Longitudinal changes in control energy of brain networks in older adults at familial risk for Alzheimers disease

Abstract

The ability of structural brain networks to control neural dynamics is affected by healthy and pathological aging, including Alzheimers disease (AD). According to network control theory, transitions between functional brain states incur energetic costs. How these costs change longitudinally with aging remains unknown. Here, in 279 older adults at familial risk for AD, control energy increased longitudinally for state maintenance and transitions between sensorimotor and attentional networks, while decreasing for another sensorimotor--attentional transition. Changes in transition costs between higher-order networks were associated with cognitive performance: greater DorsAttn--Executive cost was associated with poorer immediate memory, while lower Executive--Def+Lim cost was associated with poorer attention. Higher plasma p-tau181 was associated with lower DorsAttn--Def+Lim transition cost. These findings reveal transition-specific alterations in the energetic landscape of aging and distinct relationships with cognition and AD pathology. Together, they highlight energetic constraints on brain network reconfiguration as a complementary perspective on neurocognitive aging.