Search

INTRODUCTION: We assessed whether macro‐ and/or micro‐structural white matter properties are associated with cognitive resilience to Alzheimer's disease pathology years prior to clinical onset. METHODS: We examined whether global efficiency, an indicator of communication efficiency in brain networks, and diffusion measurements within the limbic network and default mode network moderate the association between amyloid‐β/tau pathology and cognitive decline. We also investigated whether demographic and health/risk factors are associated with white matter properties. RESULTS: Higher global efficiency of the limbic network, as well as free‐water corrected diffusion measures within the tracts of both networks, attenuated the impact of tau pathology on memory decline. Education, age, sex, white matter hyperintensities, and vascular risk factors were associated with white matter properties of both networks. DISCUSSION: White matter can influence cognitive resilience against tau pathology, and promoting education and vascular health may enhance optimal white matter properties. Highlights: Aβ and tau were associated with longitudinal memory change over ∼7.5 years.White matter properties attenuated the impact of tau pathology on memory change.Health/risk factors were associated with white matter properties. [ABSTRACT FROM AUTHOR]

Background: White matter changes are hypothesized to be among the earliest changes occurring in the course of Alzheimer's disease (AD). We investigated associations between AD pathology, namely amyloid and tau, and white matter microstructural changes in cognitively unimpaired individuals at risk of AD dementia. We expected higher levels of pathology to be related to lower free‐water‐corrected fractional anisotropy. Method: We studied 223 cognitively normal older adults with PET and multi‐shell diffusion MRI from the PREVENT‐AD cohort. We first measured standardized uptake value ratios (SUVRs) in 40 brain regions for amyloid‐β (Aβ, i.e., lateral and medial prefrontal, parietal, lateral temporal, and cingulate cortical regions in both hemispheres) and 26 brain regions for tau (i.e., Braak stages I, III, and IV). We then used TractSeg, a deep learning framework, to segment 49 well‐established white matter bundles. We mapped free‐water‐corrected fractional anisotropy (FAT) in those bundles. Lastly, to explore the relationships between AD pathology and white matter microstructure across bundles, we applied two separate partial least squares analyses: one for Aβ pathology and one for tau (Figure 1). Result: We found one significant latent variable relating Aβ pathology to FAT. Specifically, higher levels of Aβ pathology were associated with the combination of higher levels of FAT within the rostral body of the uncinate fascicle and superior longitudinal fascicle (Figure 2A). We found two significant latent variables relating tau pathology to FAT for which higher levels of tau pathology were associated with elevated FAT within the uncinate fascicle, thalamo‐occipital bundle, optic radiation, inferior occipitofrontal fascicle, inferior longitudinal fascicle, middle longitudinal fascicle, etc. (Figure 2B). Conclusion: In cognitively unimpaired individuals at risk of AD, increased Aβ and tau pathology were associated with higher levels of FAT in AD‐related bundles. These unexpected findings may suggest an inverted U‐shape pattern between AD pathology and FAT in the preclinical phase of AD for which FAT values would start by increasing before decreasing later in the course of the disease. This first increase might be a marker of early pathological processes such as neuroinflammation or swelling. [ABSTRACT FROM AUTHOR]

Background: Growing evidence suggests Alzheimer's disease (AD) develops from a complex cascade of events that vary between individuals. Multivariate approaches have the potential to capture the complexity and heterogeneity of pathologies underlying AD in a more holistic manner compared to univariate approaches. Method: The MRI data of 105 older adults with a family history of AD from the PREVENT‐AD dataset were analysed. Forty six of these individuals had the APOE4‐3 genotype (greater risk of developing AD), while 59 subjects had the APOE3‐3 genotype (normal risk). Diffusion‐weighted imaging and multi‐echo magnetization transfer, proton density and T1‐weighted data were used to compute several WM metrics (Fig1 b‐c). We computed the voxel‐wise Mahalanobis distance (D2) in WM between APOE4‐3 individuals and a reference group (APOE3‐3) (Fig1). D2 is a multivariate measure that combines several MRI metrics, accounting for their covariance, and yielding a score indicative of the degree of abnormality at each WM voxel (Fig2). Associations between WM D2 and cognitive performance were then investigated (Fig3). Independent component analysis (ICA) was performed on spatial dimensions and linear regression analyses were conducted between D2 in ICA components and scores on items of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). We then extracted the contribution of each MRI feature to D2 in the regions where significant associations were found. Result: A significant negative association (R = ‐0.36, p = 0.014) was found between D2 in a component corresponding to the splenium of the corpus callosum (CC) and scores on RBANS‐immediate memory (Fig3), indicating a link between a greater degree of WM abnormality and poorer performance. Proton density (MPM_PD) and longitudinal relaxation rate (MPM_R1), potentially indicative of WM loss, contributed most to D2 in this region. Conclusion: Our findings suggest that microstructural alterations in the posterior part of the CC may play a role in memory deficit in older adults who are at greater risk of developing AD. This is in line with other studies that showed a link between CC degeneration and memory impairments in early AD stages.1 1. Qiu et al. (2016). Inter‐hemispheric dysconnectivity in AD and aMCI. Sci.Rep. [ABSTRACT FROM AUTHOR]

Background: Alzheimer's disease (AD) has been defined as a biological construct that can be determined using in vivo biomarkers. Recent studies confirmed the clinical value of the AT(N) framework when using Aβ and tau positron emission tomography (PET) scans in predicting progression from cognitively unimpaired to MCI. Our study aims to investigate if plasma biomarkers can yield similar predictive performance to PET biomarkers, and to also compare against cerebrospinal fluid (CSF) biomarkers Method: Individuals from the PREVENT‐AD cohort were included in the study. Participants had Aβ and tau plasma measurements and at least one year of clinical follow‐up thereafter (n = 278). A subset of participants had CSF measurements (n = 102) and/or PET scans (n = 133) also with 1‐year clinical follow‐up. Participants were cognitively unimpaired at the time of all biomarker measurements. MCI diagnoses were made by clinical consensus among expert clinical and research staff blind to plasma, CSF, PET, MRI and APOE genotype. We established thresholds of plasma and CSF Aβ and tau positivity based on >80% specificity in identifying individuals with significant levels of Aβ‐PET (plasma Aβ42/40 = 0.057; plasma pTau181 = 7.53 pg/mL; CSF Aβ42 = 961.37; CSF pTau181 = 66.59 pg/mL). Aβ ([18F]NAV4694) and tau ([18F]AV1451)PET thresholds were defined by the mean plus 2 standard deviations from young participants (Aβ SUVR cut‐off = 1.17; tau SUVR cut‐off = 1.24). Result: 36.8% (7/19) of the A+T+ plasma group progressed to MCI, 50% (4/8) of the A+T+ CSF group progressed to MCI, and 46.6% (7/15) A+T+ PET group progressed to MCI (Fig. 1).We found differences in MCI progression status between the A+T+ group compared to the A‐T‐ and A+T‐ groups across all biomarkers' measurements, with the A+T+ group always showing a higher rate of progression than the other two groups (post‐hoc chi‐square analyses; p<0.05). Conclusion: Abnormal levels of Aβ and tau plasma biomarkers can help predict near‐ term progression to MCI. When plasma thresholds are derived based on PET data, however, plasma biomarkers have a slightly lower specificity than PET and CSF biomarkers, and both plasma and CSF biomarkers are less sensitive than PET biomarkers. [ABSTRACT FROM AUTHOR]

Background: Parental history of sporadic Alzheimer's disease (AD) is considered a prime risk factor since at least the late 1980s. Routinely used clinical samples of at‐risk individuals (8‐16 subjects) may conceal weaker population‐based effects, especially those of paternal inheritance. We performed a phenome‐wide examination of the dissociable influences of maternal vs. paternal AD risk on male vs. female probands in the largest single‐site family‐based at‐risk AD cohort: PREVENT‐AD. Method: We built sex‐specific PLS‐regression models in which the APOE genotypes (e.g., ɛ3/3, ɛ3/4, ɛ3/2) were estimated based on ∼1,000 patient visits, each covering ∼260 risk indicators, to derive three APOE‐driven intermediate phenotypes (IPs). We examined how much the three IPs vary with regard to sex and maternal vs. paternal AD lineage across the phenome. We assessed the sex‐specific and lineage‐specific variation of the IPs over a 4‐year follow‐up period. Lastly, we examined how matri‐ vs. patrilinear AD risk captured by the IPs were expressed in AD‐vulnerable brain structures. Result: Across IPs, we linked patrilinear AD risk to verbal‐numerical cognition and cardiovascular health in males. In contrast, we linked matrilinear AD risk to memory performance in both sexes. Over time, the mean difference in cognitive performance decreases with regard to sex but increases with regard to AD lineage. Cis‐ and trans‐generational sex effects in neocortex subregions (e.g., mPFC, PCu) were captured by the first and second IPs, respectively. The third IP mainly captured trans‐generational sex effects in hippocampus subregions (e.g., CA1, CA4). Conclusion: We uncovered IPs of AD susceptibility differently expressed in male and female probands and affected by the diagnosed parent's sex. Maternal inheritance highlighted memory performance in both sexes, whereas paternal inheritance was particularly linked to cardiovascular health in males. The inheritance of the IPs was reflected in the brain structure at both superficial and deeper layers of the cortex. As the first study of its kind, our cross‐generational analysis of matri‐ vs. patrilinear AD risk bridges the epidemiological and clinical literature by leveraging the power of ∼1,000 patient visits. Our completely data‐driven framework ultimately dissociated phenotypes of maternal and paternal AD risk single‐handedly expressed in male and female probands. [ABSTRACT FROM AUTHOR]

Background: People with subjective cognitive decline (SCD+) are thought to be at greater risk of Alzheimer's disease (AD). The Nucleus basalis of Meynert (NbM) has been shown to change early in AD (Fernández‐Cabello et al., 2020), and it is the main source of cholinergic projection to entorhinal cortex (EC) (Mesulam et al., 2013). We thus investigated if there was evidence of degeneration in the NbM, or EC in SCD+ participants compared to those without (SCD‐), and if there was a relationship between changes in these regions and SCD status using deformation‐based morphometry (DBM). Method: Data included baseline visit MRI scans of 232 cognitively healthy older adults (148 SCD‐; 84 SCD+) from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset (using only ADNI2). SCD status was extracted based on the cognitive change index (CCI) of the subjects reported in ADNIMERGE data (CCI ≥16). MRI resolution was increased to 0.5 mm isotropic voxel‐size before non‐linear registration to an ADNI‐based unbiased symmetric template. The resulting deformation fields were used to compute the Jacobian determinant map for each subject as a proxy for local volume. NbM, and EC masks were mapped to our high‐resolution template and mean determinant values inside these masks were estimated, and sex‐ and age‐corrected z‐scores were calculated. Two‐sample t‐tests were then used to compare the SCD+:SCD‐ differences. Result: Figure 1 shows the mean z‐score values for NbM and EC regions for both SCD+ and SCD‐ groups. Results for left NbM showed a significant difference between two groups (p =.007), but not for the right NbM (p =.12). However, z‐scored EC volumes did not show any statistically significant difference between groups for either left (p =.58) or right region (p =.41). Conclusion: Our results suggest that in SCD+ subjects, while the EC is still intact, neurodegeneration in the NbM appears to have already begun. Quantifying degeneration in the NbM may allow for early diagnosis of AD. Further research is needed to evaluate if such differences can be computed on an individual subject basis for use as a biomarker. [ABSTRACT FROM AUTHOR]

Background: One of the earliest pathological events in the course of AD is thought to be the degeneration of cholinergic neurons in the basal forebrain (Grothe et al. 2012). The largest cluster of cholinergic cells within the basal forebrain are found in the Nucleus basalis of Meynert (NbM) (Hampel et al. 2020). Studies show that loss of cholinergic projections due to cholinergic degeneration in NbM, is associated with the decline in cognitive abilities, specifically in memory and attention processing (Mesulam et al. 2013). However, identifying the NbM region on MR images is difficult due to limitations in the resolution and contrast of T1w images. Method: Data included the baseline MRI scans of 677 amyloid‐positive subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset from ADNI1, 2, and GO. Scan resolution was increased to 0.5 mm isotropic voxel size by super‐sampling (Manjón et al. 2010) before non‐linear registration to an ADNI‐based unbiased template. The resulting deformation fields were used to compute the Jacobian determinant map for each subject as a proxy for local volume. A voxel‐wise linear regression model analysis was performed on Jacobian maps to assess the pattern of volumetric change within an NbM mask according to diagnosis: DBM ∼ 1+Dx+AGE+Dx:AGE+SEX; where DBM are the Jacobian values for subjects, Dx is the categorical variable for disease stage (NC, eMCI, lMCI, AD) and Dx:AGE is an interaction term between diagnosis and age of each subject. Result: Figure 1 shows the statistically significant differences in local volume, comparing successive disease stages, within the NbM mask, after FDR correction (coronal section, A) NC vs eMCI, B) eMCI vs lMCI, C) lMCI vs AD). As it has been shown here, the Jacobian map (as a proxy for atrophy) becomes more pronounced as the disease advances. Conclusion: Our results show that NbM degeneration starts as early as the preliminary stages of mild cognitive impairment, and this change accelerates as the disease progresses. This points to MRI‐based measurements of NbM as potential biomarkers for early AD detection and as a marker of disease burden. [ABSTRACT FROM AUTHOR]

Background: Nucleus basalis of Meynert (NbM) is the main source of cholinergic projection to the cerebral cortex, including entorhinal cortex (EC). Both cholinergic cells of NbM and their receptive targets in EC show degeneration during early stages of Alzheimer's disease (AD). While precise delineation of NbM region on T1w scans is difficult due to limited spatial resolution and contrast, we sought to quantify the degeneration of NbM along the disease trajectory and confirm recent studies suggesting that NbM degeneration happens before degeneration in the EC using deformation‐based morphometry (DBM). Method: Data included MRI scans of 1473 subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Resolution was increased to 0.5 mm isotropic voxel‐size before non‐linear registration to an ADNI‐based unbiased symmetric template. The resulting deformation fields were used to compute the Jacobian determinant map for each subject as a proxy for local volume. NbM and EC masks were mapped to our high‐resolution template and mean determinant values inside these masks were calculated for groups with different baseline diagnosis, and sex‐ and age‐corrected z‐scores were calculated. Two‐sample t‐tests were then used to compare the scores between different groups. Result: Figure 1 shows the mean DBM score of both NbM and EC regions for Normal Controls (CN), subjects with early‐ and late‐stage mild cognitive impairment (EMCI and LMCI) and subjects with probable AD dementia. Results corrected for multiple comparisons showed a significant difference between CN and both EMCI and LMCI for left and right NbM. A significant difference was also observed between EMCI and LMCI in left NbM (p‐value = 0.0024). Z‐scored EC volumes did not show a significant difference between CN and EMCI groups; However, results from EMCI and LMCI differ significantly in both hemispheres. We also observed the expected differences between each group with AD in both NbM and EC (P<0.001). Conclusion: Our results show that while macroscopic changes in EC emerge as MCI subjects advance toward the later stages, NbM degeneration starts sooner, as early as preliminary stages of MCI. This points to MRI‐based measurements of NbM as potential biomarkers for early detection of AD and as a marker of disease burden. [ABSTRACT FROM AUTHOR]