Contributions of default mode network stability and deactivation to adolescent task engagement

During a risky decision-making task, the stability of default mode network (DMN) suppression, rather than its absolute level of deactivation, predicted how well adolescents learned from feedback. Among 65 adolescents (mean age 13.32, 21 females), those with more stable DMN activation across time showed greater task engagement. The finding suggests a new mechanism for how the brain down-regulates internally-directed cognition to support goal-directed behavior, highlighting the value of model-based network approaches for studying brain dynamics.

Study at a glance

Abstract

Out of the several intrinsic brain networks discovered through resting-state functional analyses in the past decade, the default mode network (DMN) has been the subject of intense interest and study. In particular, the DMN shows marked suppression during task engagement, and has led to hypothesized roles in internally-directed cognition that need to be down-regulated in order to perform goal-directed behaviors. Previous work has largely focused on univariate deactivation as the mechanism of DMN suppression. However, given the transient nature of DMN down-regulation during task, an important question arises: Does the DMN need to be strongly , or more stably suppressed to promote successful task learning? In order to explore this question, 65 adolescents (M_age = 13.32; 21 females) completed a risky decision-making task during an fMRI scan. We tested our primary question by examining individual differences in absolute level of deactivation against the stability of activation across time in predicting levels of feedback learning on the task. To measure stability, we utilized a model-based functional connectivity approach that estimates the stability of activation across time within a region. In line with our hypothesis, the stability of activation in default mode regions predicted task engagement over and above the absolute level of DMN deactivation, revealing a new mechanism by which the brain can suppress the influence of brain networks on behavior. These results also highlight the importance of adopting model-based network approaches to understand the functional dynamics of the brain.