Associations Between Age, Hippocampal Volume, and Spatial Working Memory in Periadolescent Children: Preliminary Findings from the PRANK Study

Abstract

AbstractBackgroundDuring childhood development, particularly the periadolescent epoch, significant changes in cognition and brain structure are known to occur. The development of spatial working memory (SWM), for example, continues throughout the teenage years. The hippocampus is necessary for normal SWM, as seen by the substantial memory deficits seen in patients with Alzheimer's disease (AD) and their associated hippocampal pathology. It is hypothesized that vulnerability to AD could be modulated by early brain development, so understanding the developmental relationship between hippocampal volume (HV) and SWM would provide a key baseline for comparison of different trajectories. Our ongoing study, Polygenetic Risk of Alzheimer’s Disease in Nebraska Kids (PRANK; R01 AG064247), measures brain structure/function, cognitive abilities, and Alzheimer’s polygenetic risk score in periadolescent children (age 8‐13). Here, we report preliminary data from PRANK measuring the association between HV, age, and SWM.MethodUsing preliminary data from the PRANK project, we investigated the association between SWM, age, and HV. Our sample included children (N=80, age 8‐13 years), recruited from the community and tested at University of Nebraska Medical Center; demographics were consistent with the recruitment area. SWM was measured using the Spatial Working Memory task between errors term from the Cambridge Neuropsychological Battery. Structural MRI data were collected using a 3T Siemens Prisma instrument. HV was measured using automated segmentation in Freesurfer.ResultSWM was associated with age, r(80) = ‐.334, p = .002, such that older children showed fewer spatial working memory errors than younger children. HV was not significantly associated with SWM, controlling for age, r(77) = ‐.027, p = .813. However, the direction of the relationship between HV and SWM (larger HV associated with fewer SWM errors) was consistent with previous literature (Faridi et al., 2015).ConclusionThese preliminary results suggest that increasing age is associated with better SWM in typical childhood; findings for HV were non‐significant. We anticipate that our study’s full sample will provide appropriate statistical power to test the association of SWM with HV. In future efforts, we will investigate this association as it relates to polygenic risk for Alzheimer’s disease and whether these results generalize to children with Down syndrome.