ALZHEIMER’S • GENETICS • BRAIN HEALTH
Alzheimer’s: 48 New Genetic Risk Regions
A larger map of disease associations provides research leads—and calls for careful interpretation.
ADDED OCTOBER 28, 2026 • PUBLISHED IN NATURE, OCTOBER 28, 2026
Numerical verification status (October 30, 2026): The counts below are transcribed from the study summary used to prepare this article. The original Nature publication and DOI have been identified, but its full methods, Table 1, and supplements could not be independently rechecked during this audit because publisher access was restricted. Treat the precise counts and gene-level interpretations as pending direct source comparison—not as independently certified figures.
What the study found
Uffelmann and colleagues analyzed 2,815,127 participants across 30 cohorts. They identified 118 loci in a multi-ancestry analysis plus nine in ancestry-specific analyses: 127 regions, including 48 new ones.
| Group | Count |
|---|---|
| Cases | 109,479 |
| Proxy cases | 74,141 |
| Controls | 2,131,799 |
| Proxy controls | 499,708 |
Proxy cases reported parental Alzheimer’s; they were not all personally diagnosed. The effective sample size was approximately 416,493, with 91% European ancestry, so diversity remains a limitation.
Biological clues
Risk associations were enriched among genes upregulated in microglia and downregulated in neurons in cases versus controls. This supports investigating immune-cell and neuronal processes together; it does not establish one universal causal chain.
Drug-target claims: what is supported?
The authors identified 17 genes interacting with drugs approved for other conditions. Earlier research offered animal-model promise for RRM2B, SYK and KEAP1; some approaches targeting CACNA1S and AXL had unsuccessful results. The discussion also cites promising mouse evidence involving AXL, so results depend on the particular intervention and model. These are research leads, not demonstrated Alzheimer’s therapies.
Gene guide: regions, pathways and research leads
Updated October 29, 2026. These examples are not the complete list of 48 loci. An “effector gene” is a proposed gene through which an association may act.
| Genes or loci | What the paper highlights |
|---|---|
| SNX2, DOK3, SYK, GGA2, CEBPA | Connections with known Alzheimer’s genes in amyloid processing and the microglial response to amyloid-β. |
| QPCT | Newly prioritized at an already known region, replacing earlier gene nominations PRKD3 and NDUFAF7; not a newly discovered locus. |
| HDAC9, C16orf95, HIP1R, MBNL1, UBE2V1–SPATA2 | Highlighted alongside known GBA1 and TMEM163 regions in the discussion of broader neurodegenerative processes. |
Shared disease associations do not automatically establish a shared causal variant. For example, the authors found evidence against shared causal variants for HIP1R and MBNL1 across Alzheimer’s and Parkinson’s, while supporting shared-variant evidence at UBE2V1–SPATA2.
Read the paper’s results and gene-prioritization discussion →
Reading “risk region” correctly
A locus is a genomic region. An association does not automatically identify a causal variant, a single responsible gene or a treatment. Genetic risk changes probability; it is not an individual diagnosis or a guarantee of disease.
From a signal to a useful therapy
MKP interpretation: A useful way to read this research is as a sequence of questions: Is the association reproducible? Which biological process explains it? Can changing that process help? Does an intervention improve outcomes in people? Success at one stage does not answer the next.
Corrections to the shared summary
The published journal is Nature. “48 new regions” should not become “48 proven causes.” Nor should five previously investigated targets all be presented as successful or equally promising. A drug’s approval for another condition does not establish its benefit for Alzheimer’s.
Why this matters for healthy longevity
MKP perspective: Preserving cognition is central to healthspan. A more detailed research map can improve the questions scientists ask, while meaningful progress toward treatment still needs functional experiments and clinical evidence.
Sources
- Uffelmann, E., Wightman, D. P., Bahrami, S. and colleagues. Alzheimer’s disease genetics converges on microglial and neuronal dysregulation. Nature (2026). The article’s Table 1 and supplementary files document the new loci and gene prioritization.
- Alzheimer’s Association: Genetics and Alzheimer’s — background on risk genes versus deterministic mutations.
Discuss the evidence
Why might researchers use parental-history proxies? What would make a drug-repurposing lead convincing? Why does ancestry representation matter for risk prediction?
Join the MKP discussion →Related research
Explore connected evidence and questions across the Murphy Knowledge Project.