Decision‐tree–testing cognition‐MRI associations to define and differentiate cognitive reserve and brain maintenance

Lídia Vaqué‐Alcázar,María R. Cabello‐Toscano,Tetiana Gorbach,Kilian Abellaneda‐Pérez,Cristina Solé‐Padullés,Athanasia M. Mowinckel,Mikael Stiernstedt,Dídac Macià,Alvaro Pascual‐Leone,Kristine B. Walhovd,Anders M. Fjell,Lars Nyberg,David Bartres‐Faz
DOI: https://doi.org/10.1002/alz.056074
2021-12-01
Abstract:Abstract Background Cognitive reserve (CR) and brain maintenance (BM) represent heuristic concepts defining general mechanisms that may underlie successful cognition in advanced age. However, a robust differentiation of both models employing a unified approach is lacking. Within the framework of a recent NIH‐funded study (R24‐AG061421) from the Reserve & Resilience collaborative network (https://reserveandresilience.com/) we have proposed an iterative process (Figure 1) to help operationalizing and testing BM and CR definitions using longitudinal episodic memory (EM) and MRI (structural and functional) longitudinal assessments. The current study presents findings after analyzing data according to the first step of the proposed pipeline (Figure 1.1). Method We analyzed longitudinal (2‐3 observations) MRI‐derived hippocampal (HC) volumes calculated with FreeSurfer 6.0 (https://surfer.nmr.mgh.harvard.edu/) and EM from 356 individuals (age range: 60‐85 years at baseline, 56.2% female) from three different cohorts comprised within the European Lifebrain (https://www.lifebrain.uio.no/) consortia. To define EM, we used the mean of scaled visual and verbal memory tests scores. All the analytical procedures have been recently described (Gorbach et al., Alzheimers Dement. 2020;28;12(1):e12110). Result According to the interarrival process step 1, we identified two groups: (i) BM pathway 1 (N = 99) in the first quadrant (positive change in EM and HC) and (ii) ‘deviant cases’ (N = 74) in the fourth quadrant (those with marked HC atrophy but stable EM, see Figure 2). There were no differences regarding education and mini mental state examination (MMSE) scores between the groups identified. Conclusion The analysis of the associations between progressive EM and HC volume changes allowed us to initially differentiate a group of individuals with greater correspondence between the two measures across time (i.e., supporting the concept of BM) versus those where memory stability appears to progressively be decoupled from HC integrity (i.e., suggesting evidence of reserve/compensatory processes). Despite initial comparisons for education and global cognitive function did not reveal differences, further analyses including lifestyle engagement and functional MRI data information shall provide useful information towards a more precise characterization of CR and BM concepts.
clinical neurology
What problem does this paper attempt to address?