Elsevier

Neuropsychologia

Volume 45, Issue 9, 2007, Pages 2125-2134
Neuropsychologia

Differential fear conditioning in Asperger's syndrome: Implications for an amygdala theory of autism

https://doi.org/10.1016/j.neuropsychologia.2007.01.012Get rights and content

Abstract

Since the first descriptions of individuals with autism spectrum disorders (ASD), abnormalities in socio-emotional behaviours have been described as amongst the most characteristic clinical features of this condition. Current evidence in this area suggests that individuals with ASD experience difficulties in the perception and expression of emotions within the social domain. The causes for these emotional difficulties are, however, still poorly understood. At the developmental level, it is unclear whether emotional disturbances constitute a primary feature of the clinical presentation of ASD or whether they are secondary to abnormalities in other areas of cognition. At the neurobiological level, it is still debated to what extent abnormalities of the limbic system, in particular the amygdala, may be responsible for the emotional disturbances characterising ASD. Here we show that a group of individuals with Asperger's syndrome exhibit a pattern of abnormality in differentially acquiring fear, which suggests that their fear responses are atypically modulated by conditioned and non-conditioned stimuli. On the basis of these results and the existing literature we suggest that ASD may be characterised by atypicalities in the integration of physiological and cognitive aspects of emotional experiences which we argue arise because of poor connectivity between the amygdala and functionally associated cortical areas.

Introduction

Autism is a developmental disorder mainly characterised by impairments in reciprocal social and emotional behaviours accompanied by varying degrees of abnormal cognitive development. Although autistic disorder is distinguished from Asperger's syndrome (American Psychiatric Association, 2000, World Health Organisation, 1992) on the basis of language development, which is significantly delayed in the former but not the latter, it is now widely accepted that both conditions form part of the same spectrum of syndromes. Research over the past half a century has provided a rather detailed description of the extent and nature of the social and cognitive difficulties experienced by individuals with autism spectrum disorders (ASD), but despite Kanner's (1943) emphasis on the role of abnormal affect in the clinical manifestations of autism, the study of emotional processes in ASD is relatively limited.

In the current paper we draw on the fear conditioning literature in order to provide further insights into emotional processes in ASD. Fear conditioning is a form of Pavlovian conditioning through which individuals learn the hedonic values of previously neutral stimuli via a process of association. In a typical fear conditioning study participants are presented with a simple visual or auditory stimulus alongside a painful or noxious stimulus such as a startling noise or mild electric shock (the unconditioned stimulus; UCS). Naturally, individuals will respond to such noxious stimuli with species-typical defence behaviours (the unconditioned response; UCR) including increased autonomic activity, which in humans can readily be measured by monitoring skin conductance responses (SCR; Frederikson, Annas, Georgiades, Hursti, & Tersman, 1993). After a few pairings of the neutral stimulus and the UCS, participants will start to exhibit such fear responses to the neutral stimulus alone (the stimulus has become a conditioned stimulus; CS), indicating that they have learned the association between the noxious and neutral stimuli.

To date, only one investigation has examined fear conditioning in ASD. Bernier, Dawson, Panagiotides, and Webb (2005) employed a potentiated startle paradigm in order to assess simple fear conditioning in a group of adolescents and adults with ASD. The authors aversively conditioned participants to a red square by pairing its presentation with an aversive puff of air to the throat. Following several pairings of the red square and the puff of air, the authors assessed participants’ eye-blink startle response to either a loud noise presented alone or accompanied by the red square. The results showed that as in typical participants, eye-blink startle responses in the ASD group were enhanced during the trials including the red square indicating that both groups had learned the aversive nature of the conditioned stimulus to a similar extent.

There are several reasons why such studies of fear conditioning are of value to our understanding of ASD. First, as Bernier et al. point out, such studies contribute to our understanding of the neuropathology underlying this spectrum of disorders. Extensive animal (see LeDoux, 1994, LeDoux, 1995, LeDoux, 1998, LeDoux, 2000 for detailed reviews) and human (Bechara et al., 1995, Büchel et al., 1999; Büchel, Morris, Dolan, & Friston, 1998; Cheng, Knight, Stein, & Smith, 2003; Knight, Smith, Cheng, & Stein, 2004; LaBar, Gatenby, Gore, LeDoux, & Phelps, 1998; LaBar, LeDoux, Spencer, & Phelps, 1995; Morris, Friston, & Dolan, 1997; Morris, Friston, & Dolan, 1998; Morris, Öhman, & Dolan, 1998; Phelps et al., 1998) research has demonstrated that the associative learning in fear conditioning paradigms is mediated by the amygdala, a limbic structure which has attracted increasing attention in relation to ASD in recent years. Although several lines of research have implicated the amygdala in the pathology underlying this disorder (e.g., Bachevalier, 1994, Bachevalier, 2000, Baron-Cohen et al., 1999, Baron-Cohen et al., 2000, Fotheringham, 1991, Howard et al., 2000, Sweeten et al., 2002), the evidence is somewhat inconsistent and the extent and nature of the proposed amygdala pathology remain unclear (see Amaral, Bauman, & Mills Schumann, 2003; Palmen, van Engeland, Hof, & Schmitz, 2004; Sweeten et al., 2002 for recent reviews). Fear conditioning paradigms are valuable in this respect because different forms of conditioned fear behaviour have been shown to rely on different amygdala nuclei or pathways. The acquisition of fear in simple conditioning paradigms such as the one employed by Bernier et al. (2005), for example, is thought to be mediated primarily by a sub-cortical amygdala system involving direct sensory afferent projections from thalamic nuclei and efferent connections to various brainstem and hypothalamic nuclei that mediate the behavioural and physiological fear responses (LeDoux, 1998, LeDoux, 2000). As Bernier et al. (2005) point out, their findings suggest that at least this sub-cortical system appears to be functionally relatively intact in ASD. Important for our current investigation are findings which suggest that fear acquisition in more complex differential fear conditioning paradigms, in which participants acquire fear to only one of several different stimuli (e.g., different colours), have been shown to rely on cortical modulation of the sub-cortical amygdala system (Jarrell, Gentile, Romanski, McCabe, & Schneiderman 1987; Morris et al., 1997). This cortical modulation is thought to be important for the regulation of fear responses according to the specific conditioning contingencies (i.e., responding to the conditioned stimulus but not to the non-conditioned stimuli). Since several lines of evidence indicate that ASD may be characterised by poor connectivity between disparate brain regions (Belmonte et al., 2004, Ben Shalom, 2000, Brock et al., 2002; Castelli, Frith, Happé, & Frith, 2002; Just, Cherkassky, Keller, & Minshew, 2004; Just, Cherkassky, Keller, Kana, & Minshew, in press; McAlonan et al., 2005, Rippon et al., in press), an investigation of differential fear conditioning in ASD may provide valuable behavioural insights into the functional integrity of cortico-amygdala connectivity in this population.

In addition to providing further insights into the functional integrity of amygdala systems, studies of fear conditioning may also inform debates about the developmental role of emotional atypicalities in the clinical presentation of ASD. To date most investigations relevant to this debate have focused on how individuals with ASD perceive and express emotions within the broader context of social behaviour. Although the evidence in this area is relatively consistent in illustrating that ASD is characterised by difficulties in the recognition (Hobson, 1986a, Hobson, 1986b, Hobson et al., 1988a, Hobson et al., 1988b; Hobson, 1991, Weeks and Hobson, 1987) and context appropriate expression of emotions (Dawson, Hill, Spencer, Galpert, & Watson, 1990; Kasari, Sigman, Baumgartner, & Stipek, 1993; Kasari, Sigman, Mundy, & Yirmiya, 1990; Sigman, Kasari, Jung-Hye, & Yirmiya 1992; Yirmiya et al., 1989, Yirmiya et al., 1992), these findings can be accommodated within competing explanatory frameworks. In line with Kanner's (1943) original conclusion, some authors have argued that emotional atypicalities constitute a primary and possibly innate feature of the autistic phenotype. Hobson (1989) for example suggests that individuals with ASD are characterised by difficulties in understanding the hedonic value of their sensory-motor environment which results in an abnormal developmental progression of interpersonal relatedness (see Mundy & Sigman, 1989 for a similar suggestion). Others, however, argue that the emotional difficulties evident in ASD are secondary to impairments in more general socio-cognitive processes. Schultz (2005), for example, argues that primary face processing atypicalities are responsible for the aberrant development of socio-emotional behaviours, whilst Baron-Cohen et al., 1999, Baron-Cohen et al., 2000 have suggested that difficulties in theory of mind (ToM) understanding give rise to the abnormal social and emotional behaviours characterising the autism spectrum. These latter accounts are supported by evidence which suggests that individuals across the autism spectrum experience difficulties in processing faces (e.g., Gross, 2005, Joseph and Tanaka, 2003; Partland, Dawson, Webb, Panagiotides, & Carver, 2004; Spezio, Adolphs, Hurley, & Piven, in press) and understanding mental states such as beliefs and desires of others (e.g., Baron-Cohen, Leslie, & Frith, 1985; Happé, 1995; but see Bowler, Briskman, Gurvidi, & Fornells-Ambrojo, 2005). Fear conditioning paradigms may provide important new insights into this issue because they assess a relatively basic and automatic emotional process that does not necessitate intact socio-cognitive processes. In addition, because fear conditioning paradigms assess the processes by which individuals learn the hedonic value of sensory stimuli, such paradigms constitute a relatively direct test of Hobson's (1989) suggestion that ASD may be characterised by atypicalities in understanding the hedonic value of their sensory-motor environment.

Finally, studies of fear conditioning in ASD will add to a small but growing number of studies that have investigated emotional processes in this population at the psychophysiological rather than the behavioural level. Since it is widely accepted that psychophysiological responses form an integral part of emotional experiences and behaviours (Cannon, 1929, James, 1984), the investigation of such responses in ASD is vital to understanding the nature of emotional atypicalities in this population. The limited evidence in this area to date suggests that like typical individuals, individuals with ASD exhibit changes in autonomic activity, such as increases in skin conductance responses (SCR) or changes in heart rate, when presented with emotionally salient pictures (Ben Shalom et al., 2003, Blair, 1999, Hillier et al., 2006; Salmond, de Haan, Friston, Gadian, & Vargha-Khadem, 2003), aversive auditory stimuli (Bernier et al., 2005; Salmond et al., 2003) or emotive words (Gaigg & Bowler, 2006). However, these physiological responses seem to be atypically modulated by specific stimulus properties in ASD. Blair (1999) for example found that although children with ASD exhibited typically increased SCRs to distress cues (e.g., crying face) as compared to neutral images, their responses to threatening images (e.g., gun) were less consistently elevated than in the comparison group (see Hillier et al., 2006 for similar findings). Similarly, SCRs to faces in ASD have been found to be abnormally modulated by the direction of gaze (Joseph, Ehrman, McNally, & Tager-Flusberg, 2005; Kylliainen & Hietanen, 2006). As Ben Shalom (2000) suggests, this pattern of results would be in line with the conjecture that ASD is characterised by atypicalities in the connectivity between cortical areas responsible for the cognitive appraisal of emotional stimuli and the amygdala which mediates our physiological reactions to such stimuli. Such a view is also supported by the finding that unlike in typical individuals, SCRs do not seem to correlate with subjective ratings of emotionality in ASD (Gaigg & Bowler, 2006; Hillier et al., 2006).

As this brief overview of the literature illustrates, there are several reasons why the study of fear conditioning is important for our understanding of ASD. In the current study we draw on a differential fear conditioning paradigm employed by Bechara et al. (1995) in order to test the hypothesis that individuals with ASD would exhibit a pattern of atypicality consistent with the suggestion that the amygdala is abnormally modulated by cortical areas. Thus, on the basis of the evidence pointing to a functionally intact sub cortical amygdala system in ASD (e.g., Bernier et al., 2005) we hypothesised that a group of ASD participants would exhibit typical patterns of physiological responses to aversive stimuli and that their autonomic activity would exhibit evidence of learning the hedonic value of a previously neutral stimulus. However, based on the view that the amygdala may not be modulated normally by cortical areas, we expected that participants with ASD would not exhibit a typical pattern of acquiring fear discriminately to conditioned and non-conditioned stimuli. If our predictions are borne out this pattern of results would lend support to Hobson's (1989) suggestion that ASD may be characterised by difficulties in understanding (in this case learning about) the hedonic value of their sensory-motor environment.

Section snippets

Participants

Fifteen individuals with Asperger's syndrome (12 male, three female) and 16 typical individuals (13 male, three female) participated in this experiment. One female Asperger and two male comparison participants were excluded from all analyses as they failed to exhibit detectable changes in skin conductance in response to the UCS (80–100 dB foghorn sound). Participants in the final sample (N = 14 per group) were matched on chronological age (Asperger mean = 29.7 years, SD = 10.2; comparison mean = 30.4 

Results

Groups did not differ significantly in terms of the UCS intensities they chose (Asperger, M = 94 dB, SD = 6; comparisons, M = 97 dB, SD = 3). Similarly, a 2 (group) × 6 (trial) mixed ANOVA of SCRs elicited during the six CS+paired trials of acquisition revealed no main effects or interactions (all Fs < 1). Thus, the UCS was similarly effective for both groups in eliciting startle responses and neither group seemed to habituate to the UCS during the acquisition phase. An analysis of participants’ declarative

Discussion

In the current experiment we examined differentially conditioned autonomic fear responses in a sample of participants with a diagnosis of Asperger's syndrome and matched typical comparison participants in order to gain further insights into emotional processing difficulties and the functional integrity of the amygdala in ASD. On the basis of the relevant literature we hypothesised that (1) participants with Asperger's syndrome would exhibit typical levels of autonomic responses to aversive

Acknowledgements

We thank N. Samways for technical assistance in setting up the hardware for collecting SCR data and J. Hampton for comments on earlier drafts. This study was in part supported by a grant from the Wellcome Trust.

References (79)

  • E.A. Phelps et al.

    Extinction learning in humans: Role of the amygdala and vmPFC

    Neuron

    (2004)
  • R.T. Schultz

    Developmental deficits in social perception in autism: The role of the amygdala and fusiform face area

    International Journal of Developmental Neuroscience

    (2005)
  • T.L. Sweeten et al.

    The amygdala and related structures in the pathophysiology of autism

    Pharmacology, Biochemistry and Behavior

    (2002)
  • ADInstruments. (1994–2004). PowerLab Data acquisition unit (ML845) with SCR amplifier (ML116). ADInstruments Pty...
  • ADInstruments. (1994–2004). Chart 5 v5.2 07. ADInstruments Pty...
  • J.P. Aggleton

    The amygdala: A functional analysis

    (2000)
  • D.G. Amaral et al.

    The amygdala and autism: Implications from non-human primate studies

    Genes Brain and Behaviour

    (2003)
  • American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (DSM-IV-TR)....
  • J. Bachevalier

    The amygdala, social behaviour, and autism

  • S. Baron-Cohen et al.

    Social intelligence in the normal and autistic brain: An fMRI

    European Journal of study Neuroscience

    (1999)
  • A. Bechara et al.

    Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans

    Science

    (1995)
  • M.K. Belmonte et al.

    Autism and abnormal development of brain connectivity

    Journal of Neuroscience

    (2004)
  • D. Ben Shalom

    Autism: Emotions without feelings

    Autism

    (2000)
  • D. Ben Shalom et al.

    Intact galvanic skin responses and impaired self-reports in response to emotional pictures in high-functioning autism

    New York Academy of Sciences

    (2003)
  • R. Bernier et al.

    Individuals with autism spectrum disorder show normal responses to a fear potential startle paradigm

    Journal of Autism and Developmental Disorders

    (2005)
  • D.M. Bowler et al.

    Understanding the mind or predicting signal-dependent action? Performance of children with and without autism on analogues of the false-belief task

    Journal of cognition and development

    (2005)
  • J. Brock et al.

    The temporal binding deficit hypothesis of autism

    Development and Psychopathology

    (2002)
  • C. Büchel et al.

    Amygdala-hippocampal involvement in human aversive trace conditioning revealed through event-related functional magnetic resonance imaging

    Journal of Neuroscience

    (1999)
  • Cannon, W. B. (1929). Bodily changes in pain, hunger, fear, and rage, Vol. 2. New York: Appleton (cited in LeDoux, J....
  • F. Castelli et al.

    Autism Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes

    Brain

    (2002)
  • D.T. Cheng et al.

    Functional MRI of human amygdala activity during Pavlovian fear conditioning: Stimulus processing versus response expression

    Behavioral Neuroscience

    (2003)
  • G. Dawson et al.

    Affective exchanges between young autistic children and their mothers

    Journal of Abnormal Child Psychology

    (1990)
  • J.B. Fotheringham

    Autism: Its primary psychological and neurological deficit

    Canadian Journal of Psychiatry

    (1991)
  • M. Frederikson et al.

    Internal consistency and temporal stability of classically conditioned skin conductance responses

    Biological Psychology

    (1993)
  • S.B. Gaigg et al.

    The effects of arousal on learning and memory in Asperger's syndrome

  • T.F. Gross

    Global–local precedence in the perception of facial age and emotional expressions by children with autism and other developmental disabilities

    Journal of Autism and Developmental Disorder

    (2005)
  • F.G. Happé

    The role of age and verbal ability in the theory of mind task performance of subjects with autism

    Child Development

    (1995)
  • A. Hillier et al.

    Electrodermal responsiveness to affective stimuli among those on the autism spectrum

  • R.P. Hobson

    The autistic child's appraisal of expressions of emotion

    Journal of Child Psychology and Psychiatry

    (1986)
  • Cited by (67)

    • Age- and sex-specific fear conditioning deficits in mice lacking Pcdh10, an Autism Associated Gene

      2021, Neurobiology of Learning and Memory
      Citation Excerpt :

      Like Pcdh10+/- mice, several mouse models associated with ASD exhibit fear conditioning deficits, including CD38, Homer1a, and Scn1a mutants (Han et al., 2012; Banerjee et al., 2016; Kim et al., 2016), as well as several others (Markram et al., 2008; Stapley et al., 2013; Howell et al., 2017; Nolan et al., 2017; Fricano-Kugler et al., 2019). Reports of fear conditioning in individuals with ASD have been varied, with some reporting impairments, and some failing to find differences compared to controls (Bernier et al., 2005; Gaigg and Bowler, 2007; Sterling et al., 2013). Recent findings have indicated that amygdala response to fear conditioning is attenuated in individuals with ASD compared with age-matched controls (Top et al., 2016).

    • Emotional Expressivity in Toddlers With Autism Spectrum Disorder

      2018, Journal of the American Academy of Child and Adolescent Psychiatry
    • A single low dose of valproic acid in late prenatal life alters postnatal behavior and glutamic acid decarboxylase levels in the mouse

      2016, Behavioural Brain Research
      Citation Excerpt :

      Thus ‘non-selective’ associative learning by VPA exposed animals may be underpinned by some ‘disconnect’ between amygdala and hippocampus hubs involved in stimulus processing. This potentially fits with reports that ASD-related difficulties in discriminating between conditioned and unconditioned fear stimuli stem from “poor connectivity” between the amygdala and functional related cortical regions [65]; which is potentially due to a disruption in GABAergic signaling [66]. For example, a GAD65 knock-out has been reported to cause a generalization of cued fear (similar to that reported here) and impaired extinction of cued fear [66]; and our VPA exposed animals had significant reduction in GAD65 levels throughout the brain, including the hippocampus.

    View all citing articles on Scopus
    View full text