Vallombrosa Consensus Statement on
Environmental contaminants and
human fertility compromise.
Purposes of meeting
Areas of scientific confidence
Areas likely but requiring scientific confirmation
Directions for future research
PDF version of statement for downloading/printing
A recent national survey indicates that 12% of the reproductive age population in the United States, or 7.3 million couples, reports experiencing difficulty conceiving and/or carrying a pregnancy to term. This is precisely termed impaired fecundity, but commonly referred to, as a general experience, as infertility.
Proximate causes of infertility vary widely, for example from impaired sperm quality or reproductive tract abnormalities, to fallopian tube obstruction, hormone/menstrual cycle irregularities and anovulation, to implantation difficulties and recurrent miscarriage. Some seek medical intervention to help them conceive, and the number of people doing so has risen sharply over the last two decades.
Latest news about reproductive health:
More news about reproductive health
In 2002, an estimated $2.9 billion was spent on infertility treatments in the US. Now, some 46,000 (or one in 100) babies born to Americans each year are conceived as a result of the most advanced advanced reproductive technologies (ART).
These increasingly effective medical procedures have helped hundreds of thousands of couples around the world achieve successful pregnancies. They can, however, also be a hardship emotionally and/or financially, and often the financial costs place these interventions beyond the reach of couples who need them.
For those who can pursue such assistance, despite its great promise, success isn’t a given: An estimated one fifth or more of treated couples do not end up with a baby after a course of ART cycles.Too, other medical and/or mental health conditions can be associated with infertility in the couple experiencing it(and research is ongoing as to whether there are increased health risks that attend treatment or conception via ART).In light of all these considerations, a high value should be placed on minimizing preventable causes of infertility as well as on the treatment of it.
Multiple interacting factors are likely to contribute to biological fertility challenges, including age, heredity, lifestyle, underlying disease, reproductive tract infections and nutritional status. Demographers have identified voluntary delays in first pregnancy as a major factor. Yet, data from the US Centers for Disease Control and Prevention show that impaired fecundity over the last two decades appears to have increased in all reproductive age groups, but most sharply in younger women (under age 25). These data, together with a growing body of epidemiological literature and many experimental research results showing male and female fertility-related impairment in laboratory animals caused by a wide array of modern chemicals, implicate environmental factors also as possible contributors to human infertility.
|Scientific understanding of the relationship between environment and human health is advancing rapidly. It reveals that a larger portion of health problems, including infertility, may be caused by environmental exposures than thought possible even a decade ago.
These exposures include but are not limited to occupational sources. For some environmental agents known to have adverse effects in experimental animal studies or wildlife, impacts on human reproductive health are being found as well, and at exposure levels within the range humans commonly experience (termed “environmentally relevant”).
If involuntary infertility is actually on the rise, and troubling insights from animal studies accurately predict human impacts, then the personal and societal costs of fertility compromise could become increasingly burdensome and significant shifts in reproductive health and norms at the level of whole populations could occur. This has profound implications for public health and strongly suggests that a more comprehensive, coordinated research agenda must be developed and funded – because adverse effects caused by environmental exposures are, in principal, preventable.
Responding to these concerns, a multidisciplinary group of experts gathered at the Vallombrosa Center, Menlo Park, CA, February 27- March 1, 2005 to assess what is known about the contribution of environmental contaminants, specifically synthetic compounds and heavy metals, to human infertility and associated health conditions. Workshop organizers chose this focus because critical recent discoveries in the field have raised many new, intriguing scientific questions and heightened interest in environmental risk factors within patient organizations and reproductive medicine/science professional societies. This was the first time researchers in reproductive epidemiology, biology, toxicology and clinical medicine convened with representatives of relevant professional societies as well as infertility support, women’s health and reproductive advocacy organizations from the United States to review the state of environmental health science as it pertains to infertility.
The purposes of the meeting were:
- To review findings from diverse research disciplines concerning environmental contaminants and the biological basis of compromised fertility, with special attention to critical recent discoveries in related basic sciences;
- To identify conclusions that could be drawn with confidence from existing data;
- To identify critical knowledge gaps and areas of uncertainty;
- To establish key elements of a coherent research agenda to help fill these gaps and resolve uncertainties;
- To consider recommendations for educational initiatives and preventive interventions if and where warranted.
Over the course of the meeting, the following core points of consensus were identified, which we offer to help scientists, medical professionals and public health advocates understand, in broad brush, the current state of scientific understanding in the field and to identify important research areas that will be crucial to further advances:
A. Based on existing evidence, we are confident of the following:
1. In the US today, at least 12% of the reproductive age population reports experiencing impaired fecundity. This appears to be a rising trend, most markedly in women under
25 years old.
2. Human biological characteristics relevant to fertility vary geographically and over time. For example, semen quality varies within and between men and geographically among populations. Hypospadias,
cryptorchidism and testicular cancer are increasing in some areas but not in others. Other fertility-related diseases, for example endometriosis and polycystic ovarian syndrome (PCOS), are diagnosed more frequently now, which may result from an increase in prevalence, better detection, or both. Current data are inadequate to analyze global trends conclusively.
||3. Specialists can identify proximate (or apparent) cause or risk factors in the male, female or couple in the majority of infertility cases. Within this “explained” category, however, sometimes ultimate (or underlying) causes and mechanisms are understood, but very often they are not.
In up to 10% of cases, absolutely no reason for the infertility can be discovered at all – and in a much higher percentage than that, only minor abnormalities that are not severe enough to account for the infertility are identified. These cases are termed “unexplained.” It is biologically plausible that environmental factors could be contributing to (or a component of) ultimate causation of infertility, in both the explained and unexplained case categories.
|4. Considerable data from experimental animal and human studies demonstrate adverse effects of cigarette smoke on a spectrum of sensitive reproductive endpoints in both men and women. Cigarette smoke contains thousands of chemicals, some of which are thought to be involved in its impact on reproduction.
These compounds are also encountered elsewhere in the environment, and there is no a priori reason to eliminate these exposure pathways from concerns about reproductive health. Effects of other environmental mixtures are likely to be similarly diverse and complex.
5. Considerable experience with the pharmaceutical diethylstilbestrol (DES) clearly demonstrated that prenatal exposure to a synthetic estrogen can adversely affect reproductive physiology and impair fertility later in life, with many endpoints altered. This compound serves as a model for environmental agents that are hormonally active, in other words, endocrine disruptors. Laboratory experiments with DES-exposed animals have repeatedly demonstrated causal effects that are congruent with data on DES offspring, particularly DES daughters. While doses of DES ingested by pregnant women were much greater than those that come from exposure to environmental estrogens, many underlying mechanisms of action appear to be similar .
6. Moreover, environmental contaminant concentrations and/or potency can be amplified because of persistence (biomagnification and bioaccumulation) and because they always occur in mixtures.
7. A wide range of wildlife populations has been shown to be adversely affected by exposure to endocrine-disrupting contaminants. Well-documented effects include: decreased fertility and increased reproductive tract abnormalities in birds, fish, shellfish and mammals; feminization and demasculinization in male fish, birds, mammals and reptiles; masculinization and defeminization in female fish, birds, mammals and reptiles.
8. Some environmental contaminants at high, occupational exposure levels were shown decades ago to impair human fertility, for example lead and the fumigant dibromochloropropane.These types of exposures, however, are unlikely to explain more than a small fraction of the infertility observed in today’s population. More recently, considerable data support the contention that exposure to certain agricultural pesticides at moderate or environmentally relevant exposure levels are associated with adverse reproductive outcomes in men and women working on or living near farms (male subfertility and sperm damage; menstrual alterations, increased time to pregnancy and spontaneous miscarriage rates).
9. Recent research with animals has demonstrated effects on specific aspects of reproductive system development at very low levels of exposure to environmental contaminants (levels within ranges experienced by the general public). This is a finding that may ultimately alter how human safety thresholds are established. In animal and cell culture experiments using these low-dose exposure levels, some contaminants, for example bisphenol A and dioxin, have been shown to interfere with cellular signaling pathways that are important to fertility and reproduction. Proposed mechanisms through which such chemicals may act include perturbation of nuclear hormone signaling nuclear hormone signaling, inappropriate activation or inactivation of transcription factors and alterations in hormone metabolism. For some contaminants, non-monotonic dose response curves have been observed when responses are examined across a wide range of exposure levels.
10. Very few relevant data from epidemiological studies are available to investigate the possible associations suggested by these studies between low level environmental exposures and reproductive health. Much more work in this area must be done given the import if animal data on low-dose effects translate to humans.
11. Genetic signaling mechanisms are highly similar across vertebrate classes, particularly with respect to the structure of key signaling molecules such as steroid hormones and their receptors. Animal models of reproductive toxicity thus offer useful guidance for identifying potential reproductive toxicants in humans. For some compounds, especially DES, there has been remarkable concordance of responses between humans and other vertebrates. A similar pattern is emerging in studies of phthalates. Although differences do exist, consistency of impact across multiple species (especially if the species are from diverse vertebrate classes, e.g. birds and mammals and fish) increases the utility of animal data for identifying human reproductive toxicants.
12. Single contaminants can affect multiple endpoints in more than one tissue through alterations in the expression of multiple genes affecting multiple pathways. Some contaminants have been shown to alter the expression of hundreds of genes, and effects can vary with timing and dose. Different contaminants can affect the same physiological endpoint by acting on the same signaling pathway.
13. Genetic variation, or DNA polymorphisms, within populations (humans, wildlife and laboratory animals) can result in greater sensitivity to specific contaminants in some individuals. While such variation/sensitivity has been linked to increased risk of specific problems such as, for example, bladder cancer and fetal alcohol syndrome, it has yet to be discovered whether there are genetic polymorphisms that affect response to environmental toxicants and cause or contribute to infertility.
14. Recent measurements of contaminants in people show that humans are exposed, starting at conception, to at least hundreds of chemicals simultaneously – and some at levels within ranges known individually (chemical by chemical) in cell culture and/or animal studies to affect physiological processes relevant to reproduction.
15. The effects of a single chemical exposure have been shown in laboratory studies to differ from the effects of the same chemical in a mixture. Experiments with single chemicals can significantly underestimate effects of the same chemical in mixtures.
16. Exposures during different stages of life (pre- and periconceptional, fetal, perinatal, peripubertal and adult) have different impacts, because developmental processes create discreet windows of vulnerability for specific effects. The consequences of exposure can manifest on different time scales, some involving long latency. For example, prenatal exposures can cause abnormalities at birth or later that have impacts on adult reproductive function (e.g. as shown with DES). The abnormalities may involve structural or functional alterations, or enhanced sensitivity to subsequent endogenous or exogenous exposures.
17. To date few if any epidemiological studies have successfully incorporated the full complement of these considerations (assessing mixtures, life stage of exposure, the possibility of differential individual genetic susceptibility, etc.) into study design. Epidemiological research that does not factor in these biological considerations will be more likely to conclude erroneously that a study is “negative” and less likely to confirm adverse impacts. Facing these limitations, when epidemiological studies do report positive associations, they should be taken seriously.
18. New scientific methods and tools can and should be developed to further scientific understanding of environmental contributions to human infertility and identify opportunities for preventive interventions. However,a currentlack of adequate research funding in the field is a significant impediment.
B. We consider the following to be likely but requiring confirmation:
1. It is likely that gene-environment interactions are involved in the etiology of many reproductive problems including impaired sperm quality; PCOS; endometriosis; uterine fibroids; premature puberty, ovarian failure and menopause; and reproductive cancers. Further, it is possible that environmental (i.e. low-level ambient) exposures having the biggest impact are those that occur before conception, in utero and neonatally.
2. A cluster of abnormalities of the male reproductive tract is associated in what is termed “testicular dysgenesis syndrome” (TDS), which is hypothesized to originate from a common causal pathway of developmental errors in the fetal testis. TDS can produce a range of outcomes including cryptorchidism and hypospadias at birth, and reduced sperm quality and testicular cancer in adulthood. Semen quality in specific populations has declined (though with no geographic uniformity), and several recent epidemiological studies suggest this may be related to environmental agents. The mechanisms have not been established.
3. It is likely that environmental endocrine disruptors contribute to some manifestations of TDS in humans. In the etiology of TDS, some evidence points to interference with testosterone metabolism mediated by disruption of genetic signaling. Given the well-known, multiple effects of DES on male and female reproductive tract development, it is likely that a syndrome analogous to TDS involving interference with estrogen signaling by environmental chemicals will be identified.
4. It is likely that a broad spectrum of women’s reproductive health endpoints is affected by environmental agents including heavy metals, polychlorinated biphenyls and other hormonally- active chemicals. Attributing risk of adverse reproductive effects from these exposures is challenging, but several female-factor secular trends in some populations lend biological plausibility to such an association and support the need for further research. For example, increases in the incidence of reproductive cancers may reflect non-hereditary genetic factors, lifestyle and/or environmental factors or exposures. Age at onset and progression of puberty have been reported to be decreasing over time in several developed countries, suggesting environmental etiology inclusive of lifestyle and diet. Similarly, as prevalence of endometriosis is reported to be increasing, earlier ages at diagnosis are also noted. While greater access to medical care may account for some of these temporal patterns, accumulating evidence suggests an etiologic role for environmental contaminants.
5. Current data contradict the assumption that “weak” environmental estrogens are not a concern because of their low estrogenic potential compared to the endogenous estrogen, estradiol. Studies of mixtures in cell cultures and animals indicate that multiple "weak" estrogens can combine to have effects even when present at levels at which singly they would have no impact. Additionally, some "weak" estrogens affect cellular signaling through recently discovered cell membrane receptors as well as through “traditional” nuclear hormone receptor mediated pathways. In the former case, “weak” estrogens like bisphenol A can be equally as powerful as estradiol at provoking cellular responses.
C. Research on a wide array of fertility-related endpoints suggests several broad themes which we believe should be pursued in future scientific investigations:
1. Information about rates of infertility/subfecundity and specific contributing health conditions in the general population is very limited. For most fertility/fecundity-related endpoints, we have no population data and must rely on women, men or couples seeking medical treatment. These data are unlikely to be representative of the total population of couples of reproductive age. For this reason, the magnitude of fertility/fecundity impairments has not been fully described and quantified. This poses a challenge to scientists when attempting to assess trends or environmental influences on human reproductive health. Standardizing definitions, identifying consistent endpoints that can be compared across studies and better public health tracking of fertility/fecundity-related endpoints would strengthen the investigation of environmental associations with reproductive health compromise. More research into geographic variations and factors contributing to differences among populations would also be highly useful.
2. Highly reproducible effects in animal studies indicate that today’s framework for evaluating environmental chemical risks to reproductive health is inadequate. Study designs should explicitly incorporate the complex causal framework that has emerged from animal research, including long latencies (of effect following preconceptional, in utero, neonatal and peripubertal exposures) and interactions among multiple factors (mixtures of contaminants; gene-contaminant interactions; pharmaceuticals; subpopulations varying in genetic susceptibility; nutrition and lifestyle; complex dose-response relationships). Study designs must also be broadened to incorporate the possibility of multigenerational, epigenetic transmission of effects; consider a multiplicity of causal pathways and endpoints; and examine impacts on population-level endpoints such as sex ratio.
3. Research on wildlife populations and mechanistic studies in animals and cell cultures have proven invaluable in identifying new categories of risk and elucidating the biological mechanisms linking cause to effect. A vigorous research agenda using these approaches should be continued and expanded. These animal studies would ideally involve multidisciplinary approaches that develop biomarker of exposure and disease in animal models and translate them for use in epidemiological and clinical studies. They should assess syndromes of impacts in addition to single effects. Human epidemiological data identifying fertility impairments can help guide the animal research.
5. Prospective studies of exposures, outcomes and covariates, with high degrees of public participation and cooperation, are likely to be most helpful. For example, the National Children's Study plans to include the recruitment of couples prior to conception to explore fecundity-related impairments in relation to a host of environmental factors including chemicals. This landmark study could also include newly proposed developmental landmarks indicative of endocrine function in infants and be extended to evaluate fertility/fecundity in adulthood, as well as population level outcomes (such as changes in sex ratios, twinning and birth rates).
6. Factors contributing to differential vulnerability to environmental exposures are diverse and include age; gender; genetic and epigenetic variation; nutritional status and obesity; infections; lifestyle behaviors; pharmaceutical use; occupation; socioeconomic and racial disparities; and physical proximity to certain industries or industrial accidents. All of these factors need to be evaluated to help identify biologically sensitive and otherwise vulnerable subgroups. More systematic attention to these subgroups is likely to improve sensitivity and accuracy of epidemiological research designed to assess risks associated with exposures.
7. Developing tools of toxicogenomics,
metabolomics and the study of genetic variation (toxicogenetics) should be integrated with biomonitoring in epidemiological studies. These tools need to be developed to the point of defining specific biomarkers of susceptibility, exposure and disease. Specific markers for ovarian and testicular responses need to be developed. Increased sensitivity, availability and affordability of assays for measuring contamination levels in people would enhance research in epidemiology and clinical settings.
8. Testicular dysgenesis syndrome is emerging as a useful construct for organizing hypotheses about some aspects of male reproductive health, including infertility. Human patterns appear to be consistent with animal data, and information about impacts of contaminants on gene expression thought to be important for male reproductive development is providing insights into molecular mechanisms. We need a comprehensive national program, coordinated with efforts underway elsewhere in the world, in order to fully evaluate the TDS hypothesis – including TDS prevalence and etiology. This research program should combine epidemiological and clinical perspectives with in vivo and in vitro experimental research that targets mechanisms.
9. Research on both prevalence trends in and environmental causes of female infertility factors is of equally high priority and must be encouraged. Premature ovarian failure (POF); premature menopause; thyroid disruption; autoimmune disorders; menstrual cycle defects; PCOS; uterine fibroids; endometriosis;
meiotic aneuploidy; and repeat pregnancy loss are examples of proximate explanations for female factor infertility that call for specific examination to develop understanding of potential environmental etiologic links.
|10. A coherent environmental reproductive health research strategy should include a pointed emphasis on high priority compounds, i.e. those that are under-investigated; those that are bioactive at low doses; and those for which potential for exposure is widespread due to persistence or continuous use.
The scientific evidence we have reviewed indicates that while environmental contaminants are unlikely to be the sole etiologic factor underlying human infertility, some exposures cause adverse reproductive health outcomes that contribute to infertility. What proportion of infertility today is environmentally induced is a question of profound human, scientific and public policy significance. Existing animal and human data suggest that a greater proportion is environmentally caused than has yet been generally realized or can be demonstrated with scientific certainty.
Nothing is more fundamental to the human prospect than the ability to reproduce. Uncertain as the science on environmental causes of infertility is, it is sufficient to raise troubling questions about the future of human reproductive health, and serious debate about how to communicate the information accumulated to date to physicians, patients and the public. This amply justifies an accelerated research program built around interdisciplinary coordination and collaboration to resolve important uncertainties that currently prevail, particularly around issues involving low-level developmental exposures. A coherent, enhanced research agenda will help identify new strategies to prevent infertility, through actions that individuals can take as well as those that public health/regulatory agencies can pursue. As these investigations progress, it will be increasingly important to engage physicians, other health professionals, patients and the public in formalized educational efforts that delineate and encourage opportunities for prevention that are elucidated by the research.
Purposes of meeting
Areas of scientific confidence
Areas likely but requiring scientific confirmation
Directions for future research
PDF version of statement for downloading/printing