Left-handedness, learning disability, autoimmune disease, and seizure history influence age at onset and phenotypical targeting of Alzheimer’s disease

Zachary Miller, Rik Ossenkoppele, Neill R. Graff‐Radford, Isabel Elaine Allen, Wendy Shwe, Lynne Rosenberg, Dustin J. Olguin, Michael Erkkinen, Paul Butler, Salvatore Spina, Jennifer S. Yokoyama, Rahul S. Desikan, Philip Scheltens, Wiesje M. van der Flier, Yolande A.l. Pijnenburg, Emma E. Wolters, Rosa Rademakers, Daniel H. Geschwind, Joel H. Kramer, Howard J. Rosen, Katherine P. Rankin, Lea T. Grinberg, William W. Seeley, Virginia E. Sturm, David C. Perry, Bruce L. Miller, Gil D. Rabinovici, Maria Luisa Gorno‐TempinimedRxiv 2022 · DOI

Abstract

While traditional Alzheimer’s disease (AD) research has long focused on late-life vascular risk factors and the APOE-ɛ4 genotype, these factors often fail to explain the clinical presentation of early-onset (EOAD) and non-amnestic variants. This study investigates a suite of novel risk factors—non-right-handedness, learning disabilities, seizures, and autoimmune disease—across a large, heterogeneous cohort. The findings suggest that these neurodevelopmental and neuroenvironmental factors significantly influence both the age of symptom onset and the specific cognitive networks targeted by the disease.

Introduction: Beyond Memory and Aging

The prevailing model of Alzheimer’s disease is largely built upon late-onset Alzheimer’s disease (LOAD), typically characterized by memory impairment in patients over age 65. However, early-onset cases and "non-amnestic" variants—which may first manifest as language difficulties (logopenic aphasia), visual processing issues (posterior cortical atrophy), or executive dysfunction—remain poorly understood. Traditional risk factors like hypertension, diabetes, and hyperlipidemia are often absent in these younger patients.

This research addresses the biological "why" behind these atypical presentations, proposing that early-life brain development and subsequent immune or environmental insults create a unique susceptibility to neurodegeneration.

Methodology

The researchers employed a two-phase retrospective case-control design:

  1. Discovery Phase: Demographic and health histories were analyzed for 1,500 patients at the UCSF Memory and Aging Center (750 EOAD and 750 LOAD).
  2. Validation Phase: To ensure generalizability, the findings were tested against external datasets including the National Alzheimer’s Coordinating Center (NACC), Amsterdam University Medical Center, and the Mayo Clinic, Jacksonville, totaling over 10,000 participants.

The study compared "typical" risk factors (vascular health, APOE-ɛ4 status) against "novel" factors: non-right-handedness, learning disability (LD), history of seizures, and autoimmune disease.

Key Insights: The "Two-Hit" Model

A central finding of the study is the grouping of these novel factors into two distinct components via Principal Component Analysis (PCA):

  • Neurodevelopmental Factors: Non-right-handedness and learning disabilities.
  • Neuroenvironmental/Immune Factors: Seizures and autoimmune diseases.

The authors propose a "two-hit" hypothesis: developmental vulnerabilities in specific brain networks (the first hit) may render those regions susceptible to chronic inflammation or hyperexcitability later in life (the second hit). This synergy appears to dictate where the disease strikes and how early it begins.

Results and Impact on Age of Onset

The data revealed a striking contrast between early-onset and late-onset risk profiles. EOAD patients showed significantly higher frequencies of all novel factors compared to LOAD patients. For instance, in the UCSF cohort, learning disabilities were present in 11.1% of EOAD patients compared to only 2.5% of LOAD patients.

Exponential Acceleration

The most dramatic finding was the compounding effect of these factors on the age of symptom onset. Survival analysis showed:

  • Zero novel factors: Median onset of 71 years.
  • One novel factor: Median onset of 68 years.
  • Two or more factors: Median onset plummeted to 59 years.

This 12-year acceleration suggests that the combination of neurodevelopmental and immune factors is a more potent driver of early disease than traditional genetic markers like APOE-ɛ4 in certain populations. Interestingly, in non-amnestic patients already possessing a novel risk factor, the presence of APOE-ɛ4 provided no additional decrease in the age of onset, suggesting these novel factors operate through independent biological pathways.

Phenotypical Targeting

The study found that these novel factors do not just make the disease happen sooner; they influence the "phenotype" or clinical manifestation. These factors were exceptionally prevalent in non-amnestic variants. This suggests that early-life differences in brain "wiring" (indicated by handedness or learning differences) may determine which neural networks are most vulnerable to amyloid and tau pathology decades later.

Redefining Clinical Cutoffs

The research also challenges the traditional age cutoff for EOAD. While 65 is the historical standard, the researchers found that a cutoff of 70 years better captures the statistical divergence between patients driven by vascular/genetic risk and those driven by neurodevelopmental/immune factors.

Limitations and Future Directions

While the study is robust, it relies on retrospective chart reviews, which may be subject to ascertainment or survivor bias (e.g., individuals with severe early-life conditions might not survive to be included in LOAD cohorts). Furthermore, while a subset of the cohort had pathologically confirmed AD, future prospective studies are needed to confirm these trends across broader populations.

Conclusion

By expanding the lens of Alzheimer’s risk to include developmental and immune histories, this study provides a new framework for understanding neurodegenerative susceptibility. Incorporating detailed histories of learning differences, handedness, and autoimmune conditions into standard clinical intake could significantly improve early detection and offer new windows for therapeutic intervention before cognitive decline begins.

Summary updated

Abstract

Abstract Background Risk factors associated with sporadic non-amnestic and early-onset Alzheimer’s disease remain underexamined. We investigated a large, clinically heterogeneous Alzheimer’s disease cohort for frequencies of established Alzheimer’s disease risk factors (hypertension, hyperlipidemia, diabetes mellitus, APOE -ɛ4 frequency, and years of education), alongside a suite of novel factors with historical theoretical association (non-right-handedness, learning disability, seizures, and autoimmune disease). Methods In this case-control study, we screened the demographic and health histories of 750 consecutive early-onset and 750 late-onset Alzheimer’s disease patients from the University of California San Francisco Memory and Aging Center for the prevalence of conventional risk and novel Alzheimer’s disease factors and compared these results with 8,859 Alzheimer’s disease individuals from the National Alzheimer’s Coordinating Center, Amsterdam University Medical Center, Amsterdam, and Mayo Clinic, Jacksonville. Results Early-onset Alzheimer’s disease was associated with significantly lower frequencies of established risk factors (hypertension, hyperlipidemia, diabetes mellitus, all p <0.001, APOE -ɛ4, p =0.03) and significantly higher frequencies of novel factors (non-right-handedness, learning disability, active seizure, all p <0.001, remote seizure, p =0.002, and autoimmune disease, p =0.007). Logistic regressions predicting EOAD vs. LOAD controlling for sex, education, APOE -ɛ4 status, typical, and novel risk factors, produced findings consistent with the above. Principal component analysis loaded novel factors into two components, non-right-handedness and learning disability versus seizure and autoimmune disease, and the combination of factors from both components resulted in an exponential decrease in age at onset from any single factor alone. APOE -ɛ4 provided no additional contribution to age at onset decreases within the non-amnestic Alzheimer’s disease cohort but shifted the age of onset 3 years earlier within amnestic presentations ( p =0.013). Conclusions We identified non-right-handedness, learning disability, seizures, and autoimmune disease as novel factors that affect both the age at onset and phenotypical targeting of Alzheimer’s disease. Together these results support a new theoretical framework of neurodegenerative disease susceptibility and that through the collection of detailed developmental and health history, neurodegenerative disease risk in some may be highly predictable, offering new opportunities towards early detection, monitoring, therapeutic intervention, and ultimately disease prevention.
BibTeX
@misc{miller2022left,
  author = {Zachary Miller and Rik Ossenkoppele and Neill R. Graff‐Radford and Isabel Elaine Allen and Wendy Shwe and Lynne Rosenberg and Dustin J. Olguin and Michael Erkkinen and Paul Butler and Salvatore Spina and Jennifer S. Yokoyama and Rahul S. Desikan and Philip Scheltens and Wiesje M. van der Flier and Yolande A.L. Pijnenburg and Emma E. Wolters and Rosa Rademakers and Daniel H. Geschwind and Joel H. Kramer and Howard J. Rosen and Katherine P. Rankin and Lea T. Grinberg and William W. Seeley and Virginia E. Sturm and David C. Perry and Bruce L. Miller and Gil D. Rabinovici and Maria Luisa Gorno‐Tempini},
  title = {{Left-handedness, learning disability, autoimmune disease, and seizure history influence age at onset and phenotypical targeting of Alzheimer’s disease}},
  year = {2022},
  doi = {10.1101/2022.12.17.22283307},
  howpublished = {medRxiv},
}
Figure