Fact Sheets

Fecundity and fertility

Issue: More people today are experiencing difficulty in becoming pregnant because of exposure to hormonally active substances.

Trend data: Because of the complexity of human reproduction, it is often difficult to determine whether or not there is an actual increase in age-specific infertility rates. Published trend data for human fecundity are sparse, however, in North America, infertility rates have remained stable while the demand for fertility services has increased most likely due to an increase in the number of women above 35 years of age and a tendency to delay childbearing until later in life. In Sweden, analysis of birth registries has shown that the population of subfertile women, defined as those who did not become pregnant after more than 1 year, has actually decreased from 12.7% in 1983 to 8.3% in 1993 in the general population. Regional differences may exist since it has been observed that the time to pregnancy was shorter among couples in Finland than in the UK.

Consistency of the data: "Time to pregnancy" is an effective tool for measuring the impact of exogenous agents that affect reproduction as studies have shown that it can clearly demonstrate a difference among nonsmokers and smokers. It measures the time taken for a couple to conceive following unprotected intercourse for all subjects in a population-based approach and does not require categorization of subjects into fertile and infertile groups.

Differences in time to pregnancy have been found in a prospective study involving seven well-defined geographical areas in Europe. The highest fecundity was observed in southern Italy and northern Sweden, the lowest fecundity was in east Germany. The differences in time to pregnancy remained significant after adjustment for regional differences in body mass, smoking, frequency of intercourse and sexually transmitted disease. The longest time to pregnancy was observed in Paris and the shortest in Rome.

Another useful approach in addition to time to pregnancy is to review the total fecundity of a population of people with no predisposition for limitation of family size. For example, there has been a decreased age-specific fertility rate in the Hutterite population, a group in which reproductive practices are unlikely to have changed over time. These retrospective cohort studies revealed a decline in the total number of children born beginning with a cohort 1931 to 1935 and a continuing decline with subsequent birth cohorts. These data appear to indicate some form of extraneous factor, although there is no clear link to an endocrine disruptor hypothesis.

Experimental evidence: Occupational exposures are often cited as evidence of external impacts on fertility but a review of the published literature fails to reveal a clear pattern of effects. In one study, 281 women with a diagnosis of infertility were compared to 216 postpartum women for chemical exposures. Women with a history of working in the agricultural industry had an elevated risk of infertility. A recent study of the exposure of female wood workers exposed to formaldehyde demonstrated a highly significant effect on time to pregnancy including an apparent dose-related effect as those with higher exposures (i.e. no gloves), had a more severe effect.

Several epidemiological studies have been undertaken to study the fecundity and fertility of farmers exposed to pesticides. A retrospective study of 43 couples in the Netherlands whose male partner was a fruit grower included 91 pregnancies from 1978 to 1990. Exposure to pesticides was determined by self-reported data. An adverse effect of pesticide exposure was found, mainly apparent in highly exposed men who tried to conceive during the spraying season. The incidence of couples consulting a physician because of a fertility problem was also much greater in the high exposure group. The same Dutch group initiated an on-going case control study on occupational exposures and semen quality among couples consulting an infertility clinic. Among 899 men who delivered a semen sample, an association between impaired semen parameters and aromatic solvent exposure was observed but no association was found with pesticide exposure. However, in another study significantly decreased fertilization rates were observed for couples with male partners exposed to pesticides and enrolled in an in vitro fertilization program. Adjustment for paternal or maternal smoking habits, caffeine use, alcohol consumption or other occupational exposures had little effect on the observed association. In contrast, a retrospective study of 2,012 farm couples demonstrated no strong or consistent pattern of association of exposure to various classes of pesticides with time to pregnancy. Similarly a large study made in Denmark and France on exposure to pesticides and a control group of agricultural workers did not demonstrate any effect of pesticide exposure on time to pregnancy. There have also been controversial results associated with the consumption of sport fish, containing what both PCBs and mercury, by males on time to pregnancy in two major studies.

Biological plausibility: Several distinct lines of evidence provide support for the biological plausibility that hormonally active chemicals can alter fecundity. Environmental contaminants, some of which possess hormone like activity, have been detected in human ovarian follicular fluid and seminal plasma of subjects attending fertility clinics. In animal studies treatment with increasing concentrations of test compounds have been shown to reduce litter size in rodents. In male rodents, test compounds have been shown to reduce daily sperm production, and alter sperm morphology and motility. From an assessment of reproductive function as assessed by regulatory style rodent studies, the most definitive outcome measures include semen quality and longer estrus cycles. However, an endocrine mechanism of action has not been conclusively demonstrated to be the causal route for these effects. Moreover the dose required to induce changes in rodent fertility are in excess of those measured in contemporary residue analyses of human tissues.

Tissue culture experiments have shown that hormonally active agents can alter steroidogenesis in granulosa and Leydig cells and alter sperm-egg interactions. Other studies have shown that test substances can also affect oocyte quality. However, translation of in vitro results to whole animals and then to humans remains problematic.

Conclusions: In summary, the relationship of changes in the time to pregnancy to endocrine disruption is highly speculative. Part of this is due to the complex array of issues that may alter normal human reproduction and result in a longer time to pregnancy.