Association of Polygenic Risk Score of Alzheimer’s disease with A/T/N Biomarkers and Cognitive Decline
Joon Hyung Jung, Min Soo Byun, Dahyun Yi, Gijung Jung, Hyejin Ahn, Shiwei Liu, Yen‐Ning Huang, Anthony J. Griswold, Margaret A. A Pericak‐Vance, Yu Kyeong Kim, Yun‐Sang Lee, Chul‐Ho Sohn, Koung Mi Kang, Jun‐Young Lee, Andrew J. Saykin, Kwangsik Nho, Dong Young Lee- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
Polygenic risk scores (PRS) quantify the combined effects of genetic variants to enable the prediction of an individual’s risk of developing the disease. Alzheimer’s disease (AD) is a complex disorder in which the relationship among PRS, neuropathological substrates, and related cognitive decline remain especially understudied in Asian populations. We investigated the associations of PRS with AD neuroimaging biomarkers and cognitive phenotypes in a Korean Cohort.
Method
The study included 501 older adults (264 cognitively normal [CN], 146 mild cognitive impairment [MCI], and 91 AD dementia) from the Korean Brain Aging Study for the Early Diagnosis and Prediction of Alzheimer’s Disease (KBASE). Comprehensive clinical and neuropsychological evaluations, as well as multi‐modal neuroimaging including 3T MRI and 11C‐PiB‐PET were administered. A subset of participants underwent 18F‐AV‐1451 PET (n = 167). Imaging phenotypes including beta‐amyloid (Aβ) positivity, global tau deposition, and adjusted hippocampal volume (HVa) were used for amyloid (A), tau (T), and neurodegeneration (N) biomarkers, respectively., which was imputed using the TOPMed reference panel. PRS was calculated using the summary statistics of a recent large‐scale GWAS study for AD by PRS‐CS. The Apolipoprotein E (APOE) region was excluded in the calculation. The associations between standardized PRS and imaging/cognitive phenotypes were examined using either multiple linear or logistic regression analysis with covariates (age and sex for neuroimaging biomarker phenotypes; age, sex and education for cognitive phenotypes).
Result
PRS was significantly higher in the AD dementia group compared to the CN group. Higher PRS was associated with greater CDR‐sum‐of‐boxes (β[SE] = 0.233[0.083], t = 2.818, p = 0.005) and lower MMSE scores (β[SE] = ‐0.648[0.198], t = ‐3.271, p = 0.001). In terms of AD neuroimaging biomarkers (Figure 1), higher PRS was significantly associated with an increased likelihood of Aβ positivity (OR = 1.233, 95%CI = 1.024‐1.491), and lower HVa (β[SE] = ‐132.79[49.624], t = ‐2.676, p = 0.008), but not with global tau deposition (β[SE] = 0.075[0.048], t = 1.573, p = 0.118).
Conclusion
These findings suggest that higher PRS predicts increased pathological Aβ deposition and accelerated neurodegeneration, as well as greater cognitive decline. Further investigation in larger and multiethnic samples are needed to determine generalizability to other ethnoracial populations.