Phenome-wide association study of male and female sex chromosome trisomies in 1.5 million participants of MVP, FinnGen, and UK Biobank

Abstract

Sex chromosome trisomies (SCTs) are the most common whole-chromosome aneuploidy in humans. Yet, our understanding of the prevalence and associated health outcomes is largely driven by observational studies of clinically diagnosed individuals, resulting in a disproportionate focus on 47,XXY and associated hypogonadism. We analyzed microarray intensity data of sex chromosomes for 1.5 million individuals enrolled in three large cohorts—the Million Veteran Program, FinnGen, and UK Biobank—to identify individuals with 47,XXY, 47,XYY, and 47,XXX. We examined disease conditions associated with each SCT by performing phenome-wide association studies using electronic health records for each cohort, followed by meta-analysis across cohorts. We identified 2,769 individuals with SCTs (47,XXY: 1,319; 47,XYY: 1,108; and 47,XXX: 342), most of whom had no documented clinical diagnosis (47,XXY: 73.8%; 47,XYY: 98.6%; and 47,XXX: 93.6%). The identified phenotypic associations with SCT spanned all examined disease categories except neoplasms. Many associations are shared among three SCT subtypes, particularly for vascular diseases (e.g., chronic venous insufficiency [odds ratio (OR) (95% confidence interval [CI]) for 47,XXY: 4.7 (3.9,5.8), 47,XYY: 5.6 (4.5,7.0), and 47,XXX: 4.6 (2.7,7.6)]; venous thromboembolism [47,XXY: 4.6 (3.7–5.6), 47,XYY: 4.1 (3.3–5.0), and 47,XXX: 8.1 (4.2–15.4)]; and glaucoma [47,XXY: 2.5 (2.1–2.9), 47,XYY: 2.4 (2.0–2.8), and 47,XXX: 2.3 (1.4–3.5)]). A third sex chromosome confers an increased risk for systemic comorbidities, even if the SCT is not documented. SCT phenotypes largely overlap, suggesting that one or more X/Y homolog genes, possibly in the pseudoautosomal region, may underlie pathophysiology and comorbidities across SCTs.