Methods in a longitudinal cohort study of late reproductive age women: the Penn Ovarian Aging Study (POAS)

Background: This report describes the methods utilized in the Penn Ovarian Aging Study (POAS), which is a longitudinal cohort study of hormone dynamics and menopausal symptoms of women in the menopause transition. Methods/Design: The cohort is a community-based sample of generally healthy women enrolled in the late reproductive years. The study population is a stratified random sample of African-American and Caucasian women, identified by random digit dialing. Of the 1427 women who were identified as potentially eligible, 578 women were eligible after full screening; 75 % of the eligible women enrolled in the study (436/578). At Period 14 (14 years after study enrollment), 67 % remained active and were fully evaluated (293/436). Attrition was non-differential with respect to the sample characteristics. The aims of the project overall are to 1) identify within-woman trends of reproductive hormones (estradiol, follicle stimulating hormone, hormone, lutinizing hormone, inhibin B, dehydroepiandrosterone, testosterone, and anti-mullerian hormone), cofactors such as race, body mass index (BMI), age, physical and behavioral symptoms, and their predictions of menopausal symptoms, and patterns around the final menstrual period; 2) identify associations of hormone dynamics with physical and behavioral symptoms that occur with ovarian aging and identify racial differences in these factors; 3) identify associations of genetic polymorphisms with levels and longitudinal trends in menopausal symptoms. The cohort consists of 436 late reproductive-age women at enrollment, and now has 18 years of approximately annual follow-up assessments. Menopausal stage based on concurrent menstrual dates is identified at each follow-up period. Discussion: Studies of the cohort have shown that hot flashes can occur well before menopause and extend 10 or more years beyond menopause for sizeable numbers of women; provide evidence for new-onset depressed mood in the menopause transition and show that the final menstrual period is pivotal in the increases in depressive symptoms prior to menopause and decreases postmenopausal; suggest that poor sleep is common in the late reproductive years but increases in relation to the final menstrual period in only a small proportion of women; and show effects of obesity on reproductive hormones in the menopause transition. To date, more than 50 studies of the cohort are published in medical journals, demonstrating the relevance of these data to the clinical care of mid-life women.