Alzheimer's Disease Neuroimaging Initiative biomarkers as quantitative phenotypes: Genetics core aims, progress, and plans

Association of APOC1 with cortical atrophy during conversion to Alzheimer's disease

Alzheimer's disease (AD) is a prevalent neurodegenerative disorder, with its progression influenced by aberrant gene expression and alterations in the brain network topology. Although APOE has been extensively studied in relation to AD, the role of APOC1 remains relatively underexplored. This study investigated the impact of APOC1 on changes in cortical thickness (CTh) during conversion to AD in a longitudinal setting. Using a normative modeling approach, we examined changes in CTh in patients with mild cognitive impairment (MCI). The spatial patterns of CTh changes were then correlated with APOC1 mRNA expression levels. We estimated the time to conversion to AD and compared progression rates between the low and high APOC1 expression groups. Finally, mediation analysis was performed to assess the indirect effects of APOC1 expression on memory function via CTh changes. In patients with MCI and AD, reduced CTh was observed in the limbic and default mode regions, with a notable impact on the entorhinal cortex, parahippocampus, and fusiform gyrus when comparing baseline and follow-up measurements. The degree of change in CTh was significantly associated with APOC1 expression, with the paralimbic regions identified as particularly vulnerable. Furthermore, the high APOC1 expression group demonstrated more rapid conversion to AD than that observed in the low expression group. Mediation analysis indicated a trend suggesting that APOC1 expression indirectly affected memory and cognitive function through its influence on CTh. These results highlight the potential of APOC1 as an additional focus of AD research, offering insights into the genetic influences on AD pathology.

Identification of Alzheimer's disease susceptibility genes by the integration of genomics and transcriptomics

BACKGROUND

Alzheimer's disease (AD) is a progressive neurodegenerative disease. With the deepening of clinical and genomic research, a series of biomarkers and risk factors related to AD have been identified. However, the exact molecular mechanism of AD is not completely understood.

METHODS

By combining expression quantitative trait loci (eQTLs) analysis with the results of genome-wide association studies (GWAS), the candidate genes (CG) related to AD were screened out accurately. We identified that intersection genes of differentially expressed genes (DEGs) and CG are the key genes. Then, GO, KEGG, and GSEA were utilized for functional enrichment analysis. Finally, we predicted AD responses to immunotherapy by the single sample gene set enrichment analysis (ssGSEA).

RESULTS

A total of 253 DEGs were identified. The three key genes (VASP, SURF2, and TARBP1) were identified by taking the intersection of DEGs and CG. Through Mendelian randomization (MR) analysis, it was found that the risk of AD was significantly increased when VASP expression increased (OR = 0.1.046), while the risk of AD was significantly decreased when SURF2 (OR = 0.897) and TARBP1(OR = 0.920) expression increased. Subsequently, the functional analysis indicated that the core genes were mainly enriched in Leukocyte Transendothelial Migration, cGMP-PKG signaling pathway, and Rap1 signaling pathway. Through ssGSEA analysis showed that all three core genes were significantly related to M2 macrophages.

CONCLUSIONS

Three core genes were screened by integrating eQTLs data, GWAS data and transfer group data, and the potential mechanism of diagnosis and treatment of AD was revealed.

Beyond the eye: A relational model for early dementia detection using retinal OCTA images

Early detection of dementia, such as Alzheimer's disease (AD) or mild cognitive impairment (MCI), is essential to enable timely intervention and potential treatment. Accurate detection of AD/MCI is challenging due to the high complexity, cost, and often invasive nature of current diagnostic techniques, which limit their suitability for large-scale population screening. Given the shared embryological origins and physiological characteristics of the retina and brain, retinal imaging is emerging as a potentially rapid and cost-effective alternative for the identification of individuals with or at high risk of AD. In this paper, we present a novel PolarNet+ that uses retinal optical coherence tomography angiography (OCTA) to discriminate early-onset AD (EOAD) and MCI subjects from controls. Our method first maps OCTA images from Cartesian coordinates to polar coordinates, allowing approximate sub-region calculation to implement the clinician-friendly early treatment of diabetic retinopathy study (ETDRS) grid analysis. We then introduce a multi-view module to serialize and analyze the images along three dimensions for comprehensive, clinically useful information extraction. Finally, we abstract the sequence embedding into a graph, transforming the detection task into a general graph classification problem. A regional relationship module is applied after the multi-view module to explore the relationship between the sub-regions. Such regional relationship analyses validate known eye-brain links and reveal new discriminative patterns. The proposed model is trained, tested, and validated on four retinal OCTA datasets, including 1,671 participants with AD, MCI, and healthy controls. Experimental results demonstrate the performance of our model in detecting AD and MCI with an AUC of 88.69% and 88.02%, respectively. Our results provide evidence that retinal OCTA imaging, coupled with artificial intelligence, may serve as a rapid and non-invasive approach for large-scale screening of AD and MCI. The code is available at https://github.com/iMED-Lab/PolarNet-Plus-PyTorch, and the dataset is also available upon request.

Predicting the progression of MCI and Alzheimer's disease on structural brain integrity and other features with machine learning

Machine learning (ML) on structural MRI data shows high potential for classifying Alzheimer's disease (AD) progression, but the specific contribution of brain regions, demographics, and proteinopathy remains unclear. Using Alzheimer's Disease Neuroimaging Initiative (ADNI) data, we applied an extreme gradient-boosting algorithm and SHAP (SHapley Additive exPlanations) values to classify cognitively normal (CN) older adults, those with mild cognitive impairment (MCI) and AD dementia patients. Features included structural MRI, CSF status, demographics, and genetic data. Analyses comprised one cross-sectional multi-class classification (CN vs. MCI vs. AD dementia, n = 568) and two longitudinal binary-class classifications (CN-to-MCI converters vs. CN stable, n = 92; MCI-to-AD converters vs. MCI stable, n = 378). All classifications achieved 70-77% accuracy and 61-83% precision. Key features were CSF status, hippocampal volume, entorhinal thickness, and amygdala volume, with a clear dissociation: hippocampal properties contributed to the conversion to MCI, while the entorhinal cortex characterized the conversion to AD dementia. The findings highlight explainable, trajectory-specific insights into AD progression.

Conversion to Mild Cognitive Impairment and Alzheimer's Disease Dementia Related to Apathy, APOE Genotype and Antidepressant Use

ObjectiveApathy and APOE ε4 genotype are risk factors for developing Alzheimer's disease dementia (ADD). Antidepressant use is known to induce apathy. This study aimed to examine associations between APOE ε4, apathy, and antidepressant use with progression from cognitively normal (CN) to mild cognitive impairments (MCI), and MCI to ADD.MethodsParticipants aged 55-90 were recruited from the Alzheimer's Disease Neuroimaging Initiative. Participants were CN or had MCI at baseline and had completed at least 3 consecutive study visits. The NPI and NPI-Q apathy subscales were used to index the presence of apathy. Antidepressants used by participants included SSRIs, SNRIs, and AYTADs. Cox proportional hazards analyses examined the combined effects of apathy, APOE ε4 genotype, and antidepressant use on conversion from CN to MCI and from MCI to ADD.ResultsApathy and APOE ε4 were associated with increased risk of conversion along the CN-MCI-ADD continuum. Antidepressant use was associated with progression from MCI to ADD, and progression from CN to MCI in non-apathetic APOE ε4 carriers.ConclusionOur findings support apathy and APOE ε4 as robust predictors of conversion to MCI and ADD, and demonstrate novel associations between antidepressant use and conversion. Future research should explore whether antidepressant use in MCI and ADD causes apathetic symptoms or serves to index apathy/depression severity.

Neuropsychiatric symptoms in cognitive decline and Alzheimer's disease: biomarker discovery using plasma proteomics

BACKGROUND AND OBJECTIVES

Neuropsychiatric symptoms (NPS) are common in older people with cognitive impairment and Alzheimer's disease (AD). No biomarkers to detect the related pathology or predict the clinical evolution of NPS are available yet. This study aimed to identify plasma proteins that may serve as biomarkers for NPS and NPS-related clinical disease progression.

METHODS

A panel of 190 plasma proteins was quantified using Luminex xMAP in the Alzheimer's Disease Neuroimaging Initiative cohort. NPS and cognitive performance were assessed at baseline and after 1 and 2 years. Logistic regression, receiver operating characteristic analysis and cross-validation were used to address the relations of interest.

RESULTS

A total of 507 participants with mild cognitive impairment (n=396) or mild AD dementia (n=111) were considered. Selected plasma proteins improved the prediction of NPS (area under the curve (AUC) from 0.61 to 0.76, p<0.001) and future NPS (AUC from 0.63 to 0.80, p<0.001) when added to a reference model. Distinct protein panels were identified for single symptoms. Among the selected proteins, ANGT, CCL1 and IL3 were associated with NPS at all three time points while CCL1, serum glutamic oxaloacetic transaminase and complement factor H were also associated with cognitive decline. The associations were independent of the presence of cerebral AD pathology as assessed using cerebrospinal fluid biomarkers.

CONCLUSIONS

Plasma proteins are associated with NPS and improve prediction of future NPS.

An imaging and genetic-based deep learning network for Alzheimer's disease diagnosis

Conventional computer-aided diagnostic techniques for Alzheimer's disease (AD) predominantly rely on magnetic resonance imaging (MRI) in isolation. Genetic imaging methods, by establishing the link between genes and brain structures in disease progression, facilitate early prediction of AD development. While deep learning methods based on MRI have demonstrated promising results for early AD diagnosis, the limited dataset size has led most AD studies to lean on statistical approaches within the realm of imaging genetics. Existing deep-learning approaches typically utilize pre-defined regions of interest and risk variants from known susceptibility genes, employing relatively straightforward feature fusion methods that fail to fully capture the relationship between images and genes. To address these limitations, we proposed a multi-modal deep learning classification network based on MRI and single nucleotide polymorphism (SNP) data for AD diagnosis and mild cognitive impairment (MCI) progression prediction. Our model leveraged a convolutional neural network (CNN) to extract whole-brain structural features, a Transformer network to capture genetic features, and employed a cross-transformer-based network for comprehensive feature fusion. Furthermore, we incorporated an attention-map-based interpretability method to analyze and elucidate the structural and risk variants associated with AD and their interrelationships. The proposed model was trained and evaluated using 1,541 subjects from the ADNI database. Experimental results underscored the superior performance of our model in effectively integrating and leveraging information from both modalities, thus enhancing the accuracy of AD diagnosis and prediction.

Multimodal diagnosis of Alzheimer's disease based on resting-state electroencephalography and structural magnetic resonance imaging

Multimodal diagnostic methods for Alzheimer's disease (AD) have demonstrated remarkable performance. However, the inclusion of electroencephalography (EEG) in such multimodal studies has been relatively limited. Moreover, most multimodal studies on AD use convolutional neural networks (CNNs) to extract features from different modalities and perform fusion classification. Regrettably, this approach often lacks collaboration and fails to effectively enhance the representation ability of features. To address this issue and explore the collaborative relationship among multimodal EEG, this paper proposes a multimodal AD diagnosis model based on resting-state EEG and structural magnetic resonance imaging (sMRI). Specifically, this work designs corresponding feature extraction models for EEG and sMRI modalities to enhance the capability of extracting modality-specific features. Additionally, a multimodal joint attention mechanism (MJA) is developed to address the issue of independent modalities. The MJA promotes cooperation and collaboration between the two modalities, thereby enhancing the representation ability of multimodal fusion. Furthermore, a random forest classifier is introduced to enhance the classification ability. The diagnostic accuracy of the proposed model can achieve 94.7%, marking a noteworthy accomplishment. This research stands as the inaugural exploration into the amalgamation of deep learning and EEG multimodality for AD diagnosis. Concurrently, this work strives to bolster the use of EEG in multimodal AD research, thereby positioning itself as a hopeful prospect for future advancements in AD diagnosis.

Association of genetic risk of Alzheimer's disease and cognitive function in two European populations

Although there is some evidence of an association between Alzheimer's disease polygenic risk score (AD PRS) and cognitive function, limited validations have been performed in large populations. We investigated the relationship between AD PRS and cognitive function in the UK Biobank in over 276,000 participants and further validated the association in the Alzheimer's Disease Neuroimaging Initiative (ADNI) sample. We developed the AD PRS (excluded the APOE variants) in the middle age UK Biobank participants (age ranged 39-72, mean age 57 years) of European ancestries by LDpred2. To validate the association of AD PRS and cognitive function internally in the UK Biobank, we linearly regressed standardized cognitive function on continuous standardized AD PRS with age at cognitive test, sex, genotyping array, first 10 principal components of genotyping, smoking, education in years, body mass index, and apolipoprotein E gene ε4 (APOE4) risk allele dosages. To validate the associations externally, we ran the linear mixed effects model in the ADNI sample free of dementia (age ranged 55-91, mean age 73), including similar covariates as fixed effects and participants' IDs as the random effect. Stratification by age, APOE4 carrier status, and cognitive status (cognitively normal or mild cognitive impairment) was also investigated. Our study validated the associations of AD PRS and cognitive function in both midlife and late-life observational cohorts. Although not all of the cognitive measures were significantly associated with AD PRS, non-verbal fluid reasoning (matrix pattern completion, β = - 0.022, P = 0.003), processing speed (such as symbol digit substitution, β = - 0.017, P = 1.08E-05), short-term memory and attention (such as pairs matching, β = - 0.014, P = 1.66E-10), and reaction time (β = - 0.010, P = 1.19E-06) were inversely associated with increasing AD PRS in the UK Biobank. Higher likelihood of cognitive impairment was also associated with higher AD PRS in the ADNI cognitive normal individuals (AD assessment scale β = 0.079, P = 0.02). In summary, we confirmed that poorer cognitive function was associated with a higher polygenic AD risk, and suggested the potential utility of the AD PRS in identifying those who may be at risk for further cognitive decline.

Machine Learning-Based Alzheimer's Disease Stage Diagnosis Utilizing Blood Gene Expression and Clinical Data: A Comparative Investigation

Background/Objectives: This study presents a comparative analysis of the multistage diagnosis of Alzheimer's disease (AD), including mild cognitive impairment (MCI), utilizing two distinct types of biomarkers: blood gene expression and clinical biomarker samples. Both of these samples, obtained from participants in the Alzheimer's Disease Neuroimaging Initiative (ADNI), were independently analyzed utilizing machine learning (ML)-based multiclassifiers. This study applied novel machine learning-based data augmentation techniques to gene expression profile data that are high-dimensional, low-sample-size (HDLSS) and inherently highly imbalanced. The investigation obtained the highest multiclassification performance to date in the multistage diagnosis of Alzheimer's disease utilizing the blood gene expression profiles of Alzheimer's Disease Neuroimaging Initiative (ADNI) participants. Based on the performance results obtained, and other factors such as early prediction capabilities, this study compares the efficacies of the two types of biomarkers for multistage diagnosis. This study presents the sole investigation in which multiclassification-based AD stage diagnosis was conducted utilizing blood gene expression data. We obtained the best multiclassification result in both modalities of the ADNI data in terms of F1-score and were able to identify new genetic biomarkers. Methods: The combination of the XGBoost and SFBS (Sequential Floating Backward Selection) methods was used to select the features. We were able to select the 95 most effective gene probe sets out of 49,386. For the clinical study data, eight of the most effective biomarkers were selected using SFBS. A deep learning (DL) classifier was used to identify the stages-cognitive normal (CN), mild cognitive impairment (MCI), and Alzheimer's disease (AD)/dementia. DL, support vector machine (SVM), gradient boosting (GB), and random forest (RF) classifiers were used for the AD stage detection from gene expression profile data. Because of the high data imbalance in genomic data, borderline oversampling/data augmentation was applied in the model training and original samples for validation. Results: Utilizing clinical data, the highest ROC AUC scores attained were 0.989, 0.927, and 0.907 for the identification of the CN, MCI, and dementia stages, respectively. The highest F1 scores achieved were 0.971, 0.939, and 0.886. Employing gene expression data, we obtained ROC AUC scores of 0.763, 0.761, and 0.706 for the CN, MCI, and dementia stages, respectively, and F1 scores of 0.71, 0.77, and 0.53 for CN, MCI, and dementia, respectively. Conclusions: This represents the best outcome to date for AD stage diagnosis from ADNI blood gene expression profile data utilizing multiclassification techniques. The results indicated that our multiclassification model effectively manages the imbalanced data of a high-dimension, low-sample-size (HDLSS) nature to identify samples of the minority class. MAPK14, PLG, FZD2, FXYD6, and TEP1 are among the novel genes identified as being associated with AD risk.

Exploring Brain Imaging and Genetic Risk Factors in Different Progression States of Alzheimer's Disease Through OSnetNMF-Based Methods

Alzheimer's disease (AD) is a neurodegenerative disease with no effective treatment, often preceded by mild cognitive impairment (MCI). Multimodal imaging genetics integrates imaging and genetic data to gain a deeper understanding of disease progression and individual variations. This study focuses on exploring the mechanisms that drive the transition from normal cognition to MCI and ultimately to AD. As an effective joint feature extraction and dimensionality reduction method, non-negative matrix factorization (NMF) and its improved variants, particularly the network-based non-negative matrix factorization (netNMF), have been widely used in multimodal analysis to mine brain imaging and genetic data by considering the interactions between different features. However, many of these methods overlook the importance of the coefficient matrix and do not address issues related to data accuracy and feature redundancy. To address these limitations, we propose an orthogonal sparse network non-negative matrix factorization (OSnetNMF) algorithm, which introduces orthogonal and sparse constraints based on netNMF. By establishing linear relationships between structural magnetic resonance imaging (sMRI) and corresponding gene expression data, OSnetNMF reduces feature redundancy and decreases correlation between data, resulting in more accurate and reliable biomarker extraction. Experiments demonstrate that the OSnetNMF algorithm can accurately identify risk regions of interest (ROIs) and key genes that characterize AD progression, revealing significant trends in ROI pairs such as l4thVen-HIF1A, rBst-MPO, and rBst-PTK2B. Comparative experiments show that the improved algorithm outperforms traditional methods, identifying more disease-related biomarkers and achieving better reconstruction performance.

Clinical Applications of PET Imaging in Alzheimer's Disease

Alzheimer's disease (AD) involves a complex pathophysiology of neurodegeneration that leads to severe cognitive deficiencies. Understanding the molecular alterations that underlie this disease is fundamental to clinical management and therapeutic innovation. Functional imaging with positron emission tomography (PET) enables a visualization of these impaired pathways, such as cerebral hypometabolism, amyloid and tau accumulation, and neurotransmitter dysfunction. This review discusses the clinical applications of PET in AD and mild cognitive impairment, focusing on relevant original research studies for disease diagnosis, progression, and treatment response.

Multi-modal Imaging-based Pseudotime Analysis of Alzheimer progression

Alzheimer's disease (AD) is a neurodegenerative disorder that results in progressive cognitive decline but without any clinically validated cures so far. Understanding the progression of AD is critical for early detection and risk assessment for AD in aging individuals, thereby enabling initiation of timely intervention and improved chance of success in AD trials. Recent pseudotime approach turns cross-sectional data into "faux" longitudinal data to understand how a complex process evolves over time. This is critical for Alzheimer, which unfolds over the course of decades, but the collected data offers only a snapshot. In this study, we tested several state-of-the-art pseudotime approaches to model the full spectrum of AD progression. Subsequently, we evaluated and compared the pseudotime progression score derived from individual imaging modalities and multi-modalities in the ADNI cohort. Our results showed that most existing pseudotime analysis tools do not generalize well to the imaging data, with either flipped progression score or poor separation of diagnosis groups. This is likely due to the underlying assumptions that only stand for single cell data. From the only tool with promising results, it was observed that all pseudotime, derived from either single imaging modalities or multi-modalities, captures the progressiveness of diagnosis groups. Pseudotime from multi-modality, but not the single modalities, confirmed the hypothetical temporal order of imaging phenotypes. In addition, we found that multi-modal pseudotime is mostly driven by amyloid and tau imaging, suggesting their continuous changes along the full spectrum of AD progression.

Application of a localized morphometrics approach to imaging-derived brain phenotypes for genotype-phenotype associations in pediatric mental health and neurodevelopmental disorders

Introduction

Quantitative global or regional brain imaging measurements, known as imaging-specific or -derived phenotypes (IDPs), are commonly used in genotype-phenotype association studies to explore the genomic architecture of the brain and how it may be affected by neurological diseases (e.g., Alzheimer's disease), mental health (e.g., depression), and neurodevelopmental disorders (e.g., attention-deficit hyperactivity disorder [ADHD]). For this purpose, medical images have been used as IDPs using a voxel-wise or global approach via principal component analysis. However, these methods have limitations related to multiple testing or the inability to isolate high variation regions, respectively.

Methods

To address these limitations, this study investigates a localized, principal component analysis-like approach for dimensionality reduction of cross-sectional T1-weighted MRI datasets utilizing diffeomorphic morphometry. This approach can reduce the dimensionality of images while preserving spatial information and enables the inclusion of spatial locality in the analysis. In doing so, this method can be used to explore morphometric brain changes across specific components and spatial scales of interest and to identify associations with genome regions in a multivariate genome-wide association study. For a first clinical feasibility study, this method was applied to data from the Adolescent Brain Cognitive Development (ABCD) study, including adolescents with ADHD (n = 1,359), obsessive-compulsive disorder (n = 1,752), and depression (n = 1,766).

Results

Meaningful associations of specific morphometric features with genome regions were identified with the data and corresponded to previous found brain regions in the respective mental health and neurodevelopmental disorder cohorts.

Discussion

In summary, the localized, principal component analysis-like approach can reduce the dimensionality of medical images while still being able to identify meaningful local brain region alterations that are associated with genomic markers across multiple scales. The proposed method can be applied to various image types and can be easily integrated in many genotype-phenotype association study setups.

Dementia risk scores, apolipoprotein E, and risk of Alzheimer's disease: One size does not fit all

INTRODUCTION

Evaluating the generalizability of dementia risk scores, primarily developed in non-Latinx White (NLW) participants, and interactions with genetic risk factors in diverse populations is crucial for addressing health disparities.

METHODS

We analyzed the association of the Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) and modified CAIDE (mCAIDE) scores with dementia risk using logistic regression models stratified by race/ethnicity in National Alzheimer's Coordinating Center (NACC) and Alzheimer's Disease Neuroimaging Initiative (ADNI), and assessed their interaction with apolipoprotein E (APOE).

RESULTS

Higher CAIDE scores were associated with an increased risk of dementia in Asian, Latinx, and NLW participants but not in Black participants. In contrast, higher mCAIDE scores were also associated with an increased risk of dementia in Black participants. Unfavorable mCAIDE risk profiles exacerbated the apolipoprotein E*ε4 (APOE*ε4) risk effect and attenuated the APOE*ε2 protective effect.

DISCUSSION

Our findings underscore the importance of evaluating the validity of dementia risk scores in diverse populations for their use in personalized medicine approaches to promote brain health.

HIGHLIGHTS

Dementia risk scores demonstrate race/ethnic-specific effects on dementia risk. Unfavorable modifiable risk profiles moderate the effect of APOE on dementia risk. Dementia risk scores need to be validated in diverse populations.

Deep joint learning diagnosis of Alzheimer's disease based on multimodal feature fusion

Alzheimer's disease (AD) is an advanced and incurable neurodegenerative disease. Genetic variations are intrinsic etiological factors contributing to the abnormal expression of brain function and structure in AD patients. A new multimodal feature fusion called "magnetic resonance imaging (MRI)-p value" was proposed to construct 3D fusion images by introducing genes as a priori knowledge. Moreover, a new deep joint learning diagnostic model was constructed to fully learn images features. One branch trained a residual network (ResNet) to learn the features of local pathological regions. The other branch learned the position information of brain regions with different changes in the different categories of subjects' brains by introducing attention convolution, and then obtained the discriminative probability information from locations via convolution and global average pooling. The feature and position information of the two branches were linearly interacted to acquire the diagnostic basis for classifying the different categories of subjects. The diagnoses of AD and health control (HC), AD and mild cognitive impairment (MCI), HC and MCI were performed with data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). The results showed that the proposed method achieved optimal results in AD-related diagnosis. The classification accuracy (ACC) and area under the curve (AUC) of the three experimental groups were 93.44% and 96.67%, 89.06% and 92%, and 84% and 81.84%, respectively. Moreover, a total of six novel genes were found to be significantly associated with AD, namely NTM, MAML2, NAALADL2, FHIT, TMEM132D and PCSK5, which provided new targets for the potential treatment of neurodegenerative diseases.

Integrating amyloid imaging and genetics for early risk stratification of Alzheimer's disease

INTRODUCTION

Alzheimer's disease (AD) initiates years prior to symptoms, underscoring the importance of early detection. While amyloid accumulation starts early, individuals with substantial amyloid burden may remain cognitively normal, implying that amyloid alone is not sufficient for early risk assessment.

METHODS

Given the genetic susceptibility of AD, a multi-factorial pseudotime approach was proposed to integrate amyloid imaging and genotype data for estimating a risk score. Validation involved association with cognitive decline and survival analysis across risk-stratified groups, focusing on patients with mild cognitive impairment (MCI).

RESULTS

Our risk score outperformed amyloid composite standardized uptake value ratio in correlation with cognitive scores. MCI subjects with lower pseudotime risk score showed substantial delayed onset of AD and slower cognitive decline. Moreover, pseudotime risk score demonstrated strong capability in risk stratification within traditionally defined subgroups such as early MCI, apolipoprotein E (APOE) ε4+ MCI, APOE ε4- MCI, and amyloid+ MCI.

DISCUSSION

Our risk score holds great potential to improve the precision of early risk assessment.

HIGHLIGHTS

Accurate early risk assessment is critical for the success of clinical trials. A new risk score was built from integrating amyloid imaging and genetic data. Our risk score demonstrated improved capability in early risk stratification.

Spectral Graph Sample Weighting for Interpretable Sub-cohort Analysis in Predictive Models for Neuroimaging

Recent advancements in medicine have confirmed that brain disorders often comprise multiple subtypes of mechanisms, developmental trajectories, or severity levels. Such heterogeneity is often associated with demographic aspects (e.g., sex) or disease-related contributors (e.g., genetics). Thus, the predictive power of machine learning models used for symptom prediction varies across subjects based on such factors. To model this heterogeneity, one can assign each training sample a factor-dependent weight, which modulates the subject's contribution to the overall objective loss function. To this end, we propose to model the subject weights as a linear combination of the eigenbases of a spectral population graph that captures the similarity of factors across subjects. In doing so, the learned weights smoothly vary across the graph, highlighting sub-cohorts with high and low predictability. Our proposed sample weighting scheme is evaluated on two tasks. First, we predict initiation of heavy alcohol drinking in young adulthood from imaging and neuropsychological measures from the National Consortium on Alcohol and NeuroDevelopment in Adolescence (NCANDA). Next, we detect Dementia vs. Mild Cognitive Impairment (MCI) using imaging and demographic measurements in subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Compared to existing sample weighting schemes, our sample weights improve interpretability and highlight sub-cohorts with distinct characteristics and varying model accuracy.

Connecting dementia risk loci to the CSF proteome identifies pathophysiological leads for dementia

Genome-wide association studies have successfully identified many genetic risk loci for dementia, but exact biological mechanisms through which genetic risk factors contribute to dementia remains unclear. Integrating CSF proteomic data with dementia risk loci could reveal intermediate molecular pathways connecting genetic variance to the development of dementia. We tested to what extent effects of known dementia risk loci can be observed in CSF levels of 665 proteins [proximity extension-based (PEA) immunoassays] in a deeply-phenotyped mixed memory clinic cohort [n = 502, mean age (standard deviation, SD) = 64.1 (8.7) years, 181 female (35.4%)], including patients with Alzheimer's disease (AD, n = 213), dementia with Lewy bodies (DLB, n = 50) and frontotemporal dementia (FTD, n = 93), and controls (n = 146). Validation was assessed in independent cohorts (n = 99 PEA platform, n = 198, mass reaction monitoring-targeted mass spectroscopy and multiplex assay). We performed additional analyses stratified according to diagnostic status (AD, DLB, FTD and controls separately), to explore whether associations between CSF proteins and genetic variants were specific to disease or not. We identified four AD risk loci as protein quantitative trait loci (pQTL): CR1-CR2 (rs3818361, P = 1.65 × 10-8), ZCWPW1-PILRB (rs1476679, P = 2.73 × 10-32), CTSH-CTSH (rs3784539, P = 2.88 × 10-24) and HESX1-RETN (rs186108507, P = 8.39 × 10-8), of which the first three pQTLs showed direct replication in the independent cohorts. We identified one AD-specific association between a rare genetic variant of TREM2 and CSF IL6 levels (rs75932628, P = 3.90 × 10-7). DLB risk locus GBA showed positive trans effects on seven inter-related CSF levels in DLB patients only. No pQTLs were identified for FTD loci, either for the total sample as for analyses performed within FTD only. Protein QTL variants were involved in the immune system, highlighting the importance of this system in the pathophysiology of dementia. We further identified pQTLs in stratified analyses for AD and DLB, hinting at disease-specific pQTLs in dementia. Dissecting the contribution of risk loci to neurobiological processes aids in understanding disease mechanisms underlying dementia.

Utility of cerebrovascular imaging biomarkers to detect cerebral amyloidosis

INTRODUCTION

The relationship between cerebrovascular disease (CVD) and amyloid beta (Aβ) in Alzheimer's disease (AD) is understudied. We hypothesized that magnetic resonance imaging (MRI)-based CVD biomarkers-including cerebral microbleeds (CMBs), lacunar infarction, and white matter hyperintensities (WMHs)-would correlate with Aβ positivity on positron emission tomography (Aβ-PET).

METHODS

We cross-sectionally analyzed data from the Alzheimer's Disease Neuroimaging Initiative (ADNI, N = 1352). Logistic regression was used to calculate odds ratios (ORs), with Aβ-PET positivity as the standard-of-truth.

RESULTS

Following adjustment, WMHs (OR = 1.25) and superficial CMBs (OR = 1.45) remained positively associated with Aβ-PET positivity (p < 0.001). Deep CMBs and lacunes exhibited a varied relationship with Aβ-PET in cognitive subgroups. The combined diagnostic model, which included CVD biomarkers and other accessible measures, significantly predicted Aβ-PET (pseudo-R2 = 0.41).

DISCUSSION

The study highlights the translational value of CVD biomarkers in diagnosing AD, and underscores the need for more research on their inclusion in diagnostic criteria.

CLINICALTRIALS

gov: ADNI-2 (NCT01231971), ADNI-3 (NCT02854033).

HIGHLIGHTS

Cerebrovascular biomarkers linked to amyloid beta (Aβ) in Alzheimer's disease (AD). White matter hyperintensities and cerebral microbleeds reliably predict Aβ-PET positivity. Relationships with Aβ-PET vary by cognitive stage. Novel accessible model predicts Aβ-PET status. Study supports multimodal diagnostic approaches.

Connecting genomic and proteomic signatures of amyloid burden in the brain

Background

Alzheimer's disease (AD) has a high heritable component characteristic of complex diseases, yet many of the genetic risk factors remain unknown. We combined genome-wide association studies (GWAS) on amyloid endophenotypes measured in cerebrospinal fluid (CSF) and positron emission tomography (PET) as surrogates of amyloid pathology, which may be helpful to understand the underlying biology of the disease.

Methods

We performed a meta-analysis of GWAS of CSF Aβ42 and PET measures combining six independent cohorts (n=2,076). Due to the opposite effect direction of Aβ phenotypes in CSF and PET measures, only genetic signals in the opposite direction were considered for analysis (n=376,599). Polygenic risk scores (PRS) were calculated and evaluated for AD status and amyloid endophenotypes. We then searched the CSF proteome signature of brain amyloidosis using SOMAscan proteomic data (Ace cohort, n=1,008) and connected it with GWAS results of loci modulating amyloidosis. Finally, we compared our results with a large meta-analysis using publicly available datasets in CSF (n=13,409) and PET (n=13,116). This combined approach enabled the identification of overlapping genes and proteins associated with amyloid burden and the assessment of their biological significance using enrichment analyses.

Results

After filtering the meta-GWAS, we observed genome-wide significance in the rs429358-APOE locus and nine suggestive hits were annotated. We replicated the APOE loci using the large CSF-PET meta-GWAS and identified multiple AD-associated genes as well as the novel GADL1 locus. Additionally, we found a significant association between the AD PRS and amyloid levels, whereas no significant association was found between any Aβ PRS with AD risk. CSF SOMAscan analysis identified 1,387 FDR-significant proteins associated with CSF Aβ42 levels. The overlap among GWAS loci and proteins associated with amyloid burden was very poor (n=35). The enrichment analysis of overlapping hits strongly suggested several signalling pathways connecting amyloidosis with the anchored component of the plasma membrane, synapse physiology and mental disorders that were replicated in the large CSF-PET meta-analysis.

Conclusions

The strategy of combining CSF and PET amyloid endophenotypes GWAS with CSF proteome analyses might be effective for identifying signals associated with the AD pathological process and elucidate causative molecular mechanisms behind the amyloid mobilization in AD.

A Risk Variant rs6922617 in TREM Is Discrepantly Associated With Defining Neuropathological Hallmarks in the Alzheimer's Continuum

The single nucleotide polymorphism (SNP)-rs6922617 in the triggering receptor expressed on myeloid cells (TREM) gene cluster is a potential risk factor for Alzheimer's disease (AD). Here, we examined whether rs6922617 is associated with AD-defining neuropathological hallmarks and memory performance. We assessed the interaction between the variant rs6922617 and levels of beta-amyloid (Aβ), tau pathology, neurodegeneration, namely amyloid-tau-neurodegeneration framework, and cognition functions in 660 healthy controls, 794 mild cognitively impaired, and 272 subjects with AD. We employed linear regression and linear mixed models to examine the association. Here we find that the SNP-rs6922617 in the TREM gene cluster is associated with a higher global amyloid-ligands positron emission tomography (Aβ-PET) burden and lower fluorodeoxyglucose positron emission tomography (FDG-PET) load. Interestingly, rs6922617 risk allele carriers exhibit a significantly reduced tau accumulation compared to the non-carriers, indicating a discrepant association with Aβ and tau pathologies. Though the participants carrying the rs6922617 risk allele do not show a correlation with poorer cognitive performance, stronger neuropathological phenotypes, and memory impairments are evident in ApoE ε4 carriers with the rs6922617 risk allele. These results support the notion that the SNP-rs6922617 in the TREM gene cluster is associated with AD-related neuropathological hallmarks, such as Aβ and FDG-mediated neurodegeneration, rather than tau accumulation. Although the direct association with memory impairment in the Alzheimer's continuum remains inconclusive, our findings suggest a potential role of rs6922617 in facilitating neuropathology hallmarks.

Multimodal analysis of disease onset in Alzheimer's disease using Connectome, Molecular, and genetics data

Alzheimer's disease (AD) and its related age at onset (AAO) are highly heterogeneous, due to the inherent complexity of the disease. They are affected by multiple factors, such as neuroimaging and genetic predisposition. Multimodal integration of various data types is necessary; however, it has been nontrivial due to the high dimensionality of each modality. We aimed to identify multimodal biomarkers of AAO in AD using an extended version of sparse canonical correlation analysis, in which we integrated two imaging modalities, functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), and genetic data in the form of single-nucleotide polymorphisms (SNPs) obtained from the Alzheimer's disease neuroimaging initiative database. These three modalities cover low-to-high-level complementary information and offer multiscale insights into the AAO. We identified multivariate markers of AAO in AD using fMRI, PET, and SNP. Furthermore, the markers identified were largely consistent with those reported in the existing literature. In particular, our serial mediation analysis suggests that genetic variants influence the AAO in AD by indirectly affecting brain connectivity by mediation of amyloid-beta protein accumulation, supporting a plausible path in existing research. Our approach provides comprehensive biomarkers related to AAO in AD and offers novel multimodal insights into AD.

Dementia Risk Scores, APOE, and risk of Alzheimer disease: one size does not fit all

Introduction

Evaluating the generalizability of dementia risk scores, primarily developed in non-Latinx White (NLW) participants, and interactions with genetic risk factors in diverse populations is crucial for addressing health disparities.

Methods

We analyzed the association of the Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) and modified CAIDE (mCAIDE) scores with dementia risk using logistic regression models stratified by race/ethnicity in NACC and ADNI, and assessed their interaction with APOE.

Results

Higher CAIDE scores were associated with an increased risk of dementia in Asian, Latinx, and NLW participants but not in Black participants. In contrast, higher mCAIDE scores were also associated with an increased risk of dementia in Black participants. Unfavorable mCAIDE risk profiles exacerbated the APOE*ε4 risk effect and attenuated the APOE*ε2 protective effect.

Discussion

Our findings underscore the importance of evaluating the validity of dementia risk scores in diverse populations for their use in personalized medicine approaches to promote brain health.

Towards cascading genetic risk in Alzheimer's disease

Alzheimer's disease typically progresses in stages, which have been defined by the presence of disease-specific biomarkers: amyloid (A), tau (T) and neurodegeneration (N). This progression of biomarkers has been condensed into the ATN framework, in which each of the biomarkers can be either positive (+) or negative (-). Over the past decades, genome-wide association studies have implicated ∼90 different loci involved with the development of late-onset Alzheimer's disease. Here, we investigate whether genetic risk for Alzheimer's disease contributes equally to the progression in different disease stages or whether it exhibits a stage-dependent effect. Amyloid (A) and tau (T) status was defined using a combination of available PET and CSF biomarkers in the Alzheimer's Disease Neuroimaging Initiative cohort. In 312 participants with biomarker-confirmed A-T- status, we used Cox proportional hazards models to estimate the contribution of APOE and polygenic risk scores (beyond APOE) to convert to A+T- status (65 conversions). Furthermore, we repeated the analysis in 290 participants with A+T- status and investigated the genetic contribution to conversion to A+T+ (45 conversions). Both survival analyses were adjusted for age, sex and years of education. For progression from A-T- to A+T-, APOE-e4 burden showed a significant effect [hazard ratio (HR) = 2.88; 95% confidence interval (CI): 1.70-4.89; P < 0.001], whereas polygenic risk did not (HR = 1.09; 95% CI: 0.84-1.42; P = 0.53). Conversely, for the transition from A+T- to A+T+, the contribution of APOE-e4 burden was reduced (HR = 1.62; 95% CI: 1.05-2.51; P = 0.031), whereas the polygenic risk showed an increased contribution (HR = 1.73; 95% CI: 1.27-2.36; P < 0.001). The marginal APOE effect was driven by e4 homozygotes (HR = 2.58; 95% CI: 1.05-6.35; P = 0.039) as opposed to e4 heterozygotes (HR = 1.74; 95% CI: 0.87-3.49; P = 0.12). The genetic risk for late-onset Alzheimer's disease unfolds in a disease stage-dependent fashion. A better understanding of the interplay between disease stage and genetic risk can lead to a more mechanistic understanding of the transition between ATN stages and a better understanding of the molecular processes leading to Alzheimer's disease, in addition to opening therapeutic windows for targeted interventions.

TA-RNN: an attention-based time-aware recurrent neural network architecture for electronic health records

MOTIVATION

Electronic health records (EHRs) represent a comprehensive resource of a patient's medical history. EHRs are essential for utilizing advanced technologies such as deep learning (DL), enabling healthcare providers to analyze extensive data, extract valuable insights, and make precise and data-driven clinical decisions. DL methods such as recurrent neural networks (RNN) have been utilized to analyze EHR to model disease progression and predict diagnosis. However, these methods do not address some inherent irregularities in EHR data such as irregular time intervals between clinical visits. Furthermore, most DL models are not interpretable. In this study, we propose two interpretable DL architectures based on RNN, namely time-aware RNN (TA-RNN) and TA-RNN-autoencoder (TA-RNN-AE) to predict patient's clinical outcome in EHR at the next visit and multiple visits ahead, respectively. To mitigate the impact of irregular time intervals, we propose incorporating time embedding of the elapsed times between visits. For interpretability, we propose employing a dual-level attention mechanism that operates between visits and features within each visit.

RESULTS

The results of the experiments conducted on Alzheimer's Disease Neuroimaging Initiative (ADNI) and National Alzheimer's Coordinating Center (NACC) datasets indicated the superior performance of proposed models for predicting Alzheimer's Disease (AD) compared to state-of-the-art and baseline approaches based on F2 and sensitivity. Additionally, TA-RNN showed superior performance on the Medical Information Mart for Intensive Care (MIMIC-III) dataset for mortality prediction. In our ablation study, we observed enhanced predictive performance by incorporating time embedding and attention mechanisms. Finally, investigating attention weights helped identify influential visits and features in predictions.

AVAILABILITY AND IMPLEMENTATION

https://github.com/bozdaglab/TA-RNN.

Utility of cerebrovascular imaging biomarkers to detect cerebral amyloidosis

INTRODUCTION

The relationship between cerebrovascular disease (CVD) and amyloid-β (Aβ) in Alzheimer disease (AD) is understudied. We hypothesized that magnetic resonance imaging (MRI)-based CVD biomarkers, including cerebral microbleeds (CMBs), ischemic infarction, and white matter hyperintensities (WMH), would correlate with Aβ positivity on positron emission tomography (Aβ-PET).

METHODS

We cross-sectionally analyzed data from the Alzheimer's Disease Neuroimaging Initiative (ADNI, N=1,352). Logistic regression was used to calculate odds ratios (ORs), with Aβ-PET positivity as the standard-of-truth.

RESULTS

Following adjustment, WMH (OR=1.25) and superficial CMBs (OR=1.45) remained positively associated with Aβ-PET positivity (p<.001). Deep CMBs and infarcts exhibited a varied relationship with Aβ-PET in cognitive subgroups. The combined diagnostic model, which included CVD biomarkers and other accessible measures, significantly predicted Aβ-PET (pseudo-R 2 =.41).

DISCUSSION

The study highlights the translational value of CVD biomarkers in diagnosing AD, and underscores the need for more research on their inclusion in diagnostic criteria. ClinicalTrials.gov: ADNI-2 ( NCT01231971 ), ADNI-3 ( NCT02854033 ).

Functional Neural Networks for High-Dimensional Genetic Data Analysis

Artificial intelligence (AI) is a thriving research field with many successful applications in areas such as computer vision and speech recognition. Machine learning methods, such as artificial neural networks (ANN), play a central role in modern AI technology. While ANN also holds great promise for human genetic research, the high-dimensional genetic data and complex genetic structure bring tremendous challenges. The vast majority of genetic variants on the genome have small or no effects on diseases, and fitting ANN on a large number of variants without considering the underlying genetic structure (e.g., linkage disequilibrium) could bring a serious overfitting issue. Furthermore, while a single disease phenotype is often studied in a classic genetic study, in emerging research fields (e.g., imaging genetics), researchers need to deal with different types of disease phenotypes. To address these challenges, we propose a functional neural networks (FNN) method. FNN uses a series of basis functions to model high-dimensional genetic data and a variety of phenotype data and further builds a multi-layer functional neural network to capture the complex relationships between genetic variants and disease phenotypes. Through simulations, we demonstrate the advantages of FNN for high-dimensional genetic data analysis in terms of robustness and accuracy. The real data applications also showed that FNN attained higher accuracy than the existing methods.

Statistical and Machine Learning Analysis in Brain-Imaging Genetics: A Review of Methods

Brain-imaging-genetic analysis is an emerging field of research that aims at aggregating data from neuroimaging modalities, which characterize brain structure or function, and genetic data, which capture the structure and function of the genome, to explain or predict normal (or abnormal) brain performance. Brain-imaging-genetic studies offer great potential for understanding complex brain-related diseases/disorders of genetic etiology. Still, a combined brain-wide genome-wide analysis is difficult to perform as typical datasets fuse multiple modalities, each with high dimensionality, unique correlational landscapes, and often low statistical signal-to-noise ratios. In this review, we outline the progress in brain-imaging-genetic methodologies starting from early massive univariate to current deep learning approaches, highlighting each approach's strengths and weaknesses and elongating it with the field's development. We conclude by discussing selected remaining challenges and prospects for the field.

Impact of APOE-ε alleles on brain structure and cognitive function in healthy older adults: A VBM and DTI replication study

BACKGROUND

The Apolipoprotein E (APOE) gene has been established in the Alzheimer's disease (AD) literature to impact brain structure and function and may also show congruent effects in healthy older adults, although findings in this population are much less consistent. The current study aimed to replicate and expand the multimodal approach employed by Honea et al. Structural magnetic resonance imaging (MRI), diffusion tensor imaging (DTI), and neuropsychological measures were used to investigate the impact of APOE-ε status on grey matter structure, white matter integrity, and cognitive functioning.

METHODS

Data were obtained from the Alzheimer's Disease Initiative Phase 3 (ADNI3) database. Baseline MRI, DTI and cognitive composite scores for memory (ADNI-Mem) and executive function (ADNI-EF) were acquired from 116 healthy controls. Participants were grouped according to APOE allele presence (APOE-ε2+ N = 17, APOE-ε3ε3 N = 64, APOE-ε4+ N = 35). Voxel-based morphometry (VBM) and tract-based spatial statistics (TBSS) were used to compare grey matter volume (GMV) and white matter integrity, respectively, between APOE-ε2+ and APOE-ε3ε3 controls, and again between APOE-ε4+ and APOE-ε3ε3 controls. Multivariate analysis of covariance (MANCOVA) was used to examine the effects of APOE polymorphism on memory and EF across all APOE groups with age, sex and education as regressors of no interest. Cognitive scores were correlated (Pearson r) with imaging metrics within groups.

RESULTS

No significant differences were seen across groups, within groups in MRI metrics, or cognitive performance (p>0.05, corrected for multiple comparisons).

CONCLUSIONS

The current study partially replicated and extended previous findings from an earlier multimodal study (Honea 2009). Future studies should clarify APOE mechanisms in healthy ageing by adding other imaging, cognitive, and lifestyle metrics and longitudinal design in larger sample sizes.

Association of β-Amyloid, Microglial Activation, Cortical Thickness, and Metabolism in Older Adults Without Dementia

BACKGROUND AND OBJECTIVES

Plasma β-amyloid42 (Aβ42)/Aβ40 levels have shown promise in identifying Aβ-PET positive individuals. This study explored the concordance and discordance of plasma Aβ42/Aβ40 positivity (Plasma±) with CSF Aβ42/Aβ40 positivity (CSF±) and Aβ-PET positivity (PET±) in older adults without dementia. Associations of Aβ deposition, cortical thickness, glucose metabolism, and microglial activation were also investigated.

METHODS

We selected participants without dementia who had concurrent plasma Aβ42/Aβ40 and Aβ-PET scans from the Alzheimer's Disease Neuroimaging Initiative cohort. Participants were categorized into Plasma±/PET± based on thresholds of composite 18F-florbetapir (FBP) standardized uptake value ratio (SUVR) ≥1.11 and plasma Aβ42/Aβ40 ≤0.1218. Aβ-PET-negative individuals were further divided into Plasma±/CSF± (CSF Aβ42/Aβ40 ≤0.138), and the concordance and discordance of Aβ42/Aβ40 in the plasma and CSF were investigated. Baseline and slopes of regional FBP SUVR were compared among Plasma±/PET± groups, and associations of regional FBP SUVR, FDG SUVR, cortical thickness, and CSF soluble Triggering Receptor Expressed on Myeloid Cell 2 (sTREM2) levels were analyzed.

RESULTS

One hundred eighty participants (mean age 72.7 years, 51.4% female, 96 cognitively unimpaired, and 84 with mild cognitive impairment) were included. We found that the proportion of Plasma+/PET- individuals was 6.14 times higher (odds ratio (OR) = 6.143, 95% confidence interval (CI) 2.740-16.185, p < 0.001) than that of Plasma-/PET+ individuals, and Plasma+/CSF- individuals showed 8.5 times larger percentage (OR = 8.5, 95% CI: 3.031-32.974, p < 0.001) than Plasma-/CSF+ individuals in Aβ-PET-negative individuals. Besides, Plasma+/PET- individuals exhibited faster (p < 0.05) Aβ accumulation predominantly in bilateral banks of superior temporal sulcus (BANKSSTS) and supramarginal, and superior parietal cortices compared with Plasma-/PET- individuals, despite no difference in baseline FBP SUVRs. In Plasma+/PET+ individuals, higher CSF sTREM2 levels correlated with slower BANKSSTS Aβ accumulation (standardized β (βstd) = -0.418, 95% CI -0.681 to -0.154, p = 0.002). Conversely, thicker cortical thickness and higher glucose metabolism in supramarginal and superior parietal cortices were associated with faster (p < 0.05) CSF sTREM2 increase in Plasma+/PET- individuals rather than in Plasma+/PET+ individuals.

DISCUSSION

These findings suggest that plasma Aβ42/Aβ40 abnormalities may predate CSF Aβ42/Aβ40 and Aβ-PET abnormalities. Higher sTREM2-related microglial activation is linked to thicker cortical thickness and higher metabolism in early amyloidosis stages but tends to mitigate Aβ accumulation primarily at relatively advanced stages.

Multiple phenotype association tests based on sliced inverse regression

BACKGROUND

Joint analysis of multiple phenotypes in studies of biological systems such as Genome-Wide Association Studies is critical to revealing the functional interactions between various traits and genetic variants, but growth of data in dimensionality has become a very challenging problem in the widespread use of joint analysis. To handle the excessiveness of variables, we consider the sliced inverse regression (SIR) method. Specifically, we propose a novel SIR-based association test that is robust and powerful in testing the association between multiple predictors and multiple outcomes.

RESULTS

We conduct simulation studies in both low- and high-dimensional settings with various numbers of Single-Nucleotide Polymorphisms and consider the correlation structure of traits. Simulation results show that the proposed method outperforms the existing methods. We also successfully apply our method to the genetic association study of ADNI dataset. Both the simulation studies and real data analysis show that the SIR-based association test is valid and achieves a higher efficiency compared with its competitors.

CONCLUSION

Several scenarios with low- and high-dimensional responses and genotypes are considered in this paper. Our SIR-based method controls the estimated type I error at the pre-specified level α .

Atrophy of hippocampal subfields and amygdala nuclei in subjects with mild cognitive impairment progressing to Alzheimer's disease

The hippocampus and amygdala are the first brain regions to show early signs of Alzheimer's Disease (AD) pathology. AD is preceded by a prodromal stage known as Mild Cognitive Impairment (MCI), a crucial crossroad in the clinical progression of the disease. The topographical development of AD has been the subject of extended investigation. However, it is still largely unknown how the transition from MCI to AD affects specific hippocampal and amygdala subregions. The present study is set to answer that question. We analyzed data from 223 subjects: 75 healthy controls, 52 individuals with MCI, and 96 AD patients obtained from the ADNI. The MCI group was further divided into two subgroups depending on whether individuals in the 48 months following the diagnosis either remained stable (N = 21) or progressed to AD (N = 31). A MANCOVA test evaluated group differences in the volume of distinct amygdala and hippocampal subregions obtained from magnetic resonance images. Subsequently, a stepwise linear discriminant analysis (LDA) determined which combination of magnetic resonance imaging parameters was most effective in predicting the conversion from MCI to AD. The predictive performance was assessed through a Receiver Operating Characteristic analysis. AD patients displayed widespread subregional atrophy. MCI individuals who progressed to AD showed selective atrophy of the hippocampal subiculum and tail compared to stable MCI individuals, who were undistinguishable from healthy controls. Converter MCI showed atrophy of the amygdala's accessory basal, central, and cortical nuclei. The LDA identified the hippocampal subiculum and the amygdala's lateral and accessory basal nuclei as significant predictors of MCI conversion to AD. The analysis returned a sensitivity value of 0.78 and a specificity value of 0.62. These findings highlight the importance of targeted assessments of distinct amygdala and hippocampus subregions to help dissect the clinical and pathophysiological development of the MCI to AD transition.

BayesKAT: bayesian optimal kernel-based test for genetic association studies reveals joint genetic effects in complex diseases

Genome-wide Association Studies (GWAS) methods have identified individual single-nucleotide polymorphisms (SNPs) significantly associated with specific phenotypes. Nonetheless, many complex diseases are polygenic and are controlled by multiple genetic variants that are usually non-linearly dependent. These genetic variants are marginally less effective and remain undetected in GWAS analysis. Kernel-based tests (KBT), which evaluate the joint effect of a group of genetic variants, are therefore critical for complex disease analysis. However, choosing different kernel functions in KBT can significantly influence the type I error control and power, and selecting the optimal kernel remains a statistically challenging task. A few existing methods suffer from inflated type 1 errors, limited scalability, inferior power or issues of ambiguous conclusions. Here, we present a new Bayesian framework, BayesKAT (https://github.com/wangjr03/BayesKAT), which overcomes these kernel specification issues by selecting the optimal composite kernel adaptively from the data while testing genetic associations simultaneously. Furthermore, BayesKAT implements a scalable computational strategy to boost its applicability, especially for high-dimensional cases where other methods become less effective. Based on a series of performance comparisons using both simulated and real large-scale genetics data, BayesKAT outperforms the available methods in detecting complex group-level associations and controlling type I errors simultaneously. Applied on a variety of groups of functionally related genetic variants based on biological pathways, co-expression gene modules and protein complexes, BayesKAT deciphers the complex genetic basis and provides mechanistic insights into human diseases.

Impaired sleep is associated with tau deposition on 18F-flortaucipir PET and accelerated cognitive decline, accounting for medications that affect sleep

BACKGROUND

Impaired sleep is commonly associated with Alzheimer's disease (AD), although the underlying mechanisms remain unclear. Furthermore, the moderating effects of sleep-affecting medications, which have been linked to AD pathology, are incompletely characterized. Using data from the Alzheimer's Disease Neuroimaging Initiative, we investigated whether a medical history of impaired sleep, informant-reported nighttime behaviors, and sleep-affecting medications are associated with beta-amyloid and tau deposition on PET and cognitive change, cross-sectionally and longitudinally.

METHODS

We included 964 subjects with 18F-florbetapir PET scans. Measures of sleep impairment and medication use were obtained from medical histories and the Neuropsychiatric Inventory Questionnaire. Multivariate models, adjusted for covariates, were used to assess associations among sleep-related features, beta-amyloid and tau, and cognition. Cortical tau deposition, categorized by Braak stage, was assessed using the standardized uptake value peak alignment (SUVP) method on 18F-flortaucipir PET.

RESULTS

Medical history of sleep impairment was associated with greater baseline tau in the meta-temporal, Braak 1, and Braak 4 regions (p = 0.04, p < 0.001, p = 0.025, respectively). Abnormal nighttime behaviors were also associated with greater baseline tau in the meta-temporal region (p = 0.024), and greater cognitive impairment, cross-sectionally (p = 0.007) and longitudinally (p < 0.001). Impaired sleep was not associated with baseline beta-amyloid (p > 0.05). Short-term use of selective serotonin reuptake inhibitors and benzodiazepines slightly weakened the sleep-tau relationship.

CONCLUSIONS

Sleep impairment was associated with tauopathy and cognitive decline, which could be linked to increased tau secretion from neuronal hyperactivity. Clinically, our results help identify high-risk individuals who could benefit from sleep-related interventions aimed to delay cognitive decline and AD.

Longitudinal trajectories of cognitive reserve in hypometabolic subtypes of Alzheimer's disease

Previous studies have demonstrated resilience to AD-related neuropathology in a form of cognitive reserve (CR). In this study we investigated a relationship between CR and hypometabolic subtypes of AD, specifically the typical and the limbic-predominant subtypes. We analyzed data from 59 Aβ-positive cognitively normal (CN), 221 prodromal Alzheimer's disease (AD) and 174 AD dementia participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) from ADNI and ADNIGO/2 phases. For replication, we analyzed data from 5 Aβ-positive CN, 89 prodromal AD and 43 AD dementia participants from ADNI3. CR was estimated as standardized residuals in a model predicting cognition from temporoparietal grey matter volumes and covariates. Higher CR estimates predicted slower cognitive decline. Typical and limbic-predominant hypometabolic subtypes demonstrated similar baseline CR, but the results suggested a faster decline of CR in the typical subtype. These findings support the relationship between subtypes and CR, specifically longitudinal trajectories of CR. Results also underline the importance of longitudinal analyses in research on CR.

Polygenic effects on the risk of Alzheimer's disease in the Japanese population

BACKGROUND

Polygenic effects have been proposed to account for some disease phenotypes; these effects are calculated as a polygenic risk score (PRS). This score is correlated with Alzheimer's disease (AD)-related phenotypes, such as biomarker abnormalities and brain atrophy, and is associated with conversion from mild cognitive impairment (MCI) to AD. However, the AD PRS has been examined mainly in Europeans, and owing to differences in genetic structure and lifestyle, it is unclear whether the same relationships between the PRS and AD-related phenotypes exist in non-European populations. In this study, we calculated and evaluated the AD PRS in Japanese individuals using genome-wide association study (GWAS) statistics from Europeans.

METHODS

In this study, we calculated the AD PRS in 504 Japanese participants (145 cognitively unimpaired (CU) participants, 220 participants with late mild cognitive impairment (MCI), and 139 patients with mild AD dementia) enrolled in the Japanese Alzheimer's Disease Neuroimaging Initiative (J-ADNI) project. In order to evaluate the clinical value of this score, we (1) determined the polygenic effects on AD in the J-ADNI and validated it using two independent cohorts (a Japanese neuropathology (NP) cohort (n = 565) and the North American ADNI (NA-ADNI) cohort (n = 617)), (2) examined the AD-related phenotypes associated with the PRS, and (3) tested whether the PRS helps predict the conversion of MCI to AD.

RESULTS

The PRS using 131 SNPs had an effect independent of APOE. The PRS differentiated between CU participants and AD patients with an area under the curve (AUC) of 0.755 when combined with the APOE variants. Similar AUC was obtained when PRS calculated by the NP and NA-ADNI cohorts was applied. In MCI patients, the PRS was associated with cerebrospinal fluid phosphorylated-tau levels (β estimate = 0.235, p value = 0.026). MCI with a high PRS showed a significantly increased conversion to AD in APOE ε4 noncarriers with a hazard rate of 2.22. In addition, we also developed a PRS model adjusted for LD and observed similar results.

CONCLUSIONS

We showed that the AD PRS is useful in the Japanese population, whose genetic structure is different from that of the European population. These findings suggest that the polygenicity of AD is partially common across ethnic differences.

sTREM2 Mediates the Correlation Between BIN1 Gene Polymorphism and Tau Pathology in Alzheimer's Disease

Background

Bridging integrator 1 (BIN1) gene polymorphism has been reported to play a role in the pathological processes of Alzheimer's disease (AD).

Objective

To explore the association of BIN1 loci with neuroinflammation and AD pathology.

Methods

Alzheimer's Disease Neuroimaging Initiative (ADNI, N = 495) was the discovery cohort, and Chinese Alzheimer's Biomarker and LifestylE (CABLE, N = 619) study was used to replicate the results. Two BIN1 gene polymorphism (rs7561528 and rs744373) were included in the analysis. Multiple linear regression model and causal mediation analysis conducted through 10,000 bootstrapped iterations were used to examine the BIN1 loci relationship with cerebrospinal fluid (CSF) AD biomarkers and alternative biomarker of microglial activation microglia-soluble triggering receptor expressed on myeloid cells 2 (sTREM2).

Results

In ADNI database, we found a significant association between BIN1 loci (rs7561528 and rs744373) and levels of CSF phosphorylated-tau (P-tau) (pc = 0.017; 0.010, respectively) and total-tau (T-tau) (pc = 0.011; 0.013, respectively). The BIN1 loci were also correlated with CSF sTREM2 levels (pc = 0.010; 0.008, respectively). Mediation analysis demonstrated that CSF sTREM2 partially mediated the association of BIN1 loci with P-tau (Proportion of rs7561528 : 20.8%; Proportion of rs744373 : 24.8%) and T-tau (Proportion of rs7561528 : 36.5%; Proportion of rs744373 : 43.9%). The analysis in CABLE study replicated the mediation role of rs7561528.

Conclusions

This study demonstrated the correlation between BIN1 loci and CSF AD biomarkers as well as microglia biomarkers. Additionally, the link between BIN1 loci and tau pathology was partially mediated by CSF sTREM2.

Comparison of Commonly Measured Plasma and Cerebrospinal Fluid Proteins and Their Significance for the Characterization of Cognitive Impairment Status

BACKGROUND

Although cerebrospinal fluid (CSF) amyloid-β42 peptide (Aβ42) and phosphorylated tau (p-tau) and blood p-tau are valuable for differential diagnosis of Alzheimer's disease (AD) from cognitively normal (CN) there is a lack of validated biomarkers for mild cognitive impairment (MCI).

OBJECTIVE

This study sought to determine how plasma and CSF protein markers compared in the characterization of MCI and AD status.

METHODS

This cohort study included Alzheimer's Disease Neuroimaging Initiative (ADNI) participants who had baseline levels of 75 proteins measured commonly in plasma and CSF (257 total, 46 CN, 143 MCI, and 68 AD). Logistic regression, least absolute shrinkage and selection operator (LASSO) and Random Forest (RF) methods were used to identify the protein candidates for the disease classification.

RESULTS

We observed that six plasma proteins panel (APOE, AMBP, C3, IL16, IGFBP2, APOD) outperformed the seven CSF proteins panel (VEGFA, HGF, PRL, FABP3, FGF4, CD40, RETN) as well as AD markers (CSF p-tau and Aβ42) to distinguish the MCI from AD [area under the curve (AUC) = 0.75 (plasma proteins), AUC = 0.60 (CSF proteins) and AUC = 0.56 (CSF p-tau and Aβ42)]. Also, these six plasma proteins performed better than the CSF proteins and were in line with CSF p-tau and Aβ42 in differentiating CN versus MCI subjects [AUC = 0.89 (plasma proteins), AUC = 0.85 (CSF proteins) and AUC = 0.89 (CSF p-tau and Aβ42)]. These results were adjusted for age, sex, education, and APOEϵ4 genotype.

CONCLUSIONS

This study suggests that the combination of 6 plasma proteins can serve as an effective marker for differentiating MCI from AD and CN.

The Alzheimer's Disease Neuroimaging Initiative and the role and contributions of the Private Partners Scientific Board (PPSB)

The Alzheimer's Disease Neuroimaging Initiative (ADNI) Private Partners Scientific Board (PPSB) encompasses members from industry, biotechnology, diagnostic, and non-profit organizations that have until recently been managed by the Foundation for the National Institutes of Health (FNIH) and provided financial and scientific support to ADNI programs. In this article, we review some of the major activities undertaken by the PPSB, focusing on those supporting the most recently completed National Institute on Aging grant, ADNI3, and the impact it has had on streamlining biomarker discovery and validation in Alzheimer's disease. We also provide a perspective on the gaps that may be filled with future PPSB activities as part of ADNI4 and beyond. HIGHLIGHTS: The Private Partners Scientific board (PPSB) continues to play a key role in enabling several Alzheimer's Disease Neuroimaging Initiative (ADNI) activities. PPSB working groups have led landscape assessments to provide valuable feedback on new technologies, platforms, and methods that may be taken up by ADNI in current or future iterations.

Association Between Small Vessel Disease and Financial Capacity: A Study Based on Cognitively Normal Older Adults

Background

Financial capacity is vital for the elderly, who possess a substantial share of global wealth but are vulnerable to financial fraud.

Objective

We explored the link between small vessel disease (SVD) and financial capacity in cognitively unimpaired (CU) older adults via both cross-sectional and longitudinal analyses.

Methods

414 CU participants underwent MRI and completed the Financial Capacity Instrument-Short Form (FCI-SF). Subsequent longitudinal FCI-SF data were obtained from 104, 240, and 141 participants at one, two, and four years, respectively. SVD imaging markers, encompassing white matter hyperintensities (WMH), cerebral microbleeds (CMB), and lacune were evaluated. We used linear regression analyses to cross-sectionally explore the association between FCI-SF and SVD severity, and linear mixed models to assess how baseline SVD severity impacted longitudinal FCI-SF change. The false discovery rate method was used to adjust multiple comparisons.

Results

Cross-sectional analysis revealed a significant association between baseline WMH and Bank Statement (BANK, β=-0.194), as well as between lacune number and Financial Conceptual Knowledge (FC, β= -0.171). These associations were stronger in APOE ɛ4 carriers, with β= -0.282 for WMH and BANK, and β= -0.366 for lacune number and FC. Longitudinally, higher baseline SVD total score was associated with severe FCI-SF total score decrease (β= -0.335). Additionally, baseline WMH burden predicted future decreases in Single Checkbook/Register Task (SNG, β= -0.137) and FC (β= -0.052). Notably, the association between baseline WMH and SNG changes was amplified in APOE ɛ4 carriers (β= -0.187).

Conclusions

Severe SVD was associated with worse FCI-SF and could predict the decline of financial capacity in CU older adults.

Genome-Wide Association Analysis across Endophenotypes in Alzheimer's Disease: Main Effects and Disease Stage-Specific Interactions

The underlying genetic susceptibility for Alzheimer's disease (AD) is not yet fully understood. The heterogeneous nature of the disease challenges genetic association studies. Endophenotype approaches can help to address this challenge by more direct interrogation of biological traits related to the disease. AD endophenotypes based on amyloid-β, tau, and neurodegeneration (A/T/N) biomarkers and cognitive performance were selected from the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort (N = 1565). A genome-wide association study (GWAS) of quantitative phenotypes was performed using an SNP main effect and an SNP by Diagnosis interaction (SNP × DX) model to identify disease stage-specific genetic effects. Nine loci were identified as study-wide significant with one or more A/T/N endophenotypes in the main effect model, as well as additional findings significantly associated with cognitive measures. These nine loci include SNPs in or near the genes APOE, SRSF10, HLA-DQB1, XKR3, and KIAA1671. The SNP × DX model identified three study-wide significant genetic loci (BACH2, EP300, and PACRG-AS1) with a neuroprotective effect in later AD stage endophenotypes. An endophenotype approach identified novel genetic associations and provided insight into the molecular mechanisms underlying the genetic associations that may otherwise be missed using conventional case-control study designs.

BayesKAT: Bayesian Optimal Kernel-based Test for genetic association studies reveals joint genetic effects in complex diseases

GWAS methods have identified individual SNPs significantly associated with specific phenotypes. Nonetheless, many complex diseases are polygenic and are controlled by multiple genetic variants that are usually non-linearly dependent. These genetic variants are marginally less effective and remain undetected in GWAS analysis. Kernel-based tests (KBT), which evaluate the joint effect of a group of genetic variants, are therefore critical for complex disease analysis. However, choosing different kernel functions in KBT can significantly influence the type I error control and power, and selecting the optimal kernel remains a statistically challenging task. A few existing methods suffer from inflated type 1 errors, limited scalability, inferior power, or issues of ambiguous conclusions. Here, we present a new Bayesian framework, BayesKAT( https://github.com/wangjr03/BayesKAT ), which overcomes these kernel specification issues by selecting the optimal composite kernel adaptively from the data while testing genetic associations simultaneously. Furthermore, BayesKAT implements a scalable computational strategy to boost its applicability, especially for high-dimensional cases where other methods become less effective. Based on a series of performance comparisons using both simulated and real large-scale genetics data, BayesKAT outperforms the available methods in detecting complex group-level associations and controlling type I errors simultaneously. Applied on a variety of groups of functionally related genetic variants based on biological pathways, co-expression gene modules, and protein complexes, BayesKAT deciphers the complex genetic basis and provides mechanistic insights into human diseases.

White matter hyperintensity longitudinal morphometric analysis in association with Alzheimer disease

INTRODUCTION

Vascular damage in Alzheimer's disease (AD) has shown conflicting findings particularly when analyzing longitudinal data. We introduce white matter hyperintensity (WMH) longitudinal morphometric analysis (WLMA) that quantifies WMH expansion as the distance from lesion voxels to a region of interest boundary.

METHODS

WMH segmentation maps were derived from 270 longitudinal fluid-attenuated inversion recovery (FLAIR) ADNI images. WLMA was performed on five data-driven WMH patterns with distinct spatial distributions. Amyloid accumulation was evaluated with WMH expansion across the five WMH patterns.

RESULTS

The preclinical group had significantly greater expansion in the posterior ventricular WM compared to controls. Amyloid significantly associated with frontal WMH expansion primarily within AD individuals. WLMA outperformed WMH volume changes for classifying AD from controls primarily in periventricular and posterior WMH.

DISCUSSION

These data support the concept that localized WMH expansion continues to proliferate with amyloid accumulation throughout the entirety of the disease in distinct spatial locations.

Unsupervised [18F]Flortaucipir cutoffs for tau positivity and staging in Alzheimer's disease

PURPOSE

Several [18F]Flortaucipir cutoffs have been proposed for tau PET positivity (T+) in Alzheimer's disease (AD), but none were data-driven. The aim of this study was to establish and validate unsupervised T+ cutoffs by applying Gaussian mixture models (GMM).

METHODS

Amyloid negative (A-) cognitively normal (CN) and amyloid positive (A+) AD-related dementia (ADRD) subjects from ADNI (n=269) were included. ADNI (n=475) and Geneva Memory Clinic (GMC) cohorts (n=98) were used for validation. GMM-based cutoffs were extracted for the temporal meta-ROI, and validated against previously published cutoffs and visual rating.

RESULTS

GMM-based cutoffs classified less subjects as T+, mainly in the A- CN (<3.4% vs >28.5%) and A+ CN (<14.5% vs >42.9%) groups and showed higher agreement with visual rating (ICC=0.91 vs ICC<0.62) than published cutoffs.

CONCLUSION

We provided reliable data-driven [18F]Flortaucipir cutoffs for in vivo T+ detection in AD. These cutoffs might be useful to select participants in clinical and research studies.

Identification of Mild Cognitive Impairment Subtypes using an Interpretable Neural Network based Clustering of Gene Expression Data and Neuroimaging Markers

In this study, an attempt is made to cluster the gene expression data and neuroimaging markers using an interpretable neural network model to identify Mild Cognitive Impairment (MCI) subtypes. For this, structural Magnetic Resonance (MR) brain images and gene expression data of early and late MCI subjects are considered from a public database. A neural network model is employed to cluster the gene expression data and regional MR volumes. To evaluate the performance of model, clustering metrics are employed and model is explained using perturbation-based method. Results indicate that the developed model is able to identify MCI subtypes. The network learns latent embeddings of disease-specific genes and MR images markers. The clustering metrics are found to be highest when both the imaging and genetic markers are employed. Volumes of lateral ventricles, hippocampus, amygdala and thalamus are found to be associated with late MCI. Significant scores suggest that genes such as StAR, CCDC108, APOO, TRMT13, RASAL2 and ZNF43 play a key role in identifying the MCI subtypes.Clinical Relevance-Identifying distinct MCI subtypes offer potential for precision diagnostics and targeted clinical recruitment.

PPAD: a deep learning architecture to predict progression of Alzheimer's disease

MOTIVATION

Alzheimer's disease (AD) is a neurodegenerative disease that affects millions of people worldwide. Mild cognitive impairment (MCI) is an intermediary stage between cognitively normal state and AD. Not all people who have MCI convert to AD. The diagnosis of AD is made after significant symptoms of dementia such as short-term memory loss are already present. Since AD is currently an irreversible disease, diagnosis at the onset of the disease brings a huge burden on patients, their caregivers, and the healthcare sector. Thus, there is a crucial need to develop methods for the early prediction AD for patients who have MCI. Recurrent neural networks (RNN) have been successfully used to handle electronic health records (EHR) for predicting conversion from MCI to AD. However, RNN ignores irregular time intervals between successive events which occurs common in electronic health record data. In this study, we propose two deep learning architectures based on RNN, namely Predicting Progression of Alzheimer's Disease (PPAD) and PPAD-Autoencoder. PPAD and PPAD-Autoencoder are designed for early predicting conversion from MCI to AD at the next visit and multiple visits ahead for patients, respectively. To minimize the effect of the irregular time intervals between visits, we propose using age in each visit as an indicator of time change between successive visits.

RESULTS

Our experimental results conducted on Alzheimer's Disease Neuroimaging Initiative and National Alzheimer's Coordinating Center datasets showed that our proposed models outperformed all baseline models for most prediction scenarios in terms of F2 and sensitivity. We also observed that the age feature was one of top features and was able to address irregular time interval problem.

AVAILABILITY AND IMPLEMENTATION

https://github.com/bozdaglab/PPAD.

Discovering Precision AD Biomarkers with Varying Prognosis Effects in Genetics Driven Subpopulations

Alzheimer's Disease (AD) is a highly heritable neurodegenerative disorder characterized by memory impairments. Understanding how genetic factors contribute to AD pathology may inform interventions to slow or prevent the progression of AD. We performed stratified genetic analyses of 1,574 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants to examine associations between levels of quantitative traits (QT's) and future diagnosis. The Chow test was employed to determine if an individual's genetic profile affects identified predictive relationships between QT's and future diagnosis. Our chow test analysis discovered that cognitive and PET-based biomarkers differentially predicted future diagnosis when stratifying on allelic dosage of AD loci. Post-hoc bootstrapped and association analyses of biomarkers confirmed differential effects, emphasizing the necessity of stratified models to realize individualized AD diagnosis prediction. This novel application of the Chow test allows for the quantification and direct comparison of genetic-based differences. Our findings, as well as the identified QT-future diagnosis relationships, warrant future investigation from a biological context.

Individual-specific functional connectivity improves prediction of Alzheimer's disease's symptoms in elderly people regardless of APOE ε4 genotype

To date, reliable biomarkers remain unclear that could link functional connectivity to patients' symptoms for detecting and predicting the process from normal aging to Alzheimer's disease (AD) in elderly people with specific genotypes. To address this, individual-specific functional connectivity is constructed for elderly participants with/without APOE ε4 allele. Then, we utilize recursive feature selection-based machine learning to reveal individual brain-behavior relationships and to predict the symptom transition in different genotypes. Our findings reveal that compared with conventional atlas-based functional connectivity, individual-specific functional connectivity exhibits higher classification and prediction performance from normal aging to AD in both APOE ε4 groups, while no significant performance is detected when the data of two genotyping groups are combined. Furthermore, individual-specific between-network connectivity constitutes a major contributor to assessing cognitive symptoms. This study highlights the essential role of individual variation in cortical functional anatomy and the integration of brain and behavior in predicting individualized symptoms.