Shared neural network interactions underlying visual cognition, attentional reorientation, and executive function across developmental stages

Previous research in developmental neuroscience has explored changes in the brain's functional connectivity during various developmental stages, offering insights into its dynamic network changes. However, it remains unclear how these alterations relate to cognitive performances.
We employed partial least squares (PLS) to uncover the latent components (LCs) that describe the relationship between visual cognition, attentional reorientation, and executive function performances and resting-state functional connectivity (RSFC) across ages. Forty-two participants participated in fMRI
sessions including a visual contrast sensitivity task, a combined Posner-Flanker attention cueing task, and a resting state. The participants were adults (21-33 years old, n = 17), adolescents (14-17 years old, n = 9), and children (7-13 years old, n =16). We extracted time-series data from Schaefer’s 1,000 parcellations cortical atlas to construct whole-brain RSFC. The behavioral scores and the whole-brain RSFC were used as inputs for PLS. Interestingly, we found only one robust LC (p-value < 0.05, FDR corrected), characterized by its specific RSFC alteration pattern and associated changes in age and cognitive performances. Specifically, this LC suggested that RSFC in limbic and peripheral vision networks were positively correlated with age. On
the other hand, RSFC in executive control, default mode, and salience/ventral attention networks were negatively correlated with age. Additionally, this pattern was associated with the improvement in cognitive performances in older participants. In summary, Our study shows age-related changes in RSFC, linked to better cognitive performances, offering insights into developmental mechanisms that
explain neural network interactions that support key cognitive functions across different ages.