Scientists build a more accurate Alzheimer’s risk score using global DNA data
Standard genetic tests for Alzheimer's often fail in non-European populations. A new study shows that blending global genetic data creates a highly accurate tool that predicts both Alzheimer's risk and underlying brain damage across diverse ancestral groups.
Scientists have developed a novel method to create a more precise Alzheimer's risk score by utilizing genetic data from various global populations. This approach overcomes the limitations of traditional genetic risk scoring methods that primarily rely on European genetic information, which can result in inaccurate predictions for individuals of African or Asian descent.
The study, published in Alzheimer’s & Dementia, involved researchers from the University of California, San Francisco, who aimed to develop a cross-ancestry risk score that could accurately predict both the clinical diagnosis and the underlying brain damage associated with Alzheimer's disease in a diverse group of older adults.
The researchers collected genetic summary data from hundreds of thousands of individuals of European, African, admixed American, East Asian, and Caribbean Hispanic descent. They constructed three different types of polygenic risk scores: a single-ancestry score using only European genetic data, a multi-ancestry score pooling all genetic data from the different groups, and a cross-ancestry score—a more sophisticated statistical model that jointly analyzed the data, adjusting the weight of each genetic variant based on shared genetic effects across populations.
Ancestry normalization was applied to ensure fairness in scaling across different genetic backgrounds.
To evaluate the effectiveness of these scores, the researchers applied them to two large U.S. cohorts of patients with varying demographic backgrounds. The first cohort, from the Alzheimer’s Disease Sequencing Project, consisted of 19,398 individuals of European, admixed American, and African descent. The second cohort, from the Health and Aging Brain Study–Health Disparities, included 2,559 individuals of the same backgrounds.
The analyses controlled for factors such as age, sex, and the APOE gene, which is strongly linked to Alzheimer’s risk.
The findings revealed that the single-ancestry score, based solely on European genetic data, performed poorly outside of European populations, particularly for African participants, failing to accurately predict Alzheimer’s risk. In contrast, the cross-ancestry score demonstrated superior predictive accuracy across all demographic groups.
For African participants, a higher cross-ancestry score was associated with a 71% increased likelihood of having an Alzheimer’s diagnosis compared to the baseline risk. Additionally, the cross-ancestry score effectively predicted both cognitive decline and physical markers of the disease, such as impaired memory and language skills, and cerebrospinal fluid protein levels, in diverse patient populations.
This innovative scoring model offers a more balanced and accurate tool for assessing Alzheimer’s risk across a wide range of global demographics.
Written by urgent.news from PsyPost's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
