Time will tell: The contributions of short and long timescales of brain network development to cognition

During development, cognitive and neural changes are present across multiple timescales (i.e., minutes, years), yet the relations between timescales are not quite understood. According to Dynamics Systems Theory (DST), prior developmental change over the course of several years should inform a child’s current neurocognitive function during a minutes-long task. According to another theory known as neuroconstructivism, cognition should emerge from underlying brain organization such that more modular, or segregated, brain networks should correspond to specialized cognitive functions (i.e., motor execution), and between-network connections, or integration, should correspond to more complex cognitive functions that require coordination of multiple specialized processes (i.e., working memory). Leveraging a longitudinal dataset that included resting-state neuroimaging data acquired between 0-6 years (11 timepoints) and task neuroimaging data acquired between 8-12 years, the current project aimed to 1) investigate how trajectories of functional segregation and integration at the short timescale of minutes relate to motor learning and working memory at 8-12 years (Studies 1 and 2), and 2) investigate how trajectories of functional segregation and integration at the long timescale of years (0-6 years) relate to the same trajectories of motor learning and working memory across minutes (Studies 3 and 4). The first study used the serial reaction time task to probe motor learning and revealed that segregation between visual and somatomotor networks marginally increased as children learned motor sequences across minutes. Further, reduced integration between visual and somatomotor networks supported faster motor sequence performance. The second study used the n-back task to probe working memory and demonstrated that somatomotor-salience network integration increased as children made working memory gains, whereas integration of the default mode network decreased across these blocks. The third study characterized years-long trajectories of early functional segregation to find that more rapid increases of visual-somatomotor segregation across the first two years of life were related to reduced motor learning rates and greater visual- somatomotor segregation increases, suggesting that accelerated brain maturation is less optimal later in childhood. The fourth and final study revealed greater levels of somatomotor-salience network integration between 3 and 6 years, but not earlier in development, were associated with greater working memory gains at 8-12 years, suggesting that protracted neural integration is more optimal for later working memory. Collectively, this project provided partial support for DST and neuroconstructivist theories, and revealed the importance of early functional network development for later neurocognitive functioning in childhood.