logo SBA


Digital archive of theses discussed at the University of Pisa


Thesis etd-10122023-111140

Thesis type
Tesi di specializzazione (4 anni)
email address
g.amatori@studenti.unipi.it, g.amatori1992@gmail.com
Thesis title
Autistic dimensions underlying social anxiety, obsessive-compulsive, and panic disorders
Course of study
relatore Prof.ssa Dell'Osso, Liliana
correlatore Dott.ssa Carpita, Barbara
  • Neurodevelopmental Continuum
  • Panic Disorder
  • Obsessive-Compulsive Disorder
  • Social Anxiety Disorder
  • Autism Spectrum Disorder
Graduation session start date
Release date
Background: From a dimensional point of view, autistic traits (AT) are distributed along a continuum in the general population, with varying expressiveness and evidence based on their quantity and quality, as well as interactions with specific environmental factors. It has been hypothesized that, in the universe of subclinical manifestations, significant autistic traits might represent the neurodevelopmental matrix of vulnerability of multiple possible psychopathological conditions. This view would be supported by the frequent comorbidities with other mental disorders observed in the psychopathological trajectory of autistic patients. The aim of the present study is to investigate the presence of autistic traits in individuals with social anxiety disorder (SAD), obsessive-compulsive disorder (OCD), and panic disorder (PD), while also assessing which specific symptom dimensions of autism are most represented in the different disorders. In addition, the study aims to investigate the distribution of social-phobic, obsessive-compulsive and panic-agoraphobic traits within the total sample.

Material and method: The total sample consists of 245 subjects distributed into five diagnostic groups: 43 patients with autism spectrum disorder (ASD), 56 patients with social anxiety disorder (SAD), 46 patients with obsessive-compulsive disorder (OCD), 40 patients with panic disorder (PD), and 60 healthy controls (HC). The subjects were recruited from the Psychiatric Clinic of the University of Pisa and assessed by means of the Structured Clinical Interview for DSM-5 (SCID-5), the Adult Autism Subtreshold Spectrum (AdAS Spectrum) questionnaire, the Social Anxiety Spectrum - Short Version (SHY-SV) questionnaire, the Obsessive - Compulsive Spectrum - Short Version (OBS-SV) questionnaire and the Panic Agoraphobic Spectrum - Short Version (PAS-SV) questionnaire. Statistical analyses included Analysis of Variance (ANOVA), Bonferroni post-hoc tests, Chi square test, linear regression and multinomial logistic regression.

Results: The total AdAS score was highest in subjects with ASD and lowest in healthy controls. Particularly high scores, immediately following those of ASD subjects, were observed in the SAD and OCD groups, significantly higher than those in the PD group.
Regarding individual domains, subjects with ASD consistently reported higher scores than all other groups, except compared to the SAD group for the Childhood/adolescence and Non-Verbal Communication domains, and to the OCD group for the Inflexibility and adherence to routine domain. Indeed, the SAD group reported high scores especially in the Childhood/adolescence, Verbal Communication and Empathy domains, while the OCD group in the Inflexibility and adherence to routine and Restricted interests and rumination domains. The PD group presented low scores overall, always inferior to the ASD group, and to the SAD and OCD groups in most domains. Furthermore, distinguishing the total sample on the basis of the total AdAS score into the three groups ASD (full-blown disorder), AT (significant autistic traits) and Non-AT (absence of significant autistic traits) it was observed that individuals with ASD were distributed predominantly in the ASD group (79.1%), individuals with SAD (53.6%) and OCD (52.2%) in the AT category, and healthy controls in the Non-AT group (100%). Linear regression analysis showed that high total scores at SHY-SV (B = 0.623, p<.01), OBS-SV (B = 0.494, p<.01), and PAS-SV (B = 0.333, p<.01) would be predictive of an high AdAS total score. Multinomial logistic regression analysis revealed that high total AdAS Spectrum scores would be predictive of a diagnosis of ASD (B = 0.433, p<.01), SAD (B = 0.372, p<.01), OCD (B = 0.361, p<.01) and PD (B = 0.293, p<.01) with respect to the HC group. In addiction, comparing SHY-SV, OBS-SV and PAS-SV total and domain scores between the diagnostic groups, we found that the higher scores were always obtained by ASD group, second only to the group with a diagnosis matching the spectrum under investigation.

Discussion: The current study revealed the presence of a continuous distribution of autistic traits within the total sample, highest in ASD subjects, immediately followed by patients with SAD and OCD, and descending toward the PD group, until reaching minimum levels in healthy controls. In each diagnosis, moreover, it was possible to identify a specific profile of autistic traits, based on the most represented domains, capable of differentiating between the different diagnoses. In addition, social-phobic, obsessive-compulsive and panic-agoraphobic spectrum symptoms were always found to be higher in ASD subjects, second only to the group with a diagnosis matching the spectrum under investigation. This would seem to support the hypothesis of a neurodevelopmental basis, identifiable in the autism spectrum, underlying the different diagnoses considered in the study. Furthermore, the predictive role of social-phobic, obsessive-compulsive, and panic-agoraphobic spectrum symptoms with respect to the presence of autistic traits, as well as the predictive role of autistic traits toward the presence of SAD, OCD, and PD, suggests the importance of investigating a potential autism spectrum hidden beneath such diagnoses - promoted by knowledge of disorder-specific profiles of autistic manifestations as revealed by the present study - and paying attention to the occurrence of these mental disorders in patients with ASD. Such cautions, in a clinical setting, could represent a valuable element in terms of prevention and individualized treatment.

Conclusion: Autistic traits appear to be distributed along a descending gradient from the diagnosis of ASD to healthy controls, passing through social anxiety and obsessive-compulsive disorders characterized by high autistic traits, and panic disorder defined by low levels of autism spectrum manifestations. Each of these diagnoses is also characterized by a specific profile of autistic symptoms. High levels of social-phobic, obsessive-compulsive, and panic-agoraphobic spectrum symptoms were found to be predictive of an elevated AdAS total score, which in turn was predictive of a diagnosis of ASD, SAD, OCD, and eventually PD, compared with healthy controls. Social-phobic, obsessive-compulsive and panic-agoraphobic spectrum symptoms were always found to be higher in ASD subjects, second only to the group with a diagnosis matching the spectrum under investigation.