Background
The Social Anxiety Disorder (SAD) was firstly described by Pierre Janet [
1] in the beginning of 20th century under the name of Social Phobia (SP). First grouped among specific phobias [
2], SAD gained diagnostic dignity in the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) [
3], and only in the fourth the name was changed in its current one, which better highlights the impairment and pervasiveness of the condition, firmly differentiating it from specific phobias [
4,
5]. SAD in characterized by persistent fear of one or more social situations or performances in which the person is exposed to non-familiar people or to a potential judgment by others that, as a result, lead the individual to avoid the feared situation or face it with excruciating anxiety or distress [
6]. Despite wanting to be with others, subjects with SAD tend to refrain from social situations and avoid expressing their opinions out of concern that they would be viewed as unreliable or stupid [
7]. The lifetime prevalence estimated in the general population varies from 1.9 to 13.7% [
8,
9]. Typically, SAD symptoms begin early in life, even in childhood [
7], persisting through the entire school career [
10] and leading to negative effects on the academic performances such as school interruption, increased possibility of failing exams [
11] and lack of graduation [
12,
13]. SAD frequently coexists with other mental disorders, in particular with mood disorders such as major depression (MDD), dysthymia, or bipolar disorder (BD) [
14,
15], as well as with other anxiety disorders like obsessive compulsive disorder (OCD) [
16], generalized anxiety disorder (GAD) [
17], panic disorder (PD) [
18], body dysmorphic disorder (BDD) [
19], or even substance abuse. In particular, many studies highlighted the high comorbidity between social anxiety and alcohol/cannabis use disorder [
20,
21] and even how subjects with subclinical traits of social anxiety (SA) have a greater risk of incur in an alcohol or cannabis use disorder than the non-clinical population [
22,
23]. Moreover, people with SAD may be more susceptible to problematic substance use in order to avoid being negatively judged by others [
22], to come with internal distress or to conform and fit in with their peers [
24‐
26].
Although being quite neglected in clinical settings, to the point of gaining the label of ‘‘the neglected anxiety disorder’’ [
27], SAD is a rather frequent and impairing condition that in the past years raised interest in many researchers [
28,
29]. Noticeably, SAD has been reported to be frequently under-recognized, due to the same nature of the disorder, which increases the tendency to avoid contacts with other subjects, including clinicians, but also due to socio-cultural factors and prejudices about the acceptability of shyness, especially among women [
30,
31].
Since its conceptualization, one of the main concerns in the field of SAD was the definition of a diagnostic threshold [
32]. Many researchers suggested that SAD may be more properly classified as a spectrum of severity rather than a discrete disorder based on subjectively determined threshold [
33] and that boundaries of SAD should be determined by its severity rather than by qualitative characteristics [
4,
34,
35]. In line with this view, the latest editions of the DSM [
6,
36] apported some changes in the chapter dedicated to SAD, reflecting a new and greater understanding of such condition in various social situations. In particular, while in previous descriptions of the disorder the presence of acute discomfort or dread when performing in front of other people used to be the primary criterion for the diagnosis of SAD [
37], the new editions of the manual removed the distinction between the generalized and specific forms, increasing the range of circumstances in which a person may have social anxiety symptoms and removed the requirement of being aged over 18 years and of recognizing the excessiveness and unreasonableness of their discomfort, prompting a re-consideration of symptoms distribution in non-clinical populations [
28]. A specifier was instead added for the subtype “performance only”, which should be used when SAD symptoms only arise when the subjects have to speak or do other performance in front of an audience. Interestingly, a major change revolved around the objectivity of the disproportionality of anxiety symptoms: this therefore allows including in the clinical evaluation even individuals who judge their reaction to be normal in certain situations due to their belief of having a “constitutional shyness” or due to the lack of complete awareness of their symptoms. Following this conceptualization, several studies hypothesized that SAD would be better categorized as dimensional continuum [
33].
According to such literature, the wide sub-threshold manifestations that may coexist with the major mental disorder can be more easily identified using a spectrum model of psychopathology [
38]. In this context, the term “spectrum” in used to describe mental health conditions that cover a range of symptoms and behavioral traits connected to a recognized DSM or ICD illness construct (like depression, panic or obsessive compulsive disorders) [
39]. While the primary symptoms of the current DSM diagnostic categories are included in the spectrum of symptoms and traits, the spectrum conceptualization also includes sub-clinical and atypical manifestations, in addition to temperamental and/or personality traits and isolated signs and symptoms, symptom clusters, and behavioral patterns [
32,
39‐
46]. In this view, the spectrum symptomatology can be compared to the part of an iceberg that is hidden under water surface, whereas the full-blown diagnostic criteria symptoms represent the visible portion [
39].
According to this model, in the early 2000’s, the “Social Phobia Spectrum Self-report” (SHY-SR) instrument was developed and validated in the context of the “spectrum project”, an international collaboration with the purpose of share light about the validity of a dimensional approach to psychopathology [
32,
38‐
46].
The SHY-SR aims to assess not only the prototypic symptoms of SAD but also atypical manifestations, temperamental traits, and other noteworthy clinical and sub-clinical aspects linked to the main symptoms [
41,
42]. The questionnaire demonstrated high internal consistency, and a good inter-rater reliability along with good discriminant validity. During the last decades, it was used in different clinical settings [
30,
31,
47,
48]. However, due to the extended time needed to complete it – up to 60 min – its implementation in regular clinical practice has remained quite challenging. Additionally, the instrument, being tailored on DSM-IV TR criteria, still included outdated and unnecessary components.
As the main authors of the SHY-SR, we aimed to develop a new revised and shortened version of the questionnaire, the Social Anxiety Spectrum– Short Version (SHY-SV) which should report a shorter compilation time as well as higher internal consistency, inter-rater reliability, and discriminant validity. The new instrument was developed including more contemporary items and excluding older ones, in order to be a more useful and update tool for clinical practice and research on both the full-blown and milder subsyndromal form of SAD. In this framework, the present work aimed to validate the SHY-SV questionnaire in a clinical population of patients with SAD patients, OCD patients and in healthy controls (HC). In particular, considering the reported presence of sub-threshold SAD traits in subjects with OCD and the overlapping features between SAD and OCD spectra, the OCD group was recruited as a potential intermediate group for SAD traits between SAD patients and HC [
48].
Discussion
The purpose of this work was to present and measure the validity and reliability of the SHY-SV, a clinical instrument prompted by a dimensional approach to psychopathology, in light of the spectrum model [
32,
40‐
46]. The SHY-SV assesses the core symptoms of SAD as well as the atypical manifestations, the temperamental traits and other remarkable clinical aspects associated with the central symptoms. The results of the study provided strong evidence of the validity and reliability of the SHY-SV, which was administered to a sample of subjects with a clinical diagnosis of SAD, OCD and individuals without current or lifetime mental disorders. We found excellent internal consistency and test-retest reliability and a significant and positive convergent validity with the alternative dimensional measures of SAD.
The questionnaire performed differently in each of the three groups studied, and the SHY-SV scores increased gradually from HC to OCD subjects up to SAD patients. The SHY-SV scores showed significant and strong positive correlations with the LSAS, one of the most popular tools used today to evaluate SAD symptoms and features [
56‐
58]. Moreover, it is noteworthy to mention that the SHY-SV questionnaire appeared to be an instrument capable of identifying even subthreshold SAD traits in the OCD group and the non-clinical population, showing an increasing gradient of social anxiety traits from the HC, passing through OCD subjects up to the SAD group (Fig.
1). The presence of social anxiety traits at intermediate levels in the OCD population is consistent with previous research in the field, which frequently noted multiple social anxiety symptoms among OCD patients as well as similarities between the two disorders, further supporting a spectrum model of psychopathology [
28,
30,
48,
59,
60]. Overall, our results support the capacity of the SHY-SV to accurately detect the whole spectrum of SAD, from the subthreshold manifestations to the full-blown clinical picture. However, some limitations concerning the study should be considered. The main limitation is the relatively small sample size, which might make our data less extensible. Furthermore, the SHY-SV, as the LSAS, are self-reported questionnaires, and subsequently may underestimate or overestimate symptoms based on the subjects’ perceptions, being less accurate than a clinician’s assessment. Interestingly, the substance abuse subscale shows several low correlations. Although the association of SAD with the use of alcohol and substances has been frequently reported in the literature, it is conceivable that, due to the nature of the SHY-SV a self-report questionnaire, has occurred a significant underestimation of the latter. This could also be explained by the fact that many subjects may not consider their use of substance as problematic. Moreover, even though in the test-retest group, no changes in drug therapy were made during the time between the first and second evaluation, specific information regarding the psychopharmacological therapy of the clinical subjects were not collected, excluding from the evaluation the possible inference of a psychopharmacological therapy. Lastly, the sample was not assessed with a measurement for depression nor for trait/state anxiety. In the context of those limitations, however, the SHY-SV demonstrated good psychometric properties and our results provide a coherent construct of the SHY-SV with strong internal consistency, high test-retest reliability and significative and positive convergent validity with alternative dimensional measures of SAD such as the LSAS. The SHY-SV has the advantage of being more time- and money-efficient and in line with the most recent descriptions of SAD when compared to the previous versions of the instrument and with face-to-face interviews [
6,
36]. In this context, it should be noted that, in addition to OCD subjects, SAD traits have also been linked to a wide range of psychiatric disorders, including neurodevelopmental disorders, mood disorders, eating and feeding disorders, and personality disorders, frequently worsening the clinical picture and impacting treatment outcomes [
14‐
19].
The availability of a tool that can identify sub-syndromic and atypical manifestations of this condition, which remain widely under-recognized, may improve diagnostic evaluation and treatment plans for the patients as well as support preventive and screening strategies in the general population. However, although the questionnaire demonstrated a good discriminating ability between the diagnostic categories and a good agreement with the diagnosis made by the clinician according to the DSM-5-TR criteria and through the SCID-5 diagnostic interview, the questionnaire alone is not sufficient for the diagnosis and should not be indicated as an alternative to the clinical interview, but rather as a supporting tool exploring the SA dimension in a dimensional way.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.