Region-specific impact of aging on cortical myelination and thickness

Healthy aging affects both grey and white matter. However, the trajectories of regional specific degeneration are not fully understood. Here we investigate the effects of aging on cortical thickness and myelin concentration in a large cohort of healthy participants (N = 610) aged between 18 and 89 years old who underwent single-site T1-weighted, T2-weighted and MTI sequences in the context of the Cam-CAN project. Participants were subdivided in three age groups representative of young, middle and late adulthood. The large size of the dataset allowed us to minimize the impact of sample variance without relying on multi-site acquisition protocols. We assessed linear changes in cortical thickness and cortical myelin concentration; the latter was assessed using both T1w/T2w ratio and MTR proxies, to evaluate which is the most stable metrics. Our results do not fit with either the anterior-posterior gradient or the last-in/first-out hypothesis. We demonstrate that aging patterns are more complex than just depending on a spatial gradient or the temporally reversed order of regional development. Moreover, we show a dissociation in aging patterns between somatosensory and motor regions both in terms of cortical thickness and myelin concentration. Finally, comparing T1w/T2w and MTR results of cortical myelination, we found the latter being a more stable and reliable proxy. HighlightsO_LICortical thickness and myelo-architecture changes must be jointly considered in investigating brain aging trajectories. C_LIO_LIWe assessed linear changes in cortical thickness and myelination in a large, homogeneous and single site MRI dataset. C_LIO_LIMotor and sensory regions show a dissociation in their aging trajectories both in terms of cortical thickness and myelin concentration. C_LIO_LISensory processing regions show similar aging trajectories in both cortical thickness and myelin concentration. C_LIO_LIMTR is a more reliable proxy for myelin concentration compared to T1w/T2w ratio. C_LI