Longevity & AgingResearch PaperOpen Access

RBFOX1 gene variants linked to when Alzheimer's disease begins in long-lived families

A study of long-lived families with early Alzheimer's flags RBFOX1 as a gene tied to age at onset, with supporting signals in nine other cohorts.

Tuesday, November 17, 2026 0 views
Published in Alzheimers Dement
Glowing neuron with RNA splicing machinery and a chromosome 16 DNA helix highlighting one gene segment, deep blue lab palette

Summary

Only 10-15% of early onset Alzheimer's disease (EOAD) cases are explained by mutations in APP, PSEN1 and PSEN2. Researchers studied six families from the Long-Life Family Study (LLFS), each with at least two members who developed Alzheimer's at age 65 or younger. Linkage analysis found three chromosome regions tied to age at onset. The strongest signal involving a known gene was RBFOX1, which encodes a neuron-specific RNA splicing factor (LOD 4.41). A rare RBFOX1 variant was associated with about a 5-year later onset in these families, and some variants were linked to better memory. Variants in the same gene were also associated with age at onset across nine independent early and late onset cohorts, though the specific variants differed. Blood gene expression data suggest RBFOX1 activity relates to onset timing. The work is hypothesis-generating and rests on very small discovery numbers.

Detailed Summary

Why it matters: Most early onset Alzheimer's disease (EOAD) has no known genetic explanation. Mutations in APP, PSEN1 and PSEN2 account for only 10-15% of cases, even though EOAD runs strongly in families. Finding other genes that shift when the disease starts could reveal new mechanisms and drug targets for delaying dementia.

What was studied: The team used the Long-Life Family Study (LLFS), a cohort of families selected for exceptional longevity and lower dementia risk. They identified six families with at least two members who had AD onset at age 65 or younger. The discovery sample was small: 77 family members with an EOAD-family structure, of whom 14 were EOAD cases. Using whole-genome sequencing data, they ran genome-wide multipoint linkage analysis of age at onset (age at last evaluation for unaffected relatives), then SNP association tests in the significant regions. They also tested links to a composite cognitive endophenotype and episodic memory. Findings were checked in nine independent cohorts (EFIGA, EFIGA-DR, ADSP EOAD, NACC, WRAP, ADNI, NIA-AD FBS, WHICAP, ROSMAP), plus blood RNA sequencing in LLFS and brain RNA data in ROSMAP.

Key results: Three regions reached significance (LOD of 3 or more), on chromosomes 4, 7 and 16. The highest peak was an intergenic region on chromosome 16 (SNHG9-RNF151, LOD 5.04). The strongest peak at a known gene was RBFOX1 (LOD 4.41). Within RBFOX1, the intronic variant rs76959607 had the strongest association with age at onset (P = 4.7 × 10^-5). Carriers of the G allele, with a frequency of about 5%, had onset about 5 years later (71 vs 66). Other RBFOX1 variants showed cognitive associations, with better memory in carriers of certain alleles. Every validation cohort showed an RBFOX1 association with age at onset (P from 3.4 × 10^-6 to 0.001), across both early and late onset cohorts. The top SNPs differed between cohorts and were only partly in linkage disequilibrium with the LLFS variant (D' from 0.29 to 1.0). The abstract reports that blood RBFOX1 overexpression was associated with earlier onset.

Implications: RBFOX1 encodes a neuron-specific splicing regulator already implicated in neurodevelopmental and psychiatric conditions. These data nominate it as a modifier of when Alzheimer's begins, and possibly as a target for delaying onset. The expression findings point toward lowering RBFOX1 activity rather than raising it, but this was not tested.

Caveats: The discovery sample is tiny and the key variants are rare. Effect directions and top SNPs were not uniform across cohorts. The paper's text is also inconsistent on expression direction, with the abstract saying overexpression and one summary section saying downregulation. The excerpt provided ends before the full transcriptomic results, so those details could not be verified. The findings need replication and functional follow-up before any clinical use.

Key Findings

  • Linkage analysis in six LLFS early onset Alzheimer's families found three significant regions; RBFOX1 reached LOD 4.41.
  • Rare RBFOX1 variant rs76959607 (G allele, ~5% frequency) was linked to about 5 years later onset (71 vs 66).
  • RBFOX1 variants were associated with age at onset in all nine independent early and late onset validation cohorts (P down to 3.4 × 10^-6).
  • Some RBFOX1 variants were nominally linked to better cognitive and episodic memory performance in the LLFS families.
  • Blood transcriptomics linked higher RBFOX1 expression to earlier onset, per the abstract; this was not fully verifiable in the excerpt.

Methodology

Genome-wide multipoint linkage analysis (SOLAR) of age at onset in six LLFS families with at least two EOAD cases, followed by SNP association (GMMAT) in significant regions and tests against cognitive endophenotypes. Replication used nine independent early and late onset cohorts, with blood (LLFS) and prefrontal cortex (ROSMAP) RNA expression analyses dichotomized into high and low groups.

Study Limitations

The discovery sample is very small (77 participants, 14 EOAD cases) and driven by rare variants. Validation cohorts identified different top SNPs with variable linkage disequilibrium to the LLFS variant, and effect directions were not uniform, so exact replication of one variant was not shown. Cognitive associations were only nominal, expression analyses used dichotomized levels, the paper is inconsistent on the expression direction, and the provided text was truncated before the full transcriptomic results.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: