Autism spectrum disorder and interoception: Abnormalities in global integration?

Abstract

Research over the past three decades has seen a revived interest in the way the human body—and the way in which it is perceived—interacts with aspects of our experience. Consequently, interoception (i.e. the perception of physiological feedback from the body) has recently been shown to be associated with a wide range of cognitive, emotional, and affective functions, making it broadly relevant to the study of autism spectrum disorder. Although limited qualitative accounts and empirical studies suggest that individuals with autism spectrum disorder encounter abnormalities when perceiving and integrating physiological feedback from their bodies, other studies have suggested that people with/without autism spectrum disorder do not differ in interoceptive ability after accounting for alexithymia. In this article, we discuss the newly recognized importance of interoception in autism spectrum disorder with a focus on how deficits in the perception of bodily feedback might relate to the core features and co-occuring psychopathology of autism spectrum disorder. Finally, a new integrated theory is advanced which posits that people with autism spectrum disorder may experience a reduced capacity to integrate interoceptive information that may result in a narrow attentional bodily focus and reduced motivational and behavioral drives.

Keywords

alexithymia autism spectrum disorder body interoception sensory experiences weak central coherence

Introduction

Sherrington’s (1906) classical taxonomy of the sensory system originally described interoception as the process of monitoring sensory inputs exclusively from the viscera, such that it was functionally distinct from the proprioceptive (i.e. body and limb position) and exteroceptive (i.e. touch, temperature, and pain) sensory modalities. However, recent advances in neurofunctional anatomy have provoked a substantial revision of the definition of interoception to describe the sense of the physiological condition of the entire body, not merely the viscera (Craig, 2002, 2003, 2011, 2014). Thus, the interoceptive sensory system is now considered to be maintained by a homeostatic afferent pathway that originates within the small-diameter sensory afferent fibers that innervate all tissues and organs, and which terminates in the posterior insula (Craig, 2002, 2011). Specifically, information from the small-diameter fibers is projected to pre-autonomic and homeostatic sites in the caudal medulla via the lamina I neurons in the spine, trigeminal dorsal horns, and the nucleus of the solitary tract (NTS), and then on to the hypothalamus and limbic sensory cortex (i.e. anterior insula) and limbic motor cortex (i.e. anterior cingulate; Craig, 2003). Following a posterior-to-anterior processing gradient, the interoceptive activity that represents sensory qualities (e.g. itch, pain, temperature) is processed in the posterior insula, then re-represented and integrated in the mid-insula, and finally integrated in the anterior insula cortex (Craig, 2011). Consequently, it has been suggested that the integration of inputs in the insula provides the template for a coherent representation of all feelings at one moment or a “global emotional moment” (Craig, 2009, 2010, 2011). Importantly, functional neuroimaging studies suggest that activation of the anterior insula is uniquely associated with various subjective feelings including internal bodily states (e.g. pain, temperature, heart rate) and the experience of emotion (Craig, 2002, 2009, 2011; Kober et al., 2008; Zaki et al., 2012).

Despite these recent neuroanatomical advances in mapping the mechanisms and systems involved in interoception, there remain some inconsistencies in how the definition and quantification of this multidimensional process have been applied. Several terms that have been used to describe different aspects of interoception (e.g. interoceptive awareness, sensitivity, accuracy, and acuity) have been inconsistently employed in the literature, and the differences between objective and subjective modes of evaluating interoception have generally not been acknowledged (Garfinkel and Critchley, 2013). Recently, a tripartite model of discrete interoceptive processes has been proposed which makes a clear distinction between the objective, subjective, and metacognitive aspects of interoception, and these have been shown to be dissociable in empirical tests (Garfinkel et al., 2015). Thus, interoceptive accuracy refers to a person’s behavioral performance on objective tests of interoceptive ability, such as the heartbeat tracking task (Schandry, 1981) and heartbeat discrimination task (Katkin et al., 1983). Interoceptive sensibility refers to a self-perceived interoceptive focus that is typically assessed using subjective questionnaires, such as the Body Perception Questionnaire (BPQ; Porges, 1993) and Multidimensional Assessment of Interoceptive Awareness (MAIA; Mehling et al., 2012). However, both these scales make reference to “interoceptive awareness” rather than “interoceptive sensibility.” Finally, interoceptive awareness refers to the metacognitive awareness of interoceptive accuracy which is the correspondence between objective interoceptive accuracy and subjective self-reporting of interoception (Garfinkel et al., 2015).

Many contemporary studies that have examined the awareness of bodily sensations highlight the importance of physiological feedback from the body to various cognitive, motivational, emotional, and affective functions (Critchley and Harrison, 2013). For example, awareness of physiological changes in the body has been shown to be associated with emotional experience (Wiens, 2005). In particular, Füstös et al. (2013) showed that the accurate perception of bodily states can enable the more effective regulation of emotion. Moreover, Fukushima et al. (2011) showed that physiological feedback from the body permits us to make inferences about the affective state of others (e.g. empathy). Altogether, these findings suggest that the perception of bodily sensations is an important part of subjective and social human experience that provides us with a sense of homeostatic instinct and self-awareness. Furthermore, bodily feelings alert us to our physiological state of well-being, prompting us to seek water if we are thirsty or warmth when we are cold.

Together, these types of enquires are broadly relevant to the understanding of autism spectrum disorder (ASD) which is characterized by deficits in social communication and social interaction across different contexts, difficulties in socio-emotional reciprocity, presence of restricted interests and repetitive behavior, and atypical sensory processing (American Psychiatric Association, 2013). While few prior studies have directly examined interoception in people with ASD, several authors have recently suggested that people with ASD may experience sensory under-responsivity to internal stimuli (Elwin et al., 2012; Fiene and Brownlow, 2015). In addition, a novel theoretical perspective has suggested that dysfunctional interoception in ASD may be the result of a disruption of the oxytocin system (Quattrocki and Friston, 2014). Specifically, the authors suggest that an oxytocin-mediated interoceptive dysfunction in early development prevents the contextualization of interoceptive signals which are necessary to acquire associations between internal and external cues. However, the manner in which a putative dysfunction in the interoceptive system of people with ASD is related to the core features and co-occurring symptomatology (e.g. anxiety and depression) of the disorder is still unclear. Moreover, the empirical findings to date have typically been viewed in isolation, without consideration of the available neuropsychological frameworks for cognitive and perceptual processing in ASD, especially the local processing framework (i.e. weak central coherence (WCC); Frith, 1989). At the least, based on the scant empirical findings, a local processing framework may offer insight into how interoception operates at the perceptual level in people with ASD. For example, as detailed below, people with ASD may preferentially process the local features of interoceptive states but fail to integrate them together so that their global significance can be ascertained. In this article, we critically evaluate the existing literature examining interoception in people with ASD and propose a new integrated theory of interoception in ASD that attempts to incorporate interoception into the WCC framework.

Autism and interoception

Historical traces

Dysfunctional bodily awareness in ASD is not a contemporary realization. Despite the relative paucity of interoception research in the ASD sensory processing literature, hints of interoceptive difficulties can be inferred from the first descriptions of the disorder. For example, Kanner’s (1943) original case studies of early infantile autism noted a father’s description of his 5-year-old son with autism: “He has never shown a normal appetite. Seeing children eating candy and ice-cream has never been a temptation to him” (p. 217). Similarly, another case describes an 8-year-old girl: “At camp she slid into avitaminosis [i.e. vitamin deficiency] and malnutrition but offered almost no verbal complaints” (p. 229). In subsequent decades, Bettelheim (1967) also explicitly made reference to childhood difficulties in bodily awareness, commenting on a child’s perceived lack of awareness during defecation: “If autistic children show no reaction whatsoever to defecation, we must assume that the process of alienation from their own feelings has reached such proportions that they do not even feel what goes on in their bodies” (p. 111). Thus, although limited, this early clinical literature suggests that the processing of interoceptive information may be impaired in people with ASD.

Interoceptive brain structures and systems in ASD

The brains of people with ASD are characterized by structural and functional abnormalities, with many theorists suggesting that a difference in brain network connectivity is responsible for many of the hallmark behavioral symptoms of the disorder (e.g. socio-emotional deficits). In particular, functional neuroimaging studies have consistently demonstrated significant low connectivity between different brain regions in people with ASD (Just et al., 2004, 2007, 2012; Koshino et al., 2008; Müller et al., 2011), although some studies have reported higher connectivity (Belmonte et al., 2004; Keown et al., 2013; Wass, 2011). Importantly, substantial differences in the functional connectivity of the networks subserving emotion processing and interoception have been demonstrated in people with ASD relative to controls. In particular, Ebisch et al. (2011) showed disrupted functional connectivity between the insula, amygdala, and somatosensory cortices in people with high-functioning ASD. They suggested that the results were consistent with a dysfunctional interoceptive awareness network which may alter subjective feelings and the ability to attribute emotional valence to external events.

Recent neuroimaging studies have also suggested that there are considerable group differences in the regions that support interoceptive processing in people with ASD relative to neurotypical controls (Barttfeld et al., 2012; Di Martino et al., 2009). For example, in a disgust recognition task, people with ASD showed less activation in a number of cortico-limbic brain regions including the left insula (Ogai et al., 2003). Silani et al. (2008) also showed that people with ASD exhibited reduced activation of the anterior insula when they were required to introspect about their feelings. However, the authors reported that it was the degree of alexithymic traits that was predictive of reductions in insula activation and that it was not attributable to ASD, per se. In the same way, other authors have demonstrated that activation of the insula in people with ASD is modulated by the level of alexithymic traits (Bird et al., 2010), indicating that atypical activation of the brain regions supporting interoception is not a necessary feature ASD, but rather may be due to the frequently concomitant alexithymia. Nonetheless, a recent meta-analysis of 24 functional neuroimaging studies that used social paradigms has identified that the right anterior insula is consistently hypoactive in people with ASD (Di Martino et al., 2009). Thus, the results of these system-level neuroscience and functional neuroimaging studies provide a persuasive starting point to assert that interoceptive processing may be dysfunctional in people with ASD due to fundamental dysregulation in the brain networks subserving interoception.

The body beneath the skin

Only in the past 5 years have researchers sought to examine the way in which individuals with ASD attend to and process internally derived signals. Within this time, there has been a surge in empirical work on interoception in people with ASD, examining objective, subjective, and metacognitive aspects of interoceptive awareness (DuBois et al., 2016). However, much of this work has overlooked previous neuropsychological frameworks of ASD, and there has been little systematic examination of the extent to which different methodologies and participant groups have contributed to differences in the study results. As such, the previous research findings on alterations of interoception in ASD have yielded inconsistent and often contradictory results.

Several studies have examined the subjective evaluations of interoceptive ability in people with and without ASD. For example, a qualitative study by Elwin et al. (2012) examined the autobiographies of people with ASD and reported that they experienced persistent interoceptive abnormalities, especially hypo-sensitivities to internal cues such as a difficulty in detecting and recognizing bodily sensations (e.g. pain). However, as the participants had written extensive autobiographies indicating normal or greater intelligence (DuBois et al., 2016), the findings cannot be generalized to larger heterogeneous ASD populations. Similar results were demonstrated in an online survey assessing the subjective experience of functional body and a thirst awareness in people with/without ASD. People with ASD showed significantly lower body and thirst awareness relative to controls (Fiene and Brownlow, 2015), using the Body Awareness Questionnaire (BAQ; Shields et al., 1989) and a novel Thirst Awareness Scale (TAS; Fiene and Brownlow, 2015), respectively. Although again, most participants with ASD were assumed to be on the higher functioning end of the spectrum.

Studies assessing behavioral performance (i.e. accuracy) on objective measures of interoception have tended to show mixed results. For example, Schauder et al. (2015) assessed interoceptive accuracy using a heartbeat detection task in children with/without ASD and found that although they performed equally in their ability to mentally track heartbeats, the children with ASD were better at tracking their heartbeats over longer time intervals. While it is not clear why the children with ASD demonstrated superior heartbeat counting performance over longer intervals, the authors proposed that people with ASD may disproportionately allocate attentional resources to internal, rather than external sensory information. However, others have suggested that it may be due to non-specific task adherence (Shah et al., 2016b). Alternately, Garfinkel et al. (2016b) reported that adults with ASD showed impaired interoceptive accuracy but an exaggerated subjective awareness of bodily sensations relative to controls. They suggested that this impaired ability to objectively detect bodily signals accompanied by an inflated subjective perception of bodily signals represents an interoceptive trait prediction error (ITPE) that manifests as deficits in emotional sensitivity and increased anxiety. However, although the ASD participants completed a measure of autistic traits, these scores were not directly assessed against the measures of interoceptive accuracy, sensibility, or awareness. Thus, a direct relationship between ASD status or symptoms and interoception was not determined.

Several authors have argued that people with/without ASD do not differ on measures of interoception after taking account of alexithymic traits. In fact, the alexithymia hypothesis suggests that where observed, the emotional symptoms of autism will be due to alexithymia rather than to the autism per se (Bird and Cook, 2013). Consistent with this hypothesis, other researchers suggest that alexithymia should be regarded as a reflection of a general impairment in interoception (Brewer et al., 2016). However, again, the empirical results to date have tended to be inconsistent. For example, in a small study by David et al. (2013), autistic traits but not alexithymic traits were positively correlated with cardiac awareness score in a heartbeat counting task in neurotypical adults. In contrast, other studies have emphasized the unique importance of alexithymia in explaining the relationship between ASD diagnosis and interoception (Shah et al., 2016a, 2016b). However, most prior studies have typically used only a one-dimensional approach when measuring interoception (Livingston and Livingston, 2016). Thus, although it is generally appreciated that interoception is atypical in people with ASD, there is little clarity as to the precise nature of the relationship between interoceptive deficits, alexithymic traits, and ASD.

Local–global information processing in ASD

The features of external sensory stimuli can be viewed as having a hierarchical organizational structure such that local features (e.g. specific details) are subordinate to global features (e.g. complete picture). In neurotypical individuals, the global features are typically processed before the local features, an effect termed global precedence (Navon, 1977). In contrast, people with ASD often display a preference for and show superiority in the local processing of sensory stimuli. For example, children with ASD have been shown to respond faster and more accurately on the Embedded Figures Task (EFT) relative to controls (Jolliffe and Baron-Cohen, 1997; Shah and Frith, 1983). The EFT requires participants to search for and detect a target figure that is hidden in a complex visual pattern as quickly as possible (Gottschaldt, 1926). In a recent meta-analysis, the pooled data demonstrated that people with ASD show superior visuospatial performance on tasks, such as the EFT and block design, and a stronger local processing preference on Navon tasks (Muth et al., 2014). Accordingly, several influential neurocognitive models have been developed to better understand the perceptual processing differences in people with ASD. In particular, the WCC model (Frith, 1989) attempts to explain the above privileged access to local parts and details, and it has come to dominate the psychological literature examining the local processing biases evident in people with ASD.

Weak Central Coherence

The term “central coherence” was first used by Frith (1989) to describe the propensity of neurotypical individuals to pull together large amounts of information with potential global significance and to process it in context, often at the expense of their memory for specific details. As such, central coherence is the typical drive to pursue meaning by extracting the gist or seeing the “big picture” of the incoming information in everyday events. In contrast, WCC refers to a detail-focused local processing bias that may lead to a failure to see the “big picture” (Happe and Frith, 2006). Frith (1989) suggested that this central processing deficit will manifest as a failure to detect global meaning, and it may underlie many of the non-social deficits and aptitudes that are known to characterize the ASD experience, such as enhanced attention-to-detail and greater sensory-perceptual discrimination.

The WCC account of ASD has received notable attention, and it has undergone considerable refinements since its original formulation. Currently, WCC is conceptualized as a cognitive style that is the result of two separate dimensions: (1) a reduced global integration of information and (2) a bias toward local processing (Happé and Booth, 2008). Consistent with this approach, local processing biases have been observed experimentally in people with ASD in visuospatial, auditory, and verbal domains, with many of the observations related directly to perceptual processing (Happe and Frith, 2006; Plaisted, 2001).

Mechanisms of local processing bias

Despite the prominence of the WCC model, the precise neural mechanism(s) underlying the local processing biases in people with ASD is unclear. Two possible mechanisms have been proposed to posit abnormalities in either the specialized brain regions or pathways used in a given task, or diffuse changes in neural connectivity throughout the brain (Happe and Frith, 2006). For instance, studies that have investigated hemispheric asymmetry have consistently shown that global integrative processing is predominantly lateralized to the right hemisphere, whereas local processing is predominantly lateralized to the left hemisphere (Boksem et al., 2012; Gable et al., 2013; Volberg and Hübner, 2004; Volberg et al., 2009).

While functional abnormalities in the right hemisphere may have potential implications for the WCC model, few prior studies have shown evidence of right hemisphere abnormalities in people with ASD (McKelvey et al., 1995; Waiter et al., 2005), and some studies have suggested there are left hemisphere abnormalities (Peterson et al., 2015). In any case, two functional magnetic resonance imaging (fMRI) studies have examined performance on a local–global processing task in people with ASD, and they failed to find clear evidence of right hemisphere dysfunction (Damarla et al., 2010; Ring et al., 1999). Alternately, abnormalities in the magnocellular pathway following the dorsal stream of visual processing have been proposed to account for the observed weak visuospatial coherence (Milne et al., 2002; Pellicano et al., 2005). However, these abnormalities cannot account for the auditory local processing biases observed in people with ASD.

In contrast to a regional approach, other researchers have suggested that the mechanism underlying WCC is reduced anatomical and functional connectivity and synchronization between distant cortical regions in the brain. For example, in an early study using fMRI during a sentence comprehension task, people with ASD showed consistently reduced functional connectivity throughout the cortical language system relative to controls (Just et al., 2004). The results were interpreted to mean that there is an underfunctioning of integrative neurocircuitry in people with ASD which results in impoverished information integration and is responsible for the occurrence of WCC. Similarly, another early study proposed that the features of ASD that are associated with WCC emerge due to an impairment of temporal binding between specialized local neural networks (Brock et al., 2002). More recently, review papers examining structural and functional connectivity studies in ASD have shown aberrant connectivity using diffusion tensor imaging (DTI; Travers et al., 2012), magnetoencephalography (MEG; Lajiness et al., 2014), and fMRI (Rane et al., 2015). Thus, although a number of different explanatory frameworks have been advanced, the mechanisms underlying WCC are still unclear. Nonetheless, it is noteworthy that neural under-connectivity is purported to give rise to WCC and that dysfunctional anterior insula connectivity is consistently shown to play an important role in the social interaction deficits of ASD (Anderson et al., 2010; Di Martino et al., 2009; Uddin and Menon, 2009).

Local processing of interoceptive stimuli

Current concepts of interoception suggest that it is a homeostatic sensory capacity that represents the ongoing status of all tissues and organs in the body (Craig, 2002, 2011). Relying on specialized peripheral and central neural substrates that represent all afferent activities, this system is thought to generate distinct feelings from the body including hunger, thirst, pain, temperature, itch, and all other bodily sensations to assist us in meeting the challenges of the environment. At the core of this conceptualization is an appreciation that it provides a composite representation of all salient body states at each moment of time which is encoded as a feeling (Craig, 2010). Thus, it is as if a person is presented with an ongoing interoceptive scene that is built on the integration of local physiological features to produce global features which may have homeostatic significance, and which represents the entire physiological condition of the body over time. This interoceptive scene is analogous to Craig’s (2010) cinemascopic model of awareness that is made up of a continuously updating series of global emotional moments for each moment in time.

Importantly, we propose that this conceptualization of interoception can be combined with the WCC model to test specific hypotheses about ASD. This integrated theory provides a novel framework for understanding how interoception may be impoverished if local bodily stimuli are not integrated into a composite whole. For instance, thirst is a homeostatic drive state that arises from multiple sensations and physiological signals such as mouth dryness, viscous saliva, hypovolemia, hypotension, and increased osmolality (Stevenson et al., 2015), and the individual elements of the thirst experience can, therefore, be viewed as local stimulus inputs into the interoceptive scene, which when integrated together with other motivational and affective elements will result in a global feeling state of thirst. However, a deficit in the integration of these signals, consistent with WCC, may lead to an impaired ability to perceive global features of homeostatic significance. As a result, thirst might not be accurately perceived consciously or subconsciously, resulting in the failure to create appropriate motivational and behavioral drive states to relieve the thirst. In short, a failure to properly integrate individual interoceptive inputs into a homeostatic feeling or global emotional moment (Craig, 2010) is analogous to not recognizing the proverbial forest for the trees. While the exact mechanism underlying this proposed dysfunction is not known, a failure could occur at the levels of the anterior and posterior insula, whereby hypo-activity and/or under-connectivity between these and other regions may reduce the likelihood of the stimuli being integrated accurately. Such a proposal is complementary to the Bayesian account of neurodevelopment in ASD which posits that an aberrant oxytocin system in infancy may disrupt the precision of interoceptive contextualization and integration (Quattrocki and Friston, 2014). Figure 1 shows an illustrated summary of the potential failure to integrate local stimuli accurately and the associated reduction in adaptive motivational and behavioral drive states.

Figure 1.

Illustration of the potential disturbance to the interoceptive system in ASD proposed in the integrated theory.

Subjective emotion can also be viewed as the unified representation of physiological activity, and consequently, both historical and contemporary models of emotional processing have suggested that the subjective experience of an emotion arises from the perception of physiological feedback from the body (Damasio and Carvalho, 2013; James, 1884; Schachter and Singer, 1962). Thus, when we are afraid, we will experience symptoms associated with autonomic arousal symptoms (e.g. elevated heart and breathing rate, sweating, piloerection, vasoconstriction, and pupil dilation; Kreibig, 2010; Levenson, 2003). Again, each of these individual physiological responses can be viewed as discrete local stimulus inputs that contribute to the ongoing interoceptive scene and which contribute to the global emotional feeling state of “fear.” Thus, if a person with ASD has both a bias toward local interoceptive information and a reduced capacity to integrate interoceptive information, then their interoceptive scene may be fragmented, and this may prevent the accurate perception of emotion.

Furthermore, a person’s failure to adequately integrate the discrete interoceptive signals into a coherent whole may lead to a narrowed attentional focus on local features in the body, and this may lead to uncertainty in the interpretation of emotions and feelings in the self and others (i.e. alexithymia). Importantly, deficits in emotion recognition are fundamental to the clinical understanding of ASD (American Psychiatric Association, 2013) and alexithymia (Brewer et al., 2016) and they are known to be related to hypo-activity in the brain regions which support interoceptive awareness (i.e. anterior insula; Silani et al., 2008). Thus, the interoceptive differences experienced by people with ASD could contribute to emotion recognition and emotion regulation difficulties, social interaction problems (e.g. failure to share perspectives and empathize), and mental health problems (e.g. anxiety due to uncertainty as to how to interpret and regulate autonomic arousal symptoms).

General discussion

The objective measurement of interoceptive ability typically relies on behavioral evaluations to determine how well a person perceives their bodily sensations. By convention, methodological approaches to measuring interoception have typically emphasized the detection of cardiac activity (Kleckner et al., 2015). Other measures (e.g. respiratory, gastrointestinal, and genitourinary) have also been used (Cameron, 2002), but the findings related to interoceptive performance accuracy across different modalities are mixed (Garfinkel et al., 2016a; Herbert et al., 2012; Suschinsky and Lalumière, 2012). Two widely used cardiac interoception paradigms are the heartbeat counting (or mental tracking) method (Schandry, 1981) and the heartbeat detection task (Katkin et al., 1983). Both tasks provide a measure of cardiac interoceptive accuracy, and they are often used interchangeably in the literature, but the heartbeat counting task is the principle objective method that has been used to assess interoception in people with ASD (Garfinkel et al., 2016b; Schauder et al., 2015; Shah et al., 2016a, 2016b).

Heartbeat detection is generally thought to provide an indicator of general interoceptive ability, but from the standpoint of WCC, cardiac activity may be considered as a local feature in a person’s ongoing interoceptive scene. In this case, WCC would estimate that people with ASD would perform better or at least the same, relative to typically developing individuals. Corroborating this assertion, three prior studies that have compared interoceptive accuracy using the heartbeat counting task between people with ASD and controls have shown no difference in their interoceptive abilities (Schauder et al., 2015; Shah et al., 2016a, 2016b). Crucially, the study by Schauder et al. (2015) showed that children with ASD were superior at mentally tracking their heartbeats over longer time intervals. Although the authors suggested that the results may be due to an increased attentional focus on interoceptive cues and a potential attentional trade-off between internal and external cues, the results are also consistent with a bias for local interoceptive information. On the contrary, Garfinkel et al. (2016b) showed that while there was objective impairment in the heartbeat counting task in adults with ASD, the ASD group did not differ significantly from the control group using heartbeat detection task. They suggested that because accuracy on the heartbeat detection task requires simultaneous multimodal internal–external integration, this sensory integrative process remains intact in people with ASD. Although this is at odds with what would be predicted by a WCC model of interoceptive information processing, it may be worth exploring performance differences across these two tasks in larger samples of people with ASD and in those with a wider range of severity in ASD traits and features. In any case, if a local processing bias exists for internally derived signals in people with ASD, then it raises questions as to the utility of the heartbeat detection task, unless it is used in combination with other objective interoceptive measures.

Qualitative accounts and subjective measures of bodily awareness may, therefore, offer the best means of disentangling local and global interoceptive states. However, the existing research examining subjective measures of interoception has produced mixed results and/or the differences in the measures do not permit direct comparison. For example, one study showed that people with ASD scored significantly higher on the BPQ (Porges, 1993) relative to controls (Garfinkel et al., 2016b), whereas others showed that people with ASD scored significantly lower on the TAS (Fiene and Brownlow, 2015) and the BAQ (Shields et al., 1989) relative to controls (Fiene and Brownlow, 2015). Thus, the conflicting results may be due, in part, to the use of different scale measures and the extent to which they assess local and global sensory processes. For example, the BPQ mostly asks about localized bodily symptoms (e.g. dry mouth, facial twitches, heartbeat, and palms sweating; Porges, 1993), whereas the BAQ frames the questions in terms of integrated global bodily states (e.g. “I know in advance if I am getting the flu,” “I am aware of a cycle in my activity level throughout the day,” and “I notice distinct body reactions when I am fatigued”). Indeed, some items in the TAS presume that a person already knows what the globally integrated sensory experience of thirst is to answer the questions about specific sensations (e.g. “I notice distinct body reactions when I am thirsty,” and “I notice distinct body reactions when I am no longer thirsty”). Thus, it may be possible to reconcile at least some of the conflicting results in the literature by combining interoception with the WCC model and examining the prior relevant literature from this standpoint. Therefore, future research may benefit from separating local bodily symptoms and global bodily feelings when examining interoceptive sensibility in ASD.

Specific predictions about emotion processing problems in people with ASD can be tested by examining interoception from the perspective of the WCC. If there is a local processing bias for interoceptive signals, then general tasks of central coherence should generate results that are inversely correlated with the results of emotion recognition tasks in oneself and others. For example, the “Mind in the Eyes Test” (Baron-Cohen et al., 2001), which has been used to test the ability of individuals to read emotions and other mental states from micro-expressions in the eyes of other people, has been shown to be inversely correlated with speed on the EFT in people with/without ASD (Baron-Cohen and Hammer, 1997; Jarrold et al., 2000). This pattern of results suggests that a perceptual bias toward the local features of the environment will make the process of recognizing emotions in others more difficult, as is the case in people with ASD. Second, and by extension, feeling states such as thirst which require the integration of multiple interoceptive inputs should be more difficult to perceive and interpret for people with ASD than simple feeling states. Pertinently, Fiene and Brownlow (2015) reported that individuals with ASD show significantly lower subjective thirst awareness relative to controls, which is consistent with such a theorized local sensory processing bias.

The relationship between interoception and anxiety is well established, with prior studies typically demonstrating positive associations between trait anxiety and measures of interoceptive accuracy (e.g. heartbeat detection; Domschke et al., 2010). Indeed, anxiety is highly prevalent in people with an ASD diagnosis across all age groups and affects about one-half (42%–56%) of people with the disorder (Lai et al., 2014; Simonoff et al., 2008). Importantly, Garfinkel et al. (2016a) showed that the divergence between interoceptive accuracy and interoceptive sensibility, or ITPE, was predictive of anxiety symptoms in people with ASD beyond the severity of the disorder. Along with others (Paulus and Stein, 2010), they suggest that a person’s interoceptive structure may represent a vulnerability factor for anxiety. A local processing bias toward interoceptive stimuli may emphasize the potential threat posed by autonomic reactivity (e.g. heart rate) relative to other interoceptive and exteroceptive information that could be integrated into the interoceptive scene to alleviate a person’s anxious feelings.

Such a hypothesis has yet to be empirically examined, although a similar mechanism has been proposed to operate in people with panic disorder such that their awareness of fear-related physiological processes (e.g. cardiac activity) is increased and their awareness of other processes is simultaneously neglected (Fairclough and Goodwin, 2007; Limmer et al., 2015). Thus, while people prone to panic and anxiety may have learned to bias their attention toward cardiac activity due to its perceived potential for threat, people with ASD may preferentially process cardiac information as it represents a salient local stimulus.

Clinical implications

The use of a local–global approach in examining interoception in people with ASD and alexithymia may be informative with regard to the use of therapy. For example, rather than being a core deficit, WCC is conceptualized as a bias that can be overcome with explicit task demands, and it is not considered to be a non-modifiable deficit (Happe and Frith, 2006). Thus, global interoceptive processing might be improved in people with ASD via the use of body-scan meditation techniques that focus on all bodily regions. A recent study of an unsupervised brief body-scan meditation intervention showed improvement in somatosensory perceptual decision-making in participants with medically unexplained symptoms (Mirams et al., 2013). While the authors suggested that body-scan type meditation could reduce the misperception of physical symptoms in individuals with medically unexplained symptom, people with ASD and alexithymia may also experience benefits from increased attention and sensitivity to their bodies. In fact, adults with ASD showed significant reductions in depression, anxiety, and rumination and an increase in positive affect in a randomized-controlled trial of mindfulness-based therapy which included a body-scan component (Spek et al., 2013). Importantly, other prior studies have also noted the potentially therapeutic pathway of interoception training and meditation in reducing distress and anxiety (Schaefer et al., 2014; Serpa et al., 2014). Applying clinical remediation or training approaches such as these is likely to effectively enhance interoception in a range of conditions in which local processing biases occur (e.g. anorexia; Lopez et al., 2008).

Conclusion

This review discusses the potential importance of interoception in better understanding the experiences and behavior of people with ASD and also the potential mechanisms underlying some of the socio-emotional and sensory features of ASD. A picture of atypical interoceptive processing in ASD is starting to emerge in the literature which incorporates recent advances in systems neuroscience and recent empirical studies investigating objective and subjective bodily awareness. However, the impairments appear to be more strongly linked to alexithymia than to ASD, although alexithymia frequently co-occurs with ASD. Furthermore, by extending the WCC framework to include interoception, the integrative theory that we have proposed may represent a valuable starting point to better understand the core features of ASD and the psychopathology associated with the disorder. For example, the perceptual abnormalities outlined in such a framework may impair the perception of internal sensory information in people with ASD and potentially influence their experience of homeostatic and emotional feeling states, and it may also contribute to commonly experienced affective symptoms (e.g. anxiety).

Current methodological limitations related to the objective measurement of interoceptive processing present a challenge in incorporating interoception into the WCC framework. To fully understand the context of interoception and awareness in ASD, it will be necessary to appreciate the subjective evaluations of bodily perception. Ultimately, it will be necessary to investigate the objective, subjective, and metacognitive aspects of interoception in relation to both ASD and alexithymia before we can fully understand the nature, quality, and extent of the putative perceptual biases. Moreover, appreciation of these experiences across the range of autism severity is necessary given that much of the literature has focused on high-functioning individuals with ASD who have a greater capacity to comment on their internal sensations, states, and emotions. Finally, the theoretical approach that we have presented has potential implications for the wider application of the WCC framework in ASD, which has historically only accounted for non-social features of ASD.

Supplemental Material

AUT738392_Lay_Abstract – Supplemental material for Autism spectrum disorder and interoception: Abnormalities in global integration?

Supplemental material, AUT738392_Lay_Abstract for Autism spectrum disorder and interoception: Abnormalities in global integration? by Timothy R Hatfield, Rhonda F Brown, Melita J Giummarra and Bigna Lenggenhager in Autism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: MJG was supported by an Australian Research Council Discovery Early Career Research Award (DE170100726). BL was funded by the Swiss National Science Foundation.

ORCID iD

Timothy R Hatfield

Melita J Giummarra

Bigna Lenggenhager

References

American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Arlington, VA: American Psychiatric Association.

Anderson

Druzgal

Froehlich

et al . (2010) Decreased interhemispheric functional connectivity in autism. Cerebral Cortex 21(5): 1134–1146.

Baron-Cohen

Hammer

(1997) Parents of children with Asperger syndrome: what is the cognitive phenotype? Journal of Cognitive Neuroscience 9: 548–554.

Baron-Cohen

Wheelwright

Spong

et al . (2001) Are intuitive physics and intuitive psychology independent? A test with children with Asperger Syndrome. Journal of Developmental and Learning Disorders 5: 47–78.

Barttfeld

Wicker

Cukier

et al . (2012) State-dependent changes of connectivity patterns and functional brain network topology in autism spectrum disorder. Neuropsychologia 50: 3653–3662.

Belmonte

Allen

Beckel-Mitchener

et al . (2004) Autism and abnormal development of brain connectivity. The Journal of Neuroscience 24: 9228–9231.

Bettelheim

(1967) Empty Fortress. New York: Simon & Schuster.

Bird

Cook

(2013) Mixed emotions: the contribution of alexithymia to the emotional symptoms of autism. Translational Psychiatry 3(7): e285.

Bird

Silani

Brindley

et al . (2010) Empathic brain responses in insula are modulated by levels of alexithymia but not autism. Brain 33(5): 1515–1525.

10.

Boksem

Kostermans

Tops

et al . (2012) Individual differences in asymmetric resting-state frontal cortical activity modulate ERPs and performance in a global-local attention task. Journal of Psychophysiology 26: 51–62.

11.

Brewer

Cook

Bird

(2016) Alexithymia: a general deficit of interoception. Royal Society Open Science 3(10): 150664.

12.

Brock

Brown

Boucher

et al . (2002) The temporal binding deficit hypothesis of autism. Development and Psychopathology 14: 209–224.

13.

Cameron

(2002) Visceral Sensory Neuroscience: Interoception. New York: Oxford University Press.

14.

Craig

(2002) How do you feel? Interoception: the sense of the physiological condition of the body. Nature Reviews. Neuroscience 3: 655–666.

15.

Craig

(2003) Interoception: the sense of the physiological condition of the body. Current Opinion in Neurobiology 13: 500–505.

16.

Craig

(2009) How do you feel—now? The anterior insula and human awareness. Nature Reviews. Neuroscience 10(1): 59–70.

17.

Craig

(2010) The sentient self. Brain Structure & Function 214: 563–577.

18.

Craig

(2011) Significance of the insula for the evolution of human awareness of feelings from the body. Annals of the New York Academy of Sciences 1225: 72–82.

19.

Craig

(2014) How Do You Feel? An Interoceptive Moment with Your Neurobiological Self. Princeton, NJ: Princeton University Press.

20.

Critchley

Harrison

(2013) Visceral influences on brain and behavior. Neuron 77: 624–638.

21.

Damarla

Keller

Kana

et al . (2010) Cortical underconnectivity coupled with preserved visuospatial cognition in autism: evidence from an fMRI study of an embedded figures task. Autism Research 3: 273–279.

22.

Damasio

Carvalho

(2013) The nature of feelings: evolutionary and neurobiological origins. Nature Reviews. Neuroscience 14: 143–152.

23.

David

Azevedo

Lenggenhager

et al . (2013) Increased heartbeat interoception is predicted by autism spectrum traits in the typical population. In: International Meeting for Autism Research, San Sebastián, May.

24.

Di Martino

Ross

Uddin

et al . (2009) Functional brain correlates of social and nonsocial processes in autism spectrum disorders: an activation likelihood estimation meta-analysis. Biological Psychiatry 65: 63–74.

25.

Domschke

Stevens

Pfleiderer

et al . (2010) Interoceptive sensitivity in anxiety and anxiety disorders: an overview and integration of neurobiological findings. Clinical Psychology Review 30: 1–11.

26.

DuBois

Ameis

Lai

M-C

et al . (2016) Interoception in autism spectrum disorder: a review. International Journal of Developmental Neuroscience 52:104–111.

27.

Ebisch

Gallese

Willems

et al . (2011) Altered intrinsic functional connectivity of anterior and posterior insula regions in high-functioning participants with autism spectrum disorder. Human Brain Mapping 32: 1013–1028.

28.

Elwin

Schroder

et al . (2012) Autobiographical accounts of sensing in Asperger syndrome and high-functioning autism. Archives of Psychiatric Nursing 26: 420–429.

29.

Fairclough

Goodwin

(2007) The effect of psychological stress and relaxation on interoceptive accuracy: implications for symptom perception. Journal of Psychosomatic Research 62: 289–295.

30.

Fiene

Brownlow

(2015) Investigating interoception and body awareness in adults with and without autism spectrum disorder. Autism Research 8(6):709–716.

31.

Frith

(1989) Autism: Explaining the Enigma. London: Wiley-Blackwell.

32.

Fukushima

Terasawa

Umeda

(2011) Association between interoception and empathy: evidence from heartbeat-evoked brain potential. International Journal of Psychophysiology 79: 259–265.

33.

Füstös

Gramann

Herbert

et al . (2013) On the embodiment of emotion regulation: interoceptive awareness facilitates reappraisal. Social Cognitive and Affective Neuroscience 8: 911–917.

34.

Gable

Poole

Cook

(2013) Asymmetrical hemisphere activation enhances global–local processing. Brain Cognition 83: 337–341.

35.

Garfinkel

Critchley

(2013) Interoception, emotion and brain: new insights link internal physiology to social behaviour. Commentary on: “Anterior insular cortex mediates bodily sensibility and social anxiety” by Terasawa et al. (2012). Social cognitive and Affective Neuroscience 8: 231–234.

36.

Garfinkel

Manassei

Hamilton-Fletcher

et al . (2016a) Interoceptive dimensions across cardiac and respiratory axes. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 371(1708): 20160014.

37.

Garfinkel

Seth

Barrett

et al . (2015) Knowing your own heart: distinguishing interoceptive accuracy from interoceptive awareness. Biological Psychology 104: 65–74.

38.

Garfinkel

Tiley

O’Keeffe

et al . (2016b) Discrepancies between dimensions of interoception in autism: implications for emotion and anxiety. Biological Psychology 114: 117–126.

39.

Gottschaldt

(1926) Über den Einfluss der Erfahrung auf die Wahrnehmung von Figuren. Psychological Research 8(1): 261–317.

40.

Happe

Frith

(2006) The weak coherence account: detail-focused cognitive style in Autism Spectrum Disorders. Journal of Autism and Developmental Disorders 36: 5–25.

41.

Happé

Booth

(2008) The power of the positive: revisiting weak coherence in autism spectrum disorders. The Quarterly Journal of Experimental Psychology 61: 50–63.

42.

Herbert

Muth

Pollatos

et al . (2012) Interoception across modalities: on the relationship between cardiac awareness and the sensitivity for gastric functions. PLoS ONE 7(5): e36646.

43.

James

(1884) What is an emotion? Mind 9(34): 188–205.

44.

Jarrold

Butler

Cottington

et al . (2000) Linking theory of mind and central coherence bias in autism and in the general population. Developmental Psychology 36: 126–138.

45.

Jolliffe

Baron Cohen

(1997) Are people with autism and Asperger syndrome faster than normal on the embedded figures test? Journal of Child Psychology and Psychiatry, and Allied Disciplines 38(5): 527–534.

46.

Just

Cherkassky

Keller

et al . (2004) Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity. Brain 127: 1811–1821.

47.

Just

Cherkassky

Keller

et al . (2007) Functional and anatomical cortical underconnectivity in autism: evidence from an FMRI study of an executive function task and corpus callosum morphometry. Cerebral Cortex 17: 951–961.

48.

Just

Keller

Malave

et al . (2012) Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity. Neuroscience and Biobehavioral Reviews 36: 1292–1313.

49.

Kanner

(1943) Autistic disturbances of affective contact. Nervous Child 2: 217–250.

50.

Katkin

Reed

Deroo

(1983) A methodological analysis of 3 techniques for the assessment of individual-differences in heartbeat detection. Psychophysiology 20: 452–452.

51.

Keown

Shih

Nair

et al . (2013) Local functional overconnectivity in posterior brain regions is associated with symptom severity in Autism Spectrum Disorders. Cell Reports 5: 567–572.

52.

Kleckner

Wormwood

Simmons

et al . (2015) Methodological recommendations for a heartbeat detection-based measure of interoceptive sensitivity. Psychophysiology 52: 1432–1440.

53.

Kober

Barrett

Joseph

et al . (2008) Functional grouping and cortical-subcortical interactions in emotion: a meta-analysis of neuroimaging studies. Neuroimage 42(2): 998–1031.

54.

Koshino

Kana

Keller

et al . (2008) fMRI investigation of working memory for faces in autism: visual coding and underconnectivity with frontal areas. Cerebral Cortex 18: 289–300.

55.

Kreibig

(2010) Autonomic nervous system activity in emotion: a review. Biological Psychology 84: 394–421.

56.

Lai

Lombardo

Baron-Cohen

(2014) Autism. The Lancet 383: 896–910.

57.

Lajiness-O’Neill

Bowyer

Moran

et al . (2014) Neurophysiological findings from magnetoencephalography in Autism Spectrum Disorder: a comprehensive review. Future Neurology 9(3): 355–384.

58.

Levenson

(2003) Blood, sweat, and fears. Annals of the New York Academy of Sciences 1000: 348–366.

59.

Limmer

Kornhuber

Martin

(2015) Panic and comorbid depression and their associations with stress reactivity, interoceptive awareness and interoceptive accuracy of various bioparameters. Journal of Affective Disorders 185: 170–179.

60.

Livingston

(2016) Commentary: alexithymia, not autism, is associated with impaired interoception. Frontiers in Psychology 7: 1103.

61.

Lopez

Tchanturia

Stahl

et al . (2008) Central coherence in eating disorders: a systematic review. Psychological Medicine 38(10): 1393–1404.

62.

McKelvey

Lambert

Mottron

et al . (1995) Right-hemisphere dysfunction in Asperger’s syndrome. Journal of Child Neurology 10: 310–314.

63.

Mehling

Price

Daubenmier

et al . (2012) The Multidimensional Assessment of Interoceptive Awareness (MAIA). PLoS ONE 7: e48230.

64.

Milne

Swettenham

Hansen

et al . (2002) High motion coherence thresholds in children with autism. Journal of Child Psychology and Psychiatry 43: 255–263.

65.

Mirams

Poliakoff

Brown

et al . (2013) Brief body-scan meditation practice improves somatosensory perceptual decision making. Consciousness and Cognition 22(1): 348–359.

66.

Müller

Shih

Keehn

et al . (2011) Underconnected, but how? A survey of functional connectivity MRI studies in Autism Spectrum Disorders. Cerebral Cortex 21: 2233–2243.

67.

Muth

Hönekopp

Falter

(2014) Visuo-spatial performance in autism: a meta-analysis. Journal of Autism and Developmental Disorders 44(12): 3245–3263.

68.

Navon

(1977) Forest before trees: the precedence of global features in visual perception. Cognitive Psychology 9: 353–383.

69.

Ogai

Matsumoto

Suzuki

et al . (2003) fMRI study of recognition of facial expressions in high-functioning autistic patients. Neuroreport 14: 559–563.

70.

Paulus

Stein

(2010) Interoception in anxiety and depression. Brain Structure & Function 214: 451–463.

71.

Pellicano

Gibson

Maybery

et al . (2005) Abnormal global processing along the dorsal visual pathway in autism: a possible mechanism for weak visuospatial coherence? Neuropsychologia 43: 1044–1053.

72.

Peterson

Mahajan

Crocetti

et al . (2015) Left-hemispheric microstructural abnormalities in children with high-functioning Autism Spectrum Disorder. Autism Research 8: 61–72.

73.

Plaisted

(2001) Reduced generalization in autism: an alternative to weak central coherence. In: Burack

Charman

Yirmiya

et al . (eds) The Development of Autism: Perspectives from Theory and Research. Mahwah, NJ: Erlbaum Publishers, pp.149–169.

74.

Porges

(1993) Body Perception Questionnaire. Baltimore, MD: Laboratory of Developmental Assessment, University of Maryland.

75.

Quattrocki

Friston

(2014) Autism, oxytocin and interoception. Neuroscience and Biobehavioral Reviews 47: 410–430.

76.

Rane

Cochran

Hodge

et al . (2015) Connectivity in autism: a review of MRI connectivity studies. Harvard Review of Psychiatry 23: 223–244.

77.

Ring

Baron-Cohen

Wheelwright

et al . (1999) Cerebral correlates of preserved cognitive skills in autism: a functional MRI study of embedded figures task performance. Brain 122(Pt 7): 1305–1315.

78.

Schachter

Singer

(1962) Cognitive, social, and physiological determinants of emotional state. Psychological Review 69: 379–399.

79.

Schaefer

Egloff

Gerlach

et al . (2014) Improving heartbeat perception in patients with medically unexplained symptoms reduces symptom distress. Biological Psychology 101: 69–76.

80.

Schandry

(1981) Heart beat perception and emotional experience. Psychophysiology 18: 483–488.

81.

Schauder

Mash

Bryant

et al . (2015) Interoceptive ability and body awareness in Autism Spectrum Disorder. Journal of Experimental Child Psychology 131: 193–200.

82.

Serpa

Taylor

Tillisch

(2014) Mindfulness-based stress reduction (MBSR) reduces anxiety, depression, and suicidal ideation in veterans. Medical Care 52: S19–S24.

83.

Shah

Frith

(1983) An islet of ability in autistic children: a research note. Journal of Child Psychology and Psychiatry 24(4): 613–620.

84.

Shah

Catmur

Bird

(2016a) Emotional decision-making in autism spectrum disorder: the roles of interoception and alexithymia. Molecular Autism 7: 43.

85.

Shah

Hall

Catmur

et al . (2016b) Alexithymia, not autism, is associated with impaired interoception. Cortex 81: 215–220.

86.

Sherrington

(1906) The Integrative Action of the Nervous System. New Haven, CT: Yale University Press.

87.

Shields

Mallory

Simon

(1989) The Body Awareness Questionnaire: reliability and validity. Journal of Personality Assessment 53: 802–815.

88.

Silani

Bird

Brindley

et al . (2008) Levels of emotional awareness and autism: an fMRI study. Social Neuroscience 3: 97–112.

89.

Simonoff

Pickles

Charman

et al . (2008) Psychiatric disorders in children with autism spectrum disorders: prevalence, comorbidity, and associated factors in a population-derived sample. Journal of the American Academy of Child and Adolescent Psychiatry 47: 921–929.

90.

Spek

Van Ham

Nyklíček

(2013) Mindfulness-based therapy in adults with an autism spectrum disorder: a randomized controlled trial. Research in Developmental Disabilities 34(1): 246–253.

91.

Stevenson

Mahmut

Rooney

(2015) Individual differences in the interoceptive states of hunger, fullness and thirst. Appetite 95: 44–57.

92.

Suschinsky

Lalumière

(2012) Is sexual concordance related to awareness of physiological states? Archives of Sexual Behavior 41(1): 199–208.

93.

Travers

Adluru

Ennis

et al . (2012) Diffusion tensor imaging in autism spectrum disorder: a review. Autism Research 5: 289–313.

94.

Uddin

Menon

(2009) The anterior insula in autism: under-connected and under-examined. Neuroscience and Biobehavioral Reviews 33: 1198–1203.

95.

Volberg

Hübner

(2004) On the role of response conflicts and stimulus position for hemispheric differences in global/local processing: an ERP study. Neuropsychologia 42: 1805–1813.

96.

Volberg

Kliegl

Hanslmayr

et al . (2009) EEG alpha oscillations in the preparation for global and local processing predict behavioral performance. Human Brain Mapping 30: 2173–2183.

97.

Waiter

Williams

Murray

et al . (2005) Structural white matter deficits in high-functioning individuals with autistic spectrum disorder: a voxel-based investigation. Neuroimage 24: 455–461.

98.

Wass

(2011) Distortions and disconnections: disrupted brain connectivity in autism. Brain and Cognition 75: 18–28.

99.

Wiens

(2005) Interoception in emotional experience. Current Opinion in Neurology 18: 442–447.

100.

Zaki

Davis

Ochsner

(2012) Overlapping activity in anterior insula during interoception and emotional experience. Neuroimage 62: 493–499.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.11 MB