Abstract/Summary

Extensive research has examined attention bias to threat in the context of anxiety in adults, but little is understood about this association in young children and there is a dearth of longitudinal research examining whether attention bias to threat predicts anxiety over time in childhood. In the current study, a sample of 180 children participated in a longitudinal study, first as preschoolers and again as they transitioned to formal schooling. At baseline, children aged 3-4 years completed a free-viewing eyetracking task with angry-neutral and happy-neutral face pairs and an assessment of behavioural inhibition (BI). At follow-up, parents provided daily reports of their child’s state anxiety over a 2-week period as their child started school and completed a measure of their child’s anxiety symptoms. Results indicated that, on average, preschool-aged children exhibit a bias for emotional faces that is stronger for angry than happy faces. There was little evidence that this bias was associated with anxiety symptoms. However, BI interacted with dwell bias for angry faces to predict trajectories of anxiety over the transition to school. An unexpected interaction between BI and dwell bias for happy faces was also found, with dwell for happy faces associated with lower anxiety for children higher in BI. The findings are consistent with recent developmental models of the BI-anxiety relationship and indicate that attention bias modification may not be suitable for young children, for whom attention bias to threat may be normative.