Infertility and environmental pollutants

Michael Joffe
2003 British Medical Bulletin  
While it has long been known that female fertility is impaired by oestrogen exposure, it is unclear whether environmental pollutants with weak oestrogenic effects are sufficiently potent and prevalent to have biological effects in humans. Male fertility, or sperm concentration at least, appears to have deteriorated, and there is substantial spatial variation at both national and global level, as well as a genetic component. Sperm morphology and motility are implicated too. There is good
more » ... ere is good evidence for an increase in testicular cancer, and possibly in other conditions that certain spatial characteristics plus evidence on heritability suggest are linked to impaired spermatogenesis. A candidate agent would need to have started increasing in the early 20th century. Weak environmental oestrogens are not responsible. Candidates include agents affecting endogenous maternal oestrogen levels, environmental anti-androgens (although these cannot explain the epidemiological findings), and dioxin and related compounds. Genetic damage should be considered as a unifying hypothesis, possibly focused on the Y-chromosome. Introduction A review of the literature on the determinants of fertility raises more questions than can be answered in the current state of knowledge. Even so, at least this is an advance on the situation that previously existed, when the topic was largely ignored. This state of affairs is mainly the result of the lack of priority that has been given to research on reproduction in basic biology, as well as in epidemiology and toxicology. Reproduction and/or development can be affected by exposure to a wide variety of agents, including dioxins, poly-chlorinated biphenyls (PCBs), phytoestrogens such as isoflavones, heavy metals (e.g. organic mercury, lead); chlorination disinfection by-products in water, organic solvents, polyaromatic hydrocarbons, particulate air pollution, substances emitted from landfill sites and caffeine. Often, effects on the reproductive system and/or infant development have been detected at lower doses of the substance in question than is the case for other endpoints. Most of these agents have featured in recent discussions of official regulatory bodies, which have highlighted the paucity of good quality epidemiological evidence. In principle, fertility can be studied in humans or in laboratory animals by using biomarkers such as measures of semen quality, and/or using a functional measure of the ability to conceive. In practice, for many exposures, little or no epidemiological research has been carried out with fertility as an endpoint, and the toxicology database is typically seriously incomplete in this respect as well. The much-publicised concern over the possibility of 'falling sperm counts' has altered the position in the past 10 years, but during that time the research initiative has been dominated by just one hypothesis, that endocrine disrupting agents are responsible. As a result, we are in the situation of having the following rather fragmentary information: (i) the effects of a few known agents on fertility in either sex; (ii) descriptive epidemiology, which suggests that there may be a problem with male fertility, and that this is likely to be linked to deterioration in related conditions such as testicular cancer; it also indicates when and where such a problem may be at its greatest, so that one can begin to outline the epidemiological characteristics of the responsible agent(s); (iii) some additional observations that allow preliminary assessment of the type of biological process underlying any such effect, in particular whether endocrine disruption or genetic damage are responsible. This paper reviews these three areas in turn, concentrating on epidemiology. Evidence from toxicology, endocrinology, genetics and research on wildlife is not presented, but informs the discussion. Priority is given to agents or processes that could affect whole populations, or substantial proportions of whole populations, rather than relatively small groups such as those with occupational exposure. The next section focuses on effects of known agents. As the main evidence for widespread impairment comes from epidemiological observations on males, the following section discusses this in detail. In view of the limited data available on male fertility itself, epidemiological information is also presented on possibly parallel variations in other disorders of the male reproductive system that arise early in life, especially on testicular cancer. Genetics is covered, as well as temporal and spatial variation. The available evidence is that the epidemiological observations are not explicable in terms either of the known agents or of agents that are currently known to have endocrine disrupting effects. Effects of known agents Infertility in females In the 1940s, reduced fertility was noticed in Australian ewes, and it was established that this was due to the clover that they were grazing. by guest on March 24, 2012 Downloaded from Infertility and environmental pollutants British Medical Bulletin 2003;68 49 This condition, which came to be known as 'Clover Disease', was traced to phytoestrogens-plant compounds that have oestrogenic propertiesin the clover. Ewes feeding on Australian clover developed abnormal plasma concentrations of endogenous hormones, with subsequent reduced fertility 1 . In humans, dietary phytoestrogen consumption is considerable in some populations, and constitutes by far the major route of exposure to exogenous 'endocrine disruptors' (Table 1) . Probably the most important of these are isoflavones such as genistein, which occur in legumes and are particularly abundant in soybeans and soy-based foodstuffs. The isoflavone Table 1 Principal exogenous substances that may affect sex hormone function Adapted with permission from Joffe 20 . This classification is an over-simplification: it conflates receptor-mediated effects with those due to other mechanisms, e.g. interference with hormone synthesis. Moreover, several of the 'oestrogens' show considerable affinity for the androgen receptor. In addition, many of these compounds have important biological actions that are not endocrine in mechanism. For the reasons given in the text, compounds in the 'Low potency' category cannot plausibly be considered responsible for the types of impairment of the male reproductive system considered in this paper, and the exposures may also be too low to affect females. A. Oestrogenic effects 1. High potency-pharmacological agents • DES (diethylstilboestrol) • ethinyl oestradiol (component of contraceptive pill) 2. Medium potency-dietary phytoestrogens • isoflavones, e.g. genistein, daidzein • coumestans, e.g. coumestrol • lignans 3. Low potency-environmental or occupational agents • bisphenol A • octylphenol and nonylphenol pesticides, including chlordecone, DDT, dieldrin, endosulphan, p,p′-methoxychlor, toxaphene B. Anti-androgenic effects • p,p′-DDE (the major breakdown product of DDT) • certain phthalates, e.g. DBP, DEHP • pesticides, including linuran, procymidone, metabolites of vinclozolin • hydroxyflutamide C. Others • dioxins, furans and 'dioxin-like' PCBs by guest on March 24, 2012 Downloaded from Impact of environmental pollution on health: balancing risk 50 British Medical Bulletin 2003;68 content of soy varies in relation to many factors, including plant species, strain, crop year and geographical location. The concentrations are sufficient to cause biological effects in humans, even after cooking or other processing. Populations that traditionally consume large quantities of soy, notably Chinese and Japanese people, tend to have relatively high phytoestrogen exposure. Less is known about other types of phytoestrogen; exposure to lignans is probably widespread, but the potency is lower and may not have biological significance 1 . In rodents, exposure to isoflavones and other phytoestrogens has been shown to alter a number of functions of the female reproductive system, including advancement of puberty, subfertility and irregular oestrus cycling. Perinatal, neonatal or prepubertal exposure appears to produce the most marked effects. It is unclear to what extent these findings are relevant to humans, owing to species differences in sexual development, experimental considerations such as route of administration, and uncertainty over comparability of plasma concentrations 1 . A different source of oestrogen exposure was the synthetic compound, diethylstilboestrol (DES), which has oestrogenic potency comparable to that of oestradiol. From the late 1940s onwards, it was widely used during pregnancy, especially in the USA, in the belief that it could prevent miscarriage and a range of pregnancy complications. It is estimated that more than two million women were exposed to this drug. Pharmacological doses were given, often at the stage of pregnancy during which the sexual organs develop. A randomized controlled trial published in 1953 2 showed that DES was ineffective for the conditions for which it was being prescribed. However, clinical use of the drug continued until it was banned in 1971, after the discovery that in utero exposure of female fetuses led to a risk of developing clear cell adenocarcinoma some 15 years later 3 . While this particular risk is fortunately rare, DES-exposed girls have reproductive tract anomalies, and they subsequently have reduced fertility and increased rates of ectopic pregnancy, spontaneous abortion and preterm delivery 3 . The sensitivity of the developing female reproductive tract to oestrogens raises the question of whether exposure to environmental chemicals having oestrogenic activity (Table 1 ) might affect fertility through the female route. There is insufficient evidence to answer this question definitively, as research on these exposures has tended to focus on the male, even though toxicological experiments consistently find stronger effects of oestrogens on females than males 1 . As their potency combined with exposure concentrations are many orders of magnitude lower than endogenous hormones, or even than phytoestrogen intake in oriental populations, a strong effect seems unlikely. Aside from the specific question of oestrogens, high maternal but not paternal consumption of sport fish from the heavily polluted Lake Ontario
doi:10.1093/bmb/ldg025 pmid:14757709 fatcat:geadkfurtre6jaci7nqezatuhi