Elsevier

Psychiatry Research

Volume 245, 30 November 2016, Pages 119-126
Psychiatry Research

Emotions mediate the relationship between autistic traits and disordered eating: A new autistic-emotional model for eating pathology

https://doi.org/10.1016/j.psychres.2016.08.021Get rights and content

Highlights

  • Eating pathology correlated with four out of five autism quotient (AQ) subscales.

  • There was no significant relationship between empathising and eating pathology.

  • Emotion dysregulation mediated the relationship between AQ and eating pathology.

  • Prevention programs should address emotion dysregulation in individuals scoring high on the AQ.

Abstract

The aim of the study was to assess the extent of overlap between autistic traits, body dissatisfaction and disordered eating and to explore the mediating effects of negative attitudes towards emotional expression and emotion dysregulation. The sample comprised 416 university students (82% females, 17-48 years [M=19.76, SD=3.85]), who completed an online questionnaire assessing eating attitudes and behaviours (including dieting, bulimia and oral control), body dissatisfaction, and autistic traits (including the Autism Quotient [AQ] and its related subscales as well as the Empathising Quotient). Attitudes towards emotional expression and emotion regulation were also assessed. Results revealed that eating pathology correlated highly with all AQ subscales, with the exception of the attention to detail subscale. However, there was no significant relationship between empathising and eating pathology. Path-analyses indicated that emotion dysregulation, but not negative attitudes towards emotional expression, was a significant mediator of the relationship between AQ, body dissatisfaction and eating pathology. Direct relationships were also obtained for the AQ-bulimia and the AQ-oral control paths. Prevention and early intervention programs for eating pathology would likely benefit from addressing abnormalities in emotion processes in individuals who score highly on measures of autistic traits.

Introduction

A number of inter-related risk factors have been associated with the development of eating disorders (EDs) (Culbert et al., 2015, Hilbert et al., 2014). One theoretical model of EDs, the cognitive-interpersonal model, asserts that traits associated with autism spectrum disorders (ASD), such as problems navigating interpersonal relationships and social-emotional difficulties, may act as both risk and maintaining factors for EDs (Treasure et al., 2012). However, the majority of studies (Westwood et al., 2015) investigating the relationship between EDs and autistic traits have focused on clinical Anorexia Nervosa (AN) populations. Therefore, the question of whether this relationship holds in the general population remains largely unexplored (Carton and Smith, 2014, Coombs et al., 2011). While there have been a few studies outlining that ASD and EDs share some familial (genetic) behavioural traits (e.g. social impairment and restricted and repetitive behaviours) and intermediate phenotypes (e.g. impaired set-shifting and theory of mind as well as weak central coherence) (e.g. Caglar-Nazali et al., 2014, Westwood et al., 2016, Zucker et al., 2007), investigations into other mechanisms linking the two disorders, have been scarce. Given that deficits in emotion regulation have been found in both EDs (Meyer et al., 2010, Svaldi et al., 2012) and ASD populations (Globerson et al., 2015), it is possible that the relationship between these disorders may be mediated by emotion-related traits. The current study therefore assessed the overlap between eating pathology and a range of autistic traits in a large sample of university students, and also assessed whether emotions mediate this relationship by exploring a new autistic-emotional model of eating pathology. Although the official ED prevalence rate is around 0.5–3%, depending on the specific ED diagnosis, ED symptoms in the general population have been found to be as high as 12%, suggesting that it is meaningful to explore risk factors, such as ASD traits, in general population samples (Nagl et al., 2016).

Research assessing autistic traits in clinical ED populations has mainly focused on common cognitive processing styles across the two disorder groups. Current evidence points to a strong role for three particular deficits, impaired theory of mind (Caglar-Nazali et al., 2014, Davies et al., 2016), weak central coherence (Lang et al., 2014), and deficits in executive functions, including problems in set-shifting (Roberts et al., 2007, Tchanturia et al., 2012, Westwood et al., 2016). Social interaction impairments and reciprocal communication deficits, which are key symptoms of ASD, have also been reported in ED patients (Wentz et al., 2009, Zucker et al., 2007). Moreover, a recent review by Huke et al. (2013), investigating autistic traits in EDs, found that nearly 23% of individuals with an ED displayed autism spectrum traits. These findings provide support for the overlap between ASD and EDs.

To assess specific autism-spectrum features in clinical ED patients, recent studies have applied a test of autistic traits, the Autism-Spectrum Quotient (AQ; (Baron-Cohen et al., 2001)). A recent meta-analysis (Westwood et al., 2015) reviewing seven studies, reported significantly higher AQ scores in AN patients compared to controls, with the difference largely attributable to greater social skills deficits, poorer communication skills, and greater inflexibility. Conversely, no significant differences were found between AN patients and controls on the attention to detail AQ subscale. However, it should be noted that this review was limited by the small number of studies included and substantial cross-study heterogeneity. These results are nonetheless supported by several studies that have found differences between ED patients and controls across a number of AQ scales (e.g. Anckarsater et al., 2012), although some studies have only found differences for one AQ subscale (e.g., attention to detail scale, (Iwasaki et al., 2013)). To our knowledge, only one study (Iwasaki et al., 2013) assessed the relationship between specific eating pathology facets and the AQ subscales in a clinical ED sample. Results of this study revealed that whereas communication deficits were associated with increased body dissatisfaction, social-skill difficulties were related to greater bulimic symptom scores on the Eating Disorder Inventory (EDI), indicating that autistic traits might differ depending on the ED symptoms assessed.

More recently, empathising, the capacity to identify thoughts and feelings in others and to react with appropriate emotions, a commonly reported ASD symptom, has been investigated in relation to EDs. The importance of assessing empathising can be seen in the impaired theory of mind (Caglar-Nazali et al., 2014) and social interaction deficits, which are key symptoms in ED patients (Wentz et al., 2009, Zucker et al., 2007). The studies assessing empathising in EDs have mainly used two distinct measures to assess empathy, the Interpersonal Reactivity Index (Davis, 1983), a multi-dimensional assessment of empathy, and the Empathising Quotient (EQ; (Baron-Cohen and Wheelwright, 2004)), a measure of empathy used to account for the social and communication barriers commonly found in individuals with ASD. The findings of the studies using the Interpersonal Reactivity Index (Beadle et al., 2013, Calderoni et al., 2013) have been mainly inconclusive, with the study by Calderoni et al. (2013) revealing that AN patients reported less cognitive, but not affective empathy than controls, whereas the study by Beadle et al. (2013) found that AN patients demonstrated greater personal distress (a domain of affective empathy) compared to the controls. Of the few more recent studies (Baron-Cohen et al., 2013, Courty et al., 2013, Hambrook et al., 2008), that have utilized the EQ, only the study by Baron-Cohen et al. (2013) found that EQ scores were lower for AN patients compared to controls, however this finding was only obtained for the younger AN age group (aged 12–15 years). It is possible that lack of expected association between poorer empathising using the EQ and AN is due to the fact that low empathising abilities, characterized by social and communication barriers, are broadly present in both AN samples as well as the general population, therefore obscuring any group differences. However, future studies in both clinical and normal population samples are required to test this hypothesis.

EDs and ASD are both largely conceptualised as representing continuums that account for both clinical and non-clinical populations (Baron-Cohen et al., 2001, Miller and Aillancourt, 2011). Consequently, assessing the relationship between autistic traits and disordered eating symptoms in normal populations is necessary to obtain useful information about the clinical components of the disorders, as well as their more general manifestation across populations. Assessing the overlap in eating pathology and autistic traits in community samples might also address previous problems encountered in clinical ED studies. In particular, it may assist in disentangling the relationship between AN and ASD, which has proven problematic as autistic traits are very likely to arise as a consequence of AN and the associated low weight (Mandy and Tchanturia, 2015), however longitudinal research to confirm this claim is not yet available.

To our knowledge, only two studies have assessed the link between eating pathology and autistic traits in non-clinical samples, one assessing school children (Coombs et al., 2011) and the other one recruiting university students (Carton and Smith, 2014). Both studies reported that greater attention-to-detail, communication deficits, and attention-switching deficits assessed through the AQ, were significantly related to increased total Eating Attitudes Test (EAT-26; (Garner et al., 1982)) scores. However, only Coombs et al. (2011) observed a significant association between the EAT-26 oral control subscale and total AQ score.

There is currently only one study (Bremser and Gallup, 2012) that has assessed the role of empathising in eating pathology using a community sample. The study found that empathising was significantly related to eating pathology. Future studies using community samples are needed to verify these initial results and to clarify the discrepant findings regarding the relationship between EAT-26 scores and AQ scales (Carton and Smith, 2014, Coombs et al., 2011).

Despite the demonstrated link between disordered eating symptoms and ASD, the processes that connect the two disorder sets are currently unknown. Emotion dysregulation and negative attitudes towards emotional expression have been found to be entwined with both EDs (Harrison et al., 2010, Svaldi et al., 2012) and ASD (Samson et al., 2013). While research has consistently found a relationship between greater emotion regulation difficulties, including restricted access to emotion regulation strategies and lack of emotional clarity in both EDs (Meyer et al., 2010) and ASD (Globerson et al., 2015), research on negative attitudes towards emotions is more limited. The few studies assessing negative attitudes towards emotions revealed significant positive associations between eating pathology and the reduced tendency to express emotions (Haslam et al., 2012), as well as the belief that expressing emotions is a sign of weakness (Meyer et al., 2010). Hence, in order to gain a more comprehensive understanding of the relationship between autistic traits and eating pathology, it is important to also take into account the mediating effect of emotion dysregulation and negative attitudes towards emotions.

The present study sought to build on the current state of knowledge by addressing the aforementioned limitations and research gaps. This study therefore investigated the associations between autistic traits and eating pathology in a non-clinical student population with an additional focus on whether negative attitudes towards emotions and emotion dysregulation might mediate this relationship. Assessment of a range of autistic traits, including empathising capacities, allowed for a more comprehensive picture of the specific autistic traits that may be associated with eating pathology. Additionally, identifying the potential mechanisms involved, including emotion regulation and attitudes towards emotions, can provide valuable insight into the co-occurrence of both autistic traits and disordered eating symptoms.

More specifically, the current study had two aims: 1.) to investigate the associations between autistic traits and disordered eating symptoms in a non-clinical sample and 2.) to assess whether negative attitudes towards emotions and emotion dysregulation mediated the relationship between autistic traits, body dissatisfaction and eating pathology. The proposed autistic-emotional model (see Fig. 1), was assessed using path-analyses to test this second aim. We decided to include body dissatisfaction as an extra mediating variable in our model, as research has shown it is the strongest predictor of eating pathology (Stice et al., 2010) and to our knowledge, with the exception of the study by Iwasaki et al. (2013), no other study has assessed the relationship between body dissatisfaction and autistic traits.

Section snippets

Participants

Participants were 416 university students (82% female) from an Australian University. Ages ranged from 17 to 48 years (M=19.76 years, SD=3.85). Two hundred and forty (57.7%) participants were Australian-born. A significant proportion of participants identified as Asian (42.1%, n=175) and Caucasian (37.5%, n=156), followed by Middle Eastern (4.1%, n=17), with other groups constituting 15.4% of the sample (n=64). As it was a student population, most participants had completed at least secondary

Data cleaning

Missingness was lower than 5% and was consistent with missing completely at random; Little's MCAR x2(df=771)=367.05, p=0.99. Expectation maximization was used to fill in these missing values (Tabachnick and Fidell, 2013). All variables met guidelines for normality (Curran et al., 1996), and therefore were entered into analyses untransformed.

Associations between the variables assessed

Associations between the variables of interest are displayed in Table 1. As can be seen, relationships were of varying magnitudes. Eating pathology

Discussion

This is the first study to investigate the mediating effect of negative attitudes towards emotions and emotion dysregulation on the relationship between autistic traits, body dissatisfaction and eating pathology. Our findings demonstrated positive correlations between most of the AQ scales, body dissatisfaction and EAT-26 scales, with the exception of attention to detail. The only scale not significantly related to any of the variables assessed was the EAT-26 oral control scale. Conversely,

References (54)

  • S. Baron-Cohen et al.

    The autism-spectrum quotient (AQ): evidence from sperger syndrome/high-functioning autism, males and females, scientists and mathematicians

    J. Autism Dev. Disord.

    (2001)
  • J.N. Beadle et al.

    Alexithymia, emotional empathy, and self-regulation in Anorexia Nervosa

    Ann. Clin. Psychiatry

    (2013)
  • J.A. Bremser et al.

    From one extreme to the other: negative evaluation anxiety and disordered eating as candidates for the extreme female brain

    Evol. Psychol.

    (2012)
  • M.H. Davis

    Measuring individual differences in empathy: evidence for a multideimensional approach

    J. Personal. Soc. Psychol.

    (1983)
  • S. Calderoni et al.

    The impact of internalizing symptoms on autistic traits in adolescents with restrictive anorexia nervosa

    Neuropsychiatr. Dis. Treat.

    (2015)
  • S. Calderoni et al.

    Selective cognitive empathy deficit in adolescents with restrictive anorexia nervosa

    Neuropsychiatr. Dis. Treat.

    (2013)
  • A.M. Carton et al.

    Assessing the relationship between eating disorder psychopathology and autistic traits in a non-clinical adult population

    Eat. Weight Disord.

    (2014)
  • E. Coombs et al.

    An investigation into the relationship between eating disorder psychopathology and autistic symptomatology in a non-clinical sample

    Br. J. Clin. Psychol.

    (2011)
  • P.J. Cooper et al.

    The development and validation of the Body Shape Questionnaire

    Int J. Eat. Disord.

    (1986)
  • A. Courty et al.

    Levels of autistic traits in anorexia nervosa: a comparative psychometric study

    BMC Psychiatry

    (2013)
  • K.M. Culbert et al.

    Research Review: What we have learned about the causes of eating disorders – A synthesis of sociocultural, psychological, and biological research

    J. Child Psychol. Psychiatry

    (2015)
  • P.J. Curran et al.

    The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis

    Psychol. Methods

    (1996)
  • C. Evans et al.

    Body Shape Questionnaire: derivation of shortened “alternate forms”

    Int J. Eat. Disord.

    (1993)
  • D. Garcia-Villamisar et al.

    Comorbid psychopathology and stress mediate the relationship between autistic traits and repetitive behaviours in adults with autism

    J. Intellect. Disabil. Res.

    (2015)
  • I. Garrido et al.

    Decision-making and impulsivity in eating disorder patients

    Psychiatry Res.

    (2012)
  • D.M. Garner et al.

    The Eating Attitudes Test: psychometric features and clinical correlates

    Psychol. Med.

    (1982)
  • E. Globerson et al.

    Prosody recognition in adults with high-functioning autism spectrum disorders: from psychoacoustics to cognition

    Autism Res.

    (2015)
  • Cited by (20)

    • Body image and autism: A scoping review

      2023, Research in Autism Spectrum Disorders
    • The relationship between tendency to attend to detail, sensory sensitivity, and affective response to food cues – A registered report

      2021, International Journal of Psychophysiology
      Citation Excerpt :

      Although this was not one of our original predictions it is interesting, as previous research has found links between scoring on the social skills sub-scale and eating behaviour. For instance, higher levels of dieting have been found to be linked with stronger social skills in children (Coombs et al., 2011) and further research links scoring on the social skills sub-scale with the Eating Attitudes Test (bulimia scale) (Mansour et al., 2016). More generally, autistic traits have been linked with altered processing of social reward (e.g., Vukusic et al., 2017) and variations in facial responses to social stimuli - emotional faces (spontaneous mimicry during joint attention conditions) (Neufeld et al., 2016); it may be that such variations also extend to primary reward (food) when considering the social skills sub-scale specifically.

    • The sex-specific association between autistic traits and eating behavior in childhood: An exploratory study in the general population

      2020, Appetite
      Citation Excerpt :

      One study reported that a substantial proportion of women with eating disorders had a pattern of autistic traits (e.g. social and flexibility difficulties) before the onset of their eating disorder (Mandy & Tchanturia, 2015). Furthermore, emotion dysregulation has shown to be a possible mediator in the relationship between autistic traits, body dissatisfaction and eating pathology in a general population (82% female) of university students (Mansour et al., 2016). Possibly, the developmental path between autistic traits and eating behavior is not only associated with sensory dysregulation but also mediated by emotional dysregulation, which is more common in girls.

    View all citing articles on Scopus
    View full text