Abstract
Background and Objectives Neuroimaging and biomarker studies in Alzheimer disease (AD) have shown well-characterized patterns of cortical thinning and altered biomarker concentrations of tau and β-amyloid (Aβ). However, earlier identification of AD has great potential to advance clinical care and determine candidates for drug trials. The extent to which AD risk markers relate to cortical thinning patterns in midlife is unknown. The first objective of this study was to examine cortical thickness change associated with genetic risk for AD among middle-aged military veterans. The second objective was to determine the relationship between plasma tau and Aβ and change in brain cortical thickness among veterans stratified by genetic risk for AD.
Methods Participants consisted of post-9/11 veterans (N = 155) who were consecutively enrolled in the Translational Research Center for TBI and Stress Disorders prospective longitudinal cohort and were assessed for mild traumatic brain injury (TBI) and posttraumatic disorder (PTSD). Genome-wide polygenic risk scores (PRSs) for AD were calculated using summary results from the International Genomics of Alzheimer's Disease Project. T-tau and Aβ40 and Aβ42 plasma assays were run using Simoa technology. Whole-brain MRI cortical thickness change estimates were obtained using the longitudinal stream of FreeSurfer. Follow-up moderation analyses examined the AD PRS × plasma interaction on change in cortical thickness in AD-vulnerable regions.
Results Higher AD PRS, signifying greater genetic risk for AD, was associated with accelerated cortical thickness change in a right hemisphere inferior parietal cortex cluster that included the supramarginal gyrus, angular gyrus, and intraparietal sulcus. Higher tau, but not Aβ42/40 ratio, was associated with greater cortical thickness change among those with higher AD PRS. Mild TBI and PTSD were not associated with cortical thickness change.
Discussion Plasma tau, particularly when combined with genetic stratification for AD risk, can be a useful indicator of brain change in midlife. Accelerated inferior parietal cortex changes in midlife may be an important factor to consider as a marker of AD-related brain alterations.
Glossary
- AD=
- Alzheimer disease;
- Aβ=
- β-amyloid;
- CV=
- coefficient of variation;
- DSM=
- Diagnostic and Statistical Manual of Mental Disorders;
- FWHM=
- full width half maximum;
- GWAS=
- genome-wide association study;
- LLOQ=
- lower limit of quantification;
- MP-RAGE=
- magnetization-prepared rapid gradient echo;
- mTBI=
- mild traumatic brain injury;
- PRS=
- polygenic risk score;
- p-tau=
- phosphorylated tau;
- PTSD=
- posttraumatic disorder;
- ROI=
- region of interest;
- TBI=
- traumatic brain injury;
- TRACTS=
- Translational Research Center for TBI and Stress Disorders;
- t-tau=
- total tau
Disease processes underlying late-onset Alzheimer disease (AD) begin decades before clinical symptoms appear,1 and few treatment options are available by the time a formal diagnosis is made. As such, identifying preclinical biomarkers and endophenotypes of AD is critical for early detection and intervention. To date, biomarker research involving in vivo measures of β-amyloid (Aβ) and pathologic tau has predominantly been conducted in older adulthood, closer to the time of conversion to AD. However, numerous failed therapeutic trials suggest that treatment needs to be implemented in high-risk individuals earlier in the disease course.2 As evidence has indicated that Aβ deposits occur as early as age 30 years and abnormal tau aggregates earlier in the lifespan than previously identified,3 there is growing recognition that markers of preclinical AD in midlife are necessary for understanding pathology at the earliest stages of disease.4 However, studying neuropathologic changes in the decades prior to dementia requires a sample at high risk for early pathology and endophenotypes that reflect AD pathology. Military service members may be a particularly vulnerable group given high comorbidity rates of disorders associated with dementia including traumatic brain injury (TBI), posttraumatic disorder (PTSD), depression, metabolic conditions, and sleep disorders.5,6 Post-9/11 era veterans with blast exposures show diffuse tau accumulation on PET,7 and the presence of multiple risk factors is associated with higher inflammation,8 lower cortical thickness,9 and functional disability,10 suggesting heightened future risk of dementia.
Our previous work identified factors associated with preclinical vulnerability to AD. In a cohort of young to middle-aged veterans, we found that the interaction of genetic risk for AD and mild TBI was associated with lower cortical thickness in brain regions that atrophy in early AD.11 Genetic risk for AD is an important vulnerability factor for conversion to AD. Because genetic variants do not change through the lifespan, analysis of risk variants offers a way to investigate AD-related variation in biomarker measurements in younger individuals. The APOE ε4 allele confers 7.5 times greater risk of developing AD in European ancestry cohorts.12 Beyond APOE, polygenic risk scores (PRSs), which are based on the cumulative effects of genetic variants across the genome, have gained traction as important predictors of AD and cognitive decline13 and explain more phenotypic variance than APOE alone.14 A concern of the prior literature, however, is that few studies have examined risk factors and intermediate outcomes longitudinally, and therefore, a key piece of evidence linking early risk markers with progressive neurodegenerative decline is lacking. Furthermore, there is a critical need to identify early biomarkers that improve on genetic prediction models. In older adulthood, CSF Aβ, phosphorylated tau (p-tau), and total tau (t-tau) are well-established markers of preclinical AD. With the invention of highly sensitive blood-based fluid assays that are less invasive and more cost-effective than CSF collection, plasma and serum-based measurements of Aβ and tau are promising in vivo precursors of neurodegenerative processes.15,16 Plasma measurements of the Aβ 42/40 ratio and tau have been associated with changes in brain regions implicated in AD pathology. However, the importance of these markers measured earlier in the neurodegenerative trajectory is unknown. Blood-based markers may be particularly useful if they can predict neuropathologic change in the brain measured by tools such as MRI, which is more accessible and less expensive and invasive than PET.
The goal of the current study was to characterize predictors of cortical thickness change over a 2-year period in a middle-aged sample who were at risk for negative brain outcomes. We first identified AD-vulnerable brain regions that declined over time as a function of genetic risk for AD, TBI, and their interaction. We next examined whether plasma markers of AD pathology (t-tau and Aβ42/40 ratio) were associated with change in cortical thickness in AD-vulnerable brain regions. Based on our previous cross-sectional work,11 we hypothesized that the combination of higher AD PRS scores and TBI would be associated with cortical thickness change in areas of the brain that are the first to reflect AD pathology, including the medial temporal lobes and parietal cortex.
Methods
Participants
Participants consisted of men (n = 142) and women (n = 13) post-9/11 veterans who were consecutively enrolled in the Translational Research Center for TBI and Stress Disorders (TRACTS) prospective longitudinal cohort, an ongoing study taking place at VA Boston Healthcare System. Detailed information about recruitment, inclusion/exclusion criteria, and measures in the battery has been previously described.19 Figure 1 provides a flow chart of inclusion and exclusion criteria and a description of missing data. Briefly, exclusion criteria included the following: (1) history of seizures or serious neurologic illnesses other than mild traumatic brain injury (mTBI); (2) moderate to severe TBI; (3) active homicidal and/or suicidal ideation with intent; (4) current diagnosis of bipolar or other psychotic disorder (except psychosis not otherwise specified that was due to trauma-related hallucinations) based on the Diagnostic and Statistical Manual of Mental Disorders (DSM20); and (5) cognitive disorder due to general medical condition other than TBI. Genetic analyses were restricted to genetically confirmed European ancestry participants.
Exclusion process for a priori exclusion criteria and missing data. Only participants with complete data for all variables were included in each analysis. Genetic analyses were restricted to genetically confirmed European ancestry participants. Aβ = β-amyloid; CV = coefficient of variation; TBI = traumatic brain injury; WB = whole brain.
All participants completed two 10-hour research sessions that occurred approximately 18 months apart. Each session consisted of psychiatric interviews, clinical interviews, self-report psychological questionnaires, and comprehensive neuropsychological and biological assessments.
Standard Protocol Approvals, Registrations, and Patient Consents
The Institutional Review Board at the VA Boston Healthcare System approved all research procedures, and participants provided written informed consent prior to participating in the study.
Demographics and Psychiatric Assessment
A demographic questionnaire was used to collect information related to sex, education, and general health. History of TBI was assessed using the Boston Assessment of TBI–Lifetime,21 which is a validated, semistructured interview. The Clinician-Administered PTSD Scale22 for DSM-IV was used to assess for a lifetime history of PTSD. The Structured Clinical Interview for DSM-IV Axis I disorders23 was used to determine mood, anxiety, substance use, psychotic, and other DSM-based mental health disorders. All clinical diagnoses were determined by consensus of at least 3 licensed clinical psychologists.
Genotyping and Polygenic Risk Score Computation
Genotyping in the TRACTS cohort has been described elsewhere and will only be briefly summarized here.24,25 The Illumina HumanOmni2.5–8 BeadChip was used for genotyping. Cleaning of genotype data, imputation based on the Thousand Genomes Phase 3 reference panel,26 and genetically based ancestry determination were performed using a consortium-developed pipeline.27,28
PRSs for a target cohort are computed as a weighted sum of single nucleotide variations (SNVs [formerly SNPs]) across the genome, with the beta (or log odds ratio in the case of dichotomous traits) used as the weights. Here, we computed PRSs for AD in the TRACTS sample based on the summary results of a European ancestry genome-wide association study (GWAS)29 in lieu of a more recent AD meta-analysis GWAS which included proxy (reported parental) cases30 that can affect odds ratio estimates.31 PRSs were computed using PRSice32 from the best-guess genotypes with an 80% certainty threshold. SNVs with a minor allele frequency of less than 1% or a missing rate of greater than 5% were excluded from PRS calculation. Clumping was performed based on PRSice's default parameters. PRSs are often computed based on a set of SNVs that pass a significance threshold. Here, to limit our multiple testing burden, we calculated our PRS using SNVs that were genome-wide significant (p < 5 × 10−8) based on a recent publication, indicating that AD risk is oligogenic rather than polygenic.33 There were 72 SNVs used in the PRS calculation and are listed in eTable 1, links.lww.com/WNL/C601. Additional characteristics of the SNVs that comprised the risk score can be obtained by qualified researchers who apply at niagads.org. PRSs were standardized (mean 0 SD 1 scaled) prior to analysis to improve the interpretability of PRS effect size estimates.
Plasma-Based Biomarkers
Fasting blood draws were conducted between 7:00 am and 7:30 am. Plasma was collected in EDTA tubes, centrifuged, and then immediately frozen and stored at −70°C. To avoid multiple freeze-thaw cycles, tau (t-tau), Aβ 40, and Aβ 42 assays were conducted in batches. Plasma assays were run in duplicate using the ultra-sensitive Simoa technology on the automated high-definition 1 analyzer (Quanterix, Billerica, MA). T-tau was run as part of the neurology 4-plex assay, and Aβ 40 and Aβ 42 were run as part of the neurology 3-plex assay. Plasma samples were aliquoted into 96-well plates using the same template for each plate and diluted according to the manufactures' specifications. Data from plasma samples were excluded if the coefficient of variation (CV) between duplicate runs was over 20% and/or the measured analyte concentrations were below the lower limit of quantification (LLOQ). The average CV for t-tau was 7.96% (SD = 5.08), and the LLOQ was 0.024. Plasma concentrations of t-tau were not normally distributed and were log transformed for statistical analysis. The average CV for Aβ 40 was 3.33% (SD = 3.76) and the LLOQ was 1.90, and the average CV for Aβ 42 was 5.50% (SD = 4.57) and the LLOQ was 0.225. The Aβ42/40 ratio was calculated and used for all subsequent analyses.
MRI Acquisition and Processing
Neuroimaging data were acquired on a 3-T Siemens (Erlangen, Germany) TIM Trio scanner and after scanner upgrade with a MAGNETOM Prismafit scanner (Trio-upgrade). Data for 78 participants were acquired with the TIM Trio for both baseline and follow-up time points, 39 were acquired with a MAGNETOM Prismafit scanner for both time points, and 38 were acquired across scanner upgrade. TIM Trio data were collected with a 12-channel phased-array head coil, and the Prismafit data were collected with a 20-channel head coil. The acquisition on both scanners included 2 high-resolution whole-brain T1-weighted images using magnetization-prepared rapid gradient echo (MP-RAGE) volumes with approximately matched parameters across upgrade (Trio: TR/TE = 2.53 s/3.32 ms, Prismafit: TR/TE = 2.53 s/3.35 ms, flip angle = 7 deg, and 1 mm isotropic). Scans were 3D sagittal acquisitions with 176 contiguous slices (imaging matrix = 256 × 176, in-plane resolution = 1 mm, and slice thickness = 1 mm). To create a single high contrast-to-noise image, the 2 MP-RAGE volumes were averaged.
The FreeSurfer image analysis suite (version 7.1; downloadable at; surfer.nmr.mgh.harvard.edu) was used to perform automated cortical reconstruction and volumetric segmentation. Participant data are run through a standard processing stream that includes motion correction, averaging volumetric T1-weighted images, removal of nonbrain tissue using a hybrid watershed/surface deformation procedure, automated Talairach transformation, segmentation of the subcortical white matter and deep gray matter volumetric structures, intensity normalization, tessellation of the gray and white matter boundary, automated topology correction, and surface deformation to optimally place the gray/white and gray/CSF borders at the location where the greatest shift in intensity defines the transition to the other tissue class. Every image was manually inspected and edited for accuracy of the gray/white and gray/pial surfaces and run through a second reconstruction that began at the point where edits were applied.
Longitudinal cortical thickness change metrics were estimated using the longitudinal stream of FreeSurfer version 7.1 as described in our previous work.34 Briefly, an unbiased within-subject template space and image was created based on cross-sectional images from the 2 time points for each participant using robust, inverse consistent registration. The within-subject template is used to impose consistent segmentation across time points, reducing confounds associated with longitudinal analysis. Surface maps were resampled, transformed via spherical registration to a common surface space, and smoothed using a circulatory symmetric Gaussian kernel with a full width half maximum (FWHM) of 20 mm.35
Statistical Analyses
Whole-Brain Analysis
Vertex-wise general linear models were performed using FreeSurfer 7.1, and statistical analyses and plots were produced using R (version 4.0.4). To limit the interindividual variability in scan intervals, longitudinal change in cortical thickness for each hemisphere was calculated as an annualized measure of the symmetrized percent change (i.e., the annual rate of change with respect to the average thickness measure across the 2 time points). Statistical correction for multiple comparisons was performed using a Monte Carlo simulation-based clusterwise procedure to determine the distribution of the maximum cluster size under the null hypothesis. Five thousand iterations of simulations were performed, using a threshold of p < 0.050 and an FWHM of 20 for each analysis. Whole-brain analyses examined the relationship between annual percent change of cortical thickness and main and interaction effects of AD PRS, mild TBI, and PTSD, adjusting for age. Regions of interest (ROIs) that survived multiple comparisons correction were extracted from FreeSurfer output and used for plasma biomarker analyses.
Plasma Biomarker Analysis
A subset of participants had blood biomarker data available (t-tau N = 101; Aβ42/40 ratio N = 122; Figure 1). Hierarchical linear regression and simple slopes analyses were conducted using the pequod package in R. Residualized mean annual percent change of the thickness within the significant ROI from the whole-brain analysis was calculated for all participants and used as the dependent variable. All predictor variables were mean centered before entering the model. Hierarchical regression models were conducted to examine the interaction between AD PRS and plasma-based biomarker (t-tau or Aβ42/40 ratio) on annual percent change of the significant ROI. For each hierarchical regression biomarker analysis, AD PRS and plasma-based biomarker were entered in step 1 to examine main effects, and the AD PRS by biomarker interaction was added in step 2 to determine whether a significant amount of additional variance was accounted for by the interaction. Mahalanobis distance scores were generated from the moderation model. Distance scores that exceeded the critical χ2 value for 3 degrees of freedom (χ2 = 16.27, a = 0.001) were considered outliers. Three participants met this criterion and were excluded from moderation analyses for t-tau, and 6 individuals met this criterion for the Aβ42/40 ratio. Simple slope analysis was conducted to decompose interaction effects.
Data Availability
GWAS summary statistics for AD can be downloaded with permission from NIAGADS: niagads.org. Deidentified data may be shared via DUA by contacting the TRACTS program.
Results
Participants
Demographic characteristics of the entire sample are reported in Table 1. Participants were divided into high (>−0.06) and low (<−0.06) AD PRS risk groups by median split to assess differences in demographic and covariate variables according to AD risk. Independent sample t tests revealed no group differences in age, t(153) = 0.72, p = 0.475, interval since baseline visit, t(153) = −0.68, p = 0.501, or years of education, t(153) = −1.84, p = 0.067. Chi-squared tests revealed no significant differences between risk groups in regard to sex, χ2 (1) = 0.49, p = 0.484, history of mTBI, χ2 (1) = 0.18, p = 0.672, lifetime PTSD diagnosis, χ2 (1) = 2.04, p = 0.153, or group by scanner pairings (Trio-Trio, Prism-Prisma, and Trio-Prisma), χ2 (2) = 0.95, p = 0.620.
Participant Demographics as a Function of Polygenic Risk for Alzheimer Disease
Whole-Brain Analysis
General linear model analysis revealed that there was no mild TBI × AD PRS or PTSD × AD PRS interaction on cortical thickness change. However, there was a main effect of AD PRS after adjusting for age, mild TBI, and PTSD in a right hemisphere inferior parietal cortex cluster that included the supramarginal gyrus (peak voxel Montreal Neurological Institute coordinates = 59.73, −37.81, 22.74), angular gyrus, and intraparietal sulcus. Figure 2 provides significance map and scatterplot of cluster scores plotted by PRS for the AD risk group. Although additional regions near the temporal pole, superior frontal gyrus, and occipital parietal boundary showed nominal associations with higher AD PRS (Figure 2A and eTable 2, links.lww.com/WNL/C602), the inferior parietal cluster was the only region to survive multiple comparisons correction (Figure 2B).
(A) Results from uncorrected significance map for whole-brain analysis examining the relationship between AD polygenic risk and annualized percent change in cortical thickness. (B) Significance map after clusterwise correction shows cortical thickness change in right IPC with higher polygenic risk for AD. AD = Alzheimer disease; IPC = inferior parietal cortex.
Plasma Biomarker Analysis
We next determined the influence of plasma biomarkers (t-tau and Aβ42/40 ratio) on the relationship between AD PRS and cortical thickness change in the region identified in the whole-brain analysis (i.e., the inferior parietal cortex cluster). Table 2 provides model statistics, and eTable 3, links.lww.com/WNL/C603, provides demographic information for participants in the biomarker analysis. There was a significant AD PRS by t-tau interaction on cortical thickness change in the inferior parietal cortex (β = −0.266, p = 0.006). To decompose the interaction, a simple slope analysis was conducted, in which the relationship between t-tau and inferior parietal thickness change was assessed at 1 SD above and below the mean of AD PRS (Figure 3). The relationship between t-tau and inferior parietal thickness was significant at high AD PRS (β = −0.674, p = 0.004) but not low AD PRS (β = 0.360, p = 0.173). These results indicate that among those with higher genetic risk for AD, greater concentrations of t-tau were associated with decreased cortical thickness over time in the inferior parietal cortex.
Regression Model for AD Polygenic Risk Score by Tau Interaction on Change in Cortical Thickness in the Inferior Parietal Cluster
Annualized % change values represent residualized scores adjusting for mTBI, PTSD, and age. AD = Alzheimer disease; PTSD = posttraumatic disorder.
By contrast, there was no main effect of Aβ42/40 (β = 0.009, p = 0.919) on inferior parietal cortex thinning change, nor was there an Aβ42/40 × AD PRS interaction (β = 0.006, p = 0.951). eTable 4, links.lww.com/WNL/C604, provides full regression results for the Aβ42/40 ratio model and the results for Aβ42 and Aβ40 separately.
Discussion
The objective of this study was to examine midlife biomarkers of longitudinal cortical thickness change in regions of the brain vulnerable to AD, toward the long-term goal of identifying early risk factors for AD. There were 2 main findings. First, in middle-aged veterans, higher genetic risk for AD was associated with accelerated cortical thickness change in the inferior parietal cortex, a region that has previously shown cortical thinning in early AD.36 Second, genetic risk for AD moderated the relationship between plasma tau and accelerated change in the inferior parietal cortex. Contrary to our expectation, mild TBI did not predict cortical thickness change.
Identifying risk factors for AD decades prior to its onset requires a longitudinal approach to follow progression of neurodegeneration. However, most longitudinal studies of dementia risk do not adequately capture disease processes in midlife.4 Here, we examined longitudinal change in cortical thickness in a midlife sample of US military veterans. After implementing a conservative whole-brain voxel-wise approach, we observed that PRS for AD predicted accelerated cortical thickness change in the inferior parietal cortex, providing evidence for this region as an early vulnerability site for AD. The inferior parietal cortex cluster found here includes the supramarginal gyrus, angular gyrus, and intraparietal sulcus. The inferior parietal cortex is a key polymodal association integration area that is involved in attention processing and semantic and episodic memory retrieval.37,38 Although it shows large effects sizes in cortical thinning in early and mild AD39 and declines with disease progression,40 the inferior parietal cortex has not often been regarded as a site for early AD pathology. Braak staging protocols emphasize tau pathology in the medial temporal lobe regions, beginning in the entorhinal cortex and progressing to the visual cortex, with more recent work suggesting that tau may originate in regions such as the locus coeruleus prior to the entorhinal cortex.3 However, the inferior parietal cortex is among the neocortical regions that are first to show Aβ deposits in studies of Aβ staging.41 Aβ deposits are hypothesized to propagate forward to regions receiving neuronal inputs from the neocortex, eventually spreading to the medial temporal lobes.41 Neuroimaging studies have provided converging evidence that the inferior parietal cortex serves as a network hub that propagates Aβ progression,42 implicating its involvement in early AD-related changes. Although we cannot ascertain the mechanism of accelerated cortical thickness change in the inferior parietal cortex observed here, it will be important for future work to determine a link between genetic risk and AD neuropathology in this region.
With the growing emphasis on minimally invasive and inexpensive biomarkers to improve early diagnosis of AD, we examined the extent to which peripheral markers of AD pathology (i.e., tau and Aβ) could inform changes in cortical thickness. Plasma tau, but not Aβ, interacted with AD PRS to predict greater cortical thickness change. Tau protein in CNS neurons can travel across the blood-brain barrier, and tau in blood appears to originate in the CNS, making it a suitable peripheral biomarker of CNS pathology.43 Although results are not consistently found,44 plasma t-tau has shown promise as a preclinical marker of AD.15,45 A recent meta-analysis showed that plasma tau was elevated among those with AD, making it a suitable biomarker for AD.46 The relationship between plasma tau and cortical thickness in midlife has not been widely studied; however, the results found here are consistent with prior work showing that higher CSF tau levels were associated with lower inferior parietal cortex thickness among those with AD.47 Of interest, the GWAS used to compute the AD PRS in this study implicated pathways associated with tau protein binding,48 suggesting a synergistic genetic pathway to accelerated cortical change. Although further work in this area is necessary, the results provide support for the value of measuring genetic risk for AD and plasma tau as predictors of preclinical AD endophenotypes.
Although we expected to observe a relationship between plasma Aβ42/40 ratio and cortical thickness change, the nonsignificant results may suggest a more complex relationship between plasma Aβ and cortical thickness in midlife. In older adults, plasma Aβ42/40 ratio has been associated with Aβ and tau deposition in the inferior parietal cortex and medial and lateral temporal lobes.18 However, growing evidence suggests that concentrations of Aβ follow a nonlinear relationship with age in preclinical AD,49 with initial higher concentrations of Aβ followed by a rapid decrease in Aβ concentrations following its deposition in the neocortex. This nonlinear relationship suggests that measuring longitudinal changes in the Aβ ratio may better predict cortical thinning. Alternatively, it is possible that plasma Aβ is an unreliable marker of Aβ in the CNS. Aβ in blood originates from both the CNS and the periphery including the liver, kidney, intestine, and platelets and is inconsistently associated with CSF Aβ.50 It is possible that the plasma Aβ ratio may primarily reflect non-CNS processes and thus less likely to be associated with cortical thinning in midlife.
Although our previous work demonstrated lower cortical thickness among those with mild TBI and AD PRS in a cross-sectional study,11 we did not observe a change in cortical thickness as a function of the mild TBI × genetic interaction. One possibility is that the 2-year timeframe was too brief to observe changes associated with mild TBI in midlife. Our ongoing work with these veterans at additional time points will be necessary to clarify the null findings.
Limitations of the study should be noted. With recent work suggesting that plasma p-tau may be a more specific marker of AD than t-tau, future work will be important to determine the optimal set of plasma markers for prediction of disease. Another limitation of examining participants in midlife is that we are unable to determine the extent to which accelerated cortical thickness will predict AD conversion, as disease symptoms will not become apparent for participants until later in their lifespan. Additional work that includes greater time since the baseline assessment will also be necessary to determine whether the change in cortical thickness continues to decline and is associated with neurodegeneration. Finally, the veteran sample examined here was primarily male and White, non-Hispanic, limiting generalizability to a broader population.
In conclusion, the current study provides support for PRS for AD and plasma tau as risk factors for accelerated cortical thinning in the inferior parietal cortex, an area that has been recognized as a neocortical hub of incipient AD pathology. The results underscore the importance of the parietal cortex in identification of higher risk for neurodegeneration, decades prior to the onset of AD. Future work considering blood biomarkers and MRI may reduce the need for highly invasive, painful, and expensive biomarker methods such as CSF collection and PET imaging.
Study Funding
This work was supported by the NIH, NIA grant number R01AG058822 to J.P. Hayes. This research was also supported in part by the Department of Veterans Affairs by the Translational Research Center for TBI and Stress Disorders (TRACTS), a VA Rehabilitation Research and Development Traumatic Brain Injury Center of Excellence (B9254-C).
Disclosure
The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/NG.
Appendix Authors
Footnotes
Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/NG.
The Article Processing Charge was funded by The Ohio State University.
Submitted and externally peer reviewed. The handling editor was Associate Editor Suman Jayadev, MD.
- Received September 8, 2022.
- Accepted in final form November 21, 2022.
- Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology.
This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
