A genetic risk score is associated with polycystic ovary syndrome-related traits

Abstract

Is a genetic risk score (GRS) associated with polycystic ovary syndrome (PCOS) and its related clinical features? The GRS calculated by genome-wide association studies (GWASs) was significantly associated with PCOS status and its related clinical features. PCOS is a heterogeneous disorder and is characterized by oligomenorrhea, hyperandrogenism and polycystic ovary morphology. Although recent GWASs have identified multiple genes associated with PCOS, a comprehensive genetic risk study of these loci with PCOS and related traits (e.g. free testosterone, menstruation number/year and ovarian morphology) has not been performed. This study was designed as a cross-sectional case-control study. We recruited 862 women with PCOS and 860 controls. Women with PCOS were divided into four subgroups: (1) oligomenorrhea + hyperandrogenism + polycystic ovary, (2) oligomenorrhea + hyperandrogenism, (3) oligomenorrhea + polycystic ovary and (4) hyperandrogenism + polycystic ovary. Genomic DNA was genotyped for the PCOS susceptibility loci using the HumanOmni1-Quad v1 array. Venous blood was drawn in the early follicular phase to measure baseline metabolic and hormonal parameters. A GRS was calculated by summing the number of risk alleles from 11 single-nucleotide polymorphisms (SNPs) that were identified in previous GWASs on PCOS. A weighted GRS (wGRS) was calculated by multiplying the number of risk alleles for each SNP by its estimated effect (beta) obtained from the association analysis. The GRS was higher in women with PCOS than in controls (8.8 versus 8.2, P < 0.01) and was significantly associated with PCOS after adjusting for age and BMI. An analysis of GRS quartiles (Q1 = 3-5, Q2 = 6-8, Q3 = 9-11, Q4 = 12-15) revealed that the subjects in the highest quartile showed a remarkable increased risk of PCOS compared with those in the lowest quartile (odds ratio = 6.28, P < 0.001). Free testosterone level, menstruation number per year, ovarian volume and ovarian follicle numbers were significantly associated with the GRS (in all cases, P < 0.01). The wGRS yielded similar results. We used 11 loci for the calculation of GRS, but a higher number of PCOS risk alleles was reported in previous studies. Therefore, further studies should assess the value of GRS including the additional SNPs related to PCOS. Although a GRS of a parts per thousand yen12 was significantly associated with PCOS, the GRS showed a poor predictive value; therefore, the use of genetic information based on current GWAS data only may present problems. The GRS could be used to identify asymptomatic individuals among people at risk and stratify them into accurate risk categories for the purpose of individualizing treatment approaches, which could potentially improve health outcomes. None of the authors have any conflicts of interest to declare. No funding was obtained for the study.