There is mounting evidence that endometriosis may have an environmental origin, possibly exposure to hormonally active agents such as dioxin and other compounds that bioaccumulate. Among candidate materials are polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs). Both animal and human studies support a role for PCBs and the development of endometriosis, an estrogen-dependent disorder. The present study examined the relationship, if any, between endometriosis and 62 individual PCBs. Eighty-four women 18 to 40 yeas of age who were scheduled for laparoscopy constituted the study group. Most were white, married, college-educated women; their mean age was approximately 32 years. Endometriosis was visually confirmed in 32 women. Women with endometriosis had fewer pregnancies and live births than did their unaffected peers. Both women with and those without endometriosis had relatively low serum concentrations of PCBs. Total serum PCBs ranged from 0.5 to 5.3 ng/g in women with endometriosis and from 0.2 to 5.6 ng/g serum for those without endometriosis. The sum of estrogenic PCB congeners ranged from 0.07 to 0.9 and from 0.4 to 1.7 ng/g serum, respectively. The respective figures for antiestrogenic congeners were 0.01 to 0.2 and from zero to 0.3 ng/g serum. On logistic regression analysis, a significantly elevated odds ratio for endometriosis of 3.8 was noted in women in the third tertile of antiestrogenic PCBs (95% confidence interval [CI], 1.1–12.7). The association remained significant at 3.3 (95% CI, 0.9–12.5) after controlling for gravidity, current cigarette smoking, and serum lipids. Gravidity and cigarette smoking were both associated with a significantly reduced risk of endometriosis. These findings suggest an association between antiestrogenic PCBs and the development of endometriosis.

BACKGROUND:Hormonally active environmental agents have recently been associated with the development of endometriosis.METHODS:We undertook a study to assess the relationship between endometriosis, an estrogen-dependent gynaecological disease, and 62 individual polychlorinated biphenyl (PCBs) congeners. We enrolled 84 eligible women aged 18-40 years undergoing laparoscopy for study, which included an interview and blood specimen (n=79; 94%). Thirty-two women had visually confirmed endometriosis at laparoscopy while 52 did not. Blood specimens were run in batches of 14 including four quality control samples for toxicological analysis. Each PCB congener was adjusted for recovery; batch-specific reagent blanks were subtracted. All PCB concentrations were log transformed and expressed in ng/g serum first as a sum and then as tertiles by purported estrogenic or anti-estrogenic activity of PCB congeners.RESULTS:Using unconditional logistic regression analysis, a significantly elevated odds ratio (OR) was observed for women in the third tertile of anti-estrogenic PCBs [OR 3.77; 95% confidence interval (CI) 1.12-12.68]. Risk remained elevated after controlling for gravidity, current cigarette smoking and serum lipids (OR 3.30; 95% CI 0.87-12.46).CONCLUSIONS:These data suggest that anti-estrogenic PCBs may be associated with the development of endometriosis.

Endometriosis is an estrogen dependent gynecologic disorder affecting approximately 10% of females of reproductive age. Its etiology remains elusive though hormonal and immunological factors have been postulated as underlying mechanisms. Of late, there is increasing speculation that hormonally active environmental agents may be associated with endometriosis. We investigated 61 polychlorinated biphenyl (PCBs) congeners and 7 pesticides (i.e., mirex, hexachlorobenzene (HCB), b-BHC, Oxychlordane, trans-nonachlor (t-Nonachlor), Aldrin, and 1,1-dichloro-2,2-bis(p-chlorophenyl) ethylene (p,p-DDE)) in relation to risk of incident endometriosis among women undergoing laparoscopy for either diagnostic or sterilization purposes, 1998–1999. Among the 100 eligible women identified, 84 agreed to be interviewed prior to laparoscopy. Blood specimens were collected from 80 (95%) consenting women. Thirty-two women had visually confirmed endometriosis while 52 did not. Among women with endometriosis, severity ranged from AFS-R Stage 1 (minimal, n = 15), Stage II (mild, n = 5). Stage III (moderate, n = 6), and Stage IV (severe, n = 6). All bloods were analyzed using gas chromatography with electron-capture. PCB congeners and pesticides were run in batches of 14 including four quality control samples (i.e., reagent control, serum control, control with 15 calibration standards at known values, and a duplicate participant sample). Congeners were calculated from standard curves for the 15 calibration standards, and the remaining congeners from response factors that were laboratory generated. Each congener and pesticide was adjusted for percent recovery and lipids; batch specific reagent blanks were subtracted as well. All PCB concentrations were expressed as ng/g serum lipid and summed into a total PCB concentration. Women with moderate/severe (Stages III and IV) endometriosis had a significantly higher mean concentration of total PCBs than women without disease (1.45 and 1.20 ng/g, respectively). No significant difference in mean PCB concentration was observed between women with mild/moderate (Stages I and II) and women without disease (1.44 and 1.20 ng/g, respectively). An elevated risk of endometriosis was observed for log of lipid adjusted total PCBs (OR = 1.98; 95% CI = 0.84,4.65) but not for DDE (OR = 0.93; 95% CI = 0.50,1.77) after controlling for household income, current cigarette smoking and parity. Risk of endometriosis remained for total PCBs after adjusting for DDE (OR = 2.13; 95% CI = 0.88,5.17). These data are suggestive of a possible association between PCBs and, possibly, more severe endometriosis but await confirmation in larger populations.