Background
Social phobia is one of the most common psychological disorders in children and adolescents [
1‐
3]. The disorder is characterized by a fear of being perceived as inadequate in social or achievement situations, resulting in considerable problems. Furthermore, social phobia in childhood and adolescence is a risk factor for the development of other psychological disorders [
4]. Although literature provides support for cognitive behavioral therapy (CBT) as an efficacious intervention for social phobia in children and adolescents [
5‐
7], more research is needed to improve treatments for children. Most of the initial investigations included children with various anxiety disorders.
Kendall [
8] developed the "Coping Cat program (Cat)" that contains education, modification of negative cognitions, exposure, social competence training, coping behavior and self-reinforcement. Different authors have used the program, making only slight changes [e.g. [
9,
10]]. Kendall [
8] reports significantly less general anxiety and improved coping behaviour as a result of the program, even in a follow-up after 3.5 years [
11].
"Cognitive-behavioral group therapy for social phobia in adolescents (CBGT-A)" [
12], is a specific group program. The first phase conveys information about social phobia, and implements cognitive restructuring and social skill training. The second phase includes in vivo exposure and applied routines. Studies have demonstrated improvements at post test [
13]. However, gains were not maintained at a 1-year follow-up [
14].
The group program "Social effectiveness therapy for children" (SET-C) [
15] puts its focus on exposure treatment, combined with social skills training and social interactions with non-anxious peers, but does so without cognitive interventions. Children and adolescents complete one introductory educational session with their parents, 1 group session, and 12 in-vivo exposure sessions over a 12 week period to help them improve their social skills. The SET-C group sessions provide instructions and practice, including activities where socially anxious participants interact with non-anxious peers. The individual in-vivo exposure component is designed to reduce anxiety in destressing social situations by making them more familiar. Concurrently, parents use positive reinforcement and shaping sequencing to effectively assist the progress of the SET-C program. Positive benefits have been achieved through use of this treatment protocol. Elements from the SET-C protocol were included in a school-based group behavioral treatment [
15‐
19]. In one of the longest follow-up assessment studies on youth, Garcia-Lopez et al. [
20] reported maintenance of treatment gains at the 5-year follow-up assessment. Masia et al. [
18] built on this new approach in their investigation of a 14-session group treatment in a school-setting which focuses primarily on education, realistic thinking, social skills training, exposure, and unstructured social situations to allow for practicing skills. In a pilot study of six children, three of them no longer met criteria for social phobia [
18]. Baer and Garland [
21] used a modified version of the SET-C program. The treatment involved twelve sessions. The authors concluded that a briefer version of group CBT was as effective as the more extensive research protocols.
Several reseachers posit that cognition plays an important role in the maintenance of social phobia [
22,
23]. In an attempt to increase the overall response rate for cognitive-behavioral treatment, Clark and Wells [
22] proposed a cognitive model of the maintenance of social phobia and used the model to develop a new cognitive therapy (CT) program for socially phobic adults. The four maintenance processes that are highlighted in the model are: (a) Increased self-focused attention; This means that in social situations, attention is shifted away from external social cues and instead is excessively self-focused. Connected with this is a linked decrease in observation of other people and their responses. (b) The use of misleading internal information (feelings and images) to make excessively negative inferences about how one appears to others. (c) Extensive use of overt and covert safety behaviors. Safety behaviors are strategies that are used to reduce anxiety or to hold off the social threat [
24]. Safety behaviors, however, are problematic because they contribute to the maintenance of fear. Anticipatory as well as post-event thoughts (i.e. thoughts prior to and after the social situation) contribute to the persistence of social phobia. It was shown that the inclusion of interventions targeting safety behavior leads to an increased effectiveness of CBT [
25]. (d) Problematic pre- and post-event processing [
26]. The therapy program has proved to be superior compared to treatment with SSRIs or placebo, even after 12 months [
26,
27]. Higher effect sizes have been found compared to previous meta-analyses of cognitive-behavioral therapy in socially phobic adults. This result indicates a significant increase of effectiveness [
26‐
28].
Very often, cognitive interventions are conceived as being inadequate for children due to their concrete thinking, time-limited perceptions and egocentric nature of thinking. It has, however, been suggested that children are quite capable of benefiting from cognitive interventions providing that educational and developmental features are considered. According to Ronen [
29] children can benefit from cognitive interventions provided that two conditions are met: (1) The therapist should be able to adapt the treatment to the child's personal cognitive style. Such adaptations include, for example, translations of abstract terms to concrete ones, utilization of simple words, use of demonstrations, metaphors, and illustrations taken from the child's own day-to-day life. (2) The treatment goals and procedures should be suited to the child's individual pace, as related to age and cognitive level.
Hodson et al. [
30] investigated the applicability of Clark and Wells' cognitive model to younger patients. High socially anxious children scored significantly higher than low socially children on all of the variables in Clark and Wells' model: negative social cognitions, self-focused attention, safety behaviours, and pre- and post-event processing. Findings suggest that Clark and Wells' model may be equally applicable to younger children with social phobia.
These findings have been confirmed by several studies [
31‐
34]. Results from a range of studies show that anxious children interpret ambiguous situations more often as being hostile [
35‐
37,
31]. Muris et al. [
38] showed a similar finding specifically with socially anxious children. Studies of attention control substantiate these findings: They confirm that the anxious child maintains a vigilant attention state for threatening cues [
39‐
41]. Bell-Dolan and Emery [
42] showed in a peer interaction task, that anxious children were as accurate as non-anxious children at identifying hostile intent in peer interactions, but they tended to misinterpret non-hostile situations as hostile. In a study by Johnson and Glass [
43] socially anxious children, in social or evaluation situations, also tended to focus their attention primarily on themselves, for instance, on their own physical reactions, instead of on the business at hand. Very few studies have examined the memory capacity of anxious children. In a study by Daleiden [
44] anxious children more often remembered negative information, so that a selective memory capacity was presumed to exist. In terms of anticipation of future events by socially anxious children, Spencer et al. [
45] found with 7- 14 year olds that, in comparison to children in the control group, the socially anxious children underestimated the probability of future positive social events. Controlled studies of cognitive treatment programs for socially phobic children are rare.
Therapy with children differs from therapy with youth and adults. First, very few children come to therapy on their own volition. They are brought to treatment, usually by their parents or caregivers. Second, unlike adult therapy, which involves the rational modification of thoughts, cognitive behavioral therapy for children focusing on cognition is more concerned with teaching appropriate skills and applying certain techniques.
The following study deals with the evaluation of a new cognitive behavioral treatment program for socially phobic children focusing on cognition according to the model of Clark & Wells [
22]. Although overlapping with other empirically validated CBT programs, CBT focusing on cognition has several distinctive features: (a) the development of Clark & Well's [
22] model by using the child's own thoughts, images, attentional strategies, safety behaviors, and symptoms, (b) experiential exercises in which self-focused attention and safety behaviors are systematically manipulated in order to demonstrate their adverse effects, (c) systematic training in externally focused attention, (d) techniques for restructuring distorted self-imagery, including a specialized way of using video feedback and (f) the structuring of planned confrontation with feared social situations as a behavioral experiment in which children test pre-specified negative predictions while dropping their habitual safety behaviors and focusing externally. A habituation rationale was not used [
26]. The aim of the present research was to examine the efficacy of this treatment program for socially phobic children with a focus on cognition. Our hypotheses include reduction of socially phobic symptoms and dysfunctional cognitions, improvements in anxiety coping, interaction frequency and comorbid symptoms.
Discussion
The objective of this therapy efficacy study was to determine whether socially phobic children in the treatment group differed from socially phobic children in the wait-list group at the end of a newly developed cognitive behavioral therapy program focusing on cognition. The innovation of the newly developed treatment consisted in the following: (a) using the child's own thoughts, images, attentional strategies, safety behaviors, and symptoms, (b) systematic manipulation of self-focused attention and safety behaviors, (c) systematic training in externally focused attention, (d) techniques for restructuring distorted self-imagery and (f) behavioral experiments in which a habituation rational was not used.
Three important conclusions can be drawn from the study:
1) The study provides preliminary evidence that the outcome of CBT focusing on cognition is better than the natural course of the condition. At post-assessment, children who received CBT treatment focusing on cognition compared to children in the wait-list group showed a significantly greater decrease of social phobia symptoms on the Social Phobia and Anxiety Inventory for Children (SPAIK). Significant improvement could also be seen on the severity ratings (DIPS-K). All children from the CBT treatment group showed a lower severity of social phobia compared to the waitlist group after the treatment. In addition, 30% of the children in the treatment group were free of diagnosis after treatment, whereas in the waitlist group all of the participants held their diagnosis. This suggests that the CBT treatment focusing on cognition was able to produce clinical improvement in our sample of socially phobic children. However, recent review articles have concluded that CBT packages result in around 56% of children being free of either the principal or any anxiety disorder after treatment [
64]. Therefore, reduction of anxiety diagnoses at posttreatment of our study was not within the range of those reported in CBT trials of children with different anxiety disorders.
2) Participation in our therapy decreased anxiety symptoms of social phobia and related symptoms such as negative feelings of self-worth. The results showed that the prevalence of comorbid symptoms like self-reported depression was not reduced as much as core symptoms by the treatment. However, we did not test whether symptoms of other anxiety disorders were also reduced. Further studies should examine whether the effect of the treatment was specific to the disorder of social phobia.
3) Decreased dysfunctional cognition as assessed by the SAKK suggests that the young children benefiting from our study were developmentally prepared to participate in a cognitive behavioral treatment focusing on cognition. Results from the Socially Anxious Cognitions Scale for Children (SAKK) with its Subscale of Negative Evaluation, Subscale of Positive Evaluation and Subscale of Coping Ideas, corroborate the overall results. Large effect sizes could be seen in this inventory (SAKK): g = 1.34 for Positive Self-Evaluation, g = 1.41 for Negative Self-evaluation and g = 0.89 for Coping Ideas.
Despite improvement in positive symptoms there was no improvement in K-GAS and behaviour diary ratings. There seems to be an inconsistency between positive symptom improvement but lack of functional improvement. However, changes of interaction may follow positive symptom improvement. The follow-up study will show whether such improvements may be observed.
Limitations
The study represents a first step to clarify whether CBT with a focus on cognition is an effective therapeutic approach in the treatment of socially phobic children. Further studies are necessary, however, to investigate whether the results can be replicated and whether the underlying theoretical model is adequate for socially phobic children. The significant results in the inventory assessing dysfunctional cognition show preliminary evidence, but have to be supported in further studies. Further studies are also needed to examine whether CBT focusing on cognition is superior or comparable to a general CBT approach and to examine which therapeutic approach is better suited to which patients.
One of the study's major limitations is that two advanced doctoral level graduate students conducted all screening interviews as well as the administration of the intervention. As the children should not be unduly burdened, assessment and intervention were thus carried out by the same person. Consequently, there is no independent assessment. Therefore, on the one hand, there is the risk that the children responded in ways to please the familiar interviewer. On the other hand, however, unfamiliar interviewers are likely to cause social anxiety. It follows that socially phobic children very often would indicate less social anxiety by avoiding to talk to interviewers who are unfamiliar to them. However, video recordings of all interviews were reviewed by an expert who was blind to the treatment condition.
Another major limitation concerns treatment design. Similar to many first trials of new CBT protocols for anxiety, we conducted this initial trial using a wait-list control condition. This approach provides preliminary evidence that the outcome of the proposed intervention is better than the natural course of the condition. It should be further evaluated against other interventions in subsequent trials.
Furthermore, the trial has not been registered.
Six patients dropped out of our study, four of whom participated in the treatment group. However, compared to drop-out rates in other studies, the rate of drop-out in the present treatment program is not noticeably high: According to Lincoln [
65] and Turner et al. [
66], only approximately 40% to 50% of the socially phobic adult patients seeking treatment actually completed and benefited from it in the end. There are further problems in the treatment of children, as not only the child must be motivated to participate in the treatment. According to the parents, therapies were discontinued for various reasons: quick initial successes, which seemed sufficiently high, time burden on the family, family misfortunes such as unemployment, parental separation or a parent's depression led to the premature termination of their child's therapy. Thus, it was not always the children who were most impaired who dropped out and did not receive treatment. It could be also possible that a 20-session intervention may be too intensive for some participants.
Considering a waiting period of many months, a selective dropout could have affected the configuration of the control group: Rejection could have been perceived before the beginning of the study as well as during the waiting period. However, the dropout rate does not confirm this argument, as there were only 2 dropouts in the control group compared to 4 drop-outs in the treatment group. Presumably, this relates to the very difficult state of care facilities that provide psychotherapy for children and adolescents.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SM, MK and JS carried out studies and drafted the manuscript. AW and US have made substantial contributions to conception and design. CS and FP have made substantial contribution to acquisition of data. All authors read and approved the final manuscript.