Authors:
Pauwels A, Schepens PJC, Hooghe TD, Delbeke L, Brouwer L, Weyler J.

Title:
The risk of endometriosis and exposure to dioxins and polychlorinated biphenyls: a case control study of infertile women

Source:
Human Reproduction 16(10):2050-2055, 2001.

Summary:
Studies of primates have suggested a dose-dependent relationship exists between dioxin exposure and endometriosis severity, yet epidemiological studies conducted to date have been inconclusive. In Belgium, dioxin exposure and the incidence and severity of endometriosis is among the highest in the world. A case-control study was conducted to determine if an association exists between endometriosis and chronic exposure to dioxins and polychlorinated biphenyls (PCBs), suggested to have dioxin-like activity.

Cases and controls were selected from the population attending one of the Centers for Reproductive Medicine in Belgium between 1996 and 1998. Both cases and controls were infertile (no pregnancy after 1 year of unprotected sexual intercourse) and had laparoscopy testing. Cases were defined as those with laparoscopy-confirmed endometriosis. Controls were women for whom no evidence of endometriosis was found. In total 42 cases and 27 controls were identified. Background information (e.g. medical history, risk factors) were determined by accessing medical records and conducting telephone interviews. The interviewer was blind to the exposure/case/control status of the study participant. Exposure was blindly assessed by examining a blood sample that was taken just prior to the laparoscopy for dioxin, coplanar and non-coplanar PCBs. These values were expressed as Toxic Equivalency Factors (TEF). A cut-off point of 100 pg TEQ/g serum lipids was chosen to differentiate between exposed and non-exposed populations.

Cases and controls were similar with respect to risk factors including age, ovulatory dysfunction, smoking pattern, as well as alcohol and caffeine intake. No significant association was observed between mean or high exposure to dioxin or PCBs and endometriosis in this small population of infertile women.

In assessing dioxin and PCB exposures the authors selected 100 pg TEQ/g serum lipids to differentiate between exposed and non-exposed populations, however the article does not clearly indicate how this value was chosen, whether this value correlates to the exposures experienced by the general population, or what proportion of the study population falls into the exposed/unexposed group.

The use of laparoscopy to identify cases and controls minimized the potential for selection bias. Laparoscopy has a sensitivity and specificity of 99%, and 77% respectively (Walter et al 2001), meaning that 99% of study subjects who have the disease will be identified correctly, but only 77% of study subjects that do not have the endometriosis will be correctly identified. In other words, using this test, there is a tendency to over diagnose endometriosis. Given the small numbers of cases and controls in this study, this may have led to a differential bias toward a null finding.

References:
Walter AJ, Hentz JG, Magtibay PM, Cornella JL, Magrina JF (2001). Endometriosis: correlation between histologic and visual findings at laparoscopy. Am J Obstet Gynecol. 184(7):1407-11.