Threat Interpretation Bias as Cognitive Marker and Treatment Target in Pediatric Anxiety

Anxiety is the most common mental health problem in children and adolescents. This two-phased study will test the effects of an experimental computerized intervention aimed at reducing threat-based thinking (i.e., interpretation bias) in anxious youth. Participants in both the R61 (N=46) and R33 (N=72) trials will be youth ages 10 to 17 with a primary anxiety disorder (Separation, Social, Generalized). In the R61 trial, youth will be randomly assigned to receive 16 sessions over 4 weeks of either a personalized cognitive bias modification program for interpretation bias (CBM-I) or a computerized control condition (ICC). If CBM-I reduces interpretation bias significantly more than the ICC, the R33 trial will commence. In the R33, youth will be randomly assigned to either CBM-I or an equal amount of time in a cognitive restructuring intervention, which also aims to reduce threat-based thinking in anxiety.

Anxiety is the most common mental health problem in youth, affecting one in four children and adolescents. Unfortunately, evidence-based treatments (pharmacotherapy, cognitive-behavioral therapy) are costly, not widely available, and ineffective for a substantial proportion of youth. In response, experts have called for novel treatments that directly target mechanisms underlying youth anxiety while simultaneously addressing barriers to care (i.e., cost, accessibility). One such promising mechanism is interpretation bias - the inaccurate interpretation of threat from ambiguity. The investigators have previously demonstrated that interpretation bias occurs in over 90% of anxious youth, is predictive of anxiety severity in clinical samples of youth, and differentiates between anxious and non-anxious youth. These data indicate that interpretation bias may be a ubiquitous phenomenon underlying anxiety expression in children and adolescents and therefore may be an ideal intervention target. Cognitive bias modification for interpretation bias (CBM-I) is a computerized intervention that attempts to reduce anxiety by directly modifying interpretation bias. CBM-I has demonstrated preliminary efficacy for reducing anxiety symptoms in adults. Yet extant CBM-I data in anxious youth are sparse, with little work addressing whether CBM-I significantly reduces interpretation bias, and whether this in turn reduces anxiety symptoms, as well as the dose necessary to reduce both bias and anxiety. This two-phased study tests personalized CBM-I in youth ages 10 to 17 who meet diagnostic criteria for a primary anxiety disorder (Separation, Social, Generalized). In the R61 Phase (N=46), a randomized clinical trial (RCT) examines whether CBM-I personalized to youth anxiety symptoms significantly reduces interpretation bias compared to a computerized interpretation control condition (ICC). The interpretation target will be measured at multiple time points (4, 8, 12, 16 sessions) to identify the optimal dose for reduction in interpretation bias. If the R61 trial results indicate that CBM-I outperforms ICC on interpretation bias reduction, the R33 phase will commence. In the R33 Phase, an RCT (N=72) will validate whether CBM-I significantly reduces interpretation bias, and conducts a mechanism test (i.e., does bias reduction precede and predict anxiety reduction?), by comparing CBM-I to cognitive restructuring, a clinically relevant psychosocial intervention that also targets anxious cognition.

Child, Adolescent, Pediatric, Youth, Anxiety, Separation anxiety, Social phobia, Social anxiety, Generalized anxiety, Interpretation bias, Cognitive bias, Cognitive bias modification, Interpretation bias modification, Randomized controlled trial, Threat bias, Threat interpretation, Intervention, Randomized, Cognitive training, National Institute of Mental Health, Personalized

Eligible youth will be randomly assigned to receive either 16 sessions of a personalized cognitive bias modification for interpretation bias (CBM-I) intervention or a computerized interpretation control condition (ICC). Regardless of randomization, participants will complete 16 sessions of their assigned condition within 4 weeks (1 in-lab training per week, and 3 at-home trainings per week).

Both participants and outcome assessors will be blind to condition. Unblinding occurs following all post-intervention assessment (i.e., after 16 CBM-I or ICC sessions).

Computerized 16-session intervention aimed at reducing interpretation bias. In this study, CBM-I is personalized to youth anxiety symptoms. During CBM-I sessions, youth indicate whether word-sentence pairs are related, and are provided with feedback aimed to reduce bias.

Computerized 16-session intervention that is not believed to significantly modify bias. In this study, youth see stimuli personalized to their anxiety symptoms. During ICC sessions, youth see word-sentence pairs and are required to indicate whether word and sentence are related, but are not provided with feedback that aims to "train" a reduction in interpretation bias.

Computerized intervention in which youth see word-sentence pairs personalized to their anxiety symptoms, and indicate whether these are related. Youth receive feedback aimed to reduce interpretation bias.

Computerized control condition in which youth see word-sentence pairs personalized to their anxiety symptoms. Youth indicate whether these are related, but are not provided with feedback that aims to "train" a reduction in interpretation bias.

Change in linguistic interpretation bias as assessed by the word-sentence association paradigm for youth (WSAP-Y)

The WSAP-Y is a computerized assessment of interpretation bias in which youth indicate whether word-sentence pairings are related. This measure provides information about the degree to which youth evidence interpretation bias, as well as a behavioral (reaction time) index for bias endorsement.

Youth view faces portraying neutral expressions or subtle emotional expressions (i.e., morphed faces ranging in intensity of emotional valence). Youth categorization of faces as neutral or threatening provides their sensitivity and bias for reporting presence of threat.

Change in self-reported interpretation bias as measured by the Children's Automatic Thoughts Scale (CATS)

The Children's Automatic Thoughts Scale is a youth self-report questionnaire which assesses presence and frequency of a variety of anxious thoughts from domains of: physical threat, social threat, personal failure, and anxious hostility. Total scores range from 0 to 160, with higher scores indicating more threat interpretations (i.e., worse score).

The Participant Acceptability Questionnaire is a 10-item youth and parent report questionnaire, accompanied by an exit interview, that assesses burden (travel, boredom), credibility of computerized intervention techniques, and youth comprehension of the intervention.

Inclusion Criteria: youth aged 10 to 17 diagnosed at study baseline with a primary "big three" anxiety disorder (Separation, Social, Generalized) standard score greater than or equal to 85 on the Wide Range Achievement Test - Word Reading Subtest (to ensure ability to read stimuli during interpretation bias assessment and CBM-I/ICC), and estimated IQ standard score of at least 80 on the Wechsler Abbreviated Scale of Intelligence youth and consenting parent/legal guardian speak sufficient English to complete consent/assent and study procedures no concurrent psychosocial services during study participation to reduce likelihood that other interventions are responsible for change in primary or secondary outcomes no psychotropic medications with no plans to start medications during study, or six weeks stable on SSRI or psychostimulant medication and dose with no plans to change medication/dose during study Exclusion Criteria: severe anxiety indicting that youth requires higher level of care (e.g., intensive treatment such as psychiatric hospitalization), significant diagnostic comorbidity (e.g., presence of psychosis or significant mood disorder), or another primary diagnosis that warrants alternate intervention significant uncorrected vision impairment (e.g., uncorrected blindness) that precludes participation in CBM-I/ICC safety concerns due to recent or acute suicidality with plan, intent, and/or attempt that warrants alternate intervention

Data resulting from this project will be de-identified and uploaded in a timely manner to the National Database for Clinical Trials Related to Mental Illness, including de-identified youth demographic and clinical characteristics, and summary scores from outcome measures, along with a data dictionary that provides variable definitions, value labels, and scoring of measures. The study protocol and dataset may be made available to other investigators based on successful completion of a data request form, following the conclusion of this study and publication of primary outcome papers.

Data will become available following conclusion of the study and publication of primary outcome papers.

Researchers will be required to complete a request form that includes: research identifying information and institutional affiliation, description of research objectives, and IRB approval. Completed request forms will be reviewed by the PI and Co-Is to ensure that the research request will not duplicate work being conducted by the study team. The requesting researcher must provide signed data sharing agreements from all users on their team outlining the details of how information will be kept secure, consistent with NIMH and University of Denver data sharing guidelines. Requesting researchers will also provide the University of Denver's IRB with an IRB approval from their institution. The data will only be released after the proposed study has been completed and primary outcomes have been accepted for publication.