Authors
Eskenazi, B., Warner, M., Mocarelli, P., Samuels, S., Needham, LL., Patterson Jr., DG., Lippan, S., Vercellini, P., Gerthoux, PM., Brambilla, P., Olive, D.

Title
Serum Dioxin Concentrations and Menstrual Cycle Characteristics

Journal
American Journal of Epidemiology: 156; 383-392

Summary
The dioxin 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is one of 75 related chemical compounds produced through incomplete combustion processes in nature, as well as through man-made incineration and industrial processes. Some animal studies have shown that TCDD has the ability to impact the estrus cycle. The human menstrual cycle is dependant on a hormonal interplay between estrogen and progesterone. Therefore, it is plausible that exogenous agents that alter hormone levels or antagonize the effects of endogenous hormone could interfere with this process. For example, under certain conditions TCDD has been shown to block the action of estrogen. The objective of this study was to measure the impact of TCDD on the menstrual function of women. This study is important because it is the first epidemiological study looking specifically at the effects of serum TCDD levels and the menstrual function.

The 1976 explosion at the OCMESA chemical plant located close to Seveso, Italy resulted in wide spread dioxin contamination in the surrounding area. Approximately 30kg of TCDD was deposited on a 18 km2 area. This resulted in the highest levels of TCDD in any human residential population. The exposed area was later divided up into zones based on exposure levels. Zone A was the most heavily contaminated area and Zone B had the next highest contamination. Blood samples were collected from Seveso residents soon after the explosion for clinical chemical testing. The remaining serum was stored for future studies. Twenty years later, the follow-up study was conducted. Women were eligible for inclusion into the Seveso Women's Health Study if they were 40 years of age or younger, had adequate amounts of stored serum that was collected between 1976 and 1980, and resided in Zone A or B at the time of the accident (n=981). For the current study on menstrual function, 301 women were enrolled. Women were excluded if they had surgical or natural menopause, Tuner's syndrome, had been pregnant in the last year, or had used an intrauterine device or hormonal medications such as oral contraceptives. The data was previously obtained by a trained interviewer who did not know the woman's serum TCDD level or zone of residence. Women were asked to describe the length and regularity of their cycles, as well as, the number of days and heaviness of their menstrual flow in the past year. The authors also obtained data from medical records for all obstetric and gynecologic conditions. The earliest serum sample of each woman was measured for TCDD levels by high resolution mass spectrometry. Those TCDD levels measured after 1977 were back-extrapolated to 1976 using the Filser model for women aged 16 years of age or younger in 1976 and the first-order kinetic model for women over 16 years.

In the analysis several factors were measured for their effect on menstrual function including age at interview, education, parity, smoking, body mass index, alcohol and caffeine consumption, exercise, chronic illness and abdominal surgeries. The average age of the women at the time of the explosion was 12.9 years. TCDD levels were found to be associated with cycle length in women who had not had their first menstrual period (premenarcheal) at the time of the explosion but not for women that were postmenarcheal. A 10-fold increase in serum TCDD levels was associated with an increase in cycle length of nearly a day [b=0.93 days (0.01, 1.86)]. In addition, TCDD levels were inversely associated with the odds of having irregular cycles after adjusting for age at interview and age at menarche (OR=0.46, CI=0.23-0.95). There was no relationship between TCDD levels and days of menstrual flow, heaviness of flow or regularity of cycle in pre or postmenarcheal women.

The women in this study were unaware of their TCDD serum levels limiting the possibility of systematic bias in their reporting. However, a large number of young women using intrauterine devices or oral contraceptives and other hormones (n=233) were excluded from this study which may have influenced the results. It is plausible that women who used hormones may have done so to regulate their menstrual cycle. When the authors included women taking hormones in the analysis as having irregular cycles, TCDD levels were no longer associated with irregular cycles. Yet, the lengthening of the menstrual cycle in women who were premenarcheal at exposure is consistent with some animal studies and warrants further investigation.