2002 Progress Report: The Michigan PBB Cohort 20 Years: Endocrine Disruption?

EPA Grant Number: R825300
Title: The Michigan PBB Cohort 20 Years: Endocrine Disruption?
Investigators: Marcus, Michele , Tolbert, Paige
Current Investigators: Marcus, Michele , Blanck, Heidi Michels , Cameron, Lorraine , Henderson, Alden , Hertzberg, Vicki , Rubin, Carol , Tolbert, Paige
Institution: Emory University
Current Institution: Emory University , Centers for Disease Control and Prevention , Michigan Department of Community Health
EPA Project Officer: Hunt, Sherri
Project Period: November 29, 1996 through November 28, 1999 (Extended to December 14, 2004)
Project Period Covered by this Report: November 29, 2001 through November 28, 2002
Project Amount: $360,005
RFA: Endocrine Disruptors (1996) RFA Text |  Recipients Lists
Research Category: Economics and Decision Sciences , Endocrine Disruptors , Health , Safer Chemicals


The overall objective of this research project is to ascertain the occurrence of a number of outcomes in women and their offspring who were exposed to polybrominated biphenyls (PBBs) as a contaminant of cattle feed in 1973, to determine if there is evidence of endocrine disruption associated with PBB exposure. The Michigan Department of Community Health has followed this cohort for more than 20 years.

The study involves the administration of a telephone interview to the approximately 1900 women currently in the cohort to ascertain the cumulative incidence since exposure of endocrine sensitive endpoints such as endometriosis, benign breast disease, abnormal Pap smears, uterine fibroids, thyroid dysfunction, and infertility. We will ascertain time to pregnancy for each pregnancy following exposure. Medical records will be obtained and reviewed to verify each of the clinical conditions. The occurrence of each of these outcomes will be modeled by multivariate methods as a function of PBB serum level in l976 controlling for potentially confounding variables.

This cohort, by virtue of its well-characterized exposure to PBBs and the long follow-up period, can provide valuable data on the potential endocrine disrupting effects of halogenated hydrocarbons.

Progress Summary:

The specific accomplishments of the project include: (1) completion of the telephone interviews; (2) data cleaning activities and data set construction; (3) implementation of the medical verification process; (4) development of a mathematical model of serum polybrominated biphenyls (PBB) decay; and (5) analyses of relationships between exposure and health outcomes.

We received an unfunded extension for this project until December 14, 2002, that was necessary to complete data analyses of health outcomes in relation to PBB exposure. We had been delayed in our efforts to obtain medical records to verify clinical conditions reported by study participants. The field office in Big Rapids, MI, that maintains contact with the PBB cohort was closed in a cost-cutting move by the Michigan Department of Health while we were in the middle of the process of obtaining permission to review medical records. Personnel in Lansing took over the task, but a great deal of time was consumed by this transition. Fortunately, the medical record verification process has been completed and we are proceeding with data analyses. We are continuing to work on analyses of health outcomes in relation to exposure. The progress made for specific health outcomes is described below.

Data Collection

Telephone Interviews. The field phase of the study was completed in March 1998. A total of 1,185 telephone interviews were completed. Our overall participation rate was 85 percent. A database was generated by the Michigan Public Health Institute and provided to the investigators for cleaning and analyses and to the Michigan Department of Community Health (MDCH) for updating of the registry.

Data Cleaning Activities and Data Set Construction. Frequency distributions were generated for variables to examine outliers and possible data-entry errors. Text responses (entered verbatim by interviewers) were evaluated and responses recoded, as appropriate. Logical checks between variables were examined. Archived data collected by the MDCH also were used to clarify responses and provide data in some instances where it was missing from the telephone interview. Exposure data were obtained from the MDCH for PBB, polychlorinated biphenyls (PCB), and pesticides. Composite and recoded variables were created for specific analyses.

Medical Verification of Selected Outcomes. Study personnel reviewed the archived MDCH medical records and entered the existing confirmed conditions into a database. Diagnostic verification of reported outcomes that had not been confirmed by MDCH started in July 1998. This process now has been completed for all clinical conditions reported by participants. We have obtained medical records for diagnoses of thyroid disorders, endometriosis, abnormal pap smears, benign breast disease, adverse pregnancy outcomes, infertility, menstrual cycle abnormalities, and uterine fibroids.


We have completed analyses of several outcomes that did not require medical verification, as well as several outcomes that did involve medical verification. We have published five research papers in peer-reviewed journals and two letters to the editor. A sixth manuscript currently is under review at Maturitas. Four manuscripts are in draft form and shortly will be submitted for publication. We currently are analyzing the data for two additional endocrine-sensitive endpoints. The preliminary results for works in progress are summarized in the remainder of this report.

Age at Menopause in Relation to PBBs, PCBs, and Smoking (manuscript submitted to Maturitas).

Objectives. Because halogenated biphenyl exposure is suspected to disrupt endocrine function, we assessed time to menopause in women aged 18 years and older who were orally exposed to PBBs and PCBs (n = 874). We also examined smoking in relation to menopause.

Methods. To define menopausal status, women were interviewed in 1997, and asked whether they had had any menstrual periods in the previous year, why their menstrual periods had stopped (for example, as a result of surgery), and age at their last menstrual period. Serum PBB and PCB taken at enrollment (1976-1978) into the Michigan PBB registry was used as the measure of halogenated biphenyl exposure. Women whose menopause occurred before their PBB exposures were excluded. Proportional hazard modeling was used to analyze the “risk” for menopause in relation to exposure. Premenopausal women contributed person-time until their interview date, at which time they were censored.

Results. We did not find an association between either PBB or PCB exposure and time to menopause. Women who were current smokers had a shorter time to menopause than never smokers (menopause ratio 1.48, 95 percent confidence interval [CI]: 1.11-1.97). Time to menopause was shortest among women who reported that they smoked at least 20 cigarettes per day, smoked for at least 20 years, or had at least 10 pack-years of smoking.

PBB Exposure and Menstrual Cycle Characteristics (manuscript in preparation)

Adult women with at least one serum PBB measurement were eligible for the analyses of menstrual cycle length. Women using oral contraceptives or other hormonal medications were excluded, leaving 335 women for analyses. We estimated serum PBB level at the time of the interview using the initial PBB serum level and the prediction equation developed previously. We analyzed the relationships between PBB exposure (initial and estimated) and average menstrual cycle length, and average menstrual cycle bleed length (in the past year) by multivariate linear regression, controlling for confounders. We found a negative association between menstrual cycle length and bleed length among women who had experienced a weight loss of more than 10 percent in the past year. Women in the highest decile of PBB exposure and weight loss experienced menstrual cycles that were 3 to 5 days shorter than women with lower exposure or women who had not experienced weight loss. Analyses using initial PBB levels and estimated PBB levels yielded similar results. Bleed length was significantly longer (0.87 days) among women with weight loss in the highest exposure category compared to women with lower exposure or without weight loss.

Uterine Fibroids (manuscript in preparation)

The relationship between serum PBB measurements taken at enrollment (1976-1979) and time of diagnosis of uterine fibroids of 942 participants was modeled using multivariate models. No association was found between serum PBB level and time to diagnosis of uterine fibroids. No association between PBB exposure and uterine fibroids was found when comparing women with moderate serum PBB measurements (detection limit to upper decile) to women with low serum PBB measurements (below detection limit) (hazard ratio [HR] = 1.22, 95 percent CI: 0.84-1.76). Also, no association between PBB exposure and uterine fibroids was found when comparing women with high serum PBB measurements (above the upper decile) to women with low serum PBB measurements (below detection limit) (HR = 0.73, 95 percent CI: 0.38-1.41). This is the first study to investigate halogenated biphenyl exposure and a possible association with uterine fibroids.

Time to Pregnancy (manuscript in preparation)

A total of 478 women from the Michigan PBB cohort became pregnant after exposure to PBBs. Time to pregnancy was retrospectively assessed for each pregnancy reported after 1973. Crude analyses using Kaplan-Meier survival curves revealed a slightly shorter time to pregnancy for the highest exposure group (qreater than or equal to11 ppb) compared to women with exposures below the level of detection (less than or equal to1 ppb). Multivariable analyses using Cox proportional hazards regression yielded a fecundability ratio of 1:26 (0.9-1.7) for the high exposure group and 1:0 (0.8-1.2) for the middle exposure group compared to the lowest exposure group.

Risk of Endometriosis (manuscript in preparation)

The purpose of this study was to examine the association between serum PBB levels and the risk of endometriosis among a cohort of Michigan women unintentionally exposed to PBBs in 1973 and interviewed in 1997. Extended Cox models were constructed to estimate the risk of endometriosis adjusting for serum PCB level and risk factors for endometriosis reported in the literature. Ninety (9 percent) out of 970 women reported endometriosis diagnosed by a physician. Women with moderate PBB exposure (> 1-12 ppb), (HR = 1.04, 95 percent CI: 0.65, 1.65) and high PBB exposure (> 12 ppb), (HR = 0.70, 95 percent CI: 0.29, 1.71) did not have an increased risk of endometriosis compared to women with low PBB exposure (less than or equal to 1 ppb). A significantly increased risk of endometriosis was found among women exposed to moderate PCB levels (> 5-11 ppb), (HR = 1.85, 95 percent CI: 1.17, 2.94) but not high PCB levels (> 11 ppb), (HR = 1.29, 95 percent CI: 0.54, 3.11) compared to women with low PCB exposure (less than or equal to 5 ppb). The results of this study add little to no evidence to support an association between exposure to polyhalogenated aromatic hydrocarbons and endometriosis.

Risk of Osteoporosis and Bone Fractures (statistical analyses underway)

Women 50 years old and older in 1997 were included in the current analyses. Multivariate survival analyses were performed to assess the relationship between incidence of osteoporosis and PBB exposure. The hazard ratio comparing women in the upper decile of PBB exposure and osteoporosis adjusted for age was 0:59. Multivariate survival analyses also were performed for the incidence of bone fractures. The hazard ratio comparing women in the upper decile of PBB exposure and bone fractures was 1:14. The numbers of women with osteoporosis or bone fractures were small, and additional analyses were performed combining these groups as indicators of bone density. These analyses continue.

Occurrence of Abnormal Pap Smears (statistical analyses underway)

Forty-one women were excluded because they reported an abnormal Pap smear before 1974, or the date of the Pap smear was unknown. Eighteen women either reported that they have never had a Pap smear or refused to answer the question: "Have you ever had a Pap smear?" Therefore, 931 women are included in the analysis. Women were asked if they have ever had a Pap smear, when their most recent Pap smear was, if they have ever had an abnormal Pap smear, how many times they have had an abnormal Pap smear, their age at each abnormal Pap smear, and if they have had any procedure or treatment because of the abnormal Pap smear. Women also were questioned about their education, annual household income, smoking status, pregnancy history, medical encounter frequency, health insurance, use of hormonal medications, and other health outcomes, including pelvic inflammatory disease.

For all consenting participants who self-reported on a telephone interview that they have had an abnormal Pap smear, medical records were requested for the approximate time period they indicated the abnormal Pap had occurred (for example, their self-reported age at abnormal Pap, with their year of birth revealed the approximate year of the abnormal Pap). Medical records were independently reviewed by two of the investigators and results were compared. Records were reviewed jointly to resolve all discrepancies. The first Pap smear and/or cervical biopsy with dysplasia was recorded. If dysplasia was not evident, the first Pap smear and/or cervical biopsy with atypia (for example, atypical squamous cells of undertermined significance) was recorded. If atypia was not evident, the first Pap smear and/or cervical biopsy with benign changes (such as inflammation) was recorded. If there were no abnormal Pap smears or biopsies in the medical record, the first normal Pap smear in the medical record was recorded.

Future Activities:

We will continue analyses of relationships between PBB exposure and health outcomes and submit manuscripts for publication. We have proposed to analyze blood samples for dioxin-like activity and perform additional analyses with health outcomes. Because PBBs are contaminated with brominated dioxins and several congeners of PBB have dioxin-like activity, analyses of outcomes in relation to dioxin-like activity are important.

Journal Articles on this Report : 7 Displayed | Download in RIS Format

Other project views: All 18 publications 8 publications in selected types All 8 journal articles
Type Citation Project Document Sources
Journal Article Blanck HM, Marcus M, Tolbert PE, Rubin C, Henderson AK, Hertzberg VS, Zhang RH, Cameron L. Age at menarche and tanner stage in girls exposed in utero and postnatally to polybrominated biphenyl. Epidemiology 2000;11(6):641-647. R825300 (2002)
R825300 (Final)
not available
Journal Article Blanck HM, Marcus M, Hertzberg V, Tolbert PE, Rubin C, Henderson AK, Zhang RH. Determinants of polybrominated biphenyl serum decay among women in the Michigan PBB cohort. Environmental Health Perspectives 2000;108(2):147-152. R825300 (2002)
R825300 (Final)
not available
Journal Article Blanck HM, Marcus M, Rubin C, Tolbert PE, Hertzberg VS, Henderson AK, Zhang RH. Growth in girls exposed in utero and postnatally to polybrominated biphenyls and polychlorinated biphenyls. Epidemiology 2002;13(2):205-210. R825300 (2002)
R825300 (Final)
not available
Journal Article Blanck HM, Marcus M, Tolbert PE, Schuch C, Rubin C, Henderson AK, Zhang RH, Hertzberg VS. Time to menopause in relation to PBBs, PCBs, and smoking. Maturitas 2004;49(2):97-106. R825300 (2002)
R825300 (Final)
not available
Journal Article Kaiser R, Marcus M, Blanck HM, Naughton M, Zhang RH, Henderson AK, Tolbert PE, Rubin CH, Hertzberg VS. Polybrominated biphenyl exposure and benign breast disease in a cohort of US women. Annals of Epidemiology 2003;13(1):16-23. R825300 (2002)
R825300 (Final)
not available
Journal Article Rogan WJ, Weil WB. Duration of breast-feeding and PBBs. Environmental Health Perspectives 2002;110(9):A503-A504. R825300 (2002)
R825300 (Final)
not available
Journal Article Thomas AR, Marcus M, Zhang RH, Blanck HM, Tolbert PE, Hertzberg V, Henderson AK, Rubin C. Breast-feeding among women exposed to polybrominated biphenyls in Michigan. Environmental Health Perspectives 2001;109(11):1133-1137. R825300 (2002)
R825300 (Final)
not available

Supplemental Keywords:

endocrine disruption, human, epidemiology, reproduction, polybrominated biphenyls, PBB, female, endometriosis, uterine fibroids., RFA, Health, Scientific Discipline, Environmental Chemistry, Health Risk Assessment, Endocrine Disruptors - Environmental Exposure & Risk, endocrine disruptors, Risk Assessments, Children's Health, Biology, Endocrine Disruptors - Human Health, adverse outcomes, thyroid dysfunction, breast cancer, uterine fibroids, endocrine disrupting chemicals, exposure studies, fertility, developmental processes, human exposure, human growth and development, biochemistry, sensitive endpoints, biological effects, endometriosis, pregnancy, polybrominated biphenyls (PBB)

Progress and Final Reports:

Original Abstract
  • 1997
  • 1998
  • 1999
  • 2000
  • 2001
  • 2003
  • 2004
  • Final Report