Greenlee, AR, TE Arbuckle and P-H Chyou. 2003.
Epidemiology 14:429-436
Greenlee et al. report a strong association between using herbicides and infertility in women. In their study population, women who were infertile were 27 times more likely to have mixed or applied herbicides in the two years prior to attempting conception than women who were fertile. Other factors, including smoking and exposure to passive smoke, steady weight gain during adult life, and consuming alcoholic beverages were also associated with infertility.
What did they do? Greenlee et al. conducted a retrospective case-control study examining the association between infertility in women and different risk factors.
They recruited infertile women to the study via electronic records of women seeking infertility treatment. Their diagnoses included endometriosis, anovulation, pituitary-hypothalamic dysfunctions, etc. Couples whose sterility was a result of any form of male infertility or surgical intervention, e.g., hysterectomy or vasectomy, were excluded from the analysis.
The controls were pregnant women from the same population. From the same age range, control women were seeking prenatal care during their first pregnancy, and had conceived in fewer than 12 months of trying. Any controls who reported ever having difficulty conceiving or maintaining a pregnancy, or whose male partner had a questionable history of infertility were also excluded.
Controls were then matched to cases on the bases of age and date of clinical service.
To determine exposure histories, cases and controls were interviewed about their activities for the 2 years prior to attempting to become pregnant. The questionnaire included questions on demographics, occupation, exposures, pesticide use, residency on a farm, tobacco and alcohol use, etc.
What did they find? A total of 1,791 potential cases were screened for the study, along with 822 potential controls. After recruitment and screening, 322 cases and 322 matched controls participated in the study. Cases and controls were well matched for most variables, including age, household income, smoking status, body mass index, age at menarche, and number of sexual partners in lifetime.
They differed somewhat in schooling (cases more likely to be a high school graduate; but not more likely to have schooling beyond high school). Cases were somewhat more likely to have exposure to passive smoke, to consume alcohol, and to have steadily gained weight. The odds-ratios for these variables were mostly under 2, with the odds-ratio of infertility rising to 6.7 for women who consumed at least 7 alcoholic drinks per week.
Of several associations that Greenlee et al. examined between infertility and exposure to agricultural chemicals, two stood out: infertile women were almost 27 times more likely to have mixed or applied herbicides (but not insecticides) than fertile women, and 3.3 times more likely to have used fungicides. Both these odds-ratios reflect adjustments for maternal level of education, passive smoke exposure and other variables. Perhaps paradoxically, living on a farm, ranch or in a rural home reduced the likelihood of infertility.
While the 27-fold increase in risk of infertility associated with having mixed or applied herbicides was very strong, the number of case women who fell in this category, 21, was relatively small. Hence the 95% confidence limits for the estimate of the odds-ratio was quite broad, from 1.9 to 348.
What does it mean? These findings are consistent with a host of previous studies, both epidemiological research and laboratory experiments, that have found associations between infertility and agricultural chemicals. The laboratory experiments with animals and cell lines are unambiguous: an array of compounds working through multiple pathways affecting a variety of specific endpoints can suppress fertility in exposed animals. In people, elevated risk of poor sperm quality in Missouri men with relatively high urinary levels of alachlor, atrazine and diazinon, reported recently by Swan et al. in 2003, is the most powerful example to date.
The collective weight of evidence is very strong, especially in light of the animal experiments. Taken together, they indicate that fertility of American women and men is being undermined by today’s use of agricultural chemicals. Greenlee et al.’s data in this study suggest that precautionary measures to avoid impairing fertility should include avoiding working with herbicides and fungicides for at least two years prior to attempting to conceive. Swan’s results, on the other hand, indicate that sufficient exposure to impair fertility can take place even without working directly with pesticides, and thus that broader measures to reduce exposures will be necessary.
