Subthreshold autism spectrum disorder in patients with eating disorders
Introduction
Increasing attention has been recently devoted to detecting the shared vulnerability between Anorexia Nervosa (AN) and Autism Spectrum Disorder (ASD) [1], [2], [3] upon the evidence of frequent AN comorbidity in ASD [4], [5], [6], [7], leading authors to highlight a possible conception of AN as a neurodevelopmental disorder [8], besides propelling interest on the possible overlap between Eating Disorders (EDs) and ASD [1], [8], [9].
In a large Swedish cohort of AN individuals (“the Göteborg AN study”) [8], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], higher rates of ASD were reported with respect to healthy participants. However, most recent studies reported conflicting results, despite instruments employed to assess symptoms often differed among studies leading to difficulties in comparing data. Pooni et al. [22] using the Developmental, Dimensional and Diagnostic Interview (3Di), found similar prevalence of ASD in healthy control participants and in participants with EDs, this latter showing only a higher frequency of repetitive behaviours. Rhind reported a 4% rate of ASD among AN participants by means of the Development and Well-being Assessment (DAWBA) [23], while other studies reported higher rates of ASD symptomatology in AN female participants by means of the Autism Diagnostic Observation Schedule-Second Edition (ADOS-2), with a prevalence range from 10% to 23% [24], [25], [26], [27]. In particular, Mandy & Tchanturia [28], testing the hypothesis that AN females with socio-communicative deficit may have comorbid ASD, found that five out of ten women with EDs had clinically suspected ASD.
On the other hand, several studies have reported that individuals with ASD and AN seem to share common cognitive profiles defined by rigidity, inflexibility, perfectionism, impaired social cognition and executive functions [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], as well as social anhedonia [39], [40]. AN indivudals have also been shown to have increased attention to details, poor results in advanced theory of mind and neurocognitive tests [20], [41], and a lack of ‘emotional intelligence’ [42], [43]. These findings, together with the spreading concept of “Broader Autism Phenotype” (that is to say, subthreshold manifestations of ASD frequently found among unaffected relatives of people with autistic conditions [44], [45], [46], [47], [48], [49]), propelled a growing body of studies aiming to investigate not only full-blown ASD but also autistic traits among AN affected participants, most of which used the Autism Spectrum Quotient (AQ) or its brief version (AQ-10) as assessment tools [37], [41], [50], [51], [52], [53], [54]. A recent meta-analysis of these studies [3] showed that individuals with AN score significantly higher than healthy controls on the AQ but they seldom exceed the cut-off score for a full diagnosis of ASD. A possible interpretation of these results could argue the fact that ED samples are mostly represented by females, while available questionnaires for ASD have been criticized for being tailored only on a male stereotype of ASD [55].
Despite evidences demonstrating the frequent diagnostic crossover among different ED and despite recent trends in literature suggesting to reconsider EDs from a trans-diagnostic perspective [56], [57], a possible relationship between autistic traits and EDs other than AN has been fairly neglected in the literature [28], [58]. Among the few studies, Vagni et al. [59] recently examined a sample of ED participants, including AN, Bulimia Nervosa (BN) and Binge Eating Disorder (BED), using the Ritvo Autism Asperger Diagnostic Scale Revised (RAADS-R) and reported that 33% of ED participants scored above the cut-off without significant differences between ED categories. The lack of studies is even more surprising in light of the literature showing that the cognitive profile found in both ASD and AN might be common to other EDs as well. For instance, Medina-Pradas et al. [60] highlighted a poorer emotional theory of mind in participants with BN or ED not otherwise specified compared to healthy controls, while other studies reported alterations in cognitive flexibility and social anhedonia not only in AN but also in BN participants [39], [61]. Such evidence suggests that the relationship between autistic traits and other EDs, such as BN and BED, may deserve further investigation.
Considering ASD, it is also noteworthy that a growing number of studies have highlighted remarkable gender differences in ASD clinical presentation, focusing on female phenotype of ASD [62], [63], [64], [65], [66]. In this regard, even without a diagnosis of ASD, higher autistic traits in AN could still be in line with the conception of AN as a possible ASD presentation in females [2], [67], [68], [69], [70] as well as the presence of autistic traits in childhood, which seems to increase the risk of developing EDs [71], [72], [73], [74]. However, ASD presence in AN populations may be over-represented due to symptomatology arising not during early development but in adolescence. Moreover, it is not clear if there is a peculiar relationship between ASD and EDs, since ASD seems to be linked also with other psychiatric disorders [25]. This over representation and the marked gender-prevalence outline the need for accurate and careful assessment of these pathological dimensions [75].
In the framework of a spectrum approach to psychopathology, proposed by the Italian-American research project named Spectrum-project (www.spectrumproject.org) [76], [77], [78], [79], [80], [81], [82], [83], [84], [85], [86], we recently developed and validated the Adult Autism Subthreshold Spectrum (AdAS Spectrum), which aims to assess both typical and atypical symptoms, but also attenuated manifestations, personality traits, and behavioural features that may be associated with ASD but which may also be present in subthreshold or partial forms [55]. Compared to other available instruments, the AdAS Spectrum, besides assessing more subtle manifestations of autism spectrum, investigates features that have been suggested as the female phenotype of ASD [62], [63], [64], [65], [66], [67]. More specifically, gender-related manifestations included in the questionnaire refer either to the tendency of women with ASD to camouflage by imitating others, often adopting an alternative persona and acting as someone who is socially successful, or to the proneness to avoid social interactions, choosing instead to engage in creative solitary activities, read fiction, or spend time with animals [67]. Moreover, since women with ASD are known to cope with their own social incompetence by remaining peripheral in social situations also developing intense anxiety, some social anxiety-like behavioural features have been included in the questionnaire [67], [87]. Lastly, the AdAS Spectrum domain Restricted interest and rumination took into account that fixated interests of ASD females are as strong as in males but they likely focus on subjects that are common among neurotypical counterparts as well [67]. Given the high prevalence of EDs among females, we expect that the AdAS Spectrum might represent a useful instrument to detect autistic traits underlying ED psycopathology.
The aim of the present study was to assess the presence of autism spectrum disorder symptoms in a sample of participants diagnosed with AN, BN and BED. Specifically, we aimed to test whether high levels of autistic traits are present in AN participants even in the absence of full ASD diagnosis and whether they also characterize BN and BED and/or they relate to specific cross-diagnostic eating behaviours (e.g., restrictive eating, binge-eating).
Section snippets
Methods
A sample of 138 participants with EDs, diagnosed in accordance to DSM-5 criteria, and 160 Healthy Controls (HCs) without any current or lifetime history of mental disorders were recruited at 3 Italian University Departments of Psychiatry (Pisa, Florence and Naples). We excluded participants who were aged below 18 years old, were unable to complete the assessments due to language or intellectual impairment, or with a diagnosis of Schizophrenia or full-blown ASD according to the SCID-5. The HC
Statistical analyses
Continuous variables were reported as mean ± standard deviation (SD), whereas categorical variables were reported as percentages. Univariate analysis of variance (ANOVA, with Tukey-B post hoc test for the pairwise comparison between the groups) was adopted to compare different ED sub-groups (AN-R, AN-BP, BN, BED) and HCs. Participants were also compared by means of the independent sample t-test, according to the presence/absence of BEB. The χ2 was used for comparisons of categorical variables.
Socio-demographic characteristics
All study groups reported more females than males with a statistically significant lower female/male ratio among healthy controls with respect to all ED subgroups (See Table 1). Restrictive participants were less likely to be engaged in a relationship and to be employed with respect to participants with BEB. As expected, participants with AN showed lower BMI with respect to the other diagnostic groups.
Clinical characteristics of the study sample
ED participants showed significantly higher AdAS Spectrum total scores than HCs. Significantly
Discussion
To the best of our knowledge this study is the first to report adult autism subthreshold spectrum symptoms in a sample of participants with different EDs compared to a group of HCs. In particular, this is the first study to explore the potential differences between ED participants with BEB compared to restrictive ones. The results of the present study corroborate previous data on high autism spectrum symptom levels among ED participants [24], [25], [26], [27], [28], [58] with respect to HC
Conclusions
Despite the described limitations, our study show that there are higher levels of autistic subthreshold symptoms among participants with AN, BN and BED compared to general population. We found that restrictive participants are affected by higher social insecurity and interpersonal distrust, as recorded using EDI-2, and higher autistic traits than participants with BEB: in particular higher inflexibility, adherence to routine and restricted interest/rumination, as demonstrated by the higher
Acknowledgment
The study was supported by the European Project TRIGGER (Transforming Institutions by Gendering Contents and Gaining Quality in Research – Grant agreement no. 611034).
References (96)
- et al.
Case study: anorexia nervosa and autistic disorder in an adolescent girl
J Am Acad Child Adolesc Psychiatry
(1996) Anorexia nervosa in 51 Swedish adolescents: premorbid problems and comorbidity
J Am Acad Child Adolesc Psychiatry
(1992)- et al.
Impairment of social function in young females with recent-onset anorexia nervosa and recovered individuals
J Adolesc Health
(2017) - et al.
Eating disorders in adolescent boys: a review of the adolescent and young adult literature
J Adolesc Health
(2003) - et al.
Emotional intelligence in anorexia nervosa: is anxiety a missing piece of the puzzle?
Psychiatry Res
(2012) - et al.
Adult autism subthreshold spectrum (AdAS Spectrum): validation of a questionnaire investigating subthreshold autism spectrum
Compr Psychiatry
(2017) - et al.
Using the Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R) disentangle the heterogeneity of autistic traits in an Italian eating disorder population
Res Autism Spectr Disord
(2016) - et al.
Fetal programming effects of testosterone on the reward system and behavioural approach tendencies in humans
Biol Psychiatry
(2012) - et al.
Sex/gender differences and autism: setting the scene for future research
J Am Acad Child Adolesc Psychiatry
(2015) - et al.
Behavioural validation, precursors, and concomitants of picky eating in childhood
J Am Acad Child Adolesc Psychiatry
(2003)