Phytoestrogens (such as isoflavones and lignans) and organochlorine compounds (such as pesticides and polychlorinated biphenyls) are two broad classes of putative endocnne disruptors -- chemicals that may have the capability to alter a woman's hormonal milieu. The U.S. population is exposed to dietary sources of phytoestrogens through intake of soy foods, grains, and to a lesser extent fruits and vegetables. In addition, residues of organochlorine compounds can be detected in the tissue of a large proportion of the population. Mthough phytoestrogens and organochlorine compounds are suspected of being important environmental determinants of hormone-related neoplasia, there are few epidemiologic studies testing these hypotheses. Uterine fibroids are benign tumors whose growth is highly dependent on ovarian hormones, and are among the most frequently diagnosed neoplasms among women. As such, fibroids represent an ideal setting in which to test the hypothesis that environmental endocrine disruptors affect the risk of important female health outcomes. We propose to conduct a population-based case-control study to determine whether biomarkers of exposure to phytoestrogens and organochlorine compounds are related to a woman's risk of uterine fibroids. The study will be an ancillary investigation to The Uterine Leiomyomata Epidemiology Project (TULEP), an on-going study ended by the National Institutes of Child Health and Human Development. TULEP is being conducted within the population of Group Health Cooperative (GHC) of Puget Sound, a large pre-paid health plan. TULEP data collection includes an extensive in-person interview to assess medical, reproductive, menstrual, contraceptive, and other characteristics. For the proposed study, we will enroll from TULEP a subset of 175 premenopausal women with new, ultrasound-confirmed uterine fibroids (cases) and 175 premenopausal controls, frequency matched on age and race. From each of these women we will obtain a serum sample (for organochlorine analyses), an overnight urine sample (for phytoestrogen analyses), and food frequency questionnaires. In addition, in a sub-sample of 20 cases and 20 controls we will conduct an inter-method reliability study to compare phytoestrogen excretion in an overnight urine sample against a 24-hour urine sample. Serum samples will be analyzed for organochlorine residues (e.g., total PCBs, DDE, heptachlor epoxide, oxychlordane, frans-nonachlor, dieldrin, HCB, and beta-BHC) and total lipids using standard methods. Urine samples will be analyzed for excretion of isoflavones (daidzein, genistein, equol, and O-desmethylangolensin) and lignans (enterodiol and enterolactone) and urinary creatinine. Phytoestrogen excretion and organochlorine residues in cases and controls will be compared using multivariate unconditional logistic regression models.
