OBJECTIVES: Polygenic hazard score (PHS) models can be used to predict the age-associated risk for complex diseases, including Alzheimer's disease (AD). In this study, we present an improved PHS model for AD that incorporates a large number of genetic variants and demonstrates enhanced predictive accuracy for age of onset in European populations compared to alternative models. METHODS: We used the genotyped European Alzheimer & Dementia Biobank (EADB) sample (n=42,120) to develop and evaluate the performance of the PHS model. We developed a PHS model building on 720 genetic variants, including Apolipoprotein E (APOE) ε2 and ε4 alleles. We used Elastic Net-regularized Cox regression approach to develop the PHS model. RESULTS: The new PHS model (EADB720) improved prediction accuracy compared to alternative models in European populations, with the Odds Ratio OR80/20 from the highest quintile of risk (80th risk percentile and above) to the lowest quintile of risk (20th risk percentile and below) varying between 5.10 and 13.15 within the range of age of onset from 65 - 85 years. Our model also improved risk stratification across ε3/3 individuals of European ancestry (OR80/20 ranges from 1.95 to 3.52). It was also successfully validated in independent datasets (HUSK, DemGene and ADNI) by achieving OR80/20 up to 10.00 in each independent dataset. CONCLUSION: Our EADB720 model significantly improves the accuracy of age-associated risk of AD across European populations (pval<0.03). Accurately predicting the age of onset of AD is of large clinical importance to implementing new AD medication and early intervention in clinical settings.

• Publication type
• Journal Article MeSH
• Preprint MeSH

Traditional statistical approaches have advanced our understanding of the genetics of complex diseases, yet are limited to linear additive models. Here we applied machine learning (ML) to genome-wide data from 41,686 individuals in the largest European consortium on Alzheimer's disease (AD) to investigate the effectiveness of various ML algorithms in replicating known findings, discovering novel loci, and predicting individuals at risk. We utilised Gradient Boosting Machines (GBMs), biological pathway-informed Neural Networks (NNs), and Model-based Multifactor Dimensionality Reduction (MB-MDR) models. ML approaches successfully captured all genome-wide significant genetic variants identified in the training set and 22% of associations from larger meta-analyses. They highlight 6 novel loci which replicate in an external dataset, including variants which map to ARHGAP25, LY6H, COG7, SOD1 and ZNF597. They further identify novel association in AP4E1, refining the genetic landscape of the known SPPL2A locus. Our results demonstrate that machine learning methods can achieve predictive performance comparable to classical approaches in genetic epidemiology and have the potential to uncover novel loci that remain undetected by traditional GWAS. These insights provide a complementary avenue for advancing the understanding of AD genetics.

• MeSH
• Algorithms MeSH
• Alzheimer Disease * genetics   MeSH
• Genome-Wide Association Study MeSH
• Genetic Predisposition to Disease MeSH
• Polymorphism, Single Nucleotide MeSH
• Humans MeSH
• Neural Networks, Computer MeSH
• GTPase-Activating Proteins genetics   MeSH
• Machine Learning * MeSH
• Check Tag
• Humans MeSH
• Publication type
• Journal Article MeSH
• Names of Substances
• GTPase-Activating Proteins MeSH

A polygenic score (PGS) for Alzheimer's disease (AD) was derived recently from data on genome-wide significant loci in European ancestry populations. We applied this PGS to populations in 17 European countries and observed a consistent association with the AD risk, age at onset and cerebrospinal fluid levels of AD biomarkers, independently of apolipoprotein E locus (APOE). This PGS was also associated with the AD risk in many other populations of diverse ancestries. A cross-ancestry polygenic risk score improved the association with the AD risk in most of the multiancestry populations tested when the APOE region was included. Finally, we found that the PGS/polygenic risk score captured AD-specific information because the association weakened as the diagnosis was broadened. In conclusion, a simple PGS captures the AD-specific genetic information that is common to populations of different ancestries, although studies of more diverse populations are still needed to better characterize the genetics of AD.

• MeSH
• Alzheimer Disease * genetics   epidemiology   cerebrospinal fluid   MeSH
• Apolipoproteins E genetics   MeSH
• White People * genetics   MeSH
• Biomarkers cerebrospinal fluid   MeSH
• Genome-Wide Association Study MeSH
• Genetic Predisposition to Disease * MeSH
• Genetic Risk Score MeSH
• Polymorphism, Single Nucleotide MeSH
• Humans MeSH
• Multifactorial Inheritance * genetics   MeSH
• Risk Factors MeSH
• Aged MeSH
• Check Tag
• Humans MeSH
• Male MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH
• Geographicals
• Europe epidemiology   MeSH
• Names of Substances
• Apolipoproteins E MeSH
• Biomarkers MeSH

Due to methodological reasons, the X-chromosome has not been featured in the major genome-wide association studies on Alzheimer's Disease (AD). To address this and better characterize the genetic landscape of AD, we performed an in-depth X-Chromosome-Wide Association Study (XWAS) in 115,841 AD cases or AD proxy cases, including 52,214 clinically-diagnosed AD cases, and 613,671 controls. We considered three approaches to account for the different X-chromosome inactivation (XCI) states in females, i.e. random XCI, skewed XCI, and escape XCI. We did not detect any genome-wide significant signals (P ≤ 5 × 10-8) but identified seven X-chromosome-wide significant loci (P ≤ 1.6 × 10-6). The index variants were common for the Xp22.32, FRMPD4, DMD and Xq25 loci, and rare for the WNK3, PJA1, and DACH2 loci. Overall, this well-powered XWAS found no genetic risk factors for AD on the non-pseudoautosomal region of the X-chromosome, but it identified suggestive signals warranting further investigations.

• MeSH
• Alzheimer Disease * genetics   MeSH
• Genome-Wide Association Study methods   MeSH
• Genetic Predisposition to Disease genetics   MeSH
• X Chromosome Inactivation genetics   MeSH
• Polymorphism, Single Nucleotide genetics   MeSH
• Humans MeSH
• Chromosomes, Human, X * genetics   MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Case-Control Studies MeSH
• Check Tag
• Humans MeSH
• Male MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH

Among the more than 90 identified genetic risk loci for late-onset Alzheimer's disease (AD) and related dementias, the apolipoprotein E gene (APOE) ε2/ε3/ε4 polymorphism remains the longstanding benchmark for genetic disease risk with a consistently large effect across studies1-10. Despite this massive signal, the exact mechanisms for how ε4 increases and for how ε2 decreases dementia risk is not well-understood. Importantly, recent trials of anti-amyloid therapies suggest less efficacy and higher risks of severe side effects in s4 carriers11-13, hampering the treatment of those with the highest unmet need. To improve our understanding of the genetic architecture of AD in the context of its main genetic driver, we performed genome-wide association studies (GWASs) stratified by ε4 and ε2 carrier status. Such insights may help to understand and overcome side effects, to impact clinical trial enrolment strategies, and to create the scientific basis for targeted mechanism-driven therapies in neurodegenerative diseases.

• Publication type
• Journal Article MeSH
• Preprint MeSH

Truncating genetic variants of SORL1, encoding the endosome recycling receptor SORLA, have been accepted as causal of Alzheimer's disease (AD). However, most genetic variants observed in SORL1 are missense variants, for which it is complicated to determine the pathogenicity level because carriers come from pedigrees too small to be informative for penetrance estimations. Here, we describe three unrelated families in which the SORL1 coding missense variant rs772677709, that leads to a p.Y1816C substitution, segregates with Alzheimer's disease. Further, we investigate the effect of SORLA p.Y1816C on receptor maturation, cellular localization, and trafficking in cell-based assays. Under physiological circumstances, SORLA dimerizes within the endosome, allowing retromer-dependent trafficking from the endosome to the cell surface, where the luminal part is shed into the extracellular space (sSORLA). Our results showed that the p.Y1816C mutant impairs SORLA homodimerization in the endosome, leading to decreased trafficking to the cell surface and less sSORLA shedding. These trafficking defects of the mutant receptor can be rescued by the expression of the SORLA 3Fn-minireceptor. Finally, we find that iPSC-derived neurons with the engineered p.Y1816C mutation have enlarged endosomes, a defining cytopathology of AD. Our studies provide genetic as well as functional evidence that the SORL1 p.Y1816C variant is causal for AD. The partial penetrance of the mutation suggests this mutation should be considered in clinical genetic screening of multiplex early-onset AD families.

• Keywords
• 3Fn-domain, SORL1-associated Alzheimer’s disease, SORLA, dimerization, retromer,
• MeSH
• Alzheimer Disease * genetics   metabolism   pathology   MeSH
• Endosomes * metabolism   MeSH
• HEK293 Cells MeSH
• Middle Aged MeSH
• Humans MeSH
• Membrane Transport Proteins * genetics   metabolism   MeSH
• Mutation, Missense MeSH
• Protein Multimerization MeSH
• LDL-Receptor Related Proteins * genetics   metabolism   MeSH
• Pedigree * MeSH
• Aged MeSH
• Protein Transport MeSH
• Check Tag
• Middle Aged MeSH
• Humans MeSH
• Male MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH
• Names of Substances
• Membrane Transport Proteins * MeSH
• LDL-Receptor Related Proteins * MeSH
• SORL1 protein, human MeSH Browser

Across multiancestry groups, we analyzed Human Leukocyte Antigen (HLA) associations in over 176,000 individuals with Parkinson's disease (PD) and Alzheimer's disease (AD) versus controls. We demonstrate that the two diseases share the same protective association at the HLA locus. HLA-specific fine-mapping showed that hierarchical protective effects of HLA-DRB1*04 subtypes best accounted for the association, strongest with HLA-DRB1*04:04 and HLA-DRB1*04:07, and intermediary with HLA-DRB1*04:01 and HLA-DRB1*04:03. The same signal was associated with decreased neurofibrillary tangles in postmortem brains and was associated with reduced tau levels in cerebrospinal fluid and to a lower extent with increased Aβ42. Protective HLA-DRB1*04 subtypes strongly bound the aggregation-prone tau PHF6 sequence, however only when acetylated at a lysine (K311), a common posttranslational modification central to tau aggregation. An HLA-DRB1*04-mediated adaptive immune response decreases PD and AD risks, potentially by acting against tau, offering the possibility of therapeutic avenues.

• Keywords
• Alzheimer’s dementia, HLA, Parkinson’s disease, autoimmunity, neurodegeneration,
• MeSH
• Alzheimer Disease * genetics   MeSH
• Histocompatibility Antigens MeSH
• HLA Antigens MeSH
• HLA-DRB1 Chains * genetics   MeSH
• Humans MeSH
• Parkinson Disease * genetics   MeSH
• Check Tag
• Humans MeSH
• Publication type
• Journal Article MeSH
• Research Support, Non-U.S. Gov't MeSH
• Research Support, N.I.H., Extramural MeSH
• Names of Substances
• Histocompatibility Antigens MeSH
• HLA Antigens MeSH
• HLA-DRB1 Chains * MeSH
• HLA-DRB1*04 antigen MeSH Browser

IMPORTANCE: An estimated 40% of dementia is potentially preventable by modifying 12 risk factors throughout the life course. However, robust evidence for most of these risk factors is lacking. Effective interventions should target risk factors in the causal pathway to dementia. OBJECTIVE: To comprehensively disentangle potentially causal aspects of modifiable risk factors for Alzheimer disease (AD) to inspire new drug targeting and improved prevention. DESIGN, SETTING, AND PARTICIPANTS: This genetic association study was conducted using 2-sample univariable and multivariable mendelian randomization. Independent genetic variants associated with modifiable risk factors were selected as instrumental variables from genomic consortia. Outcome data for AD were obtained from the European Alzheimer & Dementia Biobank (EADB), generated on August 31, 2021. Main analyses were conducted using the EADB clinically diagnosed end point data. All analyses were performed between April 12 and October 27, 2022. EXPOSURES: Genetically determined modifiable risk factors. MAIN OUTCOMES AND MEASURES: Odds ratios (ORs) and 95% CIs for AD were calculated per 1-unit change of genetically determined risk factors. RESULTS: The EADB-diagnosed cohort included 39 106 participants with clinically diagnosed AD and 401 577 control participants without AD. The mean age ranged from 72 to 83 years for participants with AD and 51 to 80 years for control participants. Among participants with AD, 54% to 75% were female, and among control participants, 48% to 60% were female. Genetically determined high-density lipoprotein (HDL) cholesterol concentrations were associated with increased odds of AD (OR per 1-SD increase, 1.10 [95% CI, 1.05-1.16]). Genetically determined high systolic blood pressure was associated with increased risk of AD after adjusting for diastolic blood pressure (OR per 10-mm Hg increase, 1.22 [95% CI, 1.02-1.46]). In a second analysis to minimize bias due to sample overlap, the entire UK Biobank was excluded from the EADB consortium; odds for AD were similar for HDL cholesterol (OR per 1-SD unit increase, 1.08 [95% CI, 1.02-1.15]) and systolic blood pressure after adjusting for diastolic blood pressure (OR per 10-mm Hg increase, 1.23 [95% CI, 1.01-1.50]). CONCLUSIONS AND RELEVANCE: This genetic association study found novel genetic associations between high HDL cholesterol concentrations and high systolic blood pressure with higher risk of AD. These findings may inspire new drug targeting and improved prevention implementation.

• MeSH
• Alzheimer Disease * epidemiology   genetics   MeSH
• Cholesterol, HDL MeSH
• Causality MeSH
• Humans MeSH
• Risk Factors MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Check Tag
• Humans MeSH
• Male MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH
• Research Support, Non-U.S. Gov't MeSH
• Names of Substances
• ethyl 4-azidophenyl-1,4-dithiobutyrimidate MeSH Browser
• Cholesterol, HDL MeSH