Cross-ancestry validation of a polygenic risk score for aortic stenosis in Japanese cohorts
Y Takeji, H Tada, S Yoshida, M Nomura, M Mori, S Takashima, S Usui, K Sakata, M TakamuraAbstract
Background
Although recent genome-wide association studies (GWAS) have identified numerous genetic variants associated with aortic stenosis (AS) and enabled the development of polygenic risk scores (PRS), the utility of AS-related PRS in Japanese populations has not been sufficiently evaluated.
Purpose
To evaluate the distribution and discriminative performance of a previously reported polygenic risk score for AS (PRSAS)1 in a Japanese cohort.
Methods
We applied a previously published GWAS-based polygenic risk score for AS (PRSAS) to a Japanese cohort comprising consecutively enrolled patients with severe AS (N=93) and a population-based control cohort derived from the Shikamachi Health Improvement Practice (SHIP) genome cohort in Japan (N=901). All participants underwent genome-wide genotyping followed by stringent quality control, genotype imputation, and assessment of population stratification and batch effects. Of the 241 variants originally reported, 143 variants were available after harmonization and were used to calculate PRSAS.The distribution of PRSAS was compared between cases and controls, and the association between PRSAS and AS was evaluated using logistic regression analyses.
Results
The mean age was 81.1± 7.7 years in the AS cohort and 61.9 ± 10.7 years in controls. PRSAS was significantly higher in patients with severe AS than in controls, with a rightward shift in its distribution among AS cases (Figure 1). After adjustment for age and sex, PRSAS remained independently associated with AS (adjusted odds ratio [OR] 1.55 per 1-SD increase, 95% CI 1.13–2.16; P=0.0075). A graded dose–response relationship was observed across PRS quartiles (adjusted OR 1.41 per quartile increase; 95% CI 1.06–1.88; P-trend=0.019) (Figure 2). In a sensitivity analysis with participants aged ≥70 years, the result was consistent with main findings (adjusted OR 1.65 per quartile increase; 95% CI 1.18–2.36; P for trend=0.004).
Conclusions
In this Japanese cohort, a GWAS-derived PRS for AS was independently associated with severe AS after adjustment for age and sex, with consistent dose–response relationships observed even among older individuals. These findings support the cross-ancestry transferability of PRSAS and its potential utility for genetic risk stratification in East Asian populations.Figure 1Figure 2