroimaging, London, UK: , Y., & Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Stat Methodol, 57(1), 289–300.Bjorklund, D. F., & Harnishfeger, K. K. (1995). The evolution of inhibition mechanisms and their role in human cognition and behavior. In: Dempster, F.N., & Brainerd, C.J. (Eds.). Interference and inhibition in cognition. San Diego: Academic Press.Bodfish, J.W., Symons, F.J., Parker, D.E., & Lewis, M.H. (2000). Varieties of repetitive behavior in autism: Comparisons to mental retardation. J Autism Devel Disord, 30, 237–243. Boyd, B.A., McDonough, S.G., & Bodfish, J.W. (2012). Evidence-based behavioural interventions for repetitive behaviours in autism. Journal of Autism and Developmental Disorders, 42, 1236 – 1248.Brix, M. K., Ersland, L., Hugdahl, K., Grüner, R., Posserud, M.B., Hammar, ?, . . . Beyer, M.K. (2015). Brain MR spectroscopy in autism spectrum disorder—the GABA excitatory/inhibitory imbalance theory revisited. Front Hum Neurosci, 9, 1–12 Carvalho Pereira, A., Violante, I.R., Mouga, S., Oliveira, G., & Castelo-Branco, M. (2017). Medial frontal lobe neurochemistry in autism spectrum disorder is marked by reduced n-acetylaspartate and unchanged gamma-aminobutyric acid and glutamate+glutamine levels. J Autism Dev Disord, 48(5), 1467 – 1482.Cascio, J.C., Foss-Feig, J.H., Heacock, J., Schauder, K.B., Loring, W.A., Rogers, B.P., . . . & Bolton, S. (2013). Affective neural response to restricted interests in autism spectrum disorders. Journal of Child Psychology and Psychiatry, 55(2), 162-171.Chevrier, A.D., Noseworthy, M.D., Schachar, R. (2007). Hum Brain Mapp, 28(12), 1347 – 1358.Cochran, D. M., Sikoglu, E. M., Hodge, S. M., Edden, R. a. E., Foley, A., Kennedy, D. N., . . . Frazier, J.A. (2015). Relationship among glutamine, γ-aminobutyric acid, and social cognition in autism spectrum disorders. Journal of Child and Adolescent Psychop- harmacology, 25(4), 314 – 322.Davis N.O., & Kollins S. H. (2012).?Treatment for co-occurring attention deficit/hyperactivity disorder and autism spectrum disorder.?Neurotherapeutics?9, 518–530.DeLong, M.R., Alexander, G.E., Georgopoulos, A.P., Crutcher, M.D., Mitchell, S.J., & Richardson, R.T. (1984). Role of basal ganglia in limb movements.?Human Neurobiology, 2(4), 235–44Desmond, J.E., & Glover, G.H. (2002). Estimating sample size in functional MRI (fMRI) neuroimaging studies: statistical power analysis. J Neurosci Methods, 118(2), 115 – 128.Dichter, G.S., Felder, J.N., Green, S.R., Rittenberg, A.M., Sasson, N.J. & Bodfish, J.W. (2012). Reward circuitry function in autism spectrum disorders. SCAN, 7, 160 – 172.Di Martino A., Zuo X. N., Kelly C., Grzadzinski R., Mennes M., Schvarcz A., . . . & Milham, M.P. (2013).?Shared and distinct intrinsic functional network centrality in autism and attention-deficit/hyperactivity disorder.?Biol Psychiatry,?74, 623–632Drenthen, G.S., Barendse, E.M., Aldenkamp, A.P., van Veenendaal, T.M., Puts, N.A.J., Edden, R.A.E., . . . & Jansen, J.F.A. (2016). Altered neurotransmitter metabolism in adolescents with high-functioning autism. Psychiatry Research Neuroimaging, 256, 44–49. Edden, R.A.E., Puts, N.A.J., Harris, A.D., Barker, P.B., & Evans, C.J. (2013). Gannet: A bath-processing tool for the quantitative analysis of gamma-aminobutyric acid-edited MR spectroscopy spectra. J Magn Reson Imaging, 40(60), 1445 – 1452.Friston, K.J., Williams, S.R., Howard, R., Frackowiak, R.S.J., & Turner, R. (1995). Movement-related effect in fMRI time-series.?Magn Reson Med,?35, 346-355.Gaetz, W., Bloy, L., Wang, D. J., Port, R. G., Blaskey, L., Levy, S. E., & Roberts, T. P. L. (2014). GABA estimation in the brains of children on the autism spectrum: Measurement precision and regional cortical variation. NeuroImage, 86, 1–9.Goji, A., Ito, H., Mori, K., Harada, M., Hisaoka, S., Toda, Y., . . . & Kagami, S. (2017). Assessment of anterior cingulate cortex (ACC) and left cerebellar metabolism in asperger’s syndrome with proton magnetic resonance spectroscopy (MRS). Plos ONE, 12(1), e0169288.Greene, D.J., Black, K.J., & Schlaggar, B.L. (2016). Considerations for MRI study design and implementation in pediatric and clinical populations. Developmental Cognitive Neuroscience, 18, 101-112.Hampshire, A., Chamberlain, S.R., Monti, M.M., Duncan, J., & Owen, A.M. (2010). The role of the right inferior frontal gyrus: inhibition and attentional control. Neuroimage, 50(3-3), 1313 – 1319.Han, S., Tai, C., Jones, C.J., Scheuer, T., & Catterall, W.A. (2014). Enhancement of inhibitory neurotransmission by GABAA receptors having A2,3- subunits ameliorates behavioural deficits in a mouse mode of autism. Neuron, 81(6), 1282 – 1289.Harada, M., Taki, M. M., Nose, A., Kubo, H., Mori, K., Nishitani, H., & Matsuda, T. (2011). Non-invasive evaluation of the GABAergic/glutamatergic system in autistic patients observed by MEGA-editing proton MR spectroscopy using a clinical 3 T instrument. J Autism Dev Disord, 41(4), 447–454.Hardy, C.J., Bottomley, P.A., Rohling, K.W., & Roemer, P.B. (1992). An NMR phased array for human cardiac 31P spectroscopy.?Magn Reson Med, 28, 54–64.Harris, A.D., Puts, N.A., & Edden, R.A. (2015). Tissue correction for GABA-edited MRS: Considerations of voxel composition, tissue segmentation, and tissue relaxations. J Magn Reson Imaging, 42, 1431–40. Howlin, P., & Moss, P.? (2012). Adults with autism spectrum disorders. Canadian Journal of Psychiatry, 57(5), 275-283.Huettel, S.A., & McCarthy, G. (2001). The effects of single-trial averaging upon the spatial extent of fMRI activation. Neuroreport, 12(11), 2411 -2416.Ito, H., Mori, K., Harada, M., Hisaoka, S., Toda, Y., Mori, T. . . & Kagami, S. (2017). A proton magnetic resonance spectroscopic study in autism spectrum disorder using a 3-tesla clinical magnetic resonance imaging (MRI) System: The anterior cingulate cortex and the left cerebellum. Journal of Child Neurology, 32(8), 731 – 739. Kim, H., Lim, C., & Kaang, B. (2016). Neuronal mechanisms and circuits underlying repetitive behaviors in mouse models of autism spectrum disorder. Behavioral and Brain Functions, 12, 3. Koos, T., & Tepper, J.M. (1999). Inhibtory control of neostriatal projection neurons by GABAergic interneurons. Nature Neurosci, 2, 467 – 472.Kleinhans, N., Akshoomoff , N., & Delis, D. C. (2005). Executive functions in autism and asperger's disorder: flexibility, fluency, and inhibition. Developmental Neuropsychology, 27(3), 379-401.Langen, M., Bos, D., Noordermeer, S.D., Nederveen, H., van Engeland, H., & Durston, S. (2013). Changes in the development of the striatum are involved in repetitive behaviours in autism. Biological Psychiatry, 76(5), 405 – 411.Langen, M., Leemans, A., Johnston, P., Ecker, C., Daly, E., Murphy, C.M., . . . Murphy, D.G. (2012). Frontostriatal circuitry and inhibitory control in autism: findings from diffusion tensor imaging tractography. Cortex, 48(2), 183 – 193. Langen, M., Kas, M., Staal, W., vanEngeland, H., & Durston, S. (2010). The neurobiology of repetitive behaviour: of mice....Neurosci Biobehav Rev, 35(3), 345 – 355.Lee, P.S., Yerys, B.E., Della Rosa, A., Foss-Feig, J., Barnes, K.A., James, J.D., VanMeter, J., Vaidya, C.J., Gaillard, W.D., & Kenworthy, L.E. (2009). Functional connectivity of the inferior frontal cortex changes with age in children with autism spectrum disorders: a fcMRI study of response inhibition. Cerebral Cortex, 19(8), 1787 – 1794.Leitner, Y. (2014). The co-occurrence of autism and attention deficit-hyperactivity disorder in children – what do we know? Front Hum Neurosci, 8, 268.Lemonnier, E., Villeneuve, N., Sonie, S., Serret, S., Rosier, A., Roue, M., . . . & Ben-Ari, Y. (2017). Effects of bumetanide on neurobehavioural function in children and adolescents with autism spectrum disorders. Translational Psychiatry, 7, e1056.Lopez, B.R., Lincoln, A.J., Ozonoff, S., & Lai, Z. (2005). Examining the relationship between executive functions and restricted, repetitive symptoms of autistic disorder. J Autism Dev Disord, 35, 445–460.Luna, B., Doll, B. S., Hegedus, S. J., Minshew, N. J., & Sweeney, J. A. (2006). Maturation of executive function in autism. Biological Psychiatry, 61(4), 474-481.Mescher, M., Merkle, H., Kirsch, J., Garwood, M., & Gruetter, R. (1998). Simultaneous in vivo spectral editing and water suppression. NMR in Biomedicine, 11(6), 266-272.Mikkelsen, M., Barker, P.B., Bhattacharyya, P.K., Brix, M.K., Buur, P.F., Cecil, K.M., . . . Edden, R.A.E. (2017). Big GABA: Edited MR spectroscopy at 24 research sites. Neuroimage, 159, 32 – 45.Mikl, M., Marecek, R., Hlustik, P., Pavlicova, M., Drastich, A., Chlebus, P., Brazdil, M., & Krupa, P. (2008). Effects of spatial smoothing on fMRI group inferences. Magn Reson Imaging, 26, 490 – 503.Mosconi, M.W., Kay, M., D’Cruz, A.M., Seidenfeld, A., Guter, S., Standford, L.D., & Sweeney, J.A. (2009) Impaired inhibitory control is associated with higher-order repetitive behaviors in autism spectrum disorders. Psychological Medicine, 39(9), 1559 – 1566.Ozonoff, S., Pennington, B.F., & Rogers, S.J. (1991). Executive function deficits in high functioning autistic individuals: Relationship to theory of mind. Journal of Child Psychology and Psychiatry, 32, 1081–1105.Ozonoff, S., Strayer,D.L., McMahon, W.M., & Filloux, F. (1994). Executive function abilities in autism and Tourette syndrome: an information-processing approach. Journal of Child Psychology and Psychiatry, 35, 1015–1032.Perlman, W.R., Weickert, C.S., Akil, M., & Kleinman, J.E. (2004). Postmortem investigations of the pathophysiology of schizophrenia: the role of susceptibility genes. Journal of Psychiatry and Neuroscience, 29, 287 – 293.Port, R. G., Gaetz, W., Bloy, L., Wang, D. J., Blaskey, L., Kuschner, E. S., . . . & Roberts, T.P.L. (2016). Exploring the relationship between cortical GABA concentrations, auditory gamma-band responses and development in ASD: Evidence for an altered maturational trajectory in ASD. Autism Res, 10(4), 593 – 607.Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., & Petersen, S.E. (2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. NeuroImage, 59, 2142–2154.Presti, M.F., Watson, C.J., Kennedy, R.T., Yang, M., Lewis, M.H. (2004). Behavior-related alterations of striatal neurochemistry in a mouse model of stereotyped movement disorder. Pharmacol Biochem Behav, 77(3), 501–507.Puts, N.A.J., Wodka, E.L., Harris, A.D., Crocetti, D., Tommerdahl, M., Mostofsky, S.H., & Edden, R.A.E. (2016). Reduced GABA and altered somatosensory function in children with autism spectrum disorder. Autism Res, 10, 608–619.Robertson, C.E., Ratai, E.M., & Kanwisher, N. (2015). Reduced GABAergic action in the autistic brain. Current Biology, 26(1), 80–85.Rojas, D.C., Singel, D., Steinmetz, S., Hepburn, S., & Brown, M.S. (2014). Decreased left perisylvian GABA concentration in children with autism and unaffected siblings. NeuroImage, 86, 28–34.Rolls, E.T. (1994). Neurophysiology and cognitive functions of the striatum. Rev Neurology, 150(8-9), 648-660.Sabatino, A., Rittenberg, A., Sasson, N.J., Turner-Brown, L., Bodfish, J.W., & Dichter, G.S. (2013). Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord, 43(12), 2903 – 2913.Schmitz, N., Rubia, K., Daly, E., Smith, A., Williams, S., & Murphy, D.G.M. (2006). Neural correlates of executive function in autistic spectrum disorders. Biological Psychiatry, 59, 7–16.Snedecor, G.W., & Cochran, W.G. (1989).?Statistical Methods Eighth Edition. Iowa: State University PressSilverman, J.L., Pride, M.C., Hayes, J.E., Puhger, K.R., Butler-Struben, H.M., Baker, S., & Crawley, J.N. (2015). GABA B receptor agonist R-baclofen reverses social deficits and reduces repetitive behavior in two mouse models of autism. Neuropsychopharmacology, 40(9), 2228–2239.Shafritz, K.M., Dichter, G.S., Baranek, G.T, & Belger, A. (2008). The neural circuitry mediating shifts in behavioural response and cognitive set in autism. Biological Psychiatry, 63(10), 974-980.Sladky, R., Friston, K.J., Trostl, J., Cunnington, R., Moser, E., & Windischberger, C. (2011). Slice-timing effects and their correction in functional MRI. Neuroimage, 59(2-2), 588-594.South, M., Ozonoff, S., & McMahon, W.M. (2007) The relationship between executive functioning, central coherence and repetitive behaviors in the high-functioning autism spectrum. Autism, 11, 437.Stone, J.M., Dietrich, C., Edden, R., Mehta, M., De Simoni, S., Reed, L.J., . . . Barker, G. J. (2012).?Ketamine effects on brain GABA and glutamate levels with 1H-MRS: relationship to ketamine-induced psychopathology.?Mol Psychiatry,?17, 664–665.Thakkar, K.N., Polli, F.E., Joseph, R., Tuch, D., Hadjikhani, N., Barton, J., & Monoach, D. (2008). Response monitoring, repetitive behaviour and anterior cingulate abnormalities in autism spectrum disorders (ASD). Brain, 131(9), 2464 – 2478. Turner, M.A. (1999) Annotation: repetitive behavior in autism: A review of psychological research. Journal of Child Psychology and Psychiatry, 40(6), 839 – 849.Tremblay, R., Lee, S., & Rudy, B. (2016). GABAergic interneurons in the neocortex: from cellular properties to circuits. Neuron, 91(2), 260 – 292.Traynor, J.M., Doyle-Thomas, K.A.R., Hanford, L.C., Foster, N.E., Tryfon, A., Hyde, K., . . . & Hall, G.B.C. (2018). Indices of repetitive behaviour are correlated with patterns of intrinsic functional connectivity in youth with autism spectrum disorder. Brain Research, 1685, 79 -90.Uzefovsky, F., Allison, C., Smith, P., & Baron-Cohen, S. (2016). Brief report: The Go/No-Go task online: inhibitory control deficits in autism in a large sample. J Autism Dev Disord, 46(8), 2774 – 2779.Villablanca, J.R. (2010). Why do we have a caudate nucleus? Acta Neurobiol Exp, 70, 95 – 105.Waddell, K.W., Avison, M.J., Joers, J.M., & Gore, J.C. (2007). A practical guide to robust detection of GABA in human brain by J-difference spectroscopy at 3 T using a standard volume coil.?Magn Reson Imaging, 25, 1032–1038.Wechsler, D. (2011). Wechsler Abbreviated Scale of Intelligence–Second Edition (WASI-II).?San Antonio, TX:?NCS PearsonWhitfield-Gabrieli, S., & Nieto-Castanon, A. (2012). Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connect, 2(3), 125–141. Wilden, A. (1980). System and structure: Essays in communication and exchange. Great Britain:Tavistock Publications Limited.Wu, D.H., Lewin, J.S. & Duerk, J.L. (1997). Inadequacy of motion correction algorithms in functional MRI: Role of susceptibility-induced artefacts. J Magn Reson Imag, 7, 365-370Xiao, T., Xiao, Z., Ke, X., Hong, S., Yang, H., Su, Y., . . . & Liu, Y. (2012). Response inhibition impairment in high functioning autism and attention deficit hyperactivity disorder: evidence from near-infrared spectroscopy data. PLoS ONE, 7(10), e46569.Zhou, Y., Shi, L., Cui, X., Wang, S., & Luo, X. (2016). Functional connectivity of the caudal anterior cingulate cortex is decreased in autism. PLoS ONE, 11(3), e0151879.Chapter 6General DiscussionThe etiology and function of repetitive behaviour in Autism Spectrum Disorder (ASD) is poorly understood. Compared to social communication deficits, the existing research on repetitive behaviour in ASD is sparse (Boyd, McDonough, & Bodfish, 2012). However, repetitive behaviour is associated with significant impairment in functioning (Bishop, Richler, Cain, & Lord, 2007; Greenberg, Seltzer, Krauss, Chou, & Orsmond, 2006; Nadig, Lee, Singh, Bosshart, & Ozonoff, 2010; Pierce & Courschesne, 2001), and is an important core symptom of ASD to study. The identification of biomarkers is an important area of research that can increase our understanding of the etiology and function of repetitive behaviour, and contribute to the development of novel interventions (Andersen, 2015). Thus far, biomarker research on repetitive behaviour in ASD has been carried out from the standpoint of understanding ASD as a disorder of executive dysfunction (Ozonoff, 1995; Russel, 1997) and disrupted functional connectivity (Belmonte, Cook, Andersen, Greenough, & Beckel-Mitchener, 2004; Just et al., 2004; 2007;2012; Vasa, Mostofsky, & Ewen, 2016). To date, results have suggested that there may be a common neurobiological substrate involved in both social deficits and repetitive behaviour in ASD (Benning et al., 2016; Cascio et al., 2014; Dichter et al., 2012; Foss-Feig et al., 2016; Pierce et al., 2015; Sabatino et al., 2013; Sasson, Turner-Brown, Holtzclaw, Lam, & Bodfish, 2008; Sasson, Elison, Turner-Brown, Dichter, & Bodfish, 2011; Watson et al., 2015). At the same time, it is becoming increasingly apparent that ASD may be described as a spectrum of multiple “autisms,” that vary in their symptom presentation and genetic loci (Persico & Napolioni, 2013). As such, the imaging field is moving toward identifying biomarkers of distinct behavioural subtypes across the autism spectrum, with the goal of developing more targeted interventions for idiosyncratic symptom presentations. However, several gaps in the literature remain, and the current body of research was carried out with the purpose of addressing these gaps. Chapter 2 presented a literature review of the structural and functional correlates of repetitive behaviour in ASD. Given the relative paucity of repetitive behaviour research, and the highly variable study findings to date, this review’s objectives were to create an organized framework of the neural correlates of repetitive behaviour and the experimental paradigms used to image these correlates. Results from the 32 studies included in this review converged on findings of structural abnormalities in frontostriatal brain regions that are correlated with an array of repetitive behaviour subtypes across the lifespan in ASD. Additionally, this review outlined the executive functioning paradigms used to elicit repetitive behaviour circuitry during functional magnetic resonance imaging (fMRI) task performance, and demonstrated that visual motor, target detection, cognitive flexibility, inhibition, rule violation, and reward processing paradigms can be used to examine repetitive behaviour using fMRI. From this outline, it was also derived that cognitive control circuitry may subserve both lower-order, motoric repetitive behaviour and higher-order, cognitively oriented repetitive behaviour. Finally, it was suggested that salience-, and reward-attribution circuitry may more exclusively underpin higher-order behaviour in ASD. A particular strength of this study was that it focused on creating an organized framework of i) overall, broad repetitive behaviour circuitry, and ii) subtype-specific circuitry; in children, adolescents, and adults. The comprehensive scope of this review highlighted overall frontostriatal involvement in repetitive behaviour, as well as more pointed, network- and sub-type specific circuitry. Further, this review provided a blueprint of the fMRI paradigms that have been used to study repetitive behaviour, which is a novel contribution to the field and offers a framework for future work.In terms of limitations, this review only included 32 available studies, which is quite a small number, relative to reviews on the neural circuitry of social deficits in ASD (review in Barak & Feng, 2016). Further, due to great heterogeneity in the methods and findings included, a more stringent review method, such as a meta-analysis, was not used. Moreover, in addition to including fMRI studies that correlated collected measures of repetitive behaviour with neural circuitry, this review also included studies that used fMRI paradigms to examine circuitry that is only hypothesized to underpin repetitive behaviour, but has not yet been correlated with collected measures of behaviour. For example, although the utility of a number of executive function paradigms was highlighted, only the circuitry elicited by inhibition and target detection paradigms have been statistically correlated with repetitive behaviour scores on diagnostic measures (Agam, Joseph, Barton, & Manoach, 2010; Shafritz, Dichter, Baranek, & Belger, 2008), and the possibility exists that the other paradigms presented in this review are not actually “repetitive behaviour proxies” (Anagnostou & Taylor, 2001). Finally, the majority of studies included in this review correlated brain-based markers with measures of behaviour collected from the Autism Diagnostic Interview- Revised (ADI-R; Lord, Rutter, & Le Couteur 1994), which is not a continuous measure of repetitive behaviour and is statistically suboptimal. Regardless of these limitations, this review provided an urgently needed framework for repetitive behaviour research moving forward. Chapter 3 presented the results of an eye-tracking study, which investigated the hypothesis of a common frontostriatal substrate in social and repetitive behaviour in ASD. The purpose of this study was to objectively quantify attention, autonomic arousal, and effort expenditure during viewing of social and restricted interest stimuli in ASD. This study found that compared to neurotypicals, participants with ASD expended significantly more effort to view stimuli of their restricted interest. Additionally, this study found a pattern of arousal to social and interest stimuli in the ASD group that was not identified in the control group, and this pattern suggested a devaluing of social stimuli and an overvaluing of interest stimuli. In terms of strengths, this is the first study to simultaneously quantify arousal to social and restricted interest stimuli, which allowed for a direct comparison. Whereas previous studies have asked participants to subjectively rate arousal (Sasson, Dichter, & Bodfish, 2012), our use of pupillometry, which is a reliable and objective measure of autonomic arousal (Bradley, Miccoli, Escrig, & Lang, 2008; Hess & Polt, 1960), was another strength. Additionally, a set of individualized, tightly controlled stimuli was used, which addressed existing criticisms of more generic sets of “high autism interest” stimuli (Parsons, Bayliss, & Remington, 2016). Further, use of a control group with intense and functionally interfering interests was another strength, and likely resulted in a conservative estimate of the actual differences between ASD and the general population.This study also had its limitations. Most pertinently, the small sample size used (n = 10 ASD, 19 controls) limits the stability of results, and their generalizability across the spectrum. More specifically, as restricted interests are more cognitively oriented (Turner, 1999) and occur more frequently in high-functioning individuals (Turner-Brown, Lam, Holtzclaw, Dichter, & Bodfish, 2011), the application of this research to low-functioning individuals, who tend to display more motoric, stereotyped behaviour (Szatmari et al., 2006), is limited. Additionally, in this study’s main analysis, pupil size and blink rate data were analyzed using within-subjects designs, and only effort expenditure data was analyzed using a between-subjects design. Therefore, although significant between-subject differences in effort expenditure were found, a between-group analysis of arousal was not thoroughly explored. In a supplementary between-subjects analysis, only trends toward significant between-group differences in pupil size were found. However, the supplementary model was not statistically appropriate due to the exclusion of an important baseline covariate (see Chapter 3, Supplementary Material), and as such, was not used in the main analysis. A larger sample size is needed to validate and extend our main findings using between-subjects analysis.In Chapter 4, the results of a resting-state fMRI study that investigated the association between indices of repetitive behaviour and widespread patterns of intrinsic functional connectivity in ASD were presented. This study identified that youth with ASD displayed broad over-connectivity in thalamocortical and corticostriatal networks, compared to a group of age-matched controls. Additionally, repetitive behaviour in the ASD group, as measured by total scores from the Repetitive Behaviour Scale – Revised (RBS-R; Bodfish, Symons, Parker, & Lewis, 2000 ), was positively correlated with connectivity between the left primary visual cortex (V1) and the right inferior frontal gyrus (IFG), pars orbitalis. Further, by using a principal component decomposition of scores on the RBS-R, this study demonstrated that a specific repetitive behaviour profile characterized by increased insistence on sameness behaviour and low repetitive motor and self-injurious behaviour was associated with increased attention-related, frontoparietal connectivity between the right IFG, pars triangularis and the right inferior parietal lobe (IPL). Taken together, these findings imply that repetitive behaviour is associated with unique neural underpinnings, characterized by inhibitory- and attention-related connectivity patterns. The current results also assert that in addition to being a distinct statistical factor, sameness behaviour may be delineated from other types of repetitive behaviour on a biomarker level, and parallel previous work that supports this suggestion (Hollander et al., 2005; Langen et al., 2013; Sears et al., 1999). A particular strength of this study was its use of the RBS-R, which is a statistically appropriate measure of repetitive behaviour; the RBS-R has been used much less frequently than the ADI-R (Bishop et al., 2013), which quantifies repetitive behaviour using a restricted range of scores (0-4) and is not optimal for correlation analysis. Another strength of this study was its use of widespread cortical and subcortical regions of interest (ROIs). It has been suggested that, given the heterogeneity in repetitive behaviour across the spectrum, conflicting findings in repetitive behaviour research to date may be the result of the restricted range of ROIs used in previous studies (Muller et al., 2011). Regarding limitations, the current study’s sample was comprised of youth and young adults with ASD, with a wide age range (10 to 21 years). Although it is helpful to examine the developmental trajectory of repetitive behaviour, our sample was not divided into age subgroups because of the low number of total participants (n = 30 ASD, 32 TD). As such, it is possible that a specific age group in the sample, rather than the entire sample may have driven our results, especially given that our correlational findings pertaining to repetitive behaviour in the ASD group included connectivity profiles in prefrontal cortical areas. More specifically, repetitive behaviour was correlated with right IFG connectivity; a region that undergoes marked developmental changes through child and early adulthood (Gogtay et al., 2004). As such, future work should focus on the developmental trajectory of these effects, as well as the functional consequences of such connectivity profiles on repetitive behaviour across the lifespan. Another limitation pertains to the sample size used (n = 30 ASD, 32 controls), which is small relative to samples in recent ASD imaging research (e.g., Cerliani et al., 2015). Although the sample size in the current study well exceeded calculations pertaining to the sample size needed to achieve sufficient power in fMRI designs (Desmond & Glover, 2012), because ASD is symptomatically variable, ASD imaging research may benefit from use of substantially larger sample sizes. However, the results of this study, and specifically the atypical thalamocortical over-connectivity found in the ASD group, are congruent with studies that have used much larger sample sizes (Cerliani et al., 2015), which supports the validity of the current findings.Finally, the results of a pilot study that combined a Go/No-Go, inhibitory control fMRI paradigm with magnetic resonance spectroscopy (MRS) were presented in Chapter 5. This study examined the association between the neural correlates of inhibitory control and cortical gamma-amino-butyric acid (GABA) in ASD. Significantly poorer behavioural performance during trials that required inhibition (i.e., No-Go trials) was found in the ASD group, relative to controls. Further, during No-Go blocks, ASD subjects displayed weaker intrastriatal connectivity, compared to controls. Finally, although no between-group difference in GABA concentration was found, an inverse correlation between frontostriatal connectivity during No-Go blocks and GABA was found in the ASD group, relative to controls. In terms of strengths, this is the first study to combine task-based fMRI and MRS in an ASD population, and is a novel and highly innovative research design. The significant results further suggest that such a design is both feasible and useful to employ in an ASD population, and the results contribute important, incremental evidence to the literature on GABA in ASD. With this said, the small sample size may have limited the stability of results. However, a pointed number of ROIs and a blocked fMRI design with a large number of trials were used to offset the power disadvantage of a small sample size (Huettell & McCarthy, 2001). Second, a blocked fMRI design precluded an event-related analysis; although ASD subjects displayed poorer behavioural performance during No-Go trials, the fMRI results pertain exclusively to performance during No-Go blocks, which included both Go and No-Go trials. Therefore, although it is likely that the pattern of circuitry detected pertains mostly to inhibitory control deficits, performance during Go trials also contributed to the blood-oxygen-level dependent (BOLD) signal. However, it is also possible that by including Go trials in the No-Go block, that the results are a more conservative estimate of the neural circuitry underpinning behavioural inhibition deficits in ASD. In future, an event-related design may demonstrate that effects pertaining to unsuccessful inhibition are even stronger than what was detected in the current study. Finally, this study did not assess repetitive behaviour, and the correlation between the current pattern of results and repetitive behaviour as indexed by a validated assessment measure would further support a GABAergic component to repetitive behaviour. Regardless, the current findings suggest that GABA is functionally related to behavioural inhibition and repetitive behaviour in ASD, and support the translation of animal work, which has strongly supported GABAergic involvement in repetitive behaviour (Han, Tai, Jones, Scheuer, & Catterall, 2014; Silverman et al., 2015), to human network dynamics.It is important to address how this body of research may assimilate with hypotheses of ASD as a disorder of executive dysfunction (Ozonoff, 1995; Russel, 1997) and disrupted connectivity (Belmonte et al., 2004; Just et al., 2004; 2007;2012; Vasa et al., 2016). First, this body of work supports the hypothesis of autism as a disorder of executive dysfunction (Ozonoff, 1995; Russel, 1997), and adds to the existing evidence linking inhibitory control deficits to repetitive behaviour in ASD by suggesting that there is a GABAergic component to these deficits, which may have functional consequences on repetitive behaviour in ASD. Second, it should be highlighted that i) our literature review found abnormal structural and functional connectivity in frontostriatal circuits across the lifespan in ASD and across categories of repetitive behaviour, ii) our eye-tracking study found altered frontostriatal-supported effort expenditure and arousal in ASD, and ii) our combined fMRI and MRS pilot study found abnormal frontostriatal connectivity during the performance of behavioural inhibition, as well as an atypical correlation between GABA and task-induced frontostriatal connectivity. As such, this body of work supports ASD as a disorder of disrupted frontostriatal connectivity, with several associations existing between more specific frontostriatal regions and behavioural subtypes. The current research program also demonstrates the involvement of additional patterns of circuitry in repetitive behaviour. Specifically, i) our literature review included studies that found associations between repetitive behaviour subtypes and insular (Uddin et al., 2013) parietal (Nordahl et al., 2007), thalamic (Duerden et al., 2013), and cerebellar (Pierce & Courchesne, 2001) correlates, and congruently ii) our resting-state fMRI study found an association between sameness behaviour and more distinct intrinsic connectivity profiles in attentional and sensory processing circuits; a finding that is consistent with work that has identified IS as a distinct statistical factor (Cuccaro et al., 2003; Lam, Bodfish, & Piven, 2008; Szatmari et al., 2006). Therefore, this body of work follows the recent direction in the imaging literature of identifying subtype-specific circuitry, and supports the conceptualization of several “autisms” existing along a spectrum of related disorders (Persico & Napolioni, 2013).Collectively, this body of research has important clinical implications. First, knowledge of a common frontostriatal substrate may be used to inform therapeutic targets in behavioural interventions. For example, interventions may be developed to simultaneously target social and repetitive behaviour deficits. Research may also focus on whether existing interventions for social communication deficits may impact repetitive behaviour, and whether new interventions that involve training in different domains of executive function (e.g., inhibitory control) may reduce repetitive behaviour and/or subtypes. Finally, our combined fMRI and MRS study implies that GABAergic medications may be used to treat repetitive behaviour in ASD. To date, there are no efficacious medications for the treatment of repetitive behaviour (Brondino et al., 2015). Further, several studies have demonstrated that GABA agonists have an excitatory effect in ASD, compared to their inhibitory effect on neurotypicals (Belsito, Law, Kirk, Landa, & Zimmerman, 2001; Lemonnier et al., 2012). For example, a recent trial of GABA and glutamate acting riluzole showed that despite no significant differences in GABA concentration between ASD and controls, riluzole increased prefrontal cortical connectivity, but decreased it in controls (Ajram et al., 2017). Similarly, our results demonstrated an inverse correlation between task-induced functional connectivity and GABA in ASD, relative to controls. As such, GABAergic medications may impact behaviour in an atypical manner in ASD, relative to individuals without ASD, and more pharmacological research pertaining to medications that act both directly and indirectly on the GABAergic system will provide insight into appropriate medications for repetitive behaviour. Clinical research of this nature is urgently needed, given the large impact of ASD on individuals, families, and the healthcare system.In conclusion, the current research program has presented a number of novel research findings pertaining to biomarkers of repetitive behaviour in ASD. This work has shown that i) across the lifespan, structural and functional frontostriatal abnormalities are implicated in repetitive behaviour in ASD, ii) there may be a common frontostriatal mechanism that underpins social deficits and repetitive behaviour in ASD, and these symptoms may be modulated by arousal and effort expenditure, iii) the thalamus and striatum are intrinsically over-connected with motor and sensory cortices in ASD, iv) sameness behaviour is distinctly associated with intrinsic attentional and sensory processing connectivity, v) the striatum demonstrates under-connectivity during behavioural inhibition in ASD, and vi) GABA is functionally implicated in behavioural inhibition deficits in ASD, and demonstrates an atypical correlation with task-induced inhibitory-related frontostriatal connectivity. Overall, this work has contributed novel and important findings to the limited body of existing research on repetitive behaviour in ASD.ReferencesAgam, Y., Joseph, R.M., Barton, J.J.S., & Manoach, D.S. (2010). Reduced cognitive control of response inhibition by the anterior cingulate cortex in autism spectrum disorders. NeuroImage, 52, 336-347. Ajram, L.A., Horder, J., Mendez, M.A., Galanopolous, A., Brennan, L.P., Wichers, R.H., . . . McAlonan, G.M. (2017). Shifting brain inhibitory balance and connectivity of the prefrontal cortex of adults with autism spectrum disorder. Translational Psychiatry, 7, e1137.Anagnostou, E., & Taylor, M. (2011). Review of neuroimaging in autism spectrum disorders: what have we learned and where do we go from here. Mol Autism, 2(1), 4. Anderson, G.M. (2015). Autism biomarkers: challenges, pitfalls, and possibilities. J Autism Dev Disord, 45, 1103 – 1113.Barak, B., & Feng, G. (2016). Neurobiology of social behavior abnormalities in autism and williams syndrome. Nat Neurosci, 19(6), 647 – 655.Belsito, K.M., Law, P.A., Kirk, K.S., Landa, R.J., Zimmerman, A.W. (2001). Lamotrigine therapy for autistic disorder: a randomized, double blind, placebo-controlled trial. J Autism Dev Disord, 31(2), 175 – 181.Benning, S.D., Kovac, M., Campbell, A., Miller, S., Hanna, E.K., Damiano, C.R., . . . Dichter, G.S. (2016). Late positive potential ERP responses to social and nonsocial stimuli in youth with autism spectrum disorder. J Autism Dev Disord, 46(9), 3068 – 3077.Bishop, S.L., Hus, V., Duncan, A., Huerta, M., Gotham, K., Pickles, . . . Lord, C. (2013). Subcategories of restricted and repetitive behaviors in children with autism spectrum disorders. J Autism Dev Disord, 43(6), 1287-1297.Bishop, S.L., Richler, J., Cain, A.C., & Lord, C. (2007). Predictors of perceived negative impact in mothers of children with autism spectrum disorder. Am J of Mental Retardation, 112(6), 450–461. Bodfish, J.W., Symons, F.J., Parker, D.E., & Lewis, M.H. (2000). Varieties of repetitive behavior in autism: Comparisons to mental retardation. J Autism Dev Disord. 30, 237–243. Boyd, B.A., McDonough, S.G., & Bodfish, J.W. (2012). Evidence-based behavioral interventions for repetitive behaviors in autism. J Autism Dev Disord, 42(6), 1236 – 1248. Bradley, M.M., Miccoli, L., Escrig, M.A., & Lang, P.J. (2008). The pupil as a measure of emotional arousal and autonomic activation. Psychophysiology, 45, 602–607.Brondino, N., Fusar-Poli, L., Panisi, C., Damiani, S., Barale, F., Politi, P. (2015). Pharmacological modulation of GABA function in autism spectrum disorders: a systematic review of human studies. J Autism Dev Disord, 46(3), 825 – 839.Cascio, J.C., Foss-Feig, J.H., Heacock, J., Schauder, K.B., Loring, W.A., Rogers, B.P., & Bolton, S. (2014). Affective neural response to restricted interests in autism spectrum disorders. J Child Psychol Psychiatry, 55(2), 162-171. ?Cerliani, L., Meenes, M., Thomas, R.M., Di Martino, A., Thioux, M., & Keysers, C. (2015). Increased functional connectivity between subcortical and cortical resting-state networks in autism spectrum disorder. JAMA Psychiatry, 72(8), 767 – 777.Cuccaro, M.L., Shao, Y., Grubber, J., Slifer, M., Wolpert, C.M., Donnelly, S.L., . . . Vance, M.A. (2003). Factor analysis of restricted and repetitive behaviors in autism using the Autism Diagnostic Interview-R. Child Psychiatry Hum Dev, 34, 3-17. Desmond, J.E., & Glover, G.H. (2002). Estimating sample size in functional MRI (fMRI) neuroimaging studies: statistical power analysis. J Neurosci Methods, 118(2), 115 – 128.Dichter, G.S, Felder, J.N., Green, S.R., Rittenberg, A.M., Sasson, N.J., Bodfish, J.W. (2012). Reward circuitry function in autism spectrum disorders. SCAN, 7, 160-172. ?Duerden, E.G., Card, D., Roberts, S.W., Mak-Fan, K.M., Chakravarty, M., Lerch, J.P., & Taylor, M.J. (2014). Self-injurious behaviours are associated with alterations in the somatosensory system in children with autism spectrum disorder. Brain Structure and Function, 219(4), 1251 – 1261.Foss-Feig, J.H., McGugin, R.W., Gauthier, I., Mash, L.E., Ventola, P., & Cascio, C.J. (2016). A functional neuroimaging study of fusiform response to restricted interests in children and adolescents with autism spectrum disorder. J Neurodev Disord, 8, 15.Gogtay, N., Giedd, J.N., Lusk, L., Hayashi, K.M., Greenstein, D., Vaituzis, C., . . . Thompson, P.L. (2004). Dynamic mapping of human cortical development during childhood through early adulthood. PNAS, 101(21), 8174 – 8179.Greenberg, J.S., Seltzer, M.M., Krauss, M.W., Chou, R.J., Orsmond, G. (2006). Bidirectional effects of expressed emotion and behavior problems and symptoms in adolescents and adults with autism. Am J Mental Retardation, 111, 229–249. Han, S., Tai, C., Jones, C.J., Scheuer, T., & Catterall, W.A. (2014). Enhancement of inhibitory neurotransmission by GABAA receptors having A2,3- subunits ameliorates behavioural deficits in a mouse mode of autism. Neuron, 81(6), 1282 – 1289.Hess, E.H., & Polt, J.M. (1960). Pupil size as related to interest value of visual stimuli. Science, 132, 349–350. Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M., Licalzi, E., . . . Buchsbaum, M. (2005). Striatal volumes on magnetic resonance imaging and repetitive behaviors in autism. Biol Psychiatry, 58(3), 226-232. Huettel, S.A., & McCarthy, G. (2001). The effects of single-trial averaging upon the spatial extent of fMRI activation. Neuroreport, 12(11), 2411 -2416.Lam, K.S.L., Bodfish, J.W., & Piven, J. (2008). Evidence for three subtypes of repetitive behavior in autism that differ in familiality and association with other symptoms. J Child Psychol Psychiatry, 49(11), 1193-1200. Lemonnier, E., Degrez, C., Phelep, M., Tyzio, R., Josse, F., Grandgeorge, M., Hadjikhani, N., & Ben-Ari, Y. (2012). A randomised controlled trial of bumetanide in the treatment of autism in children. Transl Psychiatry, 2(12), e202.Langen, M., Bos, D., Noordermeer, S.D., Nederveen, H., van Engeland, H., & Durston, S. (2013). Changes in the development of the striatum are involved in repetitive behaviors in autism. Biol Psychiatry, 75(5), 405-411. Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord, 24, 659–685. Muller, R.A., Shih, P., Keehn, B., Deyoe, J.R., Leyden, K.M., & Shukla, D.K. (2011). Underconnected, but how? A survey of functional connectivity MRI studies in autism spectrum disorders. Cereb Cortex, 21, 2233–2243.Nadig, A., Lee, I., Singh, L., Bosshart, K., & Ozonoff, S. (2010). How does the topic of conversation affect verbal exchange and eye gaze? A comparison between typical development and high-functioning autism. Neuropsychologia, 48, 2730–2739.Nordahl, C.W., Dierker, D., Mostafavi, I., Schumann, C.M., Rivera, S.M., Amaral, D.G., & Van Essen, D.C. (2007). Cortical folding abnormalities in autism revealed by surface-based morphometry. J Neurosci, 27(43), 11725 – 11735.Ozonoff, S. (1995). Executive functions in autism. In: Schopler, E., & Mesibov, G. (Eds.)?Learning and Cognition in Autism.?New York: Plenum Press, 199 – 219.Parsons, O.E., Bayliss, A., & Remington, A. (2016). A few of my favourite things: circumscribed interests in autism are not accompanied by increased attentional salience on a personalized selective attention task. Molecular Autism, 8, 20. Persico, A.M., & Napolioni, V. (2013). Autism genetics. Behav Brain Res, 15, 95 – 112.Pierce, K., & Courchesne, E. (2001). Evidence for a cerebellar role in reduced exploration and stereotyped behavior in autism. Biol Psychiatry, 49, 655–664.Pierce, K., Marinero, S., Hazin, R., McKenna, B., Carter Barnes, C., & Malige, A. (2015). Eye tracking reveals abnormal visual preference for geometric images as an early biomarker of an autism spectrum disorder subtype associated with increased symptom severity. Biol Psychiatry, 79, 657 – 666.Russel, J. (Ed.). (1997). Autism as an executive disorder. New York, NY: Oxford University Press.Sabatino, A., Rittenberg, A., Sasson, N.J., Turner-Brown, L., Bodfish, J.W., & Dichter, G.S. (2013). Functional neuroimaging of social and nonsocial cognitive control in autism. J Autism Dev Disord, 43 (12), 2903–2913. Sasson, N.J., Turner-Brown, L.M., Holtzclaw, T.N., Lam, K.S.L., & Bodfish, J.W. (2008). Children with autism demonstrate circumscribed attention during passive viewing of complex social and non-social picture arrays. Autism Res, 1: 1.Sasson, N.J., Elison, J.T., Turner-Brown, L.M., Dichter, G.S., & Bodfish, J.W. (2011). Brief report: circumscribed attention in young children with autism. J Autism Dev Disord, 41(2), 242 – 247.Sasson, N.J., Dichter, G.S., Bodfish, J.W. (2012). Affective responses by adults with autism are reduced to social images but elevated to images related to circumscribed interests.?PLoS One, 7(8):e42457.Sears, L.L., Vest, C., Mohammed, S., Bailey, J., Rason, B.J., & Piven, J. (1999). An MRI study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol Psychiatry, 23(4), 613- 624. Shafritz, K.M., Dichter, G.S., Baranek, G.T., & Belger, A. (2008). The neural circuitry mediating shifts in behavioral response and cognitive set in autism. Biol Psychiatry, 63(10), 974- 980. ?Silverman, J.L., Pride, M.C., Hayes, J.E., Puhger, K.R., Butler-Struben, H.M., Baker, S., Crawley, N.J. (2015). GABA B receptor agonist R-baclofen reverses social deficits and reduces repetitive behavior in two mouse models of autism. Neuropsychopharmacology, 40(9), 2228–2239.Szatmari, P., Georgiades, S., Bryson, S., Zwaigenbaum, L., Roberts, W., Mahoney, W., . . . Tuff, L. (2006). Investigating the structure of the restricted and repetitive behaviors and interests domain of autism. J Child Psychol Psychiatry, 47(6), 582- 590.Turner, M.A. (1999). Annotation: repetitive behavior in autism: A review of psychological research. J Child Psychol Psychiatry, 40(6), 839-849.Turner-Brown, L.M., Lam, K.S., Holtzclaw, T.N., Dichter, G.S., Bodfish, J.W. (2011). Phenomenology and measurement of circumscribed interests in autism spectrum disorders. Autism, 15(4), 437 – 456.Uddin, L.Q., Supekar, K., Lynch, C.J., Khouzam, A., Phillips, J., Feinstein, C., . . . Menon, V. (2013). Salience network-based classification and prediction of symptom severity in children with autism. JAMA Psychiatry, 70(8), 869 – 879.Vasa, R.A., Mostofsky, S.H., & Ewen, J.B. (2016). The disrupted connectivity hypothesis of autism spectrum disorders: time for the next phase of research. Biol Psychiatry Cogn Neurosci Neuroimaging, 1(3), 245 – 252.Watson, K.K., Miller, S., Hannah, E., Kovac, M., Damiano, C.R., Sabatino-DiCrisco, A., . . . Dichter, G.S. (2015). Increased reward value of non-social stimuli in children and adolescents with autism. Frontiers in Psychology, 6, 1026. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download