Polycystic ovary syndrome (PCOS) is denoted by hyperandrogenism and symptoms of ovulation
dysfunction (e.g., infertility, oligomenorrhea). Women who are diagnosed with PCOS represent
the more extreme end of abnormal ovarian function. While PCOS is rare, approximately 20
percent of women are expected to have subclinical polycystic ovaries, identified by
sonogram. Factors that are associated with PCOS, insulin resistance and dyslipidemia, are
also risk factors for coronary artery disease and there is evidence to suggest that women
with PCOS are more likely to experience coronary artery disease. This raises the question:
are subclinical cases of polycystic ovaries at increased risk of coronary artery disease?
The study is ancillary to the the Coronary Artery Risk Development in Young Adults (CARDIA)
Study, a large cohort study supported by the NHLBI. The CARDIA population provides a unique
opportunity to explore these associations because there is a 15-year period of risk factor
ascertainment in this population including stored blood + DNA specimens from which serum
androgens and relevant genetic markers can be ascertained. In addition, all participants
will have coronary artery calcium (CAC) determinations at the year 15 exam. In the ancillary
study an additional year 16 visit will occur for the purpose of obtaining a trans vaginal
ultrasound determination (TVUS) so as to obtain clinical evidence of hyperandrogenism and
look at the joint effects of the clinical evidence and serum markers as predictors of CAC.
The study is in response to an initiative on Ancillary Studies in Heart, Lung, and Blood
Disease Trials released in June, 2000.
The ancillary study in year 16 of CARDIA will recall 1,200 women who received a coronary
artery calcification (CAC) determination in year 15. These women will have an additional
blood draw three to 10 days after their last menstrual period, will have a questionnaire
administered to them regarding clinical symptoms of PCOS (polycystic ovarian syndrome) and
other aspects of hyperandrogenism. Two principal analyses will be performed. The first is a
longitudinal analysis using year 2 and year 10 data to use serum markers of hyperandrogenism
and genetic markers from DNA samples as predictors of CAC which is defined as present
(absent) and is determined at the year 15 examination. The principal method of analysis will
be logistic regression. An additional analysis will relate androgen levels at year 2 to
lipid levels and other coronary risk factors ascertained at year 2, 7, 10 and 15.
Observational Model: Defined Population, Observational Model: Natural History
University of Washington
United States: Federal Government