Associations between thalamic nuclei volumes and spatial working memory performance in adolescents and young adults

Abstract

Complex cognition is increasingly appreciated to be supported by neural networks spanning cortical and subcortical brain structures. A crucial structure within these networks is the thalamus, a bilateral, subcortical brain structure characterized by internally differentiated cytoarchitecture and discrete sub-nuclei. Some thalamic nuclei support functions traditionally thought to be cortically driven, such as spatial working memory (SWM). To date, research assessing these associations during human development is sparse. In this study, we tested associations between thalamic nuclei volumes and performance on two SWM tasks in a sample of adolescents and young adults. Participants ages 11-25 years old enrolled in a longitudinal study of adolescent brain and behavioral development were included. Brain images were acquired at 3T. Participants completed two SWM tasks: a Delayed Response Task (DRT) and the Cambridge Neuropsychological Test Automated Battery Spatial Working Memory task (CANTAB SWM). Thalamic nuclei were segmented from whole brain volumes using FreeSurfer version 7.4.1. and combined to create five groups per hemisphere based on anatomical boundaries: lateral-caudal, anterior-lateral, medial, intralaminar, and posterior. Two analytic aims were evaluated. First, associations between participant age and thalamic nuclei volumes were examined using a hierarchical modeling strategy with linear age added in the first step and volume of the whole thalamus added in the second step. Second, to evaluate associations between thalamic nuclei and SWM, hierarchical regressions predicting SWM performance were conducted, entering linear age, the relevant sub-nucleus, whole thalamic volume, and the age by nucleus interaction. Findings indicate that volumes of the bilateral anterior-lateral and lateral-caudal nuclei and behavior vary by age. In addition, volumes of the lateral-caudal, anterior-lateral, and posterior sub-nuclei are predictive of various aspects of SWM performance. In most cases, these associations suggest that higher volume is associated with better SWM performance. Functional implications and empirical limitations will be discussed.