Longitudinal clinical and neuroanatomical correlates of memory impairment in motor neuron disease

Abstract

Memory impairment in motor neuron disease (MND) is still an underrecognized feature and has traditionally been attributed to executive dysfunction. Here, we investigate the rate of memory impairment in a longitudinal cohort of MND patients, its relationship to other cognitive functions and the underlying neuroanatomical correlates. 142 patients with MND and 99 healthy controls (HC) underwent comprehensive neuropsychological testing and structural MRI at 3T up to four times over a period of 18 months. Linear-mixed effects models were fitted to identify changes at baseline and over time in episodic memory function (learning, immediate and delayed recall, recognition), composed cognitive scores (memory, verbal fluency, executive function), and memory-related structural brain regions (hippocampus, entorhinal cortex, parahippocampal gyrus). Associations between episodic memory performance and volumetric or cortical thickness changes of these regions were computed using Pearson's r. Learning, immediate and delayed recall, as well as recognition performance were significantly reduced in MND when compared to controls at baseline. Performances in these subtests improved over time although MND showed less improvement than controls. This relationship did not change when only "classical" ALS patients were considered. Patients with MND showed thinning of the right parahippocampal gyrus (PhG) in comparison to controls that was progressing over time. Bilateral hippocampal atrophy was observed in MND patients with memory impairment after splitting the group according to their overall episodic memory performance, with the right hippocampus shrinking over time. In MND patients, the bilateral hippocampal atrophy was associated with impairment in learning, recall, and recognition at baseline. In contrast, left PhG thinning was associated with a poorer learning performance. These results show that episodic memory impairment in MND is a frequent cognitive dysfunction. Since deficits are not clearly declining with disease course, an early involvement of this cognitive domain in the disease seems probable. The memory performance-dependent atrophy of the hippocampus and PhG provide evidence for a widespread involvement of these non-motor cortical areas in disease pathology.

Keywords: ALS; Entorhinal cortex; Hippocampus; Longitudinal episodic memory; MND; MRI.

Fig. 1

Flow diagram of recruitment scheme. MND: Motor neuron disease, RAVLT: Rey Auditory Verbal Learning Test, HC: Healthy controls, MRI: Magnetic resonance imaging, T: Timepoint.

Fig. 2

Component weights of the PCA derived components (comp) from neuropsychological tests. The color bar indicates the individual normalized weights.

Fig. 3

Individual evolution of episodic memory sub functions over time in HC (red) and MND patients (green). Dotted lines show the cut-off for < 1.5/ >1.5 standard deviations. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Distribution of z-standardized composite cognitive scores over time in HC (red) and MND patients (green). The y-axis displays the component density, dotted lines indicate the cut-off for < 1.5/ >1.5 standard deviations. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Individual trajectories of hippocampal volume and entorhinal and parahippocampal thickness over time in HC (red), MND without (violett) and MND with memory impairment (turquoise). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Correlation matrices (diagonal omitted) for the association between PhG and ERC thickness, hippocampal volume and episodic memory sub functions in HC (left) and MND (right). *Indicate significance at p < 0.05 following Bonferroni correction.