Patterns of Typical and Atypical Age-related Brainstem Volume losses

Background: The brainstem and its different sub-systems control essential functions such as motor agility etc. that worsen with age. The purpose of this study was: 1. To assess the impact of age-related volume loss within three brainstem sub-systems on functions supported by them. 2. To use data-driven machine learning to identify different volume loss patterns or subtypes and investigate how they are associated with function. Methods: Structural MRI and behavioral data from 674 Human Connectome Project Aging (HCA) participants was used in this project. The brainstem was extracted, internal brainstem structures segmented and the segmentations warped onto a probabilistic population atlas on which the nuclei of interest had been labeled. Jacobian deformation maps were calculated, each rois mean Jacobians extracted and converted into z-scores with and without correction for age. Linear regression analyses were used to assess volume-function (cognition, motor agility, autonomic control) associations for each roi belonging to the sub-system supporting these functions. Subtype and Stage Inference (SuStaIn) was used to identify different volume loss patterns in each sub-system. Results: Age explained larger percentage of the variation of the behavioral variables than brainstem volumes. SuStaIn identified up to 4 subtypes, one representing typical aging and the remainder atypical aging. The subtypes did not significantly differ behaviorally with the exception of grip strength and diastolic blood pressure. Conclusion: Aging affects brainstem systems which contributes to the worsening of these functions with increasing age. SuStaIn detected different patterns of volume loss or subtypes within each of the brainstem systems.