Emotion Recognition Training for Young People

Social anxiety disorder (SAD) is among the most prevalent mental disorders in adolescence, 10% of females and 5% of males are diagnosed with SAD in adolescence. SAD is characterised by marked fear or anxiety about social situations in which the individual is exposed to possible scrutiny by others. SAD is associated with social isolation, poor social skills, low self-esteem, and future comorbid mental disorders. Social anxiety typically develops during adolescence, thus there is a need to detect and treat it early prior to the development of a clinically-diagnosed SAD. Most early interventions for social anxiety have used Cognitive Behavioural Therapy (CBT), which focuses on modifying explicit verbal cognitions. However, many adolescents do not have access to CBT and many finding it demanding.

Research suggests that individuals who have high levels of social anxiety are more likely to misread facial emotions as threatening or disapproving, and researchers have proposed that the difficulty may lie in decoding low intensity facial expressions rather than identifying facial emotions which are expressed with greater intensity. In social anxiety, a bias towards perceiving ambiguous facial expressions as negative may give rise to avoidant behaviour and maintain or increase symptoms; therefore, this emotion processing bias may be a powerful bias for researchers to target with the aim of developing interventions to reduce symptoms.

ER training provides a promising new line of research which may be beneficial in altering emotion processing biases which maintain or increase symptoms of mental health disorders. Research has shown that ER training may be effective in reducing depressive symptoms in analogue populations with high levels of depression and anger and aggressive behaviour in healthy adults and adolescent youth at high risk of criminal offending and delinquency.

Present Study:

This research tests the effectiveness of ER training in reducing social anxiety in adolescents. Given that the age of onset of social anxiety is adolescence, peaking in late teens, this research focuses on 15-18 year olds.

Aims \& Objectives:

To test the effectiveness of ER training in reducing social anxiety in adolescents. The intervention aims to alter emotion processing biases through a training procedure designed to promote the perception of happiness over disgust in ambiguous facial expressions which we hypothesise will subsequently reduce symptoms of social anxiety.

Central Research Questions:

Is ER training associated with:

I. An increase in the perception of positive (happy) over negative (disgust) emotions in faces displaying ambiguous emotional expressions? II. A decrease in negative evaluations? III. A decrease in social anxiety? IV. A decrease in other forms of anxiety and/or symptoms of depression?

PHASE 1: SCREENING FOR SOCIAL ANXIETY IN ADOLESCENTS Objective: To screen for sub-clinical levels of social anxiety in adolescents. School Sampling: Approximately 8 large second-level schools in Dublin will be randomly selected from the Department of Education and Skills 2012 published list of schools.

Power Analysis: Using G*Power 3.1, the sample size was determined as 111 based on the following parameters: a medium effect size of 0.3, alpha of 0.05, power of 0.8, a repeated measures within-between interaction: 2 (Randomization) X 3 (Pre, Post, Follow-up).

Participants: Adolescents aged 15-18 years in second-level schools will be invited for the initial screening.

Procedure: Adolescents must return parental and informed consent forms. Participants will be screened using the Social Phobia and Anxiety Inventory for Children (SPAI-C). Participants scoring above the cut-off of ≥18 will be invited to participate in the intervention (Phase 2).

Approximately 440 second-level students across 8 schools will be invited to complete screening. With an approximate response rate of 50% (N=220), and using a cut-off of ≥18 120 will be identified (60 per condition).

PHASE 2: RANDOMIZED CONTROL TRIAL Objective: To compare the effectiveness of ER to placebo training in reducing social anxiety.

Participants: Adolescents from Phase 1 screening above a cut-off of ≥18 will be invited to participate in the intervention and will be randomly assigned to ER/placebo training. On the basis of screening, the PI will ensure group equivalence across ER versus placebo groups on key variables (e.g., age, gender, social anxiety) in advance of randomization. Post randomization, independent variables will be checked to ensure they are balanced across groups so as not to contaminate outcomes of the study.

Procedure:

Participants in both ER and placebo training groups will receive 4 training sessions over 4 consecutive days. The intervention will be performed on a laptop during school and delivered in a quiet room with the researcher present. Pre-intervention data will be collected on day 1 prior to training, post-intervention data will be collected on day 4 following training, with a follow-up assessment at 2-weeks post-intervention.

Training Task:

Participants will complete training over 4 consecutive days in schools. Each session will include a computerised training programme comprised of three phases, baseline, training, and test,which will take approximately 15 minutes to complete. A follow-up assessment will be completed two weeks post-intervention. During the 2-week follow-up assessment participants will complete a final assessment of emotion sensitivity using the test phase of the emotion recognition task.

• Prototypical 'happy' and 'disgust' composite images generated from 40 individual faces (20 x male, 20 x female) showing a disgust facial expression, and the same 40 individuals showing a happy expression.
• Prototypical images used as endpoints to generate a linear morph sequence that consists of images that change incrementally from unambiguously 'happy' to unambiguous 'disgust', with emotionally ambiguous images in the middle. The proposed research will include a sequence with 15 equally spaced images for each gender (male and female) for use as experimental stimuli.
• Previous research by the University of Bristol has used stimuli portraying male faces only. The task stimuli for the current research will be matched to the participant's gender, i.e. female participants will complete a task which will include female face stimuli and male participants will complete a task which will include male face stimuli.
• Baseline and test phases will consist of 45 trials, in which each of the stimuli from the morph sequence will be presented to the participant three times. The task will require the participant to make a forced choice judgement as to whether each face is displaying a happy or disgust expression. Images will be presented one at a time, in random order, for 150 milliseconds. Stimuli will be preceded by a fixation cross, which will be presented for a random period ranging from 1500 to 2500 milliseconds. Subsequent to presentation, and to prevent processing of after images, a backward mask of noise will be presented for 250 milliseconds, followed by a prompt asking the participant to respond. This will remain on screen until the participant makes response (i.e., a judgement of 'happy' or 'disgust').
• Trials in the training phase will be similar to trials in the baseline and test phases with respect to inter-trial interval and stimuli presentation, but with the addition of feedback subsequent to the participant's response. In the placebo condition, feedback will be based on the participant's baseline balance point. That is, responses will be classified as 'correct' if the participant identified images below the original balance point image as 'disgust' and above it as 'happy', and otherwise are classified as 'incorrect'. Feedback will be a message saying 'Correct / Incorrect! That face was disgust/happy'. In the training condition, feedback will be again based on the participant's baseline balance point, but the 'correct' classification will be shifted two morph steps towards the 'disgust' end of the continuum, so that the two images nearest the balance point that the participant would previously have classified as 'disgust' at baseline will be considered 'happy' when providing feedback. In each training block, each face from the 15 face continuum will be presented twice. The order of presentation will be randomized within each block. Three training blocks will be given to each participant, resulting in 90 training trials per session and a total of 360 training trials in total.

Outcome Measures:

Outcome measure will be administered to all participants in the RCT pre-intervention (baseline on day 1), post-intervention (day 4), and at 2-week follow-up, including:

Emotion Sensitivity Measurement Accuracy assessed using baseline and test measurements at each ER/placebo training session.

Social Phobia and Anxiety Inventory for Children (SPAI-C) The SPAI-C is a 26-item self-report instrument and has good psychometric properties with adolescent samples. The SPAI-C measures the somatic, cognitive, and behavioural aspects of social phobia in children and adolescents. Questions relate to how nervous the respondent feels when carrying out certain tasks and responses are rated on a 3-point scale from never, or hardly ever to most of the time, or always with higher scores indicating greater levels of social anxiety.

Brief Fear of Negative Evaluations Scale - Revised (BFNE-R) Negative evaluations are an important aspect of social anxiety. The BFNE-R is a 12-item measure which assesses fear of negative evaluation. Items are responded to on a 5-point Likert scale from 0 to 4. The BFNE-R has shown high internal consistency and correlates well with the original BFNE.

Screen for Child Anxiety Related Disorders (SCARED) Child Version The SCARED is used to screen for signs of anxiety disorders. The 41-item measure is used to measure anxiety, including panic, separation anxiety, generalised anxiety, social avoidance, and school phobia. It has shown high reliability with adolescents. Items are rated on a 3-point Likert scale. Internal consistencies have been shown to be high with adolescent samples.

Depression Subscale of the Revised Child Anxiety and Depression Scale (RCADS) The RCADS is a 47-item self-report measure and has good psychometric properties with adolescent samples. The Depression Subscale contains 10 items. Items relating to the frequency with which depressive symptoms are experienced are rated on a 4-point Likert scale.

Research Design: A mixed model 2 (Randomization) x 3 (Time) complex design combining between subjects and within subjects conditions will be employed. Each participant will be tested at pre-intervention, post-intervention, and 2-week follow-up.

Independent Variables:

Randomization (K=2; ER training/ Placebo) Time (K=3; Pre-intervention[day 1], Post-intervention [day 4], and Follow-up)

Analyses:

All analyses on change in social anxiety, anxiety, depression, and fear of negative evaluation will be conducted as intention-to-treat analyses.

Descriptive statistics Chi-square analyses and ANOVAs Primary analyses will consist of two-way repeated ANOVAs 2 (Randomization) x 3 (Time) to test the variables under study.