Development of a new social media fMRI task to better investigate bidirectional links between social media use and emotional health in youth.

PROJECT SUMMARY/ABSTRACT There has been much interest in the potential role of social media (SM) use in driving a current mental health crisis among teens, with a dire need for evidence that goes beyond self-report. One important avenue is to understand the role of the brain in mediating or moderating the effects of SM use on emotional health and vice

versa. However, there is almost no research addressing these questions, largely due to a lack of fMRI tasks that can probe the neural correlates of modern SM use. Existing tasks measure neural response to simulated threat and reward from fictitious peers but are out of date and do a poor job of simulating the SM experience of

today’s youth. We developed a preliminary version of the TeenBrainOnline (TBO) Task, an fMRI task that mimics key features of SM that are posited to enhance the salience of social feedback, such as quantifiable, public, and dynamic indicators of peer status. The goal of this project is to further develop and validate this new

developmentally-salient and ecologically-valid task. In Phase 1 (months 1-8), we will program the task for eyetracking and create stimuli for boys (as we previously only created stimuli for girls) and gender and racially/ethnically diverse youth to add to the task. We will also use a co-design participatory action approach

with a Youth Research Advisory Board (YRAB) comprised of teens ages 13-17 to add comments to the task, and to obtain other feedback on potential task improvements. We will then conduct 10 pilot scans, with iterative improvements to the task based on scan results and YRAB feedback. In Phase 2 (months 9-24), 50 teens

(ages 13-17) with depressive symptoms will complete the final version of TBO during fMRI with eye-tracking, the older Chatroom Interact (CHAT-I) Task, ecological momentary assessment (EMA) of daily SM use, and measures of depressive symptoms. In addition to computing split-half reliability, twenty youth will complete the

task again after three months for test-retest reliability. We will evaluate construct validity by testing whether the task activates brain regions that typically respond to social evaluation (e.g., amygdala, anterior insula, dorsal and subgenual anterior cingulate cortex, and ventral striatum) and elicits visual attention biases using eye-

tracking. Predictive validity will be assessed by attempting to show that ASN activity and attention biases on TBO predict variability in depressive symptoms at baseline and three months later, and are more predictive than similar indices from CHAT-I. Convergent validity will be tested by examining whether TBO neural and

eyetracking indices converge with EMA indices of SM use, focusing on indicators of a social evaluation orientation toward SM (e.g. social comparison orientation, fear of missing out, digital status-seeking). We expect that these associations will be stronger for TBO compared to CHAT-I indices. We will also explore

whether ASN activity on the TBO Task interacts with SM experiences to predict symptoms of depression. The task will be disseminated to facilitate future longitudinal and bidirectional studies of SM use, brain activity, and emotional health, consistent with the goals of this RFA.