Allele frequency trajectories across age groups reveal ongoing natural selection shaping disease susceptibility.

Understanding how selection shapes disease risk remains challenging. Variants influencing complex traits, including common diseases, can also impact fitness and thus be constrained by purifying selection. Consequently, genetic variance underlying disease susceptibility may be attributed to low-frequency, population-specific variants. We analyzed 509,817 genome-wide variants from 72,635 Han Taiwanese individuals to identify loci showing age-dependent allele frequency shifts that signal ongoing selection. After adjusting for potential age-related population structure, we detected 168 variants deviating from neutrality, with most showing declining frequencies in younger generations, consistent with purifying selection on deleterious alleles influencing disease risk. These variants were enriched for rare alleles (≤0.1%) and disease-associated variants. At BRCA1, we identified 16 rare pathogenic variants in strong linkage disequilibrium undergoing purifying selection that coexist with a positively selected haplotype, revealing temporally fluctuating selection; comparable patterns at BRCA2 and MLH1 suggest recurrent selective trade-offs in DNA repair genes. Phenome-wide association analysis across 30 hematologic and cardiometabolic traits linked a subset of candidates to increased erythrocyte volume and reduced hemoglobin concentration, suggesting subclinical physiological effects. These results demonstrate ongoing natural selection on disease-relevant variation, particularly affecting hematologic traits in the Han Taiwanese population, and highlight opportunities to refine precision-medicine risk models.