Differential involvement of limbic and paralimbic cortex in episodic memory processing in cognitive aging and neurodegeneration

Abstract

AbstractBackgroundA key component of cognitive aging is episodic memory decline. Here we examined to which degree structural integrity of (para)limbic cortex and amyloid load contribute to normal episodic memory variation in subjects between 50 and 80 years old using mediation analysis. The main hypothesis was that the effect of age on episodic memory would be mediated through hippocampal volume.Method180 community‐recruited cognitively intact older adults belonging to the Flemish Prevent Alzheimer's Disease Cohort KU Leuven (F‐PACK) cohort had cognitive test scores within norms as inclusion criterion, genotyping (APOE ε4 and BDNF), MRI and [18F]flutemetamol‐PET. Factor analysis was applied to determine the latent structure in the neuropsychological dataset. The primary factor (variance explained: 23%, Eigenvalue: 2.26) mainly contained episodic memory test scores: Buschke selective reminding (loadings: learning: 0.83, delayed‐recall: 0.96) and Rey Auditory Verbal Learning Test (0.60). Structural T1‐weigthed MRI were processed using voxel‐based morphometry. Hippocampal, entorhinal, perirhinal and parahippocampal volumes were derived by intersecting the modulated gray matter with Brainnetome parcellations. Amyloid load was quantified in a composite neocortical volume on [18F]flutemetamol SUVR with cerebellar grey matter reference. Linear regressions or Mann‐Whitney U tests investigated associations with episodic memory, which were further analyzed with bootstrapped mediation (R‐package ‘lavaan’).ResultHigher age was negatively related to lower episodic memory scores (β=‐0.066, p<0.001), to lower hippocampal volume (β=‐0.005, p<0.001) and to lower entorhinal cortex volume (β=‐0.003, p<0.001). Lower hippocampal volume (β=3.80, p=0.0004) as well as entorhinal cortex volume (β=3.67, p<0.001) were related to lower episodic memory. There was no significant correlation of amyloid nor a genetic effect regarding age or episodic memory (p>0.12). Hence, we did not include amyloid or genotypes in further analyses. In the mediation analysis, hippocampal volume did not mediate the relation between age and episodic memory (Figure 1A). By contrast, entorhinal cortex volume did mediate the relation between age and episodic memory, however the direct path coefficient was also significant (Figure 1B). No significant indirect effects were found in the path analysis for the perirhinal cortex or parahippocampus (p>0.14).ConclusionThe effect of age on episodic memory is partly mediated by entorhinal rather than hippocampal volume in cognitive aging.