Basal Ganglia Mechanical Stiffness is Associated with Fluid Cognitive Abilities in Amnestic Mild Cognitive Impairment

Introduction: Characterizing the departure from normal aging associated with neurodegenerative diseases and dementias is a critical step in developing methods for early disease identification and evaluating the success of disease-modifying interventions. Magnetic resonance elastography (MRE) is a novel MR imaging technique that has been shown to be sensitive to changes in brain health with aging and neurodegeneration. MRE uses mild, externally applied vibrations to quantify the mechanical properties of tissues in vivo, which reflect the microstructrural integrity of brain tissue even in the absence of macrostructural tissue changes, such as volumetric atrophy. Here we used MRE to characterize differences in brain mechanical properties between individuals with amnestic mild cognitive impairment (aMCI) and cognitively unimpaired subjects (CU). Individuals with aMCI have begun to experience memory impairment that has not yet progressed to clinical Alzheimer’s disease (AD), and as such are in a critical stage where imaging methods can be used to guide effective interventions. Successful implementation of intervention at this stage will be reliant on identification of the most affected regions, especially those that do not yet indicate any deviation from normal volumetric measurements. Here we examine the relationship between functional changes and changes in mechanical properties in the basal ganglia (BG) in a cohort of older adult subjects with and without aMCI. The BG has been shown to control access to fluid cognitive abilities such as working memory, so any change in the structural composition of these subcortical structures may affect this specific cognitive function.

Materials and

Methods: A cohort of 67 cognitively unimpaired subjects (21M/46F; 60-82y) and a cohort of 34 subjects with aMCI (11M/23F; 60-89y) completed an MRE scan to assess their brain mechanical properties. MRE scans were performed on a 3T Siemens Prisma MRI scanner (Erlangean, Germany) MRI scanner with a 64-channel head coil and 1.25 mm isotropic resolution, and vibrations were supplied by a pneumatic driver at 50 Hz (Resoundant, Rochester, MN). From these MRE stiffness mapsis, we quantified basal gangliaBG stiffness in each subject including as a subcortical region of interest comprising the caudate, pallidum, putamen, and nucleus accumbens structures. Regions Structures were segmented from a high-resolution T1-weighted anatomical image using FreeSurfer automatic segmentation (v7.2.0). Subjects also completed the NIH toolbox cognition battery from which we examined fluid cognition composite score, which reflects logic and reasoning skills, which have been shown to decline in early AD.

Results, Conclusions, and Discussions: We found a significant group difference in stiffness of the basal ganglia (BG) between aMCI subjects with multifactor ANOVA when controlling for age and sex. Interestingly, aMCI subjects had significantly higher BG stiffness than CUs (2.89 vs. 2.99 kPa, p< 0.05; Figure 1). This stiffness difference is independent of any volumetric differences in the BG between groups. Most MRE studies report that healthy brains are associated with higher stiffness than those with neurodegenerative pathology, with our previous work showing that aMCI participants had softer hippocampi than age-matched CUs. Our findings in the BG are surprising in this context but may indicate a compensatory mechanism that occurs during early progression of AD and results in increased stiffness, which was previously suggested by Murphy et al. Since aMCI is diagnosed clinically through memory impairment, a function which is primarily supported by the hippocampus, the decreased stiffness in this structure indicates damage that is well established at this stage of the disease. However, impairment of fluid cognition is not a notable feature of aMCI, though does become impaired in early AD, so decreased stiffness associated with lower fluid cognitive abilities seen in some aMCI subjects may indicate progression within the aMCI subtype to early AD.

Here we also provide the first evidence that higher BG stiffness in aMCI is associated with better cognitive function. Within the aMCI group, BG stiffness was positively correlated with fluid cognitive score (r=0.48, p< 0.01; Figure 2), where higher scores were associated with greater stiffness; the same relationship in the CU group was not statistically significant. This finding further points to higher BG stiffness being compensatory rather than pathological. The neurobiological basis of this compensatory increase in BG stiffness of aMCI subjects requires further study but could be influenced by acute neuroinflammatory processes. Our results suggest that MRE measures of mechanical properties can sensitively detect differences in BG stiffness which may be useful in assessing aMCI pathology and progression.

Acknowledgements (Optional): This research was supported in part by NIH grants R01-AG058853 and F31-AG086036. Recruitment of research subjects was supported by the Delaware Center for Cognitive Aging Research.