2004 Progress Report: Persistent Organic Pollutants and Endometriosis RiskEPA Grant Number: R829438
Title: Persistent Organic Pollutants and Endometriosis Risk
Investigators: Holt, Victoria L. , Barr, Dana Boyd , Chen, Chu
Institution: Fred Hutchinson Cancer Research Center
EPA Project Officer: Hunt, Sherri
Project Period: March 25, 2002 through March 24, 2005 (Extended to March 24, 2007)
Project Period Covered by this Report: March 25, 2004 through March 24, 2005
Project Amount: $966,841
RFA: Endocrine Disruptors: Epidemiologic Approaches (2001) RFA Text | Recipients Lists
Research Category: Endocrine Disruptors , Health , Safer Chemicals
Endometriosis, defined as the implantation of endometrial glands and stroma outside of the uterine cavity, is a relatively common abnormality that, in many women, may be a serious and chronic gynecologic disease with symptoms including severe pelvic and menstrual pain, dyspareunia, and infertility. There has been recent scientific and popular press speculation that exposure to chemicals affecting endocrine function may be an important risk factor for endometriosis; however, little epidemiologic research has been done to investigate this hypothesis. The goal of this project is to assess the prevalence of exposure to organic pollutants in endometriosis cases and controls in a large health maintenance organization in western Washington State and to examine the interactions of these exposures with polymorphisms in genes involved in estrogen metabolism. The specific objectives of the research project are to: (1) determine whether the risk of endometriosis is associated with lipid-adjusted serum levels of 14 organochlorine pesticides or urine levels of the pesticide methoxychlor; (2) determine whether the risk of endometriosis is associated with lipid-adjusted serum levels of polychlorinated biphenyls (total PCBs and 35 PCB congeners); (3) determine whether the risk of endometriosis resulting from organochlorine pesticide or PCB exposure differs among women with differing CYP1A1, CYP1A2, COMT, and GSTM1 genotypes; and (4) determine whether the risk of endometriosis resulting from organochlorine pesticide or PCB exposure differs among women with differing levels of other exposures affecting estrogen levels.
This study is an ancillary investigation to Women’s Risk of Endometriosis (WREN), a case-control study funded by the National Institute of Child Health and Human Development.
WREN data—including in-person interviews (reproductive, contraceptive, menstrual, behavioral, occupational, and other characteristics), dietary intake questionnaires, anthropometric measurements, and pharmacy information—and analyses of two polymorphic genes coding enzymes active in detoxification and estrogen metabolism (GSTM1, COMT) will be used in this project. Interview, measurement, and pharmacy data collection has been completed (n = 342 cases and 741 controls), the data have been entered, and initial analyses of self-reported occupational and nonoccupational exposure to potential endocrine-disrupting chemicals have been completed. Blood sample collection is complete; blood samples were obtained from 283 cases and 585 controls. The GSTM1 and COMT analyses that were part of the initial WREN project have been completed, as have the CYP1A1 and CYP1A2 polymorphism analyses that are part of the current project. All genetic analyses were performed at the Chen Laboratory of the Fred Hutchinson Cancer Research Center, Seattle, Washington. Urine sample collection is complete; urine samples were obtained from 159 cases and 301 controls. Centers for Disease Control and Prevention (CDC) human subjects approval has been obtained and analyses of blood samples for lipid-adjusted serum levels of total PCBs, PCB congeners, HCB, b-HCH, g-HCH, aldrin, hepachlor epoxide, oxychlordane, trans-nonachlor, p,p’-DDE, o,p’-DDE, dieldrin, endrin, o,p’-DDT, p,p’-DDT, and mirex residues are nearing completion at the Pesticide Laboratory, Toxicology Branch, Division of Environmental Health Laboratory Sciences, National Center for Environmental Health, CDC, Atlanta, Georgia, under the direction of Dr. Dana Barr, co-investigator. Analyses of urine samples for methoxychlor metabolite HPTE levels are in progress at the CDC.
When the laboratory analyses are completed and data are entered, all interview, anthropometric measurement, dietary, pharmacy, and laboratory data will be combined, and standard epidemiologic approaches for case-control analyses will be employed to accomplish the specific aims listed above. The measure of the association between organic pollutant exposure and endometriosis will be the relative risk as estimated by the odds ratio (OR); 95 percent confidence intervals will be used to estimate the precision of the ORs. Unconditional linear logistic regression, a method based on the linear estimation model, will be used to estimate the ORs. Separate analyses will be conducted for exposure to each of the above-listed compounds, and exposures will be modeled in at least four ways: (1) any detectable serum or urine level versus no detectable level; (2) top quartile of exposure levels (using control exposure distribution) versus the lower three quartiles combined; (3) top quartile of exposure levels (using control exposure distribution) versus the lowest quartile; and (4) in dose-response models by quartile, using the lowest quartile as the referent category. The association between endometriosis risk and CYP1A1 and CYP1A2 genotype will be evaluated in a similar manner. To determine whether there is evidence of an interaction between organic pollutant levels and CYP1A1 or CYP1A2 genotype, we will examine whether the joint relation between the chemical compound exposures and the polymorphism exceed that predicted based on a multiplicative relationship. We will also take advantage of the continuous nature of our exposure data and examine the hypothesis that the relationship between the level (or dose) of organic compound exposure and endometriosis is modified by the genetic factor. In such analyses, we will attempt to identify the best fitting dose-response relation using higher order terms and splines.