Brain maintenance may help explain why some individuals remain cognitively resilient despite aging, but its biological basis is not well understood. Here, we show that brain maintenance can be quantified from the relationship between brain structure and function. Using structural MRI and resting-state functional MRI from 1280 older adults, we built a model based on young adults to estimate the functional capacity supported by preserved brain structure, and defined brain maintenance as the difference between predicted and observed function. Brain maintenance was most evident in prefrontal, cingulate, and precuneus regions and was enriched in higher-order functional networks. Higher brain maintenance was associated with slower cognitive decline, lower amyloid-β burden, and domain-specific variation in memory, attention, and processing speed. These findings provide a biologically grounded marker of resilience in healthy and pathological aging. Using multimodal MRI and PET across healthy and pathological aging, the authors derive a structure-function brain maintenance metric that predicts cognitive resilience, slower decline, and lower amyloid burden.
Li et al. (2026) studied this question.