Author's Accepted Manuscript

Integrating video-feedback and cognitive preparation, social skills training and behavioural activation in a cognitive behavioural therapy in the treatment of childhood anxiety Cecilia A. Essau, Beatriz Olaya, Satoko Sasagawa, Jayshree Pithia, Diane Bray, Thomas H. Ollendick www.elsevier.com/locate/jad

PII: DOI: Reference:

S0165-0327(14)00351-6 http://dx.doi.org/10.1016/j.jad.2014.05.056 JAD6801

To appear in:

Journal of Affective Disorders

Received date: 1 September 2013 Revised date: 25 May 2014 Accepted date: 27 May 2014 Cite this article as: Cecilia A. Essau, Beatriz Olaya, Satoko Sasagawa, Jayshree Pithia, Diane Bray, Thomas H. Ollendick, Integrating video-feedback and cognitive preparation, social skills training and behavioural activation in a cognitive behavioural therapy in the treatment of childhood anxiety, Journal of Affective Disorders, http://dx.doi.org/10.1016/j.jad.2014.05.056 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Integrating video-feedback and cognitive preparation, social skills training and behavioural activation in a cognitive behavioural therapy in the treatment of childhood anxiety

Cecilia A. Essau1, Beatriz Olaya2,3, Satoko Sasagawa4, Jayshree Pithia1, Diane Bray1, & Thomas H. Ollendick5

1

2

Department of Psychology, Roehampton University, UK Unitat de Recerca i Desenvolupament, Sant Joan de Déu-SSM, Barcelona, Spain

3

Parc Sanitari Sant Joan de Déu, Sant Boi de Llobregat, Barcelona, Spain

4

Faculty of Human Sciences, Mejiro University, Japan

5

Child Study Center, Department of Psychology, Virginia Tech, Blacksburg, USA

Correspondence should be directed to: Cecilia A. Essau, Ph.D. Department of Psychology, Roehampton University, Whitelands College, Holybourne Avenue, London SW15 4JD, UK e-mail: [email protected] Tel: +44 (0) 20 8392 3647 ; Fax: +44 (0) 20 8392 352

2

Abstract Background: The present study examined the effectiveness of a transdiagnostic prevention programme, Super Skills for Life (SSL), in children with anxiety problems. SSL is based on the principles of cognitive-behaviour therapy (CBT), behavioural activation, social skills training, and uses video-feedback and cognitive preparation as part of the treatment. Methods: Participants were 61 primary school children, aged 8 to 10 years, who were referred by their teachers as having significant anxiety problems. Children were videorecorded during a 2-minute speech task in sessions 1 and 8, and during a social interaction task. All the children completed measures of anxiety symptoms, social skills, and self-esteem before and after participating in the 8-week SSL and at the 6-months follow-up assessment. Results: Anxiety symptoms were significantly reduced at post-test and follow-up assessments. SSL also had a positive effect on hyperactivity, conduct, and peer problems although it took longer for these effects to occur. Behavioral indicators of anxiety during the 2-minute speech task decreased, indicating that the independent raters observed behavioral change in the children from pre-treatment to follow-up. Boys had higher overall behavioral anxiety during the 2-minute speech task at all three assessment periods, specifically showing higher lip contortions and leg movement than girls. Limitations: The present study used an open clinical trial design, had small sample size, and did not use structured diagnostic interview schedules to assess anxiety disorders. Conclusions: This study provides preliminary empirical support for the effectiveness of SSL in children with anxiety problems.

Keywords: Anxiety symptoms, Children, Prevention, Super Skills for Life, Social skill

3

1.

Introduction Approximately 15–31.9% of children and adolescents in the general population are

affected by anxiety disorders (Essau et al., 2000; Lewinsohn et al., 1993; Merikangas et al., 2010). Anxiety disorders cause significant distress and impairment in major life domains, including educational underachievement (Essau et al., 2000; Feehan et al., 1993; Woodward & Fergusson, 2001). If left untreated, anxiety disorders that begin early in life tend to have a negative course and may serve as a risk factor for several other psychiatric disorders in adulthood (Bernstein et al., 1996; Essau et al., in press; Kessler et al., 1994). Cognitive behaviour therapy (CBT) is the treatment of choice for childhood anxiety, with 50 to 70% of children with an anxiety disorder responding positively to CBT (Barrett et al., 1996, 2001; Essau et al., 2012; Kendall et al., 1997; Seligman & Ollendick, 2011; Stallard et al., 2007). These moderate remission rates might be attributed to the fact that these studies have been based on interventions that are specifically designed for anxiety disorders despite strong evidence of the high comorbidity between anxiety and other internalizing (e.g., depression) (Essau, 2008) and externalizing disorders (e.g., conduct disorder). In fact, the presence of depressive disorder or depressive symptoms among children and adolescents who received treatment for their anxiety disorders had a poorer outcome than those who did not have comorbid depressive symptoms or disorders (Brent et al., 1998; O’Neil & Kendall, 2012). Furthermore, there has been discussion whether younger children could understand the basic concepts of CBT (e.g., the association between thoughts, feelings and behaviour) due to their cognitive development (Essau et al., 2004). For example, children have been reported to have difficulty in understanding a major component of CBT (i.e., cognitive restructuring), resulting in low compliance rate in completing their homework in this area (Essau et al., 2004).

4

Due to limitations of CBT (e.g., requiring patients to learn high-level skills) (Rhodes et al., 2014), a recent trend has been the implementation of behavioural activation in the treatment of anxiety disorders (e.g., Chen et al., 2013) as its rationale is easier to understand and operationalise for both patients and mental health professionals (Rhodes et al., 2014). Behavioural activation was originally designed to alleviate depression by targeting avoidance of normal activities which preclude individuals from experiencing positive reinforcement opportunities from their environment (Lewinsohn et al., 1985). By targeting avoidance, behavioural activation helps to disrupt the cycle of avoidance, decrease activity, and anxious/low mood through repeated exposure to goal oriented behaviour (Rhodes et al., 2014). The efficacy of behavioural activation in the treatment of anxiety disorders has recently been reported in adults (Chen et al., 2013), but not in children. Children with high anxiety have been reported to have poor social skills (Coplan et al., 2004; Ginsburg et al., 1998; Schneider, 2009; Spence et al., 1999) and low quality friendships which in turn put them at risk for victimization (La Greca & Lopez, 1998; Muris et al., 2001). These findings emphasise the importance of teaching children the skills that help them experience successful outcomes from social interactions, including non-verbal responses (e.g., appropriate use of eye contact and facial expression), and verbal skills (e.g., tone, rate and volume of speech) (Spence, 2003). Studies among adults with anxiety disorders, particularly those with social anxiety, tend to show that these adults have negative images of themselves, especially when they observe themselves in social interactions (Clark & Wells, 1995; Rapee & Heimberg, 1997). Since negative self-processing diverts individual’s attention from external social information which could help change their negative beliefs, Rapee and Hayman (1996) have suggested using cognitive preparation prior to viewing themselves as a method to modify such beliefs. In line with this suggestion, studies among adults have shown that cognitive preparation before

5

viewing videos of the self have proven effective in correcting the distorted self-perception and in reducing anxiety (Harvey et al., 2000; Kim et al., 2002; Rodebaugh, 2004). Although video-feedback with cognitive preparation (Clark & Wells, 1995; Harvey et al., 2000) is effective in the treatment of anxiety disorders in adults, this strategy has not been implemented heretofore with children. Studies have also reported that anxiety disorders tend to run in a family. According to a recent study by Schreier and Heinrichs (2010) the mechanism by which anxiety may be transmitted from mothers to children was maternal report of negative child evaluation (FNCE). It was argued that parents not only extend their own interpretational bias towards their child but also extend their fear of negative evaluation to a more general fear of negative evaluation such as fear of negative evaluation related to their child. It is however unknown if parental FNCE mediated the association between child anxiety and treatment outcome. Drawing from the above research, a transdiagnostic treatment procotocol (called “Super Skills for Life”; SSL, Essau & Ollendick, 2013) was developed for children with internalizing problems (i.e., both anxiety and depresssion). SSL has five core principles: (1) it is based on a transdiagnostic approach by targeting common core risk factors (e.g., low selfesteem, lack of social skills) of comorbid disorders, and as such it should be more time efficient and cost-effective (Rohde, 2012); (2) it is based on the principles of CBT to help children develop skills to cope with anxiety-provoking situations; (3) it uses video feedback with cognitive preparation to help children enhance their self-perception (Harvey et al., 2000; Rodebaugh, 2004). Children are exposed to social situation by asking them to give three 2minute speeches in front of the whole group facing a video camera during the intervention period (first and last sessions) and at follow-up session, as well as being asked to role play a social interaction in front of the class and being video-taped. Furthermore, two sessions of the SSL involve teaching children skills to enhance their behaviour in social situation such as

6

“when to start conversations”, “when to join a group conversation”, techniques to solve any social problems, and to role play these skills during the sessions, and to practise them as part of their homework; (4) it uses the principle of behavioural activation by having children increase their activity levels and participate in positive and rewarding activities, which in turn can help to improve their mood and overall self-esteem; (5) finally, it teaches children basic skills to use during social interactions to help increase their experience of successful outcomes from the interactions. The aim of the present study was to evaluate the effectiveness of SSL. Its specific aims were to examine: (1) the effects of the SSL on children’s anxiety and depressive symptoms, social skills, and self-esteem; and (2) the mediating factors that predict the treatment outcome. The hypotheses to be tested were: (a) there will be a decrease in the principal problem (i.e., anxiety) and a decrease in the symptoms of other disorders (e.g., depression, conduct problems) at post-test and at follow-up as compared to pre-treatment levels. (b) parental fear of negative child evaluation (as reported by the child) and self-esteem will mediate the treatment outcome.

2.

Method

2.1.

Participants A total of 61 children, aged between 8 and 10 years (M = 8.84 years and SD = 0.73)

participated. Of these, 43 were boys (70.49 %) and 18 were girls (29.51%). The participants were from low socio-economic status and from diverse cultural backgrounds (38.33% were White; 23.33% were Black; 16.67% were mixed; 15.00% were Asian; 6.67% were Middle Eastern). All the children were selected by their teachers from four primary schools in North and South-West London; these children were known by their teachers as showing significant

7

anxiety problems. Although the teachers did not use any objective measures of anxiety severity to make their referral, anxiety of the study cohort is as severe as those in numerous RCT cohorts (e.g., Stallard et al., 2007) and that the anxiety scores were in the clinical range as measured using the SCAS. Other inclusion criteria were that the children (a) were in years 5 and 6 in school; (b) no signs of self-harm were evident; and (c) the children were not currently undergoing psychological or pharmacological treatment for their anxiety problems.

2.2.

Implementation of the SSL The delivery of the SSL was conducted by facilitators, all with a Masters degree in

Psychology (major in Child Psychopathology). Before implementing the SSL, the facilitators received an intensive one-day workshop by the senior author (CAE). The workshop covered topics related to anxiety disorders and their risk factors, principles of prevention, organization and ethical issues in running the program with a group of children, as well as group leader and group process skills. Each session of the SSL was reviewed through dialogue, role play, and exercises. All the facilitators were given a leader’s manual which included a detailed outline of each session of the SSL. A weekly meeting was held by the senior author with the facilitators to discuss any problems in the delivery and understanding of exercises of the SSL. Ethical approval to conduct this research was approved by the Ethic Board at the University of Roehampton. All parents of these pupils were sent information that described the research project together with an informed consent form to be completed and returned by their parents. Children’s participation was voluntary; they were informed that their responses to the questionnaires would be kept confidential and they could withdraw from the research at anytime. The SSL was delivered in the afternoon (after class) at the pupils’ schools. Children participated in the 8 sessions of the SSL, with each session lasting for 45 minutes, once a

8

week. The range of group size was six to eight children, with an average size per group being 8 children. The programme included the delivery of the following skills: education about emotions and feelings, cognitive reappraisal, problem-solving, behaviour activation, relaxation techniques, self-monitoring, and social competence. These skills were taught through various formats, including small group and individual exercises, role plays, activities, and games. Homework was set at the end of each session, in which children were asked to practise the skills which they had been taught. Children who completed their homework were given a colourful sticker at the end of each session. In sessions 1 and 8, and at the follow-up session, children were asked to give a 2minute speech in front of the whole group and facing a video camera. For the 2-minute speech tasks that were conducted at the first and at the 6-month follow-up sessions, the children were asked to say anything they liked; for the 2-minute speech task at session 8 (i.e., post-treatment), children were asked to say something about what they had learned from participating in SSL and the skills they found to be most helpful. Before showing the children their video during the 2 minute speech tasks, they were instructed to pay attention to the way in which they appeared during the speech and not on how they felt. Children completed a set of questionnaires before and after the SSL programme, and at an average of 6 months after the intervention.

2.3.

Measures Social Skills Questionnaire (SSQ; Spence, 1998) was used to measure children’s

social skills. The total score was obtained by summing the 13 items, with higher scores indicating higher levels of social skills. The Cronbach alpha for the SSQ in this study was 0.82.

9

The child version of the Parental Fear of Negative Child Evaluation (PNCE; Schreier & Heinrichs, 2010), was used to measure parent’s (maternal and paternal) general worries and fears of negative child evaluation, as perceived by the child. It consists of 8 items which can be rated on a 4-point Likert-scale, ranging from 0 (never) to 4 (most of the time). For each item, children reported first about mothers (4 items) and then about their fathers (4 items). The total score was obtained by summing all the items, with higher scores indicating perceived parental high level of worries and fears of negative child evaluation. The Cronbach alpha for the mother and father subscales was 0.85 and 0.89, respectively. Rosenberg Self-Esteem Scale (Rosenberg, 1965) was used to measure the child’s selfesteem. The total score was obtained by summing the 10 items, with higher scores indicating high level of self-esteem. In the present study, the Cronbach alpha was 0.62. Child Anxiety Impact Scales (CAIS-C; Langley et al., 2004) was used to measure anxiety-related difficulties in school, social, and home/family. The CAIS-C contains 27 items, in this study, two items (i.e., “Going on a date”; “Having a boyfriend/girlfriend”) were excluded as they were considered inappropriate for the age group. The total score was obtained by summing the items, with higher scores indicating that the anxiety has an increasing negative impact on the children’s life. The Cronbach alpha was 0.94 for this study. Spence Children’s Anxiety Scale (SCAS; Spence, 1998) was used to measure symptoms of anxiety disorders: separation anxiety, social anxiety, obsessive-compulsive, panic disorder, specific fear and generalized anxiety disorder. Children respond to each item using a 4-point Likert-type scale ranging from 0 (never) to 4 (always). The total score was obtained by summing the 38 items, with higher scores indicating high anxiety symptoms. The Cronbach alpha was 0.96 in this study. Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997; Essau et al., 2012) was used to measure children’s general difficulties and positive attributes: emotional

10

symptoms (i.e., anxiety and depression), conduct problems, hyperactivity, peer problems and pro-social behaviour. To get the total of the difficulties score, all the subscales except for the pro-social behaviour subscale were added up; the higher the scores, the greater the difficulties. The Cronbach alpha of the SDQ in the present study was 0.80. The Performance Questionnaire (PQ-C; Cartwright-Hatton et al., 2003) was used by the children to rate their performance and how they felt during a speech. Three of the items were used to measure micro-behaviours that the child displayed during the speech task (e.g., How loud and clear was your voice? How much did you look at the camera? How much did you smile?). Two items were used to measure how nervous the child appeared during the task (How nervous did you look? Did you stumble over your words?), and the remaining three items were used to measure the global impression made by the child (How clever did you look? How friendly did you look? How good was your speech?). The items were scored on a four-point scale ranging from 1 (not very much) to 4 (very much). The Objective Performance Questionnaire (OPQ-C; Cartwright-Hatton et al., 2003) was used by an observer to rate the child’s performance during the child’s 2-minute speech task. This scale contained the same items as the PQ-C; the only difference was that the words ‘you’ was replaced by ‘the child’ (e.g., ‘How much did the child smile?). The items were scored on a four-point scale ranging from 1 (not very much) to 4 (very much). Behavioural Signs of Anxiety Scale was used to measure behavioural indicators of anxiety during the speech task. It contains 10 behavioural signs of anxiety (e.g., lip-licking; nail-biting). The observers rated (offline) the video of the 2-minute speech task indicating the frequency with which the children displayed that particular behaviour. Recording of each child during the 2-minute speech tasks at pre, post and follow-up were rated independently by two trained observers who independently rated all the videos on at each time point. Both raters, who held a Masters degree in Psychology, did not participate

11

in the delivery of the SSL and they were blind to the children’s anxiety scores on the SCAS. The mean score of the two raters were calculated for the analyses. The recording was divided into successive 30 seconds sample intervals in which the observers recorded whether or not each of the 10 behaviours was present. These same raters also observed the videos and rated each child on the OPQ-C.

2.4.

Statistical analyses Generalized Linear Models for Repeated measures was used to test the hypotheses

described above. Analyses included between (gender: boy was coded as 1, and girl was coded as 2) and within-group (time: pre, post and 6-months follow-up) effects. Interactions between gender*time were also tested. If the Mauchly's Test of Sphericity was significant, the univariate F statistic was reported with the epsilon correction (Box, 1954) based on the Greenhouse-Geisser estimation (Geisser & Greenhouse, 1958), which is more conservative. Estimated marginal means with their 95% confident interval were reported. Comparisons between the main effects using the estimated marginal means with Bonferroni post hoc tests were also used. At pre-treatment, missing data ranged from 0% to 11.48% for the impact of anxiety scale; at post-treatment, it ranged from 0% to 13.11% for the father negative evaluation, and at follow-up, missing data went from 8.20% (SDQ) to 16.39% (father negative evaluation, impact of anxiety, and total SCAS). Three participants had incomplete data for all of the outcome measures and were omitted from further analysis. The remaining 58 had T1, T2, and T3 data completed for at least one of the outcome measures. Missing data were excluded from analysis using pairwise deletion. The percentages of the demographic variables for the remaining participants were 70.69% boys and 29.31% girls; 36.84% White, 22.81% Black, 17.54% Mixed, 15.79% Asian,

12

and 7.02% Middle Eastern. No significant differences emerged between the participants who were omitted and those who remained. 3.

Results

3.1.

Treatment effect and gender differences For the SCAS, there was a significant within-subject effect of time for the following

subscales: separation anxiety (F (1.68, 87.23)=7.10, p=0.002), social phobia (F (1.46, 78.99)=6.55, p=0.006), generalized anxiety (F (1.80, 97.14)=4.43, p=0.018), and total SCAS score (F (1.61, 78.69)=5.58, p=0.009). For the SDQ, significant within-subject time effects were also observed for conduct problems (F (2, 106)=5.20, p=0.007), hyperactivity (F (2, 108)=5.33, p=0.006), peer problems (F (2, 108)=21.66, p=0.000), total SDQ score (F (2, 104)=3.76, p=0.026). Finally, significant effects were also found for father’s fear of child’s negative evaluation (F (2, 92)=4.34, p=0.016). Pairwise comparisons with Bonferroni tests indicated that overall scores decreased from pre-treatment to follow-up for separation anxiety disorder, social phobia, generalized anxiety, total SCAS score, peer problems, and father’s fear of child’s negative evaluation. Furthermore, there was an immediate decrease between pre- to post-treatment for separation anxiety disorder. Conduct problems and hyperactivity decreased from post-treatment to follow-up; these results suggest that changes in these variables took place after the treatment was terminated. On the contrary, children’s general difficulties as measured by the SDQ increased from pre-treatment to follow-up. However, since there was a significant interaction between gender and time, this has been followed through using simple main effect tests (see below). Significant between-subject effect of gender was obtained on conduct problems (F (1, 53)=9.22, p=0.004) and total social skills (F (1, 48)=6.51, p=0.014). Comparisons between marginal means estimations showed that boys reported greater conduct problems symptoms (mean: 3.92, SE=0.22) compared to girls (mean: 2.67, SE=0.35) (p=0.004), whereas girls

13

showed greater scores on social skills than boys (mean: 20.77, SE=1.01 versus 17.77, SE=0.60; p=0.014). There was a significant interaction between gender and time on emotional problems (F (2, 106)=3.22, p=0.044) and the total SDQ score (F (2, 104)=3.64, p=0.030). Analysis of marginal means estimations of emotional symptoms (i.e., anxiety and depression) by the three time levels showed that emotional symptoms showed a decrease between pre and follow-up in boys (p=0.007), but no significant differences were found for girls. As for total SDQ, the mean scores were similar in the three assessment period for boys, whereas in girls, the pre-treatment score was lower than boys (p=0.009) but at the post and follow-up assessment, these scores were similar to boys. We do not have a clear explanation as to why the gender difference disappeared in follow-up assessment.

3.2.

Behavioral changes in the speech task The children were evaluated behaviorally through the 2-minute speech task at pre-,

post-treatment, and 6-months follow-up (Table 2). There was a significant within-subject effect of time for the overall behavioural indicators of anxiety (F (1.57, 83.36)=7.05, p=0.003), nail-biting (F (1.12, 59.22)=5.16, p=0.023), lip-licking (F (1.40, 74.31)=7.50, p=0.003), mouth-touching (F (1.36, 72.02)=4.17, p=0.033), lip contortions (F (2, 106)=7.04, p=0.001), unnecessary hand movement to face (F (1.04, 55.23)=4.94, p=0.029), unnecessary hand movement to body (F (2, 106)=3.52, p=0.033), child-rated nervousness (F (1.70, 71.33)=11.25, p=0.001), child-rated global behavior ( F (2, 90)=9.43, p=0.001), child-rated micro-behaviour (F(1.39, 62.49)=8.08, p=0.003), observer-rated nervousness (F (2, 86)=41.73, p=0.000), observer-rated global behavior (F (1.52, 63.87)=5.58, p=0.011) and observer-rated micro-behaviour (F (2, 86)=10.32, p=0.001). Participants showed less behavioral anxiety at follow-up compared to pre-treatment for most of these outcome

14

variables. An exception to this trend was in unnecessary hand movement to body, which showed significant decrease between post-treatment and follow-up. Due to decreased statistical power, nail-biting, mouth-touching, and unnecessary hand movement to face reached only marginal significance in Bonferroni post-hoc analysis. Significant between-subject effect of gender was obtained for lip-biting (F (1, 53)=4.58, p=0.037), child-rated nervousness (F(1, 42)=6.22, p=0.017), and observer-rated nervousness (F (1, 43)=6.08, p=0.018). Comparisons of marginal means estimation showed that boys (mean=0.36, SE=0.07) bit their lips more than girls (0.08, SE=0.11) (p=0.037). As for nervousness, boys (child-rated mean: 8.61, SE=0.24; observer-rated mean: 7.84, SE=0.19) scored significantly higher than girls (child-rated mean: 7.39, SE=0.42; observer-rated mean: 7.00, SE=0.28) (p=0.017 and 0.018, respectively). Overall, the present results suggest that girls showed less anxious behavior than boys. There also was a significant interaction between gender and time in overall behavioral anxiety (F (1.573)=4.73, p=0.011), lip contortions (F(2)=4.63, p=0.012), and leg movement (F (2)=6.14, p=0.003). Further analysis of marginal means estimations for overall behavioral anxiety showed that boys (mean: 5.49, SE=0.50) scored higher than girls (mean: 1.13, SE=0.78) at follow-up (p=0.000). Similarly, at follow-up, boys (mean: 1.33, SE=0.21) scored higher than girls (mean: 0.19, SE=0.33) in the lip contortions (p=0.005) and leg movement (boys’ mean: 2.46, SE=0.25; girls’ mean: 0.75, SE=0.39; p=0.001). The interactions showed that girls had better behavioral scores at the time of the 6-month follow-up.

3.3.

Mediators of change Incremental scores of social skills, parental fear of negative child evaluation as

reported by the child, and self-esteem were examined as possible mediators of change between pre- and post-total anxiety scores and social phobia subscale. We evaluated this

15

using a program called RMediation (Tofighi & MacKinnon, 2011) that uses the distribution of the product of coefficients method (MacKinnon et al., 2007). In a conventional Sobel test (1982), the product of indirect effects is presumed to have a normal distribution. However, the product of two normally distributed random variables is known to be skewed. The distribution of the product of coefficients method is advantageous in that it calculates the 95% confidence interval for indirect effects, thereby providing a better index of significance. If the confidence interval does not include 0, one can conclude that the mediation effect is robust. 95% confidence intervals for the mediating effect of the Social Skills Questionnaire, Parental Fear of Negative Evaluation, and Rosenberg Self-Esteem Scale are shown in Table 3. Unexpectedly, none of these variables acted as mediators of change in pre- to post-total anxiety scores and social phobia subscale.

4.

Discussion The goal of the study was to examine the effectiveness of the SSL (Essau &

Ollendick, 2013). To our knowledge the SSL is the first CBT-based intervention that integrates behavioural activation, social skills training, and uses video-feedback and cognitive preparation as part of the treatment. The present study differed from previous intervention studies in two major ways: (1) children were referred by the school teachers, and (2) inclusion/exclusion criteria were kept to a minimum. In line with previous findings of studies that used CBT-based anxiety prevention programme such as the FRIENDS (Barrett & Turner, 2001; Lowry-Webster et al., 2003), anxiety symptoms were significantly reduced at post-test and follow-up assessments. Significant reductions were found for symptoms of GAD, social phobia and separation anxiety disorder. Furthermore, when using the cut-off points as proposed by Muris et al.

16

(2000), about half of the adolescents recovered after the treatment. That the children with social phobia benefit from the treatment were not surprising because of the amount of social exposure during SSL. Although SSL was designed to target anxiety and depressive problems, it also had positive effects on self-reported hyperactivity, conduct, and peer problems. However, unlike anxiety symptoms which had an immediate response following treatment, for these three problems it took somewhat longer for the children to experience the program’s positive impact (i.e., at follow-up). Previous studies have shown that children with hyperactivity, conduct and peer problems tend to have low self-esteem and social skills (Barry et al., 2003; Glass et al., 2011). It could be speculated that these children benefit from participating in the SSL as it contains activities that help to enhance children’s self-esteem and social skills, and thus in turn may have a positive impact on hyperactivity, conduct, and peer problems. Future studies are needed however, to formally test this hypothesis. Behavioral indicators of anxiety as coded using the video decreased, indicating that not only the children but independent raters noticed behavioral change in the children from pre-treatment to follow-up. There were no significant differences in the behavioral indicators of anxiety between pre- and post-intervention, indicating that it took a few months after the treatment for the effects of the SSL to have a meaningful (statistical) change. Specific changes in micro-behaviour included less lip-sucking and contortions. Here again, the treatment effect was the most profound at the follow-up period. This finding is in line with our previous study using the FRIENDS programme (Essau et al., 2012) in that reductions in anxiety were significantly greater for participants in the intervention compared to the control groups at the 12-month follow-up. Both the present and our previous studies (Essau et al., 2012) suggest that with more time to practice, children who participated in the intervention tend to get better throughout the follow-up period.

17

Our analyses of the 2-minutes task at the three assessment periods indicated that boys had higher overall behavioral anxiety, specifically showing higher lip contortions and leg movements than girls. This finding is interesting because no gender differences were found in children’s self-report of anxiety symptoms on the SCAS. The reason why gender difference was found on the behavioural indicators of anxiety, but not on the self-report of anxiety as measured by SCAS is unclear. Speculatively, girls and boys express their anxiety differently; more studies are needed to support this hypothesis. The children were also evaluated as less nervous at post-treatment and follow-up both by themselves and objective raters. What is curious is that the overall frequency of micro and global prosocial behaviours decreased from pre- to post-evaluation. This may have been due to the difference in the speech task. Whereas for pre-treatment, the children were allowed to select a topic of their own, the topic at post-treatment was limited to what they learned from the programme. It is also noteworthy that the children rated themselves better on the global ratings at follow-up than at post-treatment. Although speculative, it may be the case that the children’s self-images improved and the children were able to rate their performances higher. For example, in an early study by Rapee and Hayman (1996), adults who were socially anxious were capable of rating their own performance on a video recording that was consistent with that of independent observers. It was argued that the mechanisms responsible for this change may be due to modification of adult’s self-appraisals. The results show that social skills, child perceived parental fear of negative evaluation, and self-esteem did not act as mediators of change in pre- to post-total anxiety scores and in the social phobia subscale. Thus, the beneficial effects of the intervention were not because of better social skills or higher self-esteem, but other factors such as the use of the video, cognitive preparation and video feedback could have been in operation. As reported in several studies, individuals with anxiety disorders tend to underestimate their

18

performance during the speech task. Among adults, use of the video feedback approach has successfully helped to reduce individual’s underestimation of their performance (e.g., Harvey et al., 2000; Rapee & Hayman, 1996). As suggested by Harvey and colleagues (2000), the mechanisms that enabled cognitive preparation to enhance self-impressions include: First, asking people to clearly articulate how they think they appeared prior to viewing the video may activate their image of how they think they appeared during the speech. This will enable them to see the difference between how they appeared compared to how they felt they came across during the performance; this process is expected to disconfirm the belief that they have done poorly. Second, cognitive preparation may minimise reactivating the way they felt during the performance, which in turn may prevent these feelings from being superimposed on the video image and distorting it even further. Furthermore, Clark and Wells (1995) argued that the core maintaining factors in social anxiety are beliefs specific to social situations, self-focused attention and safety behaviours, rather than social skills or selfesteem.

5.

Limitations It is important to acknowledge the limitations of this study when interpreting its main

findings. First, the present study used an open clinical trial design. Thus, the potential problem related to the internal validity of the programme could be the passage of time and factors (e.g., being in a group of children) that are unrelated to the programme. However, given the treatment effects between pre-, post- and follow-up assessments, external influences are unlikely to explain the improvements and the maintenance during the followup period. Second, the sample size was small. Thus, more studies are needed with a larger sample size, using a RCT design. Third, the study did not use any structured diagnostic interviews because it was implemented in a school setting. Still, the study did use the SCAS

19

to examine children with anxiety problem because it has proven to be both valid and reliable in distinguishing children with and without any anxiety disorders (Stallard et al., 2007). Fourth, the school setting in this study may not be typical of a clinical setting. In order to have a diverse group of children, the inclusion and exclusion criteria were relaxed; despite this, the sample was similar to those reported in RCT studies in terms of the severity of the problems (i.e., moderate severity).

6.

Conclusion In conclusion, this study provides preliminary empirical support for the utility of the

SSL in reducing anxiety symptoms among primary school children.

Acknowledgement The authors are grateful to all those who participated in the study.

Conflict of Interest The authors have no conflict of interest to report in relation to the research presented in this manuscript. Contributors Cecilia Essau designed the study and wrote the first draft of the manuscript. Beatriz Olaya (self-report data) and Satoko Sasagawa (video data) undertook the statistical analyses. Jayshree Pithia collected and entered the data in SPSS format, and prepared the data for statistical analyses. Thomas Ollendick and Diane Bray coauthored the study protocol with Cecilia Essau. All authors contributed to and have approved the final manuscript.

20

Role of funding source This is part of a large on-going project on anxiety prevention in children, which is partially funded by The Education, Children’s and Cultural Services Directorate of the London Borough of Richmond upon Thames, and in collaboration with the Catholic Children’s Society.

References Barrett, P.M. & Turner, C. (2001). Prevention of anxiety symptoms in primary school children: Preliminary results from a universal school-based trial. British Journal of Clinical Psychology, 40, 399-410. Barrett, P.M., Lock, S., & Farrell, L.J. (2005). Developmental differences in universal preventive intervention for child anxiety. Clinical Child Psychology and Psychiatry, 10, 539-555. Cobham, V.E., Dadds, M.M., & Spence, S.H. (1999). Anxious children and their parents: What do they expect? Journal of Clinical Child Psychology, 28, 220–231. Barrett, P. M., Duffy, A. L., Dadds, M. R., & Rapee, R. M. (2001). Cognitive-behavioral treatment of anxiety disorders in children: Long-term (6 year) follow-up. Journal of Consulting and Clinical Psychology, 69, 135-141. Barry, C.T., Frick, P.J., & Killian, A.L. (2003). The relation of narcissism and self-esteem to conduct problems in children: a preliminary investigation. Journal of Clinical Child and Adolescent Psychology, 32(1), 139-152.

21

Box, G.E.P. (1954). Some theorems on quadrative forms applied to the study of analysis of variance problem; I: Effect of inequality of variance in the one way-classification. Annals of Mathematical Statistics, 25, 290–302. Cartwright-Hatton, S., Hodges, L., & Porter, J. (2003). Social anxiety in childhood: The relationship with self and observer rated social skills. Journal of Child Psychology and Psychiatry, 44, 1–6. Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In: R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 69–93). New York: Guilford Press. Dadds, M.R., Spence, S.H., Holland, D.E., Barrett, P.M., & Laurens, K.R. (1997). Prevention and early intervention for anxiety disorders: A controlled trial. Journal of Consulting and Clinical Psychology, 65, 627-635. Essau, C. A., & Ollendick, T. H. (2013). The Super Skills for Life program. University of Roehampton. Essau, C.A., Conradt, J., Sasagawa, S., & Ollendick, T.H. (2012). Prevention of Anxiety Symptoms in Children: A Universal School-Based Trial. Behavior Therapy, 43, 450464. Essau, C.A. (2005). Use of mental health services among adolescents with anxiety and depressive disorders. Depression and Anxiety, 22, 130-137. Essau, C.A., Conradt, J., & Petermann, F. (2000). Frequency, comorbidity, and psychosocial impairment of anxiety disorders in adolescents. Journal of Anxiety Disorders,14, 263279. Geisser, S., & Greenhouse, S.W. (1958). An extension of Box’s result on the use of F distribution in multivariate analysis. Annals of Mathematical Statistics, 29, 885–891.

22

Glass, K., Flory, K., Martin, A., & Hankin, B.L. (2011). ADHD and comorbid conduct problems among adolescents: associations with self-esteem and substance use. Attention Deficit and Hyperactivity Disorders, 3(1), 29-39. Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A Research Note. Journal of Child Psychology and Psychiatry, 38, 581-586. Harvey, A. G., Clark, D. M., Ehlers, A., & Rapee, R. M. (2000). Social anxiety and selfimpression. Cognitive preparation enhances the beneficial effects of video feedback following a stressful social task. Behaviour Research and Therapy, 38, 1183-1192. Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12–19. Jacobson, N. S., Follette, W. C., & Revenstorf, D. (1984). Toward a standard definition of clinically significant change. Behavior Therapy, 17, 308–311. Keller, M. B., Lavori, P., Wunder, J., Beardslee, W. R., Schwartz, C. E., & Roth, J. (1992). Chronic course of anxiety disorders in children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 595–599. Kendall, P. C., Flannery-Schroeder, E., Panichelli-Mindel, S. M., Southam-Gerow, M., Henin, A., & Warman, M. (1997). Therapy for youths with anxiety disorders: A second randomized clinical trial. Journal of Consulting and Clinical Psychology, 65, 366–380. Langley, A. K., Bergman, R. L., McCracken, J., & Piacentini, J. (2004). Child Anxiety Impact Scales. In impairment in childhood anxiety disorders: preliminary examination of the child anxiety impact scale-parent version. Journal Child Adolescent Psychopharmacology, 14 (1), 105 – 114. Lewinsohn, P. M., Hops, H., Roberts, R. E., Seeley, J. R., & Andrews, J. A. (1993). Adolescent psychopathology: I. Prevalence and incidence of depression and other

23

DSM-III-R disorders in high school students. Journal of Abnormal Psychology, 102, 133 – 144. Lowry-Webster, H.M., Barrett, P.M., & Dadds, M.R. (2001). A universal prevention trial of anxiety and depressive symptomatology in childhood: Preliminary data from an Australian study. Behaviour Change, 18, 36–50. Lowry-Webster, H.M., Barrett, P.M., & Lock, S. (2003). A universal prevention trial of anxiety symptomology during childhood: Results at one year follow-up. Behaviour Change, 20, 25-43. Merikangas, K.R., He, J.P., Burstein, M., Swanson, S.A., Avenevoli, S., Cui, L., et al. (2010). Lifetime prevalence of mental disorders in U.S. adolescents: results from the National Comorbidity Survey Replication--Adolescent Supplement (NCS-A). Journal of the American Academy of Child and Adolescent Psychiatry, 49, 980-989. Mostert, J. & Loxton, H. (2008). Exploring the effectiveness of the FRIENDS program in reducing anxiety symptoms among South African Children. Behaviour Change, 25, 85–96. Muris, P., Schmidt, H., & Merckelbach, H. (2000). Correlations among two self-report questionnaires for measuring DSM-defined anxiety disorder symptoms in children: the Screen for Child Anxiety Related Emotional Disorders and the Spence Children's Anxiety Scale. Personality and Individual Differences, 28(2), 333-346. Ollendick, T. H., & King, N. J. (1994). Assessment and treatment of internalizing problems: The role of longitudinal data. Journal of Consulting and Clinical Psychology, 62, 918-927. Ollendick, T. H., Costa, N. M., & Benoit, K. E. (2010). Interpersonal processes and the anxiety disorders of childhood. In G Beck (Ed.), Interpersonal processes in the

24

anxiety disorders: Implications for understanding psychopathology and treatment. Washington, DC: APA Books, Pp. 71-95. Rosenberg, M. (1965). Rosenberg Self-Esteem Scale. Society and the adolescent self-image. Princeton, NJ: Princeton. University Press. Schreier, S. S., & Heinrichs, N. (2010). Parental fear of negative child evaluation. In Parental fear of negative child evaluation in child social anxiety. Behaviour Research and Therapy, 48 (12), 1186 – 1193. Spence, S.H. (1998). A measure of anxiety symptoms among children. Behaviour Research and Therapy, 36, 545-566. Stallard, P., Simpson, N., Anderson, S., Hibbert, S., & Osborn, C. (2007). The FRIENDS Emotional Health Programme: Initial Findings from a School-Based Project. Child and Adolescent Mental Health, 12, 32–37. Tofighi, D. & MacKinnon, D. P. (2011). RMediation: An R package for mediation analysis confidence intervals. Behavior Research Methods, 44, 692–700.

25

Table 1. Estimated marginal mean of the outcomes and pairwise comparison between pre-, post-treatment and 6-months follow-up

Outcome

Pre-treatment

Post-treatment

6 follow-up

Contrasts*

Mean (95%

Mean (95%

Mean (95%

CI)

CI)

CI)

Separation Anxiety

6.13 (4.48 -

3.94 (2.88 -

3.83 (2.88 -

Pre > Post, Pre >

(n=54)

7.78)

4.99)

4.83)

Follow

Social phobia (n=56)

6.38 (4.80 -

4.82 (3.67 -

3.96 (2.84 -

Pre > Follow

7.95)

5.97)

5.07)

6.57 (4.89 -

5.31 (4.12 -

4.45 (3.45 -

8.24)

6.50)

5.44)

5.63 (3.44 -

5.09(3.22 -

4.22 (2.79 -

7.81)

6.96)

5.64)

3.88 (2.69 -

3.92 (2.96 -

3.26 (2.38 -

5.07)

4.89)

4.15)

6.81 (5.39 -

5.08 (3.82 -

4.74 (3.64 -

8.23)

6.34)

5.84)

35.91 (27.09 -

27.14 (20.90 -

23.34 (17.60 -

44.73)

33.38)

29.08)

Emotional problems

3.19 (2.40 -

3.15 (2.36 -

2.91 (2.23 -

(n=55)

3.98)

3.93)

3.59)

Conduct problems

3.44(2.80 -

3.85 (3.26 -

2.59 (1.97 -

SCAS

OCD (n=55)

Panic (n=55)

Fears (n=55)

GAD (n=56)

Total SCAS (n=51)

Pre > Follow

Pre > Follow

SDQ

Post > Follow

26

(n=55)

4.09)

4.43)

3.22)

Hyperactivity (n=56)

5.03 (4.46 -

5.78 (5.07 -

4.43 (3.84 -

5.60)

6.50)

5.01)

5.08 (4.59 -

3.21 (2.67 -

5.85)

5.57)

3.75)

Prosocial behaviour

7.63 (6.94 -

7.79 (7.13 -

7.41 (6.69 -

(n=55)

8.31)

8.45)

8.13)

Total SDQ (n=54)

16.85 (14.83 -

17.80 (15.85 -

20.20 (18.35 -

18.88)

19.75)

22.06)

18.19 (16.57 -

17.28 (15.35 -

17.39 (15.55 -

19.82)

19.21)

19.22)

Social Skills (SSQ)

18.94 (17.58 -

19.55 (17.94 -

19.32 (17.79 -

(n=50)

20.30).

21.16)

20.84)

Anxiety Impact

21.59 (15.30 -

16.62 (11.32 -

17.13 (11.53 -

(n=46)

27.87)

21.93)

22.72)

5.44 (4.17 -

4.76 (3.34 -

3.87 (2.45 -

6.72)

6.18)

5.30)

5.31 (3.98 -

4.34 (2.92 -

3.19 (1.93 -

6.63)

5.76)

4.45)

Peer problems (n=56) 5.29 (4.74 -

Self –esteem (n=45)

Post > Follow

Pre > Follow

Pre < Follow

Fear of negative evaluation Maternal (n=53)

Paternal (n=48)

Pre > Follow

*Significant contrasts are displayed. Contrasts are based on Bonferroni post hoc tests.

27

Table 2. Estimated marginal mean of the speech-task ratings and pairwise comparison between pre-, post-treatment and 6-months follow-up

Outcome

Pre-treatment

Post-

6 follow-up

Contrasts*

treatment Mean (95%

Mean (95%

Mean (95%

CI)

CI)

CI)

Overall behavioural

6.71 (4.61 -

4.64 (3.40 -

3.31 (2.38 -

anxiety (n=55)

8.81)

5.87)

4.24)

Nail-biting (n=55)

0.17 (0.03 -

0.01 (-0.03 -

0.00 (0.00 -

(marginal

0.31)

0.05)

0.00)

significance,

Behavioural indicators of anxiety Pre > Follow

Pre > Follow) Lip-licking (n=55)

0.44 (0.21 -

0.17 (0.02 -

0.04 (-0.03

Pre > Follow

0.67)

0.31)

- 0.11)

Fingers touching mouth

0.41 (0.09 -

0.08 (-0.02 -

0.06 (-0.04

(marginal

(n=55)

0.73)

0.19)

- 0.17)

significance, Pre > Follow)

Sucking/chewing (n=55) 0.23 (0.04 -

Lip contortions (n=55)

Lip-biting (n=55)

0.12 (0.02 -

0.08 (-0.03

0.42)

0.22)

- 0.19)

1.59 (1.13 -

0.87 (0.55 -

0.76 (0.37 -

Pre > Post, Pre >

2.04)

1.20)

1.15)

Follow

0.38 (0.09 -

0.21 (0.00 -

0.08 (-0.05

0.67)

0.42)

- 0.20)

28

Hand movement (face)

0.27 (0.04 -

0.01 (-0.03 -

0.00 (0.00 -

(marginal

(n=55)

0.51)

0.05)

0.00)

significance, Pre > Follow)

Hand movement (head)

0.32 (0.10 -

0.16 (0.02 -

0.24 (-0.05

(n=55)

0.55)

0.29)

- 0.54)

Hand movement (other)

0.96 (0.58 -

0.94 (0.62 -

0.44 (0.12 -

(n=55)

1.35)

1.26)

0.76)

Leg movement (n=55)

1.93 (1.31 -

2.07 (1.44 -

1.61 (1.14 -

2.55)

2.69)

2.07)

8.50 (7.83 -

9.14 (8.30 -

6.36 (5.32 -

Pre > Follow, Post

9.17)

9.97)

7.41)

> Follow

6.58 (5.72 -

5.87 (5.07 -

8.26 (7.49 -

Pre < Follow, Post

7.43)

6.68)

9.03)

< Follow

7.09 (6.26 -

6.21 (5.47 -

9.30 (7.68 -

Pre < Follow, Post

7.92)

6.94)

10.91)

< Follow

8.34 (7.72 -

8.27 (7.83 -

5.64 (5.17 -

Pre > Follow, Post

8.97)

8.71)

6.12)

> Follow

8.52 (8.03

6.30 (5.63 -

7.88 (6.32 -

Pre > Post

- 9.00)

6.96)

9.44)

8.48 (7.93 -

6.68 (6.10 -

7.45 (7.01 -

Post > Follow

Objective/ subjective ratings Child rating Nervous (n=44)

Global (friendly etc.) (n=47) Micro (smile, etc.) (n=47) Observer rating Nervous (n=45)

Global (friendly etc.) (n=44) Micro (smile, etc.)

Pre > Post, Pre >

29

(n=45)

9.04)

7.26)

7.89)

Follow

*Significant contrasts are displayed. Contrasts are based on Bonferroni post hoc tests.

Table 3. Results of confidence intervals for mediating effects

Mean

SD

Lower

Higher

limit

limit

Social skills (change in SCAS scores)

0.02

0.03

-0.03

0.09

(change in social phobia scores)

-0.01

0.02

-0.06

0.04

(change in SCAS scores)

0.02

0.03

-0.05

0.09

(change in social phobia scores)

0.00

0.04

-0.08

0.08

(change in SCAS scores)

0.02

0.03

-0.02

0.09

(change in social phobia scores)

0.02

0.02

-0.02

0.08

(change in SCAS scores)

-0.01

0.03

-0.08

0.04

(change in social phobia scores)

-0.03

0.04

-0.12

0.05

Parental fear of child evaluation Mother

Father

Self-esteem