Abstract (Expand)
Background: Hippocampal volume, assessed via high-resolution MRI, is associated with memory and visuospatial performance in humans (Squire, 2004) and specifically prone to develop atrophy with age … (Apostolova,2015). This process has been linked to neurodegenerative diseases, such as Alzheimer’s disease (Apostolova,2015) and a decline of cognitive functions (Bruno,2016). However, due to differences in study-design and characteristics certain heterogeneity in results remains, in particular considering subfieldspecific effects (deFlores,2015). Therefore, we aim to determine the association of volumes of the whole hippocampus and its subfields on cognition in a large population-based cohort. Methods: Subjects: 1956 healthy participants from the Leipzig Research-Center-for-Civilization-Disease, aged 19-82years with MRI and neuropsychological tests (mean-age=57.61,±15.08SD). Exclusion: stroke, major-brain-pathologies, central-nervous-medication. Independent Variables: Volume of hippocampus and its subfields (CornuAmmonis1, 2-3, 4-DentateGyrus,(Pre-)subiculum). Dependent Variables: Verbal word-list learning, verbal-fluency, TrailMakingTask-(TMT)-A&B. Covariates: sex, age, years-of-education, total grey-mattervolume Image Analysis on high-resolution T1-images assessed at 3T. Hippocampal volumes were estimated using automatic segmentation analysis implemented in FreeSurfer (www.freesurfer.net). Statistical Analysis: Independent and dependent variables were first entered into Pearson Correlations. Variables with a correlation coefficient of r>0.1 were entered into multiple linear-regressions and adjusted for potential confounding(forward inclusion-model). Results: According to bivariate correlations, better performance in verbal-learning, verbal-fluency and TMT-A&B correlated moderately with larger whole-hippocampal volume and the volumes of all subfields(all |r|>0.102, all p0.046, all p0.5). Conclusions: Using a large cross-sectional cohort of healthy adults we found that volumes of the whole-hippocampus and subfields covering the CA4/dentate-gyrus region were weakly, yet specifically associated with verbal-learning and spatial processing-speed. Our preliminary results are in line with previous studies presuming a differential involvement of the hippocampus in tasks of verbal-learning and spatial processing (Oosterman,2010). Upcoming analyses implementing parcellation along the anteriorposterior- axis and random-effect-models might help to further disentangle these effects.
Authors: S. Huhn, R. Zhang, Frauke Beyer, L. Lampe, T. Luck, S. G. Riedel-Heller, M. L. Schroeter, Markus Löffler, M. Stumvoll, A. Villringer, A. V. Witte
