Asymmetric thinning of the cerebral cortex across the adult lifespan is accelerated in Alzheimer’s disease

Aging and Alzheimer’s disease (AD) are associated with progressive brain disorganization. Although structural asymmetry is an organizing feature of the cerebral cortex it is unknown whether continuous age- and AD-related cortical degradation alters cortical asymmetry. Here, in multiple longitudinal adult lifespan cohorts we show that higher-order cortical regions exhibiting pronounced asymmetry at age ~20 also show progressive asymmetry-loss across the adult lifespan. Hence, accelerated thinning of the (previously) thicker homotopic hemisphere is a feature of aging. This organizational principle showed high consistency across cohorts in the Lifebrain consortium, and both the topological patterns and temporal dynamics of asymmetry-loss were markedly similar across replicating samples. Asymmetry-change was further accelerated in AD. Results suggest a system-wide dedifferentiation of the adaptive asymmetric organization of heteromodal cortex in aging and AD.


Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response. Methods n/a Involved in the study ChIP-seq Flow cytometry MRI-based neuroimaging https://www.gin.cnrs.fr/en/tools/lh-sym/. AIBL data is available at https://aibl.csiro.au/research/support/ pending application approval and compliance with the data usage agreement.
For all analyses, sample size was determined based on data availability. No analyses were performed to predetermine sample sizes. We gathered as much data as we could, from both cross-sectional and longitudinal observations. All MRI scans that did not fail FreeSurfer processing from each site were included. For analyses using AIBL data, as we were specifically interested in quantifying group differences in change, we used only longitudinal observations. All available observations of non-reverting individuals diagnosed with Alzheimer's disease (AD) by their final timepoint, or classified as cognitively healthy throughout, were included in the analysis.
LCBC discovery sample exclusion criteria were pre-established: participants were required to score <21 on the Beck Depression Inventory and !25 on the Mini Mental Status Exam. Based on these criteria, 13 observations were excluded from the initial sample, bringing the total number of observations to 2577. No exclusion criteria was applied to replication datasets. AIBL exclusion criteria were pre-established: AIBL participants reverting from an AD or Mild Cognitive Impairment diagnosis at any later timepoint were excluded to increase the validity of our longitudinally-defined groups.
We sought replication in 4 independent longitudinal aging cohorts. Results showed full replication in 3 cohorts and partial replication in 1 cohort. Analysis of an independent Alzheimer's disease sample yielded similar results to the healthy aging samples.
For the main analysis, randomization is not applicable as there was no group allocation. AIBL participants were assigned to groups based on diagnosis at their final available timepoint. Covariates such as age, sex and site were controlled for.
For the main analysis, blinding is not applicable as there was no group allocation. AIBL participant clinical status was decided by a clinical review panel. Blinding may not be applicable for the present study, as the groups tested were defined by longitudinally-derived diagnoses. The main discovery sample consisted of 2577 scans (1851 longitudinal) from 1084 healthy individuals aged 20.0 to 89.4 (mean age=50.0; 703 females) from the Center for Lifespan Changes in Brain and Cognition database. 1 to 6 timepoints were available per subject. Replication samples consisted of up to 2 timepoints from: Cam-CAN: 898 observations of 634 unique participants (age range=20-91; mean age=55.5; females=323) BASE-II: 768 observations of 447 unique participants (age range=24-83; mean age=62.4; females=170) BETULA: 480 observations of 310 unique participants (age range=25-84; mean age=62.7; females=159) DLBS: 763 observations of 471 unique participants (age range=20-93; mean age=59.7; females=292) AIBL sample consisted of up to 4 timepoints from: NC group: 435 observations of 128 unique participants (age range=60-90; mean age=73; females=221) AD group: 110 observations of 41 unique participants (age range=55-89; mean age=74.7; females=55) LCBC participants were recruited via newspaper and social media advertisements, and are thus not representative of the population due to non-random sampling. Follow-up observations suffer to some degree from selective attrition as returning participants tend to be healthier and show higher cognitive performance. Overall, the study population tends to be higher educated and perform higher relative to same-age peers. However, this is stable over the whole age-range, and thus it is unlikely to affect the main results and conclusions of the study (i.e. Age × Hemisphere and main effect of Hemisphere), which showed high consistency across independent samples, also in studies employing random recruitment from the population (i.e. Cam-Can and Betula; see associated references in Methods). Selective attrition biases also seem to affect the different ages to a similar degree in the LCBC sample, though it is not unreasonable that older adults are somewhat more affected (due to death, illness, dementia, etc.) in other cohorts. Potentially, this could lead to an underestimation of the loss of asymmetry as only healthier older individuals come for additional follow-ups. This could be one candidate explanation for the lack of full replication of the lifespan trajectories in DLBS. It is also possible that the lack of significant effects we observed for regional changes in thickness asymmetry upon longitudinal cognitive scores may be somewhat affected by the high number of cognitively above-average participants in the LCBC sample. Potential biases and limitations associated with sample recruitment are discussed in the manuscript.
For specific details of participant recruitment in each longitudinal replication sample, see the cohort-specific reference in Methods.
All LCBC studies were approved by the Norwegian Regional Committee for Medical and Health Research Ethics. For ethical approvals of sub-studies, see cohort-specific references in Methods.

T1-weighted anatomical scans
This field is not applicable as no experiment was conducted in the present study No task was performed in the scanner (only anatomical scans used). For cognitive analyses, we used scores on the California Verbal Learning Test (CVLT) and Matrix Reasoning subtest of Weschler Abbreviated Scale of Intelligence (WASI), acquired as part of a standard neuropsychological test battery