Background
Childhood maltreatment and peer victimization are known to be crucial risk factors for mental health [
1]. In social anxiety disorder (SAD), social learning experiences in childhood and adolescence are an important component of contemporary etiological models [
1‐
3]. Therefore, many researchers investigated the link between childhood maltreatment and SAD and repeatedly showed that exposure to such experiences in childhood are associated with SAD in adulthood [
4,
5]. However, comparing results across studies is difficult because a variety of assessments have been employed to assess childhood maltreatment.
A growing body of research investigated the association between history of childhood maltreatment and SAD by using the Childhood Trauma Questionnaire (CTQ) [
6], which assesses emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect. These studies linked specific maltreatment types to SAD symptom severity in individuals with SAD [
4,
5,
7,
8] and/or compared histories of child maltreatment between individuals with SAD and non-clinical control groups [
5,
9]. Findings showed that emotional abuse [
5,
9] and emotional neglect [
5] were more often reported by subjects suffering from SAD in comparison with healthy controls. Moreover, emotional abuse and emotional neglect have been significantly associated with greater SAD symptom severity [
4,
5,
7,
8], even in non-clinical samples [
10,
11]. One study [
5] further found that emotional abuse and neglect were linked to greater severity of depressive symptoms. Fewer studies found significant associations of physical neglect [
4] and sexual abuse [
7,
9] with SAD.
In addition to childhood maltreatment types, where perpetrators usually include parents and other adults, peer victimization is considered to be a further adverse social learning experience contributing to the development of SAD. In the interactional model of SAD, Spence and Rapee propose that exposure to peer victimization may increase the risk of developing SAD in individuals “who are intrinsically vulnerable “ ([
1], p., 8) (e.g., due to a genetic predisposition and/or inhibited temperament), through their impact upon behavioral and cognitive factors, including avoidance behavior or maladaptive schemes.
Therefore, a child exposed to peer victimization is more likely to experience social interactions as harmful, which may reinforce negative beliefs about themself and relationships with peers. This may lead to the avoidance of social interactions, and thereby increasing levels of social anxiety [
1,
12]. Indeed, studies have shown that peer victimization, including threats or acts of physical aggression (
overt victimization), relational manipulation or social exclusion (
relational victimization), and damaging a peers’ reputation (
reputational victimization) is associated with social anxiety symptoms in adolescence and adulthood [
13‐
15]. Growing evidence from prospective investigations implies that peer victimization puts children and adolescents at risk for developing social anxiety [
12,
14,
16,
17], with findings differing across types of victimization. Especially relational victimization, which is typically initiated by friends [
18] and comprises behavior such as excluding someone or withholding a relationship, is most strongly associated with social anxiety compared to overt or reputational victimization [
14]. However, cross-sectional studies further suggest that socially anxious youth are also more likely to become targets for peer victimization [
17,
19‐
21], so that peer victimization seems to constitute both, a predictor as well as a consequence of social anxiety [
12,
14,
17,
22].
Preliminary studies, which investigated peer victimization and childhood maltreatment simultaneously, report inconsistent findings. While cross-sectional data [
23] implies that emotional child maltreatment and peer victimization are independently linked to SAD symptom severity, longitudinal data [
24] suggest that emotional peer victimization, but not parental emotional abuse, increases social anxiety symptoms. In this longitudinal study [
24], emotional peer victimization was assessed with the items on relational victimization from the victimization scale of the Peer Reactions Questionnaire [
25].
In sum, the balance of evidence to date demonstrates that particular forms of recalled childhood adversities, namely emotional abuse, emotional neglect, and peer victimization may have more impact on SAD than other forms of childhood adversities. However, one of the key questions concerning effects of childhood adversities on SAD is the specificity of these effects. In other words, does having experienced emotional abuse, emotional neglect, or peer victimization increase the risk for SAD specifically, any anxiety disorder, any affective disorder, or any psychopathology? The above summarized studies on child maltreatment usually recruited participants for specific research projects, e.g. a brain imaging studies [
5] or intervention trials [
4,
7,
8], in which high internal validity is requested. Comorbidity was therefore allowed only on a limited basis. In some studies, depressive disorders were allowed [
4,
8,
9], in others, these were excluded [
5,
7]. Assessing comorbidity, however, may be highly relevant for the conclusions of such investigations.
For example, Rapee [
26] reviewed the evidence for specificity of sexual or physical abuse as a risk factor for anxiety disorders and concluded that “sexual abuse is shown to be a risk factor for a variety of forms of psychopathology “(p. 73). In fact, some of the effects of sexual abuse may rather unfold in patients with comorbidity of anxiety and affective disorders rather than in “pure” anxiety disorders only (p. 40) [
27]. Beyond that, exposure to childhood adversities is associated with a range of mental health problems in later life. For example, child maltreatment has shown to be associated with mood and anxiety disorders, substance abuse, psychotic symptoms, and personality disorders [
28]. Similarly, peer victimization does increase the risk for several dimensions of psychopathology, specifically internalizing problems [
29‐
31]. Only preliminary evidence suggests that social anxiety and not depression may be specifically linked to peer victimization [
13].
Taken together, key limitations in previous studies include a) the unclear specificity of the effects of childhood adversities on SAD. Most previous studies investigated associations with SAD symptom severity in individuals with SAD but did not examine whether these links are specific to a SAD diagnosis compared to other disorders. b) The neglected role of comorbid disorders in these effects, and c) limited generalizability of previous results to clinical treatment-seeking samples with high external validity (individuals not participating in a randomized controlled trial or recruited for a specific research project). Given the evidence that patients exposed to childhood adversities show poorer treatment outcomes [
32], research in representative treatment-seeking samples is key in informing practitioners and developing treatment interventions for patients commonly seen in out-patient clinics.
Therefore, the primary aim of the present study is to examine whether particular forms of recalled childhood adversities, namely emotional abuse, emotional neglect, and peer victimization are specifically associated with SAD in adulthood or whether we find similar links in other anxiety or depressive disorders by using a clinical sample who is seeking psychotherapy more in a routine care setting instead of a randomized controlled trial.
In order to investigate the specificity of links between recalled childhood adversities and SAD, we hypothesize that (1) recalled childhood emotional abuse, emotional neglect, and peer victimization are more likely to be associated with SAD than with any other mental disorder while controlling for comorbidity, (2) childhood adversity severities will differ across SAD, specific phobia (SP), and generalized anxiety disorder (GAD) without comorbidities, and (3) childhood adversity severities will differ between anxiety disorders and depressive disorders without comorbidities. The secondary aim of this study is to clarify the role of comorbid depressive disorders in the assumed effects of recalled childhood adversities. Therefore, (4) we expect that patients with SAD and comorbid depressive disorder will report more childhood adversities than patients with SAD only.
Discussion
The primary aim of the study was to investigate whether effects of different forms of recalled child maltreatment and peer victimization on SAD are specific to SAD or whether we find similar effects in other disorders as well. Four key findings emerged. Contrary to our expectations, none of the different child maltreatment types or peer victimization were found to be predictive for a SAD diagnosis in adulthood in an exclusively clinical sample. Thus, none of these childhood adversities seem to be more likely associated with SAD than with other disorders in the present sample. These findings seem inconsistent with previous findings, which may be partially explained by differences in the study design. Previous studies usually investigated the associations between childhood adversities and symptom severity in SAD samples [
4,
8] and/or compared SAD patients with healthy controls [
5], whereas we investigated links between childhood adversities and a categorical diagnosis, assessed with a clinical interview.
Secondly, neither any form of child maltreatment nor peer victimizations significantly differed among patients with SAD, SP, or GAD without comorbidities. Although previous studies repeatedly showed that at least childhood emotional abuse and emotional neglect seems to be strongly linked to SAD severity in adulthood [
4,
5,
7‐
9], these effects may not be specific to SAD, but rather apply for other anxiety disorders as well. Indeed, preliminary research indicates that child maltreatment as well as peer victimization are associated with an increased risk for any anxiety disorder, including SP and GAD [
15,
30,
41‐
43]. However, studies investigating links between other anxiety disorders than SAD or PTSD and child maltreatment assessed with the CTQ are scarce, limiting the comparability with our results.
The third key finding implies that effects may not only be non-specific to SAD, but broader non-specific to anxiety disorders. Comparing patients with anxiety disorders and depressive disorders without comorbidities showed that patients did not differ in severities of recalled emotional neglect and peer victimization. Patients with a depressive disorder reported significantly more severe emotional abuse, physical abuse, and sexual abuse. Our results support previous findings that child maltreatment constitute a risk factor for both, any anxiety disorder and any depressive disorder, although somewhat stronger association emerge for child maltreatment with depressive disorders than with anxiety disorders [
26,
44,
45]. Moreover, a recent study [
46] investigated which forms of child adversities are the best predictors for the development of lifetime major depressive disorder in women with or without depression (thus with other disorders or healthy ones). Especially emotional neglect and parental non-verbal emotional abuse assessed with the German interview version of the MACE (KERF-I) [
47] were the best predictors of lifelong depression. Taken together, our results are in line with the balance of evidence to date, implicating that consequences of particular maltreatment forms, i.e. the emotional forms and peer victimization are not specific to SAD, but expand to a wide range of other mental disorders, including internalizing problems, depression, risky health behavior, or eating disorders [
16,
26,
30,
48,
49].
In line with our hypotheses, the fourth key finding of our study showed that patients with SAD and a comorbid depressive disorder reported significantly more severe child maltreatment and peer victimization on all scales than patients with SAD only. Considering that several studies did not sufficiently control for comorbidities, this finding may contribute to the explanation of inconsistencies in the literature on effects of childhood adversities on SAD. We propose that the effects of emotional adversities on SAD stated in the literature are not specific to the disorder and moreover, may be better explained by the even stronger associations between childhood adversities and depressive disorders. Given the stated group differences in patients with anxiety and depressive disorders, these findings do not seem surprising. Indeed, emotional neglect has been shown to be associated with higher depression severity and lower self-esteem in patients with SAD [
5]. Moreover, preliminary findings showed that the association between sexual abuse and anxiety disorders only emerged in patients with a comorbid depressive disorder [
27]. Results on peer victimization are further in line with findings from Ranta and colleagues [
13], who found that among boys, social anxiety with comorbid depressive symptoms were more strongly linked to all forms of peer victimization than depression or social anxiety alone. Among girls, only the relational victimization was more frequent in the comorbid group than in the social anxiety or depression only groups. Other findings suggest that a relationship between peer victimization and anxiety is not attributable to a diagnostic overlap between anxiety and depression [
15]. However, measures to assess peer victimization vary immensely across studies and as yet there is insufficient evidence to draw valid implications on how peer victimization may contribute to the development of SAD beyond transdiagnostic effects.
Given the high comorbidities between SAD and depressive disorders, more research is warranted to investigate whether SAD is a cause or effect of comorbid disorders, or as previously discussed, “whether these patterns of comorbidity reflect common underlying causal factors” (p. 51) [
1]. Epkins and Heckler [
50] described that several models of both disorders in youth incorporate family-related problems as well as dysfunctional relationships with peers or adults, including social isolation, rejection, or criticism, associated with low self-esteem, social withdrawal, loneliness, and difficulties in interpersonal problem-solving. Therefore, both SAD and depressive disorders comprise interpersonal processes in the development as well as interpersonal consequences, indicating overlapping constructs in models of both disorders. Initial theories [
51] expected that adverse emotional events related to loss were more specific for depression, while direct threatening events such as physical or sexual abuse were more related to anxiety [
52]. However, we suppose that specific pathways from childhood adversities to SAD or depression may not be specific to the disorder, but rather lead to specific symptoms in specifically predisposed individuals. For example, in the frame of schema-based cognitive model of depression and anxiety [
53], it has been shown that emotional neglect was strongly associated with two out of three symptom dimensions (general distress and anhedonic depression), whereas sexual abuse was associated with also general distress and anxious arousal independent of an anxiety or depressive diagnosis [
54]. Further nonclinical findings suggest that emotional adversities are more strongly linked to internally-focused symptoms, while sexual and physical abuse are more associated with externally focused symptoms [
55].
Teicher and Samson [
56] point at the difference between survivors of early maltreatment and other individuals with the same mental disorder: they characterize disorders in maltreated individuals to be of greater severity, with more comorbidity, and less favorable treatment response. In fact, these authors are suggesting a “critically distinct subtype across depressive, anxiety, and substance use disorders” (p. 1114) for individuals with childhood maltreatment, defined as an “ecophenotype”. Therefore, an interesting approach for future research may be to investigate the effects of childhood adversities on a symptom level rather than a diagnostic level. A network perspective of psychopathology [
57], by conceptualizing disorders as casual networks of mutually reinforcing symptoms, may be a promising approach to investigate how childhood adversities and their potentially intercorrelated pathways affect transdiagnostic symptoms, to identify clusters of potential ecophenotypes, and to shed light on how these risk factors may contribute to the development of SAD and other disorders.
Strengths and limitations
To our knowledge this is the first study that made efforts to compare the recalled childhood adversities for patients diagnosed with anxiety and depression with or without comorbidities according to the gold standard in routine clinical care. In contrast to most previous studies, generalizability of our results to treatment-seeking samples can be assumed. Our investigations in routine clinical care provide valuable and representative information on prevalence rates and impacts of childhood adversities for treatment-seeking patients in Germany. Therefore, this information can guide practitioners in understanding the role of childhood adversities in the development of mental disorders.
Given that our sample was not recruited for a specific treatment study, anxiety patients without comorbidities were rare in this sample. However, this is truly characteristic of the out-patient population in Germany. Although comparing “pure” diagnostic groups may be considered a strength of our study, the cell number reductions due to anxiety patients with comorbidities were detrimental, limiting our statistical validity. Therefore, the null results may be attributable to low statistical power arising from small sample sizes. The small sample size of SAD only patients (
n = 25) may further challenge if this group is truly representative of this subpopulation. While child maltreatment severities and rates in SAD patients with comorbidities were comparable to other SAD+/− samples [
4,
7] and a depressive sample [
35], SAD only patients reported even lower severities of physical abuse and sexual abuse than a representative sample of the German population [
40]. However, treatment-seeking samples in routine care settings come with these kinds of data limitations. To our knowledge, maltreatment rates (assessed with the CTQ) in SAD only groups from other studies are lacking. Our result may encourage future research to fill this gap in the literature and further investigate effects in other anxiety disorders besides SAD and PTSD, with larger and pure diagnostic groups.
Although most previous studies investigated associations with SAD symptom severity in patients with SAD, we did not use symptom severity ratings in our analysis. Due to our representative treatment-seeking sample with different mental disorders, we did not assess SAD symptom severity in every patient. Our study further focused solely on the comorbidity with depressive disorders. It remains unclear, how other comorbidities may affect associations with childhood adversities. Preliminary studies showed that patients with SAD and comorbid ADHD report more emotional abuse and emotional neglect than patients with SAD only [
58]. Meanwhile, no difference emerged in patients with SAD with and without schizophrenia [
9]. Future studies may pay more attention to effects on transdiagnostic symptoms by systematically controlling for several comorbidities.
Finally, causal inferences about the links between childhood adversities and SAD are limited by the cross-sectional nature of our study. Thus, childhood adversities have been retrospectively assessed, which is prone for a recall bias and social desirability effects. Indeed, 22% of patients endorsed at least one minimization/denial item on the CTQ and our missing data analyses revealed that patients with missing values on the emotional abuse, physical abuse, and sexual abuse scales reported significant higher maltreatment severities on other maltreatment scales compared to patients without missing values. Therefore, maltreatment rates and severities reported in our sample are likely to be underestimated. Given that preliminary evidence indicates that individuals with depression show more negative bias in memory and attention compared to individuals with SAD [
59], depressed patients in our study may have recalled more adverse childhood memories and thus, reported more severe maltreatment. The retrospective assessment further prevents implications on interplays among adversities across the life span. For instance, maltreatment by parents in childhood seems to be associated with a higher risk for later peer victimization [
60,
61]. Longitudinal studies assessing the detailed temporal order of adversities are needed to identify potential mediating mechanisms across adversities to explain potential early pathways leading to SAD in later life.
Conclusion
Although links between forms of childhood adversities and SAD in adulthood have been established, our findings indicate that these effects are not specific to SAD, but rather apply to other anxiety and depressive disorders as well. Moreover, our findings implicate that most recalled childhood adversities are stronger associated with depressive disorders than with anxiety disorders. Finally, we conclude that some effects of specific childhood adversity types on SAD stated in the literature may be better explained by comorbid clinical or non-clinical depressive symptoms or may be solely based on severity of anxiety symptoms than on disorder type.
Taken together our findings support that recalled childhood adversities, including emotional abuse, emotional neglect, and peer victimization constitute transdiagnostic risk factors for a range of mental disorders in adulthood. In order to explain this multifinality, future studies may investigate effects of particular forms of childhood adversities on transdiagnostic outcomes in individuals with specific vulnerabilities, instead of continuing to examine links with single disorders. Identifying potentially moderating individual differences and vulnerabilities, as well as mediating psychological and interpersonal mechanisms, would be of particular value in informing the development of early interventions for SAD, but also other mental disorders by targeting transdiagnostic constructs.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.