# Identification of genetic modifiers of autosomal dominant Alzheimer's disease: a genome-wide association study.

- **DOI:** [10.1016/s1474-4422(26)00123-7](https://doi.org/10.1016/s1474-4422(26)00123-7)
- **Rating:** Limited (2.8 out of 5)
- **Verdict:** A hypothesis-generating ADAD GWAS identifies promising modifier candidates, but its carrier-versus-non-carrier design cannot establish mutation-independent effects.
- **Framework version:** v5.9
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-10-08
- **Evaluation page:** https://www.nabu.science/paper/10.1016%2Fs1474-4422(26)00123-7

## Key claims

1. Three genome-wide loci with significant risk were associated with ADAD risk, irrespective of the specific ADAD gene mutation.
   - Support: Tentative
   - Type: Primary finding
   - Basis: non-carrier controls and gene-concentrated alleles do not establish mutation-independent modifier effects
   - Location: Abstract, Findings
2. The CCNG1 variant is associated with earlier onset, potentially through the acceleration of pathogenic processes such as TDP-43 aggregation
   - Support: Tentative
   - Type: Secondary
   - Basis: the onset association rests on nine carriers and the proposed TDP-43 mechanism remains unestablished
   - Location: Discussion, summary paragraph
3. The RHOJ variant is associated with CSF Alzheimer’s disease biomarkers
   - Support: Tentative
   - Type: Secondary
   - Basis: the small biomarker analysis lacks disease-stage adjustment and described multiplicity control
   - Location: Discussion, summary paragraph

## Strengths

- **Broad ADAD modifier research question** The study combines Knight-ADRC, DIAN, and ADSP data to investigate modifier signals across APP, PSEN1, and PSEN2 mutation carriers rather than focusing on one family alone.
  - Dimension: contribution
  - Location: Introduction, paragraphs 2–4; Methods, Study Design and Participants
- **Multiple sensitivity and replication analyses** The GWAS uses Firth regression, adjusts for cohort and ancestry components, repeats sentinel-variant analyses with related participants using SAIGE, and tests an independent ADSP R5 cohort.
  - Dimension: methodological rigour
  - Location: Methods, Genome-Wide Association Analysis
- **Layered functional phenotype follow-up** The study evaluates age at onset, plasma proteins, cis-regulatory evidence, CSF biomarkers, and neuroimaging phenotypes, generating specific hypotheses around CCNG1, RHOJ, and CNIH4.
  - Dimension: contribution
  - Location: Methods, Protein QTL through Neuroimaging Analyses; Results, functional-characterisation paragraphs

## Limitations

- **Modifier interpretation remains structurally confounded** Cases are symptomatic ADAD mutation carriers, while controls are non-carriers; the design therefore cannot cleanly distinguish disease modification from mutation or lineage background. The uneven allele distribution by ADAD gene makes this concern material.
  - Dimension: methodological rigour
  - Location: Methods, Study Design and Participants; Results, GWAS paragraph reporting carrier counts
- **Replication and carrier counts are limited** CCNG1 and CNIH4 each involve nine discovery carriers, and only RHOJ reaches nominal significance in the 34-case replication cohort. CCNG1 remains only nearly genome-wide significant in the combined meta-analysis.
  - Dimension: contribution
  - Location: Results, GWAS and replication paragraph
- **Key limitation omitted from discussion** The limitations section addresses power, winner’s curse, and ancestry but not the use of non-carrier controls or the possibility of mutation- or lineage-linked signals. This omission leaves the central modifier claim insufficiently qualified.
  - Dimension: positioning
  - Location: Discussion, limitations paragraph

## Reliability checks

Caveats (4 of 4 checks).

- **Results check:** Clean. The reported numbers and statistics are consistent with the stated results.
- **Reporting and transparency:** Noted. Principal association probabilities are reported as p<0·0001, which is insufficiently precise to confirm the stated genome-wide significance threshold but does not contradict it.
  - Conflict of interest: not stated · Ethics approval: declared · Data availability: declared · Code availability: not stated
- **Reference integrity:** Clean
  - 31 references in manuscript
  - 31 of 31 checkable references found in an index
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Clean
- **Summary:** No retraction or expression of concern found in Crossref (includes Retraction Watch).
- **Last checked:** 2026-10-17
