Biological subtypes of Alzheimer disease: A systematic review and meta-analysis

Survival Differences Between Individuals With Typical and Atypical Phenotypes of Alzheimer Disease

BACKGROUND AND OBJECTIVES

Survival estimates for individuals with Alzheimer disease (AD) are informative to understand the disease trajectory, but precise estimates for atypical AD variants are scarce. Atypical AD variants are characterized by nonamnestic phenotypes, an early onset, and lower prevalence of APOEε4 carriership, which affect the AD trajectory. We aimed to provide survival estimates for posterior cortical atrophy (PCA), logopenic variant primary progressive aphasia (lvPPA), and behavioral AD (bvAD) and to evaluate the effect of these atypical AD diagnoses beyond known mortality determinants.

METHODS

From the Amsterdam Dementia Cohort, we retrospectively selected patients with biomarker-confirmed sporadic AD presenting at the memory clinic in the mild cognitive impairment or dementia stage. Patients were classified into atypical AD phenotypes (PCA, lvPPA, bvAD; multidisciplinary consensus and retrospective case finding) and a typical AD reference group (excluding unclassifiable atypical presentations or unconfirmed future AD dementia). Survival estimates from the first visit to death/censoring (Central Public Administration) were determined (Kaplan-Meier analysis) and compared (log-rank tests) across diagnostic groups. To assess associations of atypical AD with mortality, Cox proportional hazard models were constructed including age, sex, education, MMSE score, and APOEε4 carriership (model 1), followed by adding the atypical AD group (model 2) or atypical AD variants (model 3). A likelihood ratio test was performed to compare the fit of model 1 and model 2.

RESULTS

A total of 2,081 patients (aged 65 ± 8 years, 52% female) were classified as typical AD (n = 1,801) or atypical AD (n = 280; PCA [n = 112], lvPPA [n = 86], and bvAD [n = 82]). The estimated median survival time for atypical AD of 6.3 years (95% CI [5.8-6.9]) was shorter than for typical AD (7.2 [7.0-7.5], p = 0.02). Median survival durations across the atypical AD variants were consistently, albeit nonsignificantly, shorter (PCA: 6.3 [5.5-7.3], p = 0.055; lvPPA: 6.6 [5.7-7.7], p = 0.110; bvAD: 6.3 [5.0-9.1], p = 0.121, 48% deceased). Including atypical AD improved the model fit (model 2; χ2 = 8.88, p = 0.003) and was associated with 31% increased mortality risk compared with typical AD (hazard ratio [HR] = 1.31 [1.10-1.56], p = 0.002). In model 3, contributions of the variants were as follows: HRPCA = 1.35 (1.05-1.73), p = 0.019; HRlvPPA = 1.27 (0.94-1.69), p = 0.114; HRbvAD = 1.31 (0.94-1.83), p = 0.105.

DISCUSSION

Survival in atypical AD (PCA, lvPPA, bvAD) was shorter compared with typical AD. These atypical variants are associated with increased mortality beyond age, sex, education, APOEε4 carriership, and disease severity. Future studies are required to address generalizability of these findings and to identify factors that explain the observed survival differences.

LACE-UP: An ensemble machine-learning method for health subtype classification on multidimensional binary data

Disease and behavior subtype identification is of significant interest in biomedical research. However, in many settings, subtype discovery is limited by a lack of robust statistical clustering methods appropriate for binary data. Here, we introduce LACE-UP [latent class analysis ensembled with UMAP (uniform manifold approximation and projection) and PCA (principal components analysis)], an ensemble machine-learning method for clustering multidimensional binary data that does not require prespecifying the number of clusters and is robust to realistic data settings, such as the correlation of variables observed from the same individual and the inclusion of variables unrelated to the underlying subtype. The method ensembles latent class analysis, a model-based clustering method; principal components analysis, a spectral signal processing method; and UMAP, a cutting-edge model-free dimensionality reduction algorithm. In simulations, LACE-UP outperforms gold-standard techniques across a variety of realistic scenarios, including in the presence of correlated and extraneous data. We apply LACE-UP to dietary behavior data from the UK Biobank to demonstrate its power to uncover interpretable dietary subtypes that are associated with lipids and cardiovascular risk.

Atrophy trajectories in Alzheimer's disease: how sex matters

INTRODUCTION

Longitudinal subtypes in Alzheimer's disease (AD) have been identified based on their distinct brain atrophy trajectories, encompassing mediotemporal and cortical pathways. These subtypes include minimal atrophy, limbic predominant, limbic predominant plus, diffuse atrophy and hippocampal sparing. The impact of sex on the progression of these subtypes remains a crucial area of investigation.

METHODS

We analysed MRI data from 320 amyloid-β positive individuals with AD from three international cohorts (ADNI, J-ADNI and AIBL). Longitudinal clustering was conducted to identify atrophy trajectories over eight years from the clinical disease onset, with separate trajectories delineated for women and men.

RESULTS

Women consistently exhibited earlier hippocampal atrophy and a higher burden of white matter abnormalities compared to men, yet women displayed less cognitive decline over time. Additionally, specific risk factors and distinct neuropsychiatric symptoms were associated with sex within specific trajectories.

CONCLUSIONS

AD subtypes show sex-specific differences in disease progression, highlighting the need to account for these differences from the early disease stages. Integrating imaging biomarkers with sex differences can enable the identification of more precise treatments for each patient, ensuring that both women and men have equal access to tailored care.

Congruency between publicly available pictorial displays of medial temporal lobe atrophy

The medial temporal lobe atrophy (MTA) score is used for visual assessment of MTA on radiological images in suspected neurodegenerative dementia. Although volumetric tools are available, many radiologists still use visual scoring and compare to reference images. Numerous such example images are found online on educational websites and in scientific articles. The aim of this study was to compare congruencies between MTA scores of publicly available sample images with normalized heights and areas of relevant brain structures, measured in the same images.

METHOD

Systematic online searches yielded 148 individual sample images. The height and area of relevant brain structures were manually delineated, normalized, and compared with regard to the displayed MTA score.

RESULTS

The normalized heights and areas showed correlation with MTA but with considerable overlap between adjacent scores, especially when comparing heights. Also, displays of the MTA score were more consistent with the area of the temporal horn than with the hippocampal area.

CONCLUSION

There is considerable overlap between adjacent scores in publicly available pictorial displays of the MTA grading system. Insufficient congruency leads to confusion and reduces inter-rater reliability. We also found that publicly available images are more consistent with temporal horn area than the hippocampus, which means that ventricular size may bias the grading. This can impede relevant differential diagnostics, especially regarding normal pressure hydrocephalus. Here, we present lectotype images selected specifically with regard to the hippocampal area.

KEY POINTS

Question Overlap between publicly available example images of medial temporal atrophy causes confusion and limits reliability. Findings Available images are more consistent with ventricular dilatation than hippocampal atrophy; this article provides lectotype images selected specifically regarding the hippocampal area. Clinical relevance Visual assessment of medial temporal atrophy is used daily and worldwide in radiological examinations regarding suspected dementia. In clinical routine, many radiologists experience uncertainty, and hydrocephalus is often overlooked. This may be caused by insufficient congruency between educational sample images.

Voxel-wise insights into early Alzheimer's disease pathology progression: the association with APOE and memory decline

Longitudinal investigation of the Apolipoprotein E (APOE) genotype's impact on Alzheimer's disease (AD) biomarker progression, focusing on amyloid beta (Aβ) accumulation and gray matter (GM) atrophy, integrating cognitive decline and baseline levels. Longitudinal florbetapir-PET and T1-weighted MRI data from 100 cognitively normal (CN) and mild cognitive impaired (MCI) participants both with considerable global Aβ accumulation ("high Aβ accumulators") were analyzed using a voxel-wise approach. Associations of APOE genotype and memory decline with Aβ accumulation and GM atrophy were examined separately for each neuroimaging modality, controlling for baseline Aβ levels and diagnosis. Alternatively, the effect of baseline diagnosis, while controlling for memory decline, was investigated. A multimodal analysis evaluated interactions between genotype, memory decline, and GM atrophy on Aβ accumulation. High Aβ accumulators displayed extensive Aβ pathology predominantly in the medial orbito-frontal cortex, cingulate cortex, and precuneus, along with GM atrophy in temporal, occipital, orbito-frontal, and parietal areas. ɛ4 carriers with memory decline exhibited greater Aβ accumulation and GM atrophy in selective regions compared to non-carriers with memory decline, while no genotype difference was observed in individuals without decline. No interaction effect was observed for MCI diagnosis. Regional associations between the two biomarkers were similarly dependent on genotype and memory decline. ɛ4 carriers exhibiting memory decline present an accelerated neurobiological pattern at predementia stages, supporting early ɛ4 carrier monitoring and interventions in this at-risk group. Importantly, memory decline might be more informative than MCI regarding AD pathology progression emphasizing the importance of repeated cognitive assessments.

Alzheimer-typical temporo-parietal atrophy and hypoperfusion are associated with a more significant cholinergic impairment in amnestic neurodegenerative syndromes

BackgroundTo date, cholinomimetics remain central in the pharmacotherapy of Alzheimer's disease (AD) dementia. However, postmortem investigations indicate that the AD-typical progressive amnestic syndrome may also result from predominantly limbic non-AD neuropathology such as TDP-43 proteinopathy and argyrophilic grain disease. Experimental evidence links a beneficial response to cholinomimetics in early AD to reduced markers of cholinergic neurotransmission. However, the cholinergic impairment varies among patients with a clinical AD presentation, likely due to non-AD (co)-pathologies.ObjectiveThis study examines whether AD-typical atrophy and hypoperfusion can provide information about the cholinergic system in clinically diagnosed AD.MethodsThirty-two patients with amnestic mild cognitive impairment or mild dementia due to AD underwent positron emission tomography (PET) with the tracer N-methyl-4-piperidyl-acetate (MP4A) to estimate acetylcholinesterase (AChE) activity, neurological examinations, cerebral magnetic resonance imaging (MRI) and neuropsychological assessment. The 'cholinergic deficit' was computed as the deviation of AChE activity from cognitively normal controls across the cerebral cortex and correlated gray matter (GM) and perfusion of temporo-parietal cortices typically affected by AD and basal forebrain (BF) GM.ResultsTemporo-parietal perfusion and GM, as well as the inferior temporal to medial temporal ratio of perfusion correlated negatively with the 'cholinergic deficit'. A smaller Ch4p area of the BF was associated with a more significant 'cholinergic deficit', albeit to a lesser degree than cortical measures.ConclusionsIn clinically diagnosed AD, temporo-parietal GM and perfusion are more closely associated with the 'cholinergic deficit' than BF volumes, making them possible markers for cholinergic treatment response in amnestic neurodegeneration.

Neuroimaging-AI endophenotypes reveal underlying mechanisms and genetic factors contributing to progression and development of four brain disorders

Recent work leveraging artificial intelligence has offered promise to dissect disease heterogeneity by identifying complex intermediate brain phenotypes, called dimensional neuroimaging endophenotypes (DNEs). We advance the argument that these DNEs capture the degree of expression of respective neuroanatomical patterns measured, offering a dimensional neuroanatomical representation for studying disease heterogeneity and similarities of neurologic and neuropsychiatric diseases. We investigate the presence of nine DNEs derived from independent yet harmonized studies on Alzheimer's disease, autism spectrum disorder, late-life depression, and schizophrenia in the UK Biobank study. Phenome-wide associations align with genome-wide associations, revealing 31 genomic loci (P-value<5×10-8/9) associated with the nine DNEs.The nine DNEs, along with their polygenic risk scores, significantly enhanced the predictive accuracy for 14 systemic disease categories, particularly for conditions related to mental health and the central nervous system, as well as mortality outcomes. These findings underscore the potential of the nine DNEs to capture the expression of disease-related brain phenotypes in individuals of the general population and to relate such measures with genetics, lifestyle factors, and chronic diseases.

Death in advance or people living with dementia? Extending the philosophical discourse of Schweda and Jongsma through the persistence of self and other strengths

This article presents an extension of an article previously featured in History and Philosophy of the Life Sciences by Schweda and Jongsma (History and Philosophy of the Life Sciences, 2022), who aptly (1) critiqued the "Zombification" of people living with dementia by reviewing the historic and philosophic origins of this damaging metaphor and (2) offered a life course perspective to highlight the ethical implications related to biomedicine and the life sciences. Herein, we aim to build upon and constructively critique the important discourse offered by Schweda and Jongsma by (1) presenting a transdisciplinary perspective highlighting many important remaining social and cognitive abilities of people living with dementia that (2) further informs philosophical discussion and (3) provides ways of helping people diagnosed as well as formal and informal caregivers to live with dementia rather than enduring the damaging and incorrect "living death" notion, and its ramifications, of the syndrome. In the process, we will explore many inherent harms associated with the "zombie-like" construction of the syndrome: harms that entail dysfunctional treatment of people living with dementia. Specifically, we will draw upon evidence from psychology, sociology, philosophy, neurology, and neuroscience, to provide an integrated, whole-person perspective that adds specific dimensions to the life-course perspective and support the necessary multifaceted interdisciplinary and transdisciplinary research and clinical collaborations for this complex issue.

AI-driven fusion of neurological work-up for assessment of biological Alzheimer's disease

Alzheimer's disease (AD) diagnosis hinges on detecting amyloid beta (Aβ) plaques and neurofibrillary tau (τ) tangles. While amyloid PET imaging is now clinically approved, tau PET remains largely restricted to research settings. These imaging techniques, though valuable, are expensive and often difficult to access, limiting their widespread use in routine clinical practice. Here, we introduce a computational framework that leverages multimodal data from seven distinct cohorts comprising 12, 185 participants to estimate individual PET profiles, both global and regional, using more accessible data modalities, such as demographics, medical history, medication use, fluid measurements, functional and neuropsychological assessments, and structural MRIs. Our approach achieved an area under the receiver operating characteristic curve of 0.79 and 0.84 in classifying persons with positive Aβ and τ status, respectively. Model predictions were consistent with various biomarker and cognitive profiles, as well as with different degrees of protein abnormalities observed in post-mortem examinations. Furthermore, the regional volumes identified by the model as important aligned with the spatial distributions of the standardized uptake value ratio for regional τ labels. Our model offers a practical approach to identify potential candidates for newly approved anti-amyloid treatments and AD clinical trials for combined amyloid and tau therapies by utilizing standard neurological evaluation data.

Disease stage-specific atrophy markers in Alzheimer's disease

INTRODUCTION

Structural MRI often lacks diagnostic, prognostic, and monitoring value in Alzheimer's disease (AD), particularly in early disease stages. To improve its utility, we aimed to identify optimal MRI readouts for different use cases.

METHODS

We included 363 older adults; healthy controls (HC) who were negative or positive for amyloidbeta (Aβ) and Aβ-positive patients with subjective cognitive decline (SCD), mild cognitive impairment, or dementia of the Alzheimer type. MRI and neuropsychological assessments were administered annually for up to three years.

RESULTS

Accelerated atrophy of distinct MTL subregions was evident already during preclinical AD. Symptomatic disease stages most notably differed in their hippocampal and parietal atrophy signatures. Associations of atrophy markers and cognitive inventories varied by intended use and disease stage.

DISCUSSION

With the appropriate readout, MRI can detect abnormal atrophy already during preclinical AD. To optimize performance, MRI readouts should be tailored to the targeted disease stage and intended use.

A Proposed Role for Lymphatic Supermicrosurgery in the Management of Alzheimer's Disease: A Primer for Reconstructive Microsurgeons

The relatively recent discovery of a novel lymphatic system within the brain meninges has spurred interest in how waste products generated by neurons and glial cells-including proteins associated with Alzheimer's disease (AD) pathology such as amyloid beta (Aβ) and tau-are disposed of. Evidence is building that suggests disease progression in AD and other cognitive impairments could be explained by dysfunction in the brain's lymphatic system or obstruction of drainage. An interesting implication of this hypothesis is that, by relieving the obstruction of flow, lymphatic reconstruction along the drainage pathway could serve as a potential novel treatment. Should this concept prove true, it could represent a surgical solution to a problem for which only medical solutions have thus far been considered. This study is meant to serve as a primer for reconstructive microsurgeons, introducing the topic and current hypotheses about the potential role of lymphatic drainage in AD. A preview of current research evaluating the feasibility of lymphatic reconstruction as a surgical approach to improving Aβ clearance is provided, with the aim of inspiring others to design robust preclinical and clinical investigations into this intriguing hypothesis.

Subtypes of brain change in aging and their associations with cognition and Alzheimer's disease biomarkers

Structural brain changes underlie cognitive changes and interindividual variability in cognition in older age. By using structural MRI data-driven clustering, we aimed to identify subgroups of cognitively unimpaired older adults based on brain change patterns and assess how changes in cortical thickness, surface area, and subcortical volume relate to cognitive change. We tested (1) which brain structural changes predict cognitive change (2) whether these are associated with core cerebrospinal fluid (CSF) Alzheimer's disease biomarkers, and (3) the degree of overlap between clusters derived from different structural modalities in 1899 cognitively healthy older adults followed up to 16 years. We identified four groups for each brain feature, based on the degree of a main longitudinal component of decline. The minimal overlap between features suggested that each contributed uniquely and independently to structural brain changes in aging. Cognitive change and baseline cognition were associated with cortical area change, whereas higher baseline levels of phosphorylated tau and amyloid-β related to changes in subcortical volume. These results may contribute to a better understanding of different aging trajectories.

Driving research on successful aging and neuroprotection in Latin America: Insights from the inaugural symposium on brain resilience and healthy longevity

INTRODUCTION

Global life expectancy has steadily increased in recent decades, resulting in a significant rise in the number of individuals aged 80 years and older. This trend is also evident in Latin America, where life expectancy is improving, though at varying rates across countries and regions.

METHODS

Partnering with the Neurosciences Group of Antioquia (GNA), we launched a Colombian study on resilience in families with autosomal dominant Alzheimer's disease and the oldest-old population. Over the past 2 years, the project has expanded to include participants from Peru, Chile, and Costa Rica.

RESULTS

This research led to the first symposium on Brain Resilience and Healthy Longevity, held in Medellín, Colombia, in August 2024.

DISCUSSION

The article summarizes key discussions from the symposium, highlighting the most promising opportunities for brain resilience and prevention research in the region and offering recommendations for future research to promote healthy aging and dementia-free communities.

HIGHLIGHTS

Uncovering the genetic and physiological drivers of cognitive resilience, neurodegeneration resistance, and healthy longevity is essential for maintaining brain function as we age. "Superagers" and cognitively resilient individuals from Latin American families with Alzheimer's disease offer valuable insights into brain protection mechanisms. Studying the interplay of socio-environmental and genetic factors in the oldest-old is key to understanding healthy longevity and improving dementia prevention. The inaugural Brain Resilience and Healthy Longevity Symposium highlights the need for global collaboration to uncover factors that drive cognitive resilience and healthy aging in Latin America, advancing dementia prevention.

Precision Estimates of Longitudinal Brain Aging Capture Unexpected Individual Differences in One Year

Individual differences in human brain aging are difficult to estimate over short intervals because of measurement error. Using a cluster scanning approach that reduces error by densely repeating rapid structural scans, we measured brain aging in individuals in one year. Expected differences between young and older individuals were evident, as were differences between cognitively unimpaired and impaired individuals. Each person's brain change trajectory was compared to modeled expectations from a large cohort of age-matched UK Biobank participants. Cognitively unimpaired older individuals variably revealed relative brain maintenance, unexpectedly rapid change, and asymmetrical change. These atypical brain aging trajectories were found across structures and verified in independent within-individual test and retest data. Precision estimates of brain change are possible over short intervals and reveal marked variability including among cognitively unimpaired individuals.

Integrative Epigenomic Landscape of Alzheimer's Disease Brains Reveals Oligodendrocyte Molecular Perturbations Associated with Tau

Alzheimer's disease (AD) brains are characterized by neuropathologic and biochemical changes that are highly variable across individuals. Capturing epigenetic factors that associate with this variability can reveal novel biological insights into AD pathophysiology. We conducted an epigenome-wide association study of DNA methylation (DNAm) in 472 AD brains with neuropathologic measures (Braak stage, Thal phase, and cerebral amyloid angiopathy score) and brain biochemical levels of five proteins (APOE, amyloid-β (Aβ)40, Aβ42, tau, and p-tau) core to AD pathogenesis. Using a novel regional methylation (rCpGm) approach, we identified 5,478 significant associations, 99.7% of which were with brain tau biochemical measures. Of the tau-associated rCpGms, 93 had concordant associations in external datasets comprising 1,337 brain samples. Integrative transcriptome-methylome analyses uncovered 535 significant gene expression associations for these 93 rCpGms. Genes with concurrent transcriptome-methylome perturbations were enriched in oligodendrocyte marker genes, including known AD risk genes such as BIN1 , myelination genes MYRF, MBP and MAG previously implicated in AD, as well as novel genes like LDB3 . We further annotated the top oligodendrocyte genes in an additional 6 brain single cell and 2 bulk transcriptome datasets from AD and two other tauopathies, Pick's disease and progressive supranuclear palsy (PSP). Our findings support consistent rCpGm and gene expression associations with these tauopathies and tau-related phenotypes in both bulk brain tissue and oligodendrocyte clusters. In summary, we uncover the integrative epigenomic landscape of AD and demonstrate tau-related oligodendrocyte gene perturbations as a common potential pathomechanism across different tauopathies.

The Dementia SomaSignal Test (dSST): A plasma proteomic predictor of 20-year dementia risk

INTRODUCTION

There is an unmet need for tools to quantify dementia risk during its multi-decade preclinical/prodromal phase, given that current biomarkers predict risk over shorter follow-up periods and are specific to Alzheimer's disease.

METHODS

Using high-throughput proteomic assays and machine learning techniques in the Atherosclerosis Risk in Communities study (n = 11,277), we developed the Dementia SomaSignal Test (dSST).

RESULTS

In addition to outperforming existing plasma biomarkers, the dSST predicted mid-life dementia risk over a 20-year follow-up across two independent cohorts with different ethnic backgrounds (areas under the curve [AUCs]: dSST 0.68-0.70, dSST+age 0.75-0.81). In a separate cohort, the dSST was associated with longitudinal declines across multiple cognitive domains, accelerated brain atrophy, and elevated measures of neuropathology (as evidenced by positron emission tomography and plasma biomarkers).

DISCUSSION

The dSST is a cost-effective, scalable, and minimally invasive protein-based prognostic aid that can quantify risk up to two decades before dementia onset.

HIGHLIGHTS

The Dementia SomaSignal Test (dSST) predicts 20-year dementia risk across two independent cohorts. dSST outperforms existing plasma biomarkers in predicting multi-decade dementia risk. dSST predicts cognitive decline and accelerated brain atrophy in a third cohort. dSST is a prognostic aid that can predict dementia risk over two decades.

A Multi-Label Deep Learning Model for Detailed Classification of Alzheimer's Disease

BACKGROUND

Accurate diagnosis and classification of Alzheimer's disease (AD) are crucial for effective treatment and management. Traditional diagnostic models, largely based on binary classification systems, fail to adequately capture the complexities and variations across different stages and subtypes of AD, limiting their clinical utility.

METHODS

We developed a deep learning model integrating a dot-product attention mechanism and an innovative labeling system to enhance the diagnosis and classification of AD subtypes and severity levels. This model processed various clinical and demographic data, emphasizing the most relevant features for AD diagnosis. The approach emphasized precision in identifying disease subtypes and predicting their severity through advanced computational techniques that mimic expert clinical decision-making.

RESULTS

Comparative tests against a baseline fully connected neural network demonstrated that our proposed model significantly improved diagnostic accuracy. Our model achieved an accuracy of 83.1% for identifying AD subtypes, compared to 72.9% by the baseline. In severity prediction, our model reached an accuracy of 83.3%, outperforming the baseline (73.5%).

CONCLUSIONS

The incorporation of a dot-product attention mechanism and a tailored labeling system in our model significantly enhances the accuracy of diagnosing and classifying AD. This improvement highlights the potential of the model to support personalized treatment strategies and advance precision medicine in neurodegenerative diseases.

Clinical and biological underpinnings of longitudinal atrophy pattern progression in Alzheimer's disease

BACKGROUND

Magnetic resonance imaging (MRI) has recently enabled to identify four distinct Alzheimer's disease (AD) subtypes: hippocampal sparing (HpSp), typical AD (tAD), limbic predominant (Lp), and minimal atrophy (MinAtr). To date, however, the natural history of these subtypes, especially regarding the presence of subjects switching to other MRI patterns and their clinical and biological differences, remains poorly understood.

OBJECTIVE

To investigate the clinical and biological underpinnings of longitudinal atrophy pattern progression in AD.

METHODS

251 AD patients (16 with significant memory concern, 66 with early mild cognitive impairment (MCI), 125 with late MCI, and 44 with AD dementia) from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database were assigned to their baseline MRI atrophy subtype using Freesurfer-derived cortical:hippocampal volumes ratio. Switching to other MRI patterns was investigated on longitudinal scans, and patients were accordingly classified as "switching" and "stable". Logistic regression models were applied to identify predictors of switching to other MRI patterns.

RESULTS

40% of Lp, 26% of HpSp, and 35% of MinAtr cases switched to other MRI patterns, with tAD representing the destination subtype of all switching HpSp and Lp, and the majority of MinAtr. At baseline significant clinical, cognitive and biomarkers differences were observed across the four subtypes. Only clinical and cognitive variables, however, were significantly associated with switch to other MRI patterns.

CONCLUSIONS

Our results suggest convergent directions of disease progression across atypical and typical AD forms, at least in a subset of AD subjects, and highlight the importance of deep-phenotyping approaches to understand AD heterogeneity.

A Multiform Heterogeneity Framework for Alzheimer's Disease Based on Multimodal Neuroimaging

Understanding the heterogeneity of Alzheimer's disease (AD) is crucial for advancing precision medicine specifically tailored to this disorder. Recent research has deepened our understanding of AD heterogeneity; however, translating these insights from bench to bedside via neuroimaging heterogeneity frameworks presents significant challenges. In this review, we systematically revisit prior studies and summarize the existing methodology of data-driven neuroimaging studies for AD heterogeneity. We organized the current methodology into 1) a subtyping clustering strategy for patients with AD, and we also subdivided it into subtyping analysis based on cross-sectional multimodal neuroimaging profiles and the identification of long-term disease progression from short-term datasets; 2) a stratified strategy that integrates neuroimaging measures with biomarkers; and 3) individual-specific abnormal patterns based on the normative model. Then, we evaluated the characteristics of these studies along 2 dimensions: 1) the understanding of pathology and 2) clinical application. We systematically address the limitations, challenges, and future directions of research into AD heterogeneity. Our goal is to enhance the neuroimaging heterogeneity framework for AD, thereby facilitating its transition from bench to bedside.

Brain MRI volumetry and atrophy rating scales as predictors of amyloid status and eligibility for anti-amyloid treatment in a real-world memory clinic setting

Predicting amyloid status is crucial in light of upcoming disease-modifying therapies and the need to identify treatment-eligible patients with Alzheimer's disease. In our study, we aimed to predict CSF-amyloid status and eligibility for anti-amyloid treatment in a memory clinic by (I) comparing the performance of visual/automated rating scales and MRI volumetric analysis and (II) combining MRI volumetric data with neuropsychological tests and APOE4 status. Two hundred ninety patients underwent a comprehensive assessment. The cNeuro cMRI software (Combinostics Oy) provided automated computed rating scales and volumetric analysis. Amyloid status was determined using data-driven CSF biomarker cutoffs (Aβ42/Aβ40 ratio), and eligibility for anti-Aβ treatment was assessed according to recent recommendations published after the FDA approval of the anti-Aβ drug aducanumab. The automated rating scales and volumetric analysis demonstrated higher performance compared to visual assessment in predicting Aβ status, especially for parietal-GCA (AUC = 0.70), MTA (AUC = 0.66) scores, hippocampal (AUC = 0.68), and angular gyrus (AUC = 0.69) volumes, despite low global accuracy. When we combined hippocampal and angular gyrus volumes with RAVLT immediate recall and APOE4 status, we achieved the highest accuracy (AUC = 0.82), which remained high even in predicting anti-Aβ treatment eligibility (AUC = 0.81). Our study suggests that automated analysis of atrophy rating scales and brain volumetry outperforms operator-dependent visual rating scales. When combined with neuropsychological and genetic information, this computerized approach may play a crucial role not only in a research context but also in a real-world memory clinic. This integration results in a high level of accuracy for predicting amyloid-CSF status and anti-Aβ treatment eligibility.

Comparative mapping of single-cell transcriptomic landscapes in neurodegenerative diseases

INTRODUCTION

Alzheimer's disease (AD), Dementia with Lewy bodies (DLB), and Parkinson's disease (PD) represent a spectrum of neurodegenerative disorders (NDDs). Here, we performed the first direct comparison of their transcriptomic landscapes.

METHODS

We profiled the whole transcriptomes of NDD cortical tissue by snRNA-seq. We used computational analyses to identify common and distinct differentially expressed genes (DEGs), biological pathways, vulnerable and disease-driver cell subtypes, and alteration in cell-to-cell interactions.

RESULTS

The same vulnerable inhibitory neuron subtype was depleted in both AD and DLB. Potentially disease-driving neuronal cell subtypes were present in both PD and DLB. Cell-cell communication was predicted to be increased in AD but decreased in DLB and PD. DEGs were most commonly shared across NDDs within inhibitory neuron subtypes. Overall, we observed the greatest transcriptomic divergence between AD and PD, while DLB exhibited an intermediate transcriptomic signature.

DISCUSSION

These results help explain the clinicopathological spectrum of this group of NDDs and provide unique insights into the shared and distinct molecular mechanisms underlying the pathogenesis of NDDs.

Brain change trajectories in healthy adults correlate with Alzheimer's related genetic variation and memory decline across life

Throughout adulthood and ageing our brains undergo structural loss in an average pattern resembling faster atrophy in Alzheimer's disease (AD). Using a longitudinal adult lifespan sample (aged 30-89; 2-7 timepoints) and four polygenic scores for AD, we show that change in AD-sensitive brain features correlates with genetic AD-risk and memory decline in healthy adults. We first show genetic risk links with more brain loss than expected for age in early Braak regions, and find this extends beyond APOE genotype. Next, we run machine learning on AD-control data from the Alzheimer's Disease Neuroimaging Initiative using brain change trajectories conditioned on age, to identify AD-sensitive features and model their change in healthy adults. Genetic AD-risk linked with multivariate change across many AD-sensitive features, and we show most individuals over age ~50 are on an accelerated trajectory of brain loss in AD-sensitive regions. Finally, high genetic risk adults with elevated brain change showed more memory decline through adulthood, compared to high genetic risk adults with less brain change. Our findings suggest quantitative AD risk factors are detectable in healthy individuals, via a shared pattern of ageing- and AD-related neurodegeneration that occurs along a continuum and tracks memory decline through adulthood.

Utility of Tau PET in the diagnostic work up of neurodegenerative dementia among Indian patients

BACKGROUND AND OBJECTIVES

Tau PET is being increasingly appraised as a novel diagnostic modality for dementia work up. Given limited data among South Asians, we assessed the frequency, patterns, phenotypic associations and incremental value of positive Tau PET scans in clinically diagnosed neurodegenerative dementia.

METHODS

This cross-sectional study recruited consecutive patients of Alzheimer's disease (AD) and non-AD syndromes (September 2021 to October 2022, India). Participants underwent clinical interview, cognitive assessment, MRI brain and tau PET scan ([F-18]ML-104). Visual read in a priori regions of interest was used to identify patterns of tau deposition in the brain.

RESULTS

We recruited 54 participants (mean age: 63.2 ± 9.2 years, 64.8 % men, 77.8 % dementia, 70.4 % early onset cases, 37.8 % APOE4+). The analysis identified abnormal tau uptake in 40/54 (74.1 %) participants; with uptake in AD signature areas in 27/40 (67.5 %) cases [cortical subtype (74.1 %), limbic (14.8 %), combined cortical/limbic (11.1 %)], and patterns not conforming to AD in 13/40 (32.5 %) cases. Tau PET substantiated the diagnosis of AD among 17/19 (89.5 %) cases with clinically diagnosed AD dementia, 8/23 (34.8 %) cases with suspected non-AD cause, and 2/12 (16.7 %) cases with mild cognitive impairment. A trend for increasing proportion of early onset cases, and worsening cognition, behavior and functional ability was seen, from 'limbic' to 'combined cortical/limbic' to 'cortical' subgroups.

CONCLUSION

Tau PET is a useful modality to differentiate AD dementia from other neurodegenerative causes in the Indian setting where amyloid biomarkers are not widely available. Biological subtypes of AD map well onto clinical phenotypes and need study in larger cohorts.

MRI-guided clustering of patients with mild dementia due to Alzheimer's disease using self-organizing maps

Introduction

Alzheimer's disease (AD) is a phenotypically and pathologically heterogenous neurodegenerative disorder. This heterogeneity can be studied and disentangled using data-driven clustering techniques.

Methods

We implemented a self-organizing map clustering algorithm on baseline volumetric MRI measures from nine brain regions of interest (ROIs) to cluster 1041 individuals enrolled in the placebo arm of the EXPEDITION3 trial. Volumetric MRI differences were compared among clusters. Demographics as well as baseline and longitudinal cognitive performance metrics were used to evaluate cluster characteristics.

Results

Three distinct clusters, with an overall silhouette coefficient of 0.491, were identified based on MRI volumetrics. Cluster 1 (N = 400) had the largest baseline volumetric measures across all ROIs and the best cognitive performance at baseline. Cluster 2 (N = 269) had larger hippocampal and medial temporal lobe volumes, but smaller parietal lobe volumes in comparison with the third cluster (N = 372). Significant between-group mean differences were observed between Clusters 1 and 2 (difference, 2.38; 95% CI, 1.85 to 2.91; P < 0.001), Clusters 1 and 3 (difference, 1.93; 95% CI, 1.41 to 2.44; P < 0.001), but not between Clusters 2 and 3 (difference, 0.45; 95% CI, -0.11 to 1.02; P = 0.146) in ADAS-14.

Conclusions

Volumetric MRI can be used to identify homogenous clusters of amyloid positive individuals with mild dementia. The groups identified differ in baseline and longitudinal characteristics. Cluster 1 shows little ADAS-14 change over the first 40 weeks of study on placebo treatment and may be unsuitable for identifying early benefits of treatment.

The phospho-tau cascade, basal forebrain neurodegeneration, and dementia in Alzheimer's disease: Anti-neurodegenerative benefits of acetylcholinesterase inhibitors

A conundrum in Alzheimer's disease (AD) is why the long-term use of acetylcholinesterase (AChE) inhibitors, intended for treatment of dementia, results in slowing neurodegeneration in the cholinergic basal forebrain, hippocampus, and cortex. The phospho-tau cascade hypothesis presented here attempts to answer that question by unifying three hallmark features of AD into a specific sequence of events. It is proposed that the hyperphosphorylation of tau protein leads to the AD-associated deficit of nerve growth factor (NGF), then to atrophy of the cholinergic basal forebrain and dementia. Because the release of pro-nerve growth factor (pro-NGF) is activity-dependent and is controlled by basal forebrain projections to the hippocampus and cortex, our hypothesis is that AChE inhibitors act by increasing acetylcholine-dependent pro-NGF release and, thus, augmenting the availability of mature NGF and improving basal forebrain survival. If correct, improved central nervous system-selective AChE inhibitor therapy started prophylactically, before AD-associated basal forebrain atrophy and cognitive impairment onset, has the potential to delay not only the onset of dementia but also its rate of advancement. The phospho-tau hypothesis thus suggests that preventing hyperphosphorylation of tau protein, early should be a high priority as a strategy to help reduce dementia and its associated widespread social and economic suffering.

Divergent neurodegenerative patterns: Comparison of [18F] fluorodeoxyglucose-PET- and MRI-based Alzheimer's disease subtypes

[18F] fluorodeoxyglucose (FDG)-PET and MRI are key imaging markers for neurodegeneration in Alzheimer's disease. It has been well established that parieto-temporal hypometabolism on FDG-PET is closely associated with medial temporal atrophy on MRI in Alzheimer's disease. Substantial biological heterogeneity, expressed as distinct subtypes of hypometabolism or atrophy patterns, has been previously described in Alzheimer's disease using data-driven and hypothesis-driven methods. However, the link between these two imaging modalities has not yet been explored in the context of Alzheimer's disease subtypes. To investigate this link, the current study utilized FDG-PET and MRI scans from 180 amyloid-beta positive Alzheimer's disease dementia patients, 339 amyloid-beta positive mild cognitive impairment and 176 amyloid-beta negative cognitively normal controls from the Alzheimer's Disease Neuroimaging Initiative. Random forest hierarchical clustering, a data-driven model for identifying subtypes, was implemented in the two modalities: one with standard uptake value ratios and the other with grey matter volumes. Five hypometabolism- and atrophy-based subtypes were identified, exhibiting both cortical-predominant and limbic-predominant patterns although with differing percentages and clinical presentations. Three cortical-predominant hypometabolism subtypes found were Cortical Predominant (32%), Cortical Predominant+ (11%) and Cortical Predominant posterior (8%), and two limbic-predominant hypometabolism subtypes found were Limbic Predominant (36%) and Limbic Predominant frontal (13%). In addition, little atrophy (minimal) and widespread (diffuse) neurodegeneration subtypes were observed from the MRI data. The five atrophy subtypes found were Cortical Predominant (19%), Limbic Predominant (27%), Diffuse (29%), Diffuse+ (6%) and Minimal (19%). Inter-modality comparisons showed that all FDG-PET subtypes displayed medial temporal atrophy, whereas the distinct MRI subtypes showed topographically similar hypometabolic patterns. Further, allocations of FDG-PET and MRI subtypes were not consistent when compared at an individual level. Additional analysis comparing the data-driven clustering model with prior hypothesis-driven methods showed only partial agreement between these subtyping methods. FDG-PET subtypes had greater differences between limbic-predominant and cortical-predominant patterns, and MRI subtypes had greater differences in severity of atrophy. In conclusion, this study highlighted that Alzheimer's disease subtypes identified using both FDG-PET and MRI capture distinct pathways showing cortical versus limbic predominance of neurodegeneration. However, the subtypes do not share a bidirectional relationship between modalities and are thus not interchangeable.

Frontoparietal atrophy trajectories in cognitively unimpaired elderly individuals using longitudinal Bayesian clustering

INTRODUCTION

Frontal and/or parietal atrophy has been reported during aging. To disentangle the heterogeneity previously observed, this study aimed to uncover different clusters of grey matter profiles and trajectories within cognitively unimpaired individuals.

METHODS

Structural magnetic resonance imaging (MRI) data of 307 Aβ-negative cognitively unimpaired individuals were modelled between ages 60-85 from three cohorts worldwide. We applied unsupervised clustering using a novel longitudinal Bayesian approach and characterized the clusters' cerebrovascular and cognitive profiles.

RESULTS

Four clusters were identified with different grey matter profiles and atrophy trajectories. Differences were mainly observed in frontal and parietal brain regions. These distinct frontoparietal grey matter profiles and longitudinal trajectories were differently associated with cerebrovascular burden and cognitive decline.

DISCUSSION

Our findings suggest a conciliation of the frontal and parietal theories of aging, uncovering coexisting frontoparietal GM patterns. This could have important future implications for better stratification and identification of at-risk individuals.

Associations between the multitrajectory neuroplasticity of neuronavigated rTMS-mediated angular gyrus networks and brain gene expression in AD spectrum patients with sleep disorders

INTRODUCTION

The multifactorial influence of repetitive transcranial magnetic stimulation (rTMS) on neuroplasticity in neural networks is associated with improvements in cognitive dysfunction and sleep disorders. The mechanisms of rTMS and the transcriptional-neuronal correlation in Alzheimer's disease (AD) patients with sleep disorders have not been fully elucidated.

METHODS

Forty-six elderly participants with cognitive impairment (23 patients with low sleep quality and 23 patients with high sleep quality) underwent 4-week periods of neuronavigated rTMS of the angular gyrus and neuroimaging tests, and gene expression data for six post mortem brains were collected from another database. Transcription-neuroimaging association analysis was used to evaluate the effects on cognitive dysfunction and the underlying biological mechanisms involved.

RESULTS

Distinct variable neuroplasticity in the anterior and posterior angular gyrus networks was detected in the low sleep quality group. These interactions were associated with multiple gene pathways, and the comprehensive effects were associated with improvements in episodic memory.

DISCUSSION

Multitrajectory neuroplasticity is associated with complex biological mechanisms in AD-spectrum patients with sleep disorders.

HIGHLIGHTS

This was the first transcription-neuroimaging study to demonstrate that multitrajectory neuroplasticity in neural circuits was induced via neuronavigated rTMS, which was associated with complex gene expression in AD-spectrum patients with sleep disorders. The interactions between sleep quality and neuronavigated rTMS were coupled with multiple gene pathways and improvements in episodic memory. The present strategy for integrating neuroimaging, rTMS intervention, and genetic data provide a new approach to comprehending the biological mechanisms involved in AD.

Population clustering of structural brain aging and its association with brain development

Structural brain aging has demonstrated strong inter-individual heterogeneity and mirroring patterns with brain development. However, due to the lack of large-scale longitudinal neuroimaging studies, most of the existing research focused on the cross-sectional changes of brain aging. In this investigation, we present a data-driven approach that incorporate both cross-sectional changes and longitudinal trajectories of structural brain aging and identified two brain aging patterns among 37,013 healthy participants from UK Biobank. Participants with accelerated brain aging also demonstrated accelerated biological aging, cognitive decline and increased genetic susceptibilities to major neuropsychiatric disorders. Further, by integrating longitudinal neuroimaging studies from a multi-center adolescent cohort, we validated the 'last in, first out' mirroring hypothesis and identified brain regions with manifested mirroring patterns between brain aging and brain development. Genomic analyses revealed risk loci and genes contributing to accelerated brain aging and delayed brain development, providing molecular basis for elucidating the biological mechanisms underlying brain aging and related disorders.

Fluid biomarkers in the context of amyloid-targeting disease-modifying treatments in Alzheimer's disease

Clinical management and therapeutics development for Alzheimer's disease (AD) have entered a new era, with recent approvals of monoclonal antibody therapies targeting the underlying pathophysiology of the disease and modifying its trajectory. Imaging and fluid biomarkers are becoming increasingly important in the clinical development of AD therapeutics. This review focuses on the evidence of fluid biomarkers from recent amyloid-β-targeting clinical trials, summarizing biomarker data across 12 trials. It further proposes a simple framework to put biomarker guidance in the context of amyloid-pathway-targeted disease modification, delineates factors that impact biomarker data in clinical trials, and highlights knowledge gaps and future directions. Increased knowledge and data on biomarkers in the context of disease progression and disease modification will help to better design future AD trials and guide the clinical management of patients on AD-modifying therapies, bringing us closer to the implementation of precision medicine in AD.

Genetic and clinical correlates of two neuroanatomical AI dimensions in the Alzheimer's disease continuum

Alzheimer's disease (AD) is associated with heterogeneous atrophy patterns. We employed a semi-supervised representation learning technique known as Surreal-GAN, through which we identified two latent dimensional representations of brain atrophy in symptomatic mild cognitive impairment (MCI) and AD patients: the "diffuse-AD" (R1) dimension shows widespread brain atrophy, and the "MTL-AD" (R2) dimension displays focal medial temporal lobe (MTL) atrophy. Critically, only R2 was associated with widely known sporadic AD genetic risk factors (e.g., APOE ε4) in MCI and AD patients at baseline. We then independently detected the presence of the two dimensions in the early stages by deploying the trained model in the general population and two cognitively unimpaired cohorts of asymptomatic participants. In the general population, genome-wide association studies found 77 genes unrelated to APOE differentially associated with R1 and R2. Functional analyses revealed that these genes were overrepresented in differentially expressed gene sets in organs beyond the brain (R1 and R2), including the heart (R1) and the pituitary gland, muscle, and kidney (R2). These genes were enriched in biological pathways implicated in dendritic cells (R2), macrophage functions (R1), and cancer (R1 and R2). Several of them were "druggable genes" for cancer (R1), inflammation (R1), cardiovascular diseases (R1), and diseases of the nervous system (R2). The longitudinal progression showed that APOE ε4, amyloid, and tau were associated with R2 at early asymptomatic stages, but this longitudinal association occurs only at late symptomatic stages in R1. Our findings deepen our understanding of the multifaceted pathogenesis of AD beyond the brain. In early asymptomatic stages, the two dimensions are associated with diverse pathological mechanisms, including cardiovascular diseases, inflammation, and hormonal dysfunction-driven by genes different from APOE-which may collectively contribute to the early pathogenesis of AD. All results are publicly available at https://labs-laboratory.com/medicine/ .

AI-based differential diagnosis of dementia etiologies on multimodal data

Differential diagnosis of dementia remains a challenge in neurology due to symptom overlap across etiologies, yet it is crucial for formulating early, personalized management strategies. Here, we present an artificial intelligence (AI) model that harnesses a broad array of data, including demographics, individual and family medical history, medication use, neuropsychological assessments, functional evaluations and multimodal neuroimaging, to identify the etiologies contributing to dementia in individuals. The study, drawing on 51,269 participants across 9 independent, geographically diverse datasets, facilitated the identification of 10 distinct dementia etiologies. It aligns diagnoses with similar management strategies, ensuring robust predictions even with incomplete data. Our model achieved a microaveraged area under the receiver operating characteristic curve (AUROC) of 0.94 in classifying individuals with normal cognition, mild cognitive impairment and dementia. Also, the microaveraged AUROC was 0.96 in differentiating the dementia etiologies. Our model demonstrated proficiency in addressing mixed dementia cases, with a mean AUROC of 0.78 for two co-occurring pathologies. In a randomly selected subset of 100 cases, the AUROC of neurologist assessments augmented by our AI model exceeded neurologist-only evaluations by 26.25%. Furthermore, our model predictions aligned with biomarker evidence and its associations with different proteinopathies were substantiated through postmortem findings. Our framework has the potential to be integrated as a screening tool for dementia in clinical settings and drug trials. Further prospective studies are needed to confirm its ability to improve patient care.

Inflammatory Cytokines and Cognition in Alzheimer's Disease and Its Prodrome

OBJECTIVE

The aim of this study was to investigate the association between blood levels of tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) and cognitive impairments among elderly individuals.

METHODS

Peripheral concentration of TNF-α and IL-6 were measured in all subjects. To assess individual cognitive function, the Consortium to Establish a Registry for Alzheimer's Disease Neuropsychological Assessment Battery (CERAD-NP) was used, and standardized scores (z-scores) were calculated for each test. Cytokine levels were compared between the diagnostic groups, and correlations between blood inflammatory factor levels and z-scores were analyzed.

RESULTS

The 37 participants included 8 patients with Alzheimer's disease (AD), 15 subjects with mild cognitive impairment (MCI), and 14 cognitively healthy controls. TNF-α and IL-6 levels were higher in patients with AD than in healthy controls. TNF-α levels were higher in the AD group than in the MCI group. However, after adjusting for age, the associations between diagnosis and TNF-α and IL-6 were not significant. The higher the plasma IL-6 level, the lower the z-scores on the Boston Naming Test, Word List Learning, Word List Recognition, and Constructional Recall. The higher the serum TNF-α level, the lower the z-scores on the Word List Learning and Constructional Recall. Negative correlation between serum TNF-α level and the z-score on Word List Learning remained significant when age was adjusted.

CONCLUSION

The difference in the blood levels of TNF-α and IL-6 between the diagnostic groups may be associated with aging. However, elevated TNF-α levels were associated with worse immediate memory performance, even after adjusting for age.

Personalizing progressive changes to brain structure in Alzheimer's disease using normative modeling

INTRODUCTION

Neuroanatomical normative modeling captures individual variability in Alzheimer's disease (AD). Here we used normative modeling to track individuals' disease progression in people with mild cognitive impairment (MCI) and patients with AD.

METHODS

Cortical and subcortical normative models were generated using healthy controls (n ≈ 58k). These models were used to calculate regional z scores in 3233 T1-weighted magnetic resonance imaging time-series scans from 1181 participants. Regions with z scores < -1.96 were classified as outliers mapped on the brain and summarized by total outlier count (tOC).

RESULTS

tOC increased in AD and in people with MCI who converted to AD and also correlated with multiple non-imaging markers. Moreover, a higher annual rate of change in tOC increased the risk of progression from MCI to AD. Brain outlier maps identified the hippocampus as having the highest rate of change.

DISCUSSION

Individual patients' atrophy rates can be tracked by using regional outlier maps and tOC.

HIGHLIGHTS

Neuroanatomical normative modeling was applied to serial Alzheimer's disease (AD) magnetic resonance imaging (MRI) data for the first time. Deviation from the norm (outliers) of cortical thickness or brain volume was computed in 3233 scans. The number of brain-structure outliers increased over time in people with AD. Patterns of change in outliers varied markedly between individual patients with AD. People with mild cognitive impairment whose outliers increased over time had a higher risk of progression from AD.

Dimensional Neuroimaging Endophenotypes: Neurobiological Representations of Disease Heterogeneity Through Machine Learning

Machine learning has been increasingly used to obtain individualized neuroimaging signatures for disease diagnosis, prognosis, and response to treatment in neuropsychiatric and neurodegenerative disorders. Therefore, it has contributed to a better understanding of disease heterogeneity by identifying disease subtypes with different brain phenotypic measures. In this review, we first present a systematic literature overview of studies using machine learning and multimodal magnetic resonance imaging to unravel disease heterogeneity in various neuropsychiatric and neurodegenerative disorders, including Alzheimer's disease, schizophrenia, major depressive disorder, autism spectrum disorder, and multiple sclerosis, as well as their potential in a transdiagnostic framework, where neuroanatomical and neurobiological commonalities were assessed across diagnostic boundaries. Subsequently, we summarize relevant machine learning methodologies and their clinical interpretability. We discuss the potential clinical implications of the current findings and envision future research avenues. Finally, we discuss an emerging paradigm called dimensional neuroimaging endophenotypes. Dimensional neuroimaging endophenotypes dissects the neurobiological heterogeneity of neuropsychiatric and neurodegenerative disorders into low-dimensional yet informative, quantitative brain phenotypic representations, serving as robust intermediate phenotypes (i.e., endophenotypes), presumably reflecting the interplay of underlying genetic, lifestyle, and environmental processes associated with disease etiology.

Identifying older adults at risk for dementia based on smartphone data obtained during a wayfinding task in the real world

Alzheimer's disease (AD), as the most common form of dementia and leading cause for disability and death in old age, represents a major burden to healthcare systems worldwide. For the development of disease-modifying interventions and treatments, the detection of cognitive changes at the earliest disease stages is crucial. Recent advancements in mobile consumer technologies provide new opportunities to collect multi-dimensional data in real-life settings to identify and monitor at-risk individuals. Based on evidence showing that deficits in spatial navigation are a common hallmark of dementia, we assessed whether a memory clinic sample of patients with subjective cognitive decline (SCD) who still scored normally on neuropsychological assessments show differences in smartphone-assisted wayfinding behavior compared with cognitively healthy older and younger adults. Guided by a mobile application, participants had to find locations along a short route on the medical campus of the Magdeburg university. We show that performance measures that were extracted from GPS and user input data distinguish between the groups. In particular, the number of orientation stops was predictive of the SCD status in older participants. Our data suggest that subtle cognitive changes in patients with SCD, whose risk to develop dementia in the future is elevated, can be inferred from smartphone data, collected during a brief wayfinding task in the real world.

Disruption of sleep macro- and microstructure in Alzheimer's disease: overlaps between neuropsychology, neurophysiology, and neuroimaging

Alzheimer's disease (AD) is the leading cause of dementia, often associated with impaired sleep quality and disorganized sleep structure. This study aimed to characterize changes in sleep macrostructure and K-complex density in AD, in relation to neuropsychological performance and brain structural changes. We enrolled 30 AD and 30 healthy control participants, conducting neuropsychological exams, brain MRI, and one-night polysomnography. AD patients had significantly reduced total sleep time (TST), sleep efficiency, and relative durations of non-rapid eye movement (NREM) stages 2 (S2), 3 (S3), and rapid eye movement (REM) sleep (p < 0.01). K-complex (KC) density during the entire sleep period and S2 (p < 0.001) was significantly decreased in AD. We found strong correlations between global cognitive performance and relative S3 (p < 0.001; r = 0.86) and REM durations (p < 0.001; r = 0.87). TST and NREM stage 1 (S1) durations showed a moderate negative correlation with amygdaloid and hippocampal volumes (p < 0.02; r = 0.51-0.55), while S3 and REM sleep had a moderate positive correlation with cingulate cortex volume (p < 0.02; r = 0.45-0.61). KC density strongly correlated with global cognitive function (p < 0.001; r = 0.66) and the thickness of the anterior cingulate cortex (p < 0.05; r = 0.45-0.47). Our results indicate significant sleep organization changes in AD, paralleling cognitive decline. Decreased slow wave sleep and KCs are strongly associated with cingulate cortex atrophy. Since sleep changes are prominent in early AD, they may serve as prognostic markers or therapeutic targets.

WIMOAD: Weighted Integration of Multi-Omics data for Alzheimer's Disease (AD) Diagnosis

As the most common subtype of dementia, Alzheimer's disease (AD) is characterized by a progressive decline in cognitive functions, especially in memory, thinking, and reasoning ability. Early diagnosis and interventions enable the implementation of measures to reduce or slow further regression of the disease, preventing individuals from severe brain function decline. The current framework of AD diagnosis depends on A/T/(N) biomarkers detection from cerebrospinal fluid or brain imaging data, which is invasive and expensive during the data acquisition process. Moreover, the pathophysiological changes of AD accumulate in amino acids, metabolism, neuroinflammation, etc., resulting in heterogeneity in newly registered patients. Recently, next generation sequencing (NGS) technologies have found to be a non-invasive, efficient and less-costly alternative on AD screening. However, most of existing studies rely on single omics only. To address these concerns, we introduce WIMOAD, a weighted integration of multi-omics data for AD diagnosis. WIMOAD synergistically leverages specialized classifiers for patients' paired gene expression and methylation data for multi-stage classification. The resulting scores were then stacked with MLP-based meta-models for performance improvement. The prediction results of two distinct meta-models were integrated with optimized weights for the final decision-making of the model, providing higher performance than using single omics only. Remarkably, WIMOAD achieves significantly higher performance than using single omics alone in the classification tasks. The model's overall performance also outperformed most existing approaches, highlighting its ability to effectively discern intricate patterns in multi-omics data and their correlations with clinical diagnosis results. In addition, WIMOAD also stands out as a biologically interpretable model by leveraging the SHapley Additive exPlanations (SHAP) to elucidate the contributions of each gene from each omics to the model output. We believe WIMOAD is a very promising tool for accurate AD diagnosis and effective biomarker discovery across different progression stages, which eventually will have consequential impacts on early treatment intervention and personalized therapy design on AD.

Beyond the usual suspects: multi-factorial computational models in the search for neurodegenerative disease mechanisms

From Alzheimer's disease to amyotrophic lateral sclerosis, the molecular cascades underlying neurodegenerative disorders remain poorly understood. The clinical view of neurodegeneration is confounded by symptomatic heterogeneity and mixed pathology in almost every patient. While the underlying physiological alterations originate, proliferate, and propagate potentially decades before symptomatic onset, the complexity and inaccessibility of the living brain limit direct observation over a patient's lifespan. Consequently, there is a critical need for robust computational methods to support the search for causal mechanisms of neurodegeneration by distinguishing pathogenic processes from consequential alterations, and inter-individual variability from intra-individual progression. Recently, promising advances have been made by data-driven spatiotemporal modeling of the brain, based on in vivo neuroimaging and biospecimen markers. These methods include disease progression models comparing the temporal evolution of various biomarkers, causal models linking interacting biological processes, network propagation models reproducing the spatial spreading of pathology, and biophysical models spanning cellular- to network-scale phenomena. In this review, we discuss various computational approaches for integrating cross-sectional, longitudinal, and multi-modal data, primarily from large observational neuroimaging studies, to understand (i) the temporal ordering of physiological alterations, i(i) their spatial relationships to the brain's molecular and cellular architecture, (iii) mechanistic interactions between biological processes, and (iv) the macroscopic effects of microscopic factors. We consider the extents to which computational models can evaluate mechanistic hypotheses, explore applications such as improving treatment selection, and discuss how model-informed insights can lay the groundwork for a pathobiological redefinition of neurodegenerative disorders.

Capturing biomarkers associated with Alzheimer disease subtypes using data distribution characteristics

Late-onset Alzheimer disease (AD) is a highly complex disease with multiple subtypes, as demonstrated by its disparate risk factors, pathological manifestations, and clinical traits. Discovery of biomarkers to diagnose specific AD subtypes is a key step towards understanding biological mechanisms underlying this enigmatic disease, generating candidate drug targets, and selecting participants for drug trials. Popular statistical methods for evaluating candidate biomarkers, fold change (FC) and area under the receiver operating characteristic curve (AUC), were designed for homogeneous data and we demonstrate the inherent weaknesses of these approaches when used to evaluate subtypes representing less than half of the diseased cases. We introduce a unique evaluation metric that is based on the distribution of the values, rather than the magnitude of the values, to identify analytes that are associated with a subset of the diseased cases, thereby revealing potential biomarkers for subtypes. Our approach, Bimodality Coefficient Difference (BCD), computes the difference between the degrees of bimodality for the cases and controls. We demonstrate the effectiveness of our approach with large-scale synthetic data trials containing nearly perfect subtypes. In order to reveal novel AD biomarkers for heterogeneous subtypes, we applied BCD to gene expression data for 8,650 genes for 176 AD cases and 187 controls. Our results confirm the utility of BCD for identifying subtypes of heterogeneous diseases.

Gray Matter Volumes Mediate the Relationship Between Disease Duration and Balance Control Performance in Chronic Ankle Instability

The relationship between structural changes in the cerebral gray matter and diminished balance control performance in patients with chronic ankle instability (CAI) has remained unclear. This paper aimed to assess the difference in gray matter volume (GMV) between participants with CAI and healthy controls (HC) and to characterize the role of GMV in the relationship between disease duration and balance performance in CAI. 42 participants with CAI and 33 HC completed the structural brain MRI scans, one-legged standing test, and Y-balance test. Regional GMV was measured by applying voxel-based morphometry methods. The result showed that, compared with HC, participants with CAI exhibited lower GMV in multiple brain regions (familywise error [FWE] corrected p < 0.021). Within CAI only, but not in HC, lower GMV in the thalamus (β = -0.53, p = 0.003) and hippocampus (β = -0.57, p = 0.001) was associated with faster sway velocity of the center of pressure (CoP) in eyes closed condition (i.e., worse balance control performance). The GMV in the thalamus (percentage mediated [PM] = 32.02%; indirect effect β = 0.119, 95% CI = 0.003 to 0.282) and hippocampus (PM = 33.71%; indirect effect β = 0.122, 95% CI = 0.005 to 0.278) significantly mediated the association between the disease duration and balance performance. These findings suggest that the structural characteristics of the supraspinal elements is critical to the maintenance of balance control performance in individuals suffering from CAI, which deserve careful consideration in the management and rehabilitation programs in this population.

Alzheimer's disease genetic pathways impact cerebrospinal fluid biomarkers and imaging endophenotypes in non-demented individuals

INTRODUCTION

Unraveling how Alzheimer's disease (AD) genetic risk is related to neuropathological heterogeneity, and whether this occurs through specific biological pathways, is a key step toward precision medicine.

METHODS

We computed pathway-specific genetic risk scores (GRSs) in non-demented individuals and investigated how AD risk variants predict cerebrospinal fluid (CSF) and imaging biomarkers reflecting AD pathology, cardiovascular, white matter integrity, and brain connectivity.

RESULTS

CSF amyloidbeta and phosphorylated tau were related to most GRSs. Inflammatory pathways were associated with cerebrovascular disease, whereas quantitative measures of white matter lesion and microstructure integrity were predicted by clearance and migration pathways. Functional connectivity alterations were related to genetic variants involved in signal transduction and synaptic communication.

DISCUSSION

This study reveals distinct genetic risk profiles in association with specific pathophysiological aspects in predementia stages of AD, unraveling the biological substrates of the heterogeneity of AD-associated endophenotypes and promoting a step forward in disease understanding and development of personalized therapies.

HIGHLIGHTS

Polygenic risk for Alzheimer's disease encompasses six biological pathways that can be quantified with pathway-specific genetic risk scores, and differentially relate to cerebrospinal fluid and imaging biomarkers. Inflammatory pathways are mostly related to cerebrovascular burden. White matter health is associated with pathways of clearance and membrane integrity, whereas functional connectivity measures are related to signal transduction and synaptic communication pathways.

Association between surgical admissions, cognition, and neurodegeneration in older people: a population-based study from the UK Biobank

BACKGROUND

Previous studies have shown that major surgical and medical hospital admissions are associated with cognitive decline in older people (aged 40-69 years at recruitment), which is concerning for patients and caregivers. We aimed to validate these findings in a large cohort and investigate associations with neurodegeneration using MRI.

METHODS

For this population-based study, we analysed data from the UK Biobank collected from March 13, 2006, to July 16, 2023, linked to the National Health Service Hospital Episode Statistics database, excluding participants with dementia diagnoses. We constructed fully adjusted models that included age, time, sex, Lancet Commission dementia risk factors, stroke, and hospital admissions with a participant random effect. Primary outcomes were hippocampal volume and white matter hyperintensities, both of which are established markers of neurodegeneration, and exploratory analyses investigated the cortical thickness of Desikan-Killiany-Tourville atlas regions. The main cognitive outcomes were reaction time, fluid intelligence, and prospective and numeric memory. Surgeries were calculated cumulatively starting from 8 years before the baseline evaluation.

FINDINGS

Of 502 412 participants in the UK Biobank study, 492 802 participants were eligible for inclusion in this study, of whom 46 706 underwent MRI. Small adverse associations with cognition were found per surgery: reaction time increased by 0·273 ms, fluid intelligence score decreased by 0·057 correct responses, prospective memory (scored as correct at first attempt) decreased (odds ratio 0·96 [95% CI 0·95 to 0·97]), and numeric memory maximum correct matches decreased by 0·025 in fully adjusted models. Surgeries were associated with smaller hippocampal volume (β=-5·76 mm³ [-7·89 to -3·64]) and greater white matter hyperintensities volume (β=100·02 mm³ [66·17 to 133·87]) in fully adjusted models. Surgeries were also associated with neurodegeneration of the insula and superior temporal cortex.

INTERPRETATION

This population-based study corroborates that surgeries are generally safe but cumulatively are associated with cognitive decline and neurodegeneration. Perioperative brain health should be prioritised for older and vulnerable patients, particularly those who have multiple surgical procedures.

FUNDING

The Australian and New Zealand College of Anaesthetists (ANZCA) Foundation and the University of Sydney.

A core proteome profile unites mouse models and patients in Alzheimer disease

Levites et al. demonstrate that mouse models of Alzheimer disease (AD), exhibiting amyloid-beta (Αβ) plaque formation, share Αβ responsome proteins with humans. Their work underscores the value of these models in studying Αβ aggregation, cellular vulnerability, and early-stage AD pathology.

MRI subtypes in Parkinson's disease across diverse populations and clustering approaches

Parkinson's disease (PD) is clinically heterogeneous, which suggests the existence of subtypes; however, there has been no consensus regarding their characteristics. This study included 633 PD individuals across distinct cohorts: unmedicated de novo PD, medicated PD, mild-moderate PD, and a cohort based on diagnostic work-up in clinical practice. Additionally, 233 controls were included. Clustering based on cortical and subcortical gray matter measures was conducted with and without adjusting for global atrophy in the entire PD sample and validated within each cohort. Subtypes were characterized using baseline and longitudinal demographic and clinical data. Unadjusted results identified three clusters showing a gradient of neurodegeneration and symptom severity across the entire sample and the individual cohorts. When adjusting for global atrophy eight clusters were identified in the entire sample, lacking consistency in individual cohorts. This study identified atrophy-based subtypes in PD, emphasizing the significant impact of global atrophy on subtype number, patterns, and interpretation in cross-sectional analyses.

Disentangling brain atrophy heterogeneity in Alzheimer's disease: A deep self-supervised approach with interpretable latent space

Alzheimer's disease (AD) is heterogeneous, but existing methods for capturing this heterogeneity through dimensionality reduction and unsupervised clustering have limitations when it comes to extracting intricate atrophy patterns. In this study, we propose a deep learning based self-supervised framework that characterizes complex atrophy features using latent space representation. It integrates feature engineering, classification, and clustering to synergistically disentangle heterogeneity in Alzheimer's disease. Through this representation learning, we trained a clustered latent space with distinct atrophy patterns and clinical characteristics in AD, and replicated the findings in prodromal Alzheimer's disease. Moreover, we discovered that these clusters are not solely attributed to subtypes but also reflect disease progression in the latent space, representing the core dimensions of heterogeneity, namely progression and subtypes. Furthermore, longitudinal latent space analysis revealed two distinct disease progression pathways: medial temporal and parietotemporal pathways. The proposed approach enables effective latent representations that can be integrated with individual-level cognitive profiles, thereby facilitating a comprehensive understanding of AD heterogeneity.

CAPS: a simple clinical tool for β-amyloid positivity prediction in clinical Alzheimer syndrome

Introduction

With the advent of anti-β-amyloid therapies, clinical distinction between Aβ + and Aβ- in cognitive impairment is becoming increasingly important for stratifying referral and better utilization of biomarker assays.

Methods

Cognitive profile, rate of decline, neuropsychiatric inventory questionnaire (NPI-Q), and imaging characteristics were collected from 52 subjects with possible/probable AD.

Results

Participants with Aβ+ status had lower baseline MMSE scores (24.50 vs. 26.85, p = 0.009) and higher total NPI-Q scores (2.73 vs. 1.18, p < 0.001). NPI-Q score was found to be the only independent predictor for β-amyloid positivity (p = 0.008). A simple scoring system, namely Clinical β-Amyloid Positivity Prediction Score (CAPS), was developed by using the following parameters: NPI-Q, rapidity of cognitive decline, and white matter microangiopathy. Data from 48 participants were included in the analysis of accuracy of CAPS. CAP Score of 3 or 4 successfully classified Aβ + individuals in 86.7% cases.

Discussion

Clinical β-Amyloid Positivity Prediction Score is a simple clinical tool for use in primary care and memory clinic settings to predict β-amyloid positivity in individuals with clinical Alzheimer Syndrome can potentially facilitate referral for Anti Aβ therapies.

Deciphering ApoE Genotype-Driven Proteomic and Lipidomic Alterations in Alzheimer's Disease Across Distinct Brain Regions

Alzheimer's disease (AD) is a neurodegenerative disease with a complex etiology influenced by confounding factors such as genetic polymorphisms, age, sex, and race. Traditionally, AD research has not prioritized these influences, resulting in dramatically skewed cohorts such as three times the number of Apolipoprotein E (APOE) ε4-allele carriers in AD relative to healthy cohorts. Thus, the resulting molecular changes in AD have previously been complicated by the influence of apolipoprotein E disparities. To explore how apolipoprotein E polymorphism influences AD progression, 62 post-mortem patients consisting of 33 AD and 29 controls (Ctrl) were studied to balance the number of ε4-allele carriers and facilitate a molecular comparison of the apolipoprotein E genotype. Lipid and protein perturbations were assessed across AD diagnosed brains compared to Ctrl brains, ε4 allele carriers (APOE4+ for those carrying 1 or 2 ε4s and APOE4- for non-ε4 carriers), and differences in ε3ε3 and ε3ε4 Ctrl brains across two brain regions (frontal cortex (FCX) and cerebellum (CBM)). The region-specific influences of apolipoprotein E on AD mechanisms showcased mitochondrial dysfunction and cell proteostasis at the core of AD pathophysiology in the post-mortem brains, indicating these two processes may be influenced by genotypic differences and brain morphology.

Employing Informatics Strategies in Alzheimer's Disease Research: A Review from Genetics, Multiomics, and Biomarkers to Clinical Outcomes

Alzheimer's disease (AD) is a critical national concern, affecting 5.8 million people and costing more than $250 billion annually. However, there is no available cure. Thus, effective strategies are in urgent need to discover AD biomarkers for disease early detection and drug development. In this review, we study AD from a biomedical data scientist perspective to discuss the four fundamental components in AD research: genetics (G), molecular multiomics (M), multimodal imaging biomarkers (B), and clinical outcomes (O) (collectively referred to as the GMBO framework). We provide a comprehensive review of common statistical and informatics methodologies for each component within the GMBO framework, accompanied by the major findings from landmark AD studies. Our review highlights the potential of multimodal biobank data in addressing key challenges in AD, such as early diagnosis, disease heterogeneity, and therapeutic development. We identify major hurdles in AD research, including data scarcity and complexity, and advocate for enhanced collaboration, data harmonization, and advanced modeling techniques. This review aims to be an essential guide for understanding current biomedical data science strategies in AD research, emphasizing the need for integrated, multidisciplinary approaches to advance our understanding and management of AD.

Antigen-specific age-related memory CD8 T cells induce and track Alzheimer's-like neurodegeneration

Cerebral (Aβ) plaque and (pTau) tangle deposition are hallmarks of Alzheimer's disease (AD), yet are insufficient to confer complete AD-like neurodegeneration experimentally. Factors acting upstream of Aβ/pTau in AD remain unknown, but their identification could enable earlier diagnosis and more effective treatments. T cell abnormalities are emerging AD hallmarks, and CD8 T cells were recently found to mediate neurodegeneration downstream of tangle deposition in hereditary neurodegeneration models. The precise impact of T cells downstream of Aβ/pTau, however, appears to vary depending on the animal model. Our prior work suggested that antigen-specific memory CD8 T ("hiT") cells act upstream of Aβ/pTau after brain injury. Here, we examine whether hiT cells influence sporadic AD-like pathophysiology upstream of Aβ/pTau. Examining neuropathology, gene expression, and behavior in our hiT mouse model we show that CD8 T cells induce plaque and tangle-like deposition, modulate AD-related genes, and ultimately result in progressive neurodegeneration with both gross and fine features of sporadic human AD. T cells required Perforin to initiate this pathophysiology, and IFNγ for most gene expression changes and progression to more widespread neurodegenerative disease. Analogous antigen-specific memory CD8 T cells were significantly elevated in the brains of human AD patients, and their loss from blood corresponded to sporadic AD and related cognitive decline better than plasma pTau-217, a promising AD biomarker candidate. We identify an age-related factor acting upstream of Aβ/pTau to initiate AD-like pathophysiology, the mechanisms promoting its pathogenicity, and its relevance to human sporadic AD.