New Genetic Insights Suggest Inherited Risk for Alzheimer’s Disease May Be Underestimated

A large international team of researchers has published a comprehensive genetic analysis in the journal Nature, shedding new light on the hereditary underpinnings of Alzheimer’s disease. Drawing on genetic summary data from nearly 2.8 million individuals across diverse ancestral groups, the investigators explored how inherited risk connects to alterations in neurons and the brain’s immune cells. The findings suggest that common genetic variants may account for a larger share of disease susceptibility than previous estimation methods indicated.

Expanding the Genetic Map of Alzheimer’s Disease

While twin studies have long estimated that between 60% and 80% of susceptibility to Alzheimer’s disease is heritable, earlier genome-wide association studies had identified roughly 80 risk loci. Furthermore, previous single-nucleotide polymorphism approaches suggested that common genetic variations accounted for only about 5% of disease liability.

In the recent multi-ancestry investigation, researchers combined de-identified genetic data from 30 cohorts, encompassing more than 109,000 clinically diagnosed Alzheimer’s cases, over 74,000 proxy cases where a parent was reported to have the condition, and more than 2.1 million controls. By applying advanced analytical methods, the team identified 127 significant genetic loci overall, including 48 newly discovered risk regions. The authors noted that European ancestry accounted for approximately 91% of the effective sample size, while other ancestry groups had smaller sample sizes, leading to less precise estimates.

Cellular Pathways and Potential Targets

The study’s gene-set and cell-type analyses highlighted specific biological pathways and cellular mechanisms associated with the condition. The results pointed to significant genetic enrichment in microglia—the brain’s resident immune cells—particularly within inflammatory microglial states and genes related to phagocytosis, such as SYK, AXL, and PICALM. Additionally, certain neuronal populations showed expression differences linked to genetic associations, though the authors emphasized that it remains uncertain whether these neuronal expression variations actively contribute to the disease process or occur downstream of neurodegeneration.

Among the newly identified loci, the researchers prioritized several genes that interact with existing pharmacological agents, highlighting potential biological mechanisms for future exploration:

• RRM2B, SYK, and KEAP1, which have shown promising experimental findings in earlier research.
• CACNA1S and AXL, noting that previous human and mouse interventions involving these targets did not improve clinical outcomes.
• QPCT, which connects to an amyloid-beta modification targeted by specific monoclonal antibody treatments.

The investigators also observed genetic overlaps between Alzheimer’s disease and other neurodegenerative conditions, such as Parkinson’s disease and amyotrophic lateral sclerosis, pointing to shared causal variants at specific loci like UBE2V1-SPATA2.

Heritability Estimates and Risk Prediction

Utilizing a specialized method known as SBayesRC, the researchers calculated an estimated single-nucleotide polymorphism-based heritability on the liability scale of 16% for European ancestry cohorts, compared to 6% using traditional linkage disequilibrium score regression. Estimates stood at 13% for East Asian ancestry and 10% for African ancestry, though the differences between ancestry groups were not statistically established, and the African-ancestry estimate did not achieve statistical significance.

When constructing polygenic scores, a multi-ancestry score explained an average of 12.8% of liability variance in European cohorts beyond baseline demographic models, reaching up to 16.9% in one clinical cohort. Among European participants aged 65 and older in the UK Biobank, observed disease prevalence was notably higher in the top polygenic score decile compared to the bottom decile. However, the study authors cautioned that prediction was significantly weaker in non-European groups and emphasized that these findings do not establish the clinical utility of polygenic score-based screening. Larger, more diverse datasets are required to ensure equitable and reliable risk assessment across all populations.