Abnormalities in androgens and ovarian markers in reproductive-age racially and ethnically diverse women in a prospective longitudinal cohort

University Of Michigan At Ann Arbor, Ann Arbor MI

Investigators

Linked publications & trials

Abstract

Polycystic ovary syndrome (PCOS) is the most common endocrinopathy among reproductive-age women and characterized by hyperandrogenemia, irregular menses, and enlarged polycystic ovaries. The genetic architecture of PCOS also includes obesity and insulin resistance, and thus women with PCOS often have elevated body mass index (BMI) and glucose intolerance. In October 2021, the NIH convened a PCOS workshop; co-I Siscovick reported on knowledge gaps regarding PCOS and cardiovascular disease (CVD). First, best practices for identification of abnormal androgens, menses, and ovarian markers are unclear, particularly as women age. Cutpoints for abnormal values are arbitrary and based upon expert opinion; the predictive value of serum ovarian markers compared to radiographic markers is not accepted; and optimal tests for blood androgens is uncertain. Second, it is not understood how PCOS characteristics and subphenotypes differ by race and ethnicity; socioeconomic status (SES); and potentially modifiable factors such as BMI and oral contraceptive use (OCP). Third, it is unclear to what extent PCOS characteristics predict risk for cardiometabolic disease beyond standard risk factors. Similarly, it is unknown if PCOS subphenotypes confer different risks of cardiometabolic disease. These issues are best addressed using population-based prospective cohorts spanning reproductive age to older age, as such studies minimize biases in referral and ascertainment of outcomes. However, such studies are few. The Coronary Artery Risk Development in Young Adults Study (CARDIA) and the Dallas Heart Study (DHS) are two population-based, prospective, longitudinal cohorts which first sampled Black and white participants during their reproductive years (age 18-30 years in CARDIA, age 35-49 years in DHS) and have followed their cohorts to late middle age. The studies used identical questionnaires for ascertainment of menstrual history and exogenous estrogen use. Both studies have deeply phenotyped women with repeated measures of traditional and novel CVD risk factors, intermediate atherosclerotic markers including coronary artery calcification (CAC), adjudication of CVD events, and radiographic assessment of ovaries. Our team, consisting of PIs of these studies, and experts well-versed in the controversies of PCOS, will build upon this rich infrastructure. We will use existing data and samples from reproductive-age women in both studies (n=1163 in CARDIA, approximately half of whom were Black; n=827 in DHS, approximately two-thirds of whom were Black, 25% non-Hispanic white, and 10% Latina). We will measure androgenic metabolites including 11 oxy-androgens using state-of-the-art mass spectrometry and anti-Müllerian hormone (AMH) and sex hormone binding globulin (SHBG). We will apply machine learning strategies to generate cutpoints and PCOS groupings, compare these to conventional definitions and phenotypes, and examine associations by racial and ethnic group, SES, and OCP use. Finally, we will examine prospective associations with cardiometabolic risk, particularly events. No additional data collection is needed.