(PS9-D82) Anxiety-related Attention Bias Heterogeneity Across Adolescence

A novel cognitive mechanism in anxiety disorders is attentional bias (AB) to threat. AB is exaggerated across a range of anxiety disorders and in subclinical high-trait anxiety. However, there is heterogeneity in AB among anxious adults and youth, including AB towards or away from threat, as well as an absence of AB. Interestingly, emerging research documents the expression of AB may change with age (Dudeney et al., 2015), but findings remain scarce and inconclusive. Studying anxiety subtypes in relation to AB heterogeneity may help to clarify the role of AB in anxiety and personalize AB-based treatment approaches. This study examined AB heterogeneity in a group of younger and older anxious adolescents and its association with several anxiety disorders.

This study included adolescents ages 12-14 (M_age =12.89, N = 78; 54% female, 45% White) with mild to severe anxiety. Anxiety symptoms were assessed using the SCARED (53% above clinical threshold; Birmaher et al., 1995). AB was assessed using a standardized attention bias task (dot-probe) and measured by trial-level bias scores (TLBS) which were calculated using reaction times. TLBS scores were used to capture fluctuation of AB during the dot-probe task: variability, mean positive, and mean negative scores (Zvielli et al., 2014).

ANOVAs were conducted with age (young/old) and one of four SCARED subscales (GAD, panic, social, and school avoidance above/below clinical cutoff) as independent variables and TLBS indicators as dependent variable. There were no significant interactions between age and anxiety subscales on TLBS scores. There were significant main effects of age on TLBS scores, with older adolescents showing less variability compared to younger in all models except social anxiety (p values < .001). Analyses revealed lower mean negative (p values < .001) and mean positive TLBS scores (p values < .001) for older adolescents compared to younger in all models except social anxiety. There were two trending main effects of anxiety subgroup on TLBS scores. Those above versus below the social anxiety clinical cutoff showed dampened mean negative AB, F(1,71) = 3.909, p =. 052, and those above versus below the school avoidance cutoff showed greater mean positive AB, F(1,71) = 3.694, p = .059.

These findings suggest differences in anxiety-related AB across adolescence depending on age and anxiety subtype. A pattern of age-related dampening of AB variability and magnitude emerged, evidenced by lower variability, mean positive, and negative scores for older adolescents, suggesting neurodevelopmental increases in regulatory capacity across early adolescence. The effects of anxiety subtype suggest AB heterogeneity (towards versus away from threat) depending on anxiety subtype, with a trend such that those evidencing elevated symptoms of school avoidance exhibited a greater bias towards threat. Future work should examine distinct patterns of AB by clinical subtypes to elucidate mechanisms of anxiety in adolescence and inform treatment approaches. Studying these processes is especially pertinent during high-threat, emergency events such as the COVID-19 pandemic, when rates of anxiety increased in youth, to understand the etiology and maintenance of anxiety symptoms (Racine et al., 2021).