About

Adam Christian Naj, PhD, is an Associate Professor of Epidemiology in the Department of Biostatistics, Epidemiology, and Informatics, and the Department of Pathology and Laboratory Medicine at the University of Pennsylvania Perelman School of Medicine. His research primarily focuses on the genetics of Alzheimer’s Disease (AD) and neurodegeneration. He has contributed to genome-wide association analyses in the Alzheimer’s Disease Genetics Consortium (ADGC), resulting in a highly cited first-author paper in 2011. Since joining Penn in 2012, Dr. Naj has extended his roles in analysis and data management within the ADGC and has been actively involved in co-leading quality control and case-control analysis working groups in the Alzheimer’s Disease Sequencing Project (ADSP), which has collected data on nearly 600 whole genomes and over 10,500 whole exomes of AD cases and controls to identify rare genomic variants associated with AD risk. His recent work includes guiding the development of quality control pipelines for next-generation sequencing data as part of the Genomic Center on Alzheimer’s Disease (GCAD). Dr. Naj is also a co-founder and organizer of the annual Symposium on Advances in Genetic Epidemiology and Statistics (SAGES), promoting methodological development for genomic data analysis. His research portfolio has expanded to include studies on genetic loci contributing to multiple neurodegenerative diseases and phenotypes, including Parkinson’s disease and progressive supranuclear palsy, with the aim of identifying key genetic contributors to neurodegeneration.

Research topics

• Biology
• Genetics
• Medicine
• Internal medicine
• Bioinformatics
• Gerontology
• Computer Science
• Machine Learning
• Neuroscience
• Pathology
• Psychiatry
• Psychology

Selected publications

• Rare coding variants from ADSP R5 whole-genome sequencing implicate novel genes in Alzheimer’s disease

  Research Square · 2026-03-31

  preprintOpen access
  PublisherOA PDFDOI

• Sex and APOE Genotype Differentially Modify the Association Between Hippocampal Volume and Memory Performance

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  Abstract Background The hippocampus, vital for memory processing, is one of the first regions affected by Alzheimer's disease (AD). APOE‐ε4 (risk) and APOE ‐ε2 (protective) alleles are key genetic drivers of AD risk and cognitive decline. This study aimed to assess the modifying effect of the APOE genotype on the association between hippocampal volume and memory performance. Furthermore, we considered the modifying effects of sex, self‐reported race, and clinical diagnosis. Methods Data were obtained from 6,895 participants (mean age at baseline=75.2 years; 21% AD, 43% male, 81% non‐Hispanic white (NHW), 12% APOE‐ε2 carriers, and 40% APOE‐ε4 carriers) from four cohorts of aging and AD: ADNI, NACC, ROS/MAP/MARS, and WRAP. MRI data, processed with deep learning MUSE, included left and right hippocampal volumes, harmonized for batch differences using Longitudinal Combat. Memory composite scores were harmonized across cohorts using latent variable modeling. Linear regression on baseline memory assessed two‐way interactions between total hippocampal volume×sex, total hippocampal volume× APOE (modeled dominantly), and three‐way interactions that included sex, race, or diagnosis. Mixed effects regression models assessed these interactions on memory trajectories and included fixed and random effects for intercept and the slope (years from baseline). Results The association between hippocampal volume and memory performance is stronger in females than in males, whereby females with larger hippocampi outperform males with larger hippocampi ( p = 7.47×10 ‐13 ; R 2 =0.007; Figure 1). APOE haplotype also interacted with hippocampal volume on baseline memory, whereby the association is stronger in ε4 ( p = 6.07×10 −13 ; R 2 =0.004) carriers and attenuated in ε2 carriers ( p = 3.22×10 −5 ; R 2 =0.001), particularly among individuals with mild cognitive impairment ( p = 0.048; b=5.41×10 ‐5 ). Notably, APOE interactions with hippocampal volume were consistent across NHW and non‐Hispanic black (NHB) participants. In longitudinal analyses, we found a significant three‐way interaction between APOE‐ε2 , AD diagnosis, and hippocampal volume on memory ( p = 0.0001; R 2 c=0.8; R 2 m=0.5), driven by a stronger interaction among participants with AD ( p = 0.036; R 2 c=0.5; R 2 m=0.04). Conclusion While hippocampal volume is a potent predictor of memory performance, sex, and APOE modify this association with the most pronounced effects observed among APOE ‐ε4 carriers, women, and participants with AD. These results emphasize the importance of considering genetic, clinical, and demographic factors in AD research.

  PublisherOA PDFDOI

• Evaluation of variant‐to‐gene annotation strategies in Alzheimer's Disease pathway‐specific polygenic risk score construction

  Alzheimer s & Dementia · 2025-12-01

  articleOpen accessSenior author

  Abstract Background Pathway‐specific polygenic risk scores (pathway‐PRS) are a recently developed tool to measure genetic susceptibility to complex diseases along specific biological pathways. Positional strategies annotating risk variants to pathways may not capture genetic susceptibility in non‐coding regions. Risk variant localization to non‐coding regions is characteristic of many complex diseases, including Alzheimer's Disease (AD). We evaluated the impact of three annotation strategies on AD pathway‐PRS performance in the UK Biobank (UKB). Methods We performed pathway enrichment analysis on Kunkle et al ( Nat Genet 2019) summary statistics. Pathways meeting an adjusted p ‐value threshold were clustered based on overlapping gene content. Variants were annotated to pathway‐cluster genes based on variant position within 35kb upstream to 10kb downstream of gene boundaries (“S‐1”); S‐1 adding chromatin interaction and expression quantitative trait loci (eQTL)‐based annotations on prioritized variants (“S‐2”); and variant position within gene exon or promoter regions along with genome‐wide chromatin interaction and eQTL‐based annotations (“S‐3”). Variants annotated to pathway‐cluster genes were included in pathway‐PRS following clumping and thresholding, with tuning in an independent UKB training set (true/proxy cases=33,370; controls=229,486) and validation in a UKB testing set (true/proxy cases=8,309; controls=57,516). Results We identified 20 pathway‐clusters representing 37 Gene Ontology (GO) pathways meeting adjusted p ≤0.25. S‐2 annotated 0.5% more and S‐3 56% fewer variants to pathway‐cluster genes compared to S‐1. There was little change in odds‐ratios (OR) and incremental R 2 (Inc.R 2 ) between S‐1 and S‐2. Larger differences were observed between S‐3 and S‐1, with both increases and decreases in OR and Inc.R 2 (Figures 1 & 2). The pathway‐PRS with the largest OR under S‐1 and S‐2 (OR=1.071, p = 4.5×10 ‐09 ) contained four pathways related to regulation of amide metabolism (GO: 0034249), amyloid‐beta formation (GO: 1902430), and amyloid precursor catabolism (GO: 1902991 and GO: 1902992). The pathway‐PRS with the largest OR under S‐3 (OR=1.073, p = 2.7×10 ‐09 ) contained four protein and cellular localization pathways (GO:0060341, GO:2000009, GO:0070201, and GO:0032880) (Figure 2). Conclusion The inclusion of regulatory variants in AD pathway‐PRS lead to the prioritization of protein and cellular localization pathways over amyloid pathways. Additional strategies for pathway‐PRS construction, such as incorporating functional annotations as priors, will be tested in future work.

  PublisherOA PDFDOI

• Monocyte inflammation and resilience to Alzheimer’s disease: novel genetic risk genes identified by transcriptome-wide association study

  Translational Psychiatry · 2025-11-18 · 1 citations

  articleOpen access

  Inflammation is a key driver of Alzheimer’s disease (AD) and may connect all known AD risk factors. Recently, AD resilience outcomes have been developed which have helped to uncover mechanisms that enable some individuals to withstand significant AD pathology or genetic risk, while retaining cognitive function. We conducted a series of transcriptome-wide association studies (TWAS) focusing on monocytes, key innate immune cells that respond to pathogens and invade the CNS. Monocyte expression data under various immune stimulation states (naive, LPS 2 h, LPS 24 h, IFN 24 h) and corresponding genotype data from 432 individuals (Fairfax et al. [1]) were analyzed. We developed cis-genetic expression models using both elastic-net, and MASH combined with LD-pruning; capturing polygenic structures and independent inflammatory eQTLs across conditions, respectively. These models were applied to GWAS summary statistics of three AD resilience phenotypes: cognitive and global AD-resilience, and Amish cognitive preservation. We identified 180 TWAS results surpassing a suggestive significance threshold of PFDR < 0.20, including 92 unique genes. APP, a well-known AD gene, showed the strongest overall association, which may inform ongoing efforts targeting its action in the brain. Whole-blood RNA-seq data from a separate AD cohort confirmed differential expression between AD cases and controls in 35 putative targets, including: SURF1, ACKR3, LILRA5, FBXO2, ITPR1, and HRH4. We also demonstrate that the regulation of these genes is specific to monocytes. Finally, in-silico cell sorting (CIBERSORTx) revealed differential monocyte abundance by AD status, supporting monocyte-driven inflammation as a distinct, complementary pathway of myeloid cell involvement in AD. Together, these findings highlight monocytes as a critical and understudied cellular component for AD resilience mechanisms, with potential implications for novel immunotherapeutic strategies and precision medicine approaches in AD.

  PublisherOA PDFDOI

• NIAGADS: A data repository for Alzheimer's disease and related dementia genomics

  Alzheimer s & Dementia · 2025-06-01 · 3 citations

  articleOpen access

  The National Institute on Aging Genetics of Alzheimer's Disease Data Storage Site (NIAGADS) is the National Institute on Aging-designated national data repository for human genetics research on Alzheimer's disease and related dementias (ADRD). NIAGADS maintains a high-quality data collection for ADRD genetic/genomic research and supports genetics data production and analysis, including whole genome and exome sequence data from the Alzheimer's Disease Sequencing Project and other genotype/phenotype data, encompassing 211,000 samples. NIAGADS shares these data with hundreds of research groups around the world via the Data Sharing Service, a Federal Information Security Modernization Act moderate compliant cloud-based platform that fully supports the National Institutes of Health Genomic Data Sharing Policy. NIAGADS Open Access consists of multiple knowledge bases with genome-wide association summary statistics and rich annotations on the biological significance of genetic variants and genes across the human genome. As a one-stop access portal for Alzheimer's disease (AD) genetics, NIAGADS stands as a keystone in promoting collaborations to advance the understanding and treatment of AD. HIGHLIGHTS: The National Institute on Aging Genetics of Alzheimer's Disease Data Storage Site (NIAGADS) is a data repository for the storage of genetics and genomics data. NIAGADS houses data for Alzheimer's disease, related dementias, and healthy aging. NIAGADS offers open and qualified access data and knowledgebases to explore open access data. The Alzheimer's Disease Sequencing Project dataset is the largest Alzheimer's disease and related dementias joint called whole genome sequencing dataset (≈ 58,000 whole genomes).

  PublisherOA PDFDOI

• Evaluating the association of <i>APOE</i> genotype and cognitive resilience in SuperAgers

  medRxiv · 2025-01-07 · 1 citations

  preprintOpen access

  Abstract INTRODUCTION “SuperAgers” are oldest-old adults (ages 80+) whose memory performance more closely resembles middle-aged adults. The present study examined APOE allele frequency in non-Hispanic Black (NHB) and non-Hispanic White (NHW) SuperAgers compared to controls and Alzheimer’s disease dementia cases. METHODS In 18,080 participants from eight cohorts, harmonized clinical diagnostics and memory, executive function, and language domain scores were used to identify SuperAgers, cases, and controls across age-defined bins. RESULTS NHW SuperAgers had significantly lower frequency of APOE- ε4 alleles and higher frequency of APOE -ε2 alleles compared to all cases and controls, including oldest-old controls. Similar patterns were found in a small yet substantial sample of NHB SuperAgers; however, not all comparisons with controls reached significance. DISCUSSION We demonstrated strong evidence that APOE allele frequency relates to SuperAger status. Further research is needed with a larger sample of NHB SuperAgers to determine if mechanisms conferring resilience differ across race groups.

  PublisherOA PDFDOI

• Novel early‐onset Alzheimer‐associated genes influence risk through dysregulation of glutamate, immune activation, and intracellular signaling pathways

  Alzheimer s & Dementia · 2025-06-01 · 7 citations

  articleOpen access

  INTRODUCTION: Most genetic studies for Alzheimer's disease (AD) have been focused on late-onset AD (LOAD). There are no large genetic studies on early-onset AD (EOAD). METHODS: We performed a multi-ancestry (non-Hispanic European, African, and East Asian) genome-wide association study (GWAS) including a total of 7,349 cases and 17,887 control. Cases with age at onset younger than 70 years were included. Sensitivity analysis including cases with onset <65 was performed. Only controls older than 70 were included to decrease the risk of developing LOAD. RESULTS: We identified eight novel significant loci: six in the ancestry-specific analyses and two in the trans-ancestry analysis. By integrating gene-based analysis, expression quantitative trait loci (eQTL), protein quantitative trait loci (pQTL), and functional annotations, we nominate eight novel genes that are involved in microglia activation, glutamate production, and signaling pathways. DISCUSSION: EOAD, although sharing genes with LOAD, harbors unique genes and pathways that could be used to create better prediction models or target identification. HIGHLIGHTS: We performed the largest and first multi-ethnic genetic screening for early-onset Alzheimer's disease (AD). We identified eight novel significant loci: six in the ancestry-specific analyses and two in the trans-ancestry analysis. The novel genes are implicated microglia activation, glutamate production, and signaling pathways. EOAD, although sharing many genes with LOAD, harbors unique genes and pathways that could be used to create better prediction models or target identification for this type of AD.

  PublisherOA PDFDOI

• Analysis of 36,361 whole genome sequenced samples from the Alzheimer's Disease Sequencing Project (ADSP)

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  Abstract Background The Alzheimer's Disease Sequencing Project (ADSP) aims to uncover genomic variants that increase risk for or protect against Alzheimer's disease (AD) across various ancestral populations. The most recent ADSP whole genome sequencing (WGS) data release (R4) includes data from 36,000+ individuals across 37 study cohorts. Method Extensive quality control was conducted at the genotype, variant, and sample levels. Analyses focused on AD and related dementias (ADRDs) with cases and controls restricted to age ≥55 years. Two main genetic association approaches were applied: 1) Population‐specific analysis ‐ Genetic similarity clusters for defining populations were identified using a Gaussian mixture model. Analyses included generalized linear mixed model single‐variant tests for common variants with SAIGE; meta‐analysis of common variants across populations using METASOFT; and set‐based rare variant analyses for coding and noncoding regions with STAAR, followed by meta‐analysis using MetaSTAAR. 2) Pooled population analysis ‐ Gene‐based association tests including common and rare variants were conducted using GENESIS/R, accounting for sex, technical covariates, population structure, and empirical relationships. Rare variant analyses aggregated predicted loss‐of‐function and deleterious missense variants, with Bonferroni correction. Result For the population‐specific analysis, we defined nine clusters of individuals (8,697 ADRD cases and 14,758 controls) based on genetic similarity using high‐quality variants. The meta‐analysis of common variants identified two known ADRD‐associated loci ( APOE and CR1 ) and four novel genomic loci ( LMO1, FHIT, RPS3AP52 , and AC013762.1 ) that reached genome‐wide significance ( p &lt;5×10 ‐8 ). For the pooled analysis, common variant associations confirmed the signals in APOE and CR1 . Additionally, rare variant gene‐based analysis identified KIRREL1 ( p = 1.01×10 ‐6 ) and TNFRSF10B ( p = 2.25×10 ‐6 ) as significantly associated in population‐specific meta‐analysis and TNS1 as suggestively associated with ADRD ( p = 3.05×10 ‐6 ) in pooled analysis. Conclusion Analyses of ADSP WGS data identified a novel genome‐wide significant association at LMO1 (11p15), a gene involved in transcriptional regulation with potential relevance to neuronal health, and a suggestive association at TNS1 (2q35), previously linked to cognitive decline in non‐demented older adults. Further studies in diverse populations from the ADSP are expected to uncover additional common and rare variant associations, offering new insights into ADRD pathogenesis.

  PublisherOA PDFDOI

• Genetic variants linked to cognitive longevity in SuperAgers

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  Abstract Background “SuperAgers” are oldest‐old (ages 80+) adults with memory performance resembling adults in their 50s to mid‐60s. This study investigates the genetic drivers of SuperAging using a genome wide association study (GWAS). Method Harmonized, longitudinal memory, executive function, and language scores for participants with European ancestry were obtained from the ADSP Phenotype Harmonization Consortium. In addition to exceptional memory, SuperAgers ( N = 1,171) must have language and executive function domain scores within normal limits and remain cognitively normal across visits if longitudinal datapoints were available. We compared SuperAgers to Alzheimer's disease (AD) cases ( N = 5,372) and controls ( N = 4,012) in age‐defined subgroups (middle‐aged=ages 50‐64, old=ages 65‐79, oldest‐old=age 80+). We performed a logistic regression based GWAS comparing SuperAgers and their counterparts, adjusting for age, sex, education, and the first five principal components for population substructure. Result Comparing SuperAgers with middle‐aged cases (ages 50‐64, Figure 1), multiple variants in the confirmed AD loci APOE and BIN1 regions were associated with genome‐wide significance (GWS; indexes p = 1.12×10 ‐41 and p = 5.48×10 ‐9 , respectively). Additionally, we observed GWS association on chromosome 4 loci (rs79973832, index p = 2.95×10 ‐8 , Figure 2), which centers on a ring finger protein gene, RNF150. This family of genes are involved in the ubiquitin‐proteasome system and regulate antiviral immune responses. While several members of the RNF gene family have been liked to AD and cognitive performance, there is currently no established association between RNF150 and AD. Analyses comparing SuperAgers to old and oldest‐old cases (ages 65‐79 and 80+, respectively) found GWS only in the APOE region (indexes p = 1.90×" role="presentation" style="‐webkit‐user‐drag: none; ‐webkit‐tap‐highlight‐color: transparent; margin: 0px; padding: 0px; user‐select: text; display: inline‐table; font‐style: normal; font‐weight: normal; line‐height: normal; font‐size: 16px; font‐size‐adjust: none; text‐indent: 0px; text‐align: center; text‐transform: none; letter‐spacing: normal; word‐spacing: normal; overflow‐wrap: normal; white‐space: pre !important; float: none; direction: ltr; max‐width: none; max‐height: none; min‐width: 0px; min‐height: 0px; border: 0px; position: relative;" tabindex="0"&gt;×10 ‐80 and p = 2.53×10 ‐13 , respectively). No controls had GWS associations, with the strongest associations inconsistent among age group comparisons. Conclusion Our extreme‐phenotype GWAS comparing SuperAgers to middle‐aged cases identified both the established APOE and BIN1 genes and the novel loci rs79973832 for AD. Both Old Cases and Oldest‐old Cases reached GWS only in the APOE region. With additional harmonization efforts, larger sample sizes will allow for better comprehension of the genetic architecture of SuperAging. Future work will extend to rare variant analyses of SuperAging using Whole Genome Sequencing (WGS).

  PublisherOA PDFDOI

• NIA Genetics of Alzheimer's Disease Data Storage Site (NIAGADS): 2025 Update

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  BACKGROUND: NIAGADS is a national genomics data repository that facilitates access of genotypic and sequencing data to qualified investigators for the study of the genetics of Alzheimer's disease (AD) and related neurological diseases. Collaborations with large consortia and centers such as the Alzheimer's Disease Genetics Consortium (ADGC), Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, the Alzheimer's Disease Sequencing Project (ADSP), and the Genome Center for Alzheimer's Disease (GCAD) allow NIAGADS to lead the effort in managing large AD datasets that can be easily accessed and fully utilized by the research community. METHOD: NIAGADS is supported by National Institute on Aging (NIA) under a cooperative agreement. All data derived from NIA funded AD genetics studies are expected to be deposited in NIAGADS or another NIA approved site. NIAGADS manages a Data Sharing Service (DSS) that facilitates the deposition and sharing of genomic data and association results with approved users in the neurodegenerative disease research community. In addition, researchers are able to freely use the NIAGADS Alzheimer's Genomics Database (www.niagads.org/genomics/) to search annotation resources that link published AD studies to AD-relevant sequence features and genome-wide annotations. RESULT: As of January 2025, NIAGADS houses 130 datasets comprised of >237,000 samples including array data, sequencing, gene expression, annotations, deep phenotypes, summary statistics, among others. Qualified investigators can retrieve ADSP sequencing data with ease and flexibility through the NIAGADS DSS. To date, the ADSP and other contributing studies have completed whole exome sequencing (WES) of 20,499 samples and whole genome sequencing (WGS) of 58,507 samples. Raw WES and WGS files, quality controlled VCF files, and phenotype data files are available via qualified access. The next round of sequencing currently underway will generate around 15,000 additional genomes to be released in 2026. CONCLUSION: NIAGADS is a rich resource for AD researchers, with the goal of facilitating advances in Alzheimer's genetics research. By housing datasets from many projects and institutions, NIAGADS enables AD researchers to meet their research goals more efficiently. Datasets, guidelines, and features are available on our website at https://www.niagads.org.

  PublisherOA PDFDOI

Recent grants

• Pleiotropy GWAS of Alzheimer's Disease

  NIH · $2.1M · 2016–2022

• Core C- Biostatistics and Data Analysis Core

  NIH · $23.7M · 2016–2027

Frequent coauthors

• Margaret A. Pericak‐Vance

  Dr. John T. Macdonald Foundation

  716 shared

• Eden R. Martin

  University of Miami

  708 shared

• Jonathan L. Haines

  Case Western Reserve University

  668 shared

• Brian W. Kunkle

  University of Miami

  666 shared

• Lindsay A. Farrer

  Framingham Heart Study

  666 shared

• Gary W. Beecham

  University of Miami

  646 shared

• Richard Mayeux

  Columbia University

  638 shared

• Badri N. Vardarajan

  Columbia University Irving Medical Center

  582 shared

Education

• B.A., Biology (Genetics)/Psychology

  University of Chicago

  2000

• Ph.D., Epidemiology (Human Genetics)

  The Johns Hopkins University Bloomberg School of Public Health

  2008