|Home | About | Journals | Submit | Contact Us | Français|
Perspective on the papers by Abdullah et al and Nassar et al (see pages 576 and 580)
In this issue, Abdullah et al examine routine Hospital Episode Statistics (HES) from the 1990s for northern England and report a decreasing birth prevalence of cryptorchidism and orchidopexy. In addition, they report a small increase in birth prevalence of hypospadias but no change in the rate of hypospadias surgical repair (corrective procedure rates). The latter is more likely to reflect true prevalence as the vast majority of cases are referred for surgery. In contrast, Nassar et al, also in this issue, report on a population‐based study to determine trends in hypospadias in the Australian population, 1980–2000. Data were obtained from a congenital anomaly registry (with multiple sources of notification) for aborted and still births and for live births up to 6 years old. Overall birth prevalence was found to be 3.5 over the 20‐year study period with birth prevalence rates increasing significantly by 2% per annum.
Both these papers reporting contrasting trends in male fertility‐related disorders are part of a wider series of publications examining the evidence for adverse trends in male reproductive health, in particular, in association with exposure to ubiquitous environmental chemicals mimicking oestrogen.
Much public interest in this body of research came from popular science, most notably Carson's Silent spring in 1962 warning us of the dangers posed by man‐made pesticides. This was followed in 1996 by Colborn et al's scientific detective story Our stolen future, which suggested that we were still ignoring the danger of man‐made chemicals and thereby threatening our fertility and human survival. In the academic domain, these ideas were formalised into the controversial “endocrine disruptor hypothesis” notably by Skakkebaek, Sharpe and colleagues. Much evidence of chemical‐hormonal activity in vitro is available, together with both convincing toxicological data at high doses and observational evidence of the sexuality‐modifying effects of these chemicals in wildlife. Likely candidate human outcomes for endocrine disruption include testicular cancer, decreasing sperm counts, and the congenital anomalies of cryptorchidism and hypospadias. Skakkebaek et al have suggested that these disorders are all related in what they term “testicular dysgenesis syndrome”1 and Sharp and colleagues2,3 have put forward putative animal models for this syndrome.
But what epidemiological evidence is available for adverse trends in congenital anomalies of the male reproductive system?
For hypospadias, several authors reported increasing trends in birth prevalence from the 1960s to the 1990s; however, recent studies from Scotland and California have cast doubt on these reports. Moreover, the contrasting articles in this issue appear simply to add to the already inconclusive evidence. Previous epidemiological studies have reported widely varying estimates of hypospadias birth prevalence. Estimates have ranged from 0.9 to 9 per 1000 male live births. The uncertainty of these figures reflects the mode and the context of ascertainment as well as true differences in different populations.
For cryptorchidism, secular trends suggested large increases in rates in Great Britain and the USA between the 1950s and the 1980s. Published data on recent trends have been scarce, although Toledano et al4 using data from HES and the General Practice Research Database (GPRD) and Abdullah et al suggest, in contrast, declining trends in orchidopexy rates during the 1990s in Britain. Once again, the evidence appears unclear and inconclusive.
Reported trends are often hard to interpret due to a number of factors. These include: the use of routinely collected data and their inherent inaccuracies and complexities of use and interpretation, for example, hospital activity in the UK depends not only on the underlying prevalence of disease but also on the diagnostic accuracy and referral practice of the primary care clinician, patient‐specific factors such as individual preferences, distance of residence from hospital and socio‐economic class and on hospital‐specific factors, such as quality of the hospital data collected, supply of hospital beds, admission policies and hospital access5; incomplete ascertainment by congenital anomaly registries (which varies nationally and internationally) and differences in the information they choose to collect for surveillance, for example, the exclusion of “minor anomalies” in 1990 by the UK National Congenital Anomaly Registry (NCAR) including glandular hypospadias and cryptorchidism cases, necessitating the use of orchidopexy rates from HES as a proxy for cryptorchidism prevalence; inconsistency in the methodology employed between studies and the denominator data they use, for example, Abdullah et al in this issue have used male live births as their denominator, whereas other studies have used total births; and, finally, the choice of inclusion criteria used both nationally and internationally for hypospadias, for example, the EUROCAT definition for hypospadias excludes glandular (mild) cases.
The epidemiological data available to answer the questions posed are therefore necessarily highly complex and difficult to interpret. Meanwhile, congenital anomalies are rare, environmental exposures of interest occur at very low levels, and we are trying to detect very small increases in risk. Unsurprisingly, the resulting body of evidence continues to be inconsistent. This inconsistency may reflect the fact that epidemiology as a tool lacks resolution to answer the question.
Whilst it is likely that hypospadias and cryptorchidism have an environmental component to their aetiology and that endocrine disturbance plays a role in the pathophysiology, it is less likely that exposure to very low concentrations of chemicals is a physiologically relevant causal factor, even where such chemicals have endocrine activity in vitro. Moreover, it is unlikely that current epidemiological research will provide sufficiently firm evidence of a clinically significant environmental component amenable to public health action.
As scientists, we are thus left with a highly speculative and unprovable proposition regarding the human public health relevance of endocrine disruption. The media have a subject with all the hallmark ingredients for a continuing news story – uncertainty, disagreement between scientists (conspiracy theory) and high emotive factor, as babies are affected. We have an anxious public and no satisfactory answer for them. Public perception of risk in this area, therefore, has the potential to become exaggerated. Moreover, media‐driven public interest can lead to additional research, more equivocal data, an ever accelerating spiral of anxiety and, ultimately, public policy detrimental to health.
We discuss first the case of dichloro‐diphenyl‐trichloroethane (DDT) which we judge was an example of international policy driven mainly by exaggerated public anxiety and which resulted in damage to public health. Secondly, we highlight the case of chlorination disinfection by‐products (DBPs) where there maybe ongoing selective interpretation of epidemiological and other evidence leading to exaggerated public perception of risk. This may drive policies which are ultimately adverse to public health.
DDT was the first modern pesticide developed early in World War II. Initially it was used with great effect to combat mosquitoes spreading malaria, typhus and other insect‐borne human diseases, and as an agricultural insecticide. However, following Carson's Silent spring, there was large public outcry which eventually led to the insecticide being banned for agricultural use in the USA, and subsequently worldwide. In addition, many restrictions on the use of DDT in vector control followed, imposed by various national governments, donor countries and international aid agencies, in response to pressure from environmentalists. Such action was taken despite scientific evidence showing both a strong safety record for DDT when used in small quantities for indoor spraying in endemic regions and that, of the dozen insecticides WHO approves as safe for house spraying, the most effective was and still is DDT. In parallel, the campaign against malaria has arguably been failing6 and the WHO estimates that malaria continues to afflict 300–500 million people every year and that around 1 million people die of malaria and malaria‐related illness every year. Remarkably, in an important policy shift, the WHO announced last September that it was promoting the indoor use of DDT to fight malaria. Dr Arata Kochi, director of the WHO's Global Malaria Department, said “We must take a position based on the science and the data…. One of the best tools we have against malaria is indoor residual house spraying… the most effective is DDT. Extensive research and testing has… demonstrated that well‐managed indoor residual spraying programmes using DDT pose no harm to wildlife or to humans”.7
Chlorination has been the major disinfectant process for domestic drinking water for many years. Concern about the potential health effects of the by‐products of chlorination has prompted a flurry of investigations into the possible association between exposure to these by‐products and the incidence of human cancer, and more recently, with adverse birth outcomes including low birth weight, stillbirth and various congenital anomalies. These studies are looking for very small increases in risk associated with very low levels of exposure. Findings to date have been inconsistent.8 As with the male fertility‐related disorders, however, such research has frequently made headline news, increasing fears amongst women about swimming and drinking tap water during pregnancy. Parallel with this, there have been continuous pressures on the water industry to lower DBP levels whilst trying not to compromise microbial safety in their water supply (an ever increasingly difficult task) as well as moves to use alternative primary disinfectants in water treatment plants instead of chlorine, for example, the use of ozone in some parts of Canada and Europe, with its own ozonation by‐products. The latter may also carry health risks which to date have been little investigated. When assessing any potential risks associated with DBPs, however, it is essential for scientists to recognise the substantial benefits to health associated with disinfection by chlorination. The use of chlorine has virtually eliminated water‐borne microbial diseases because of its ability to kill or inactivate essentially all enteric pathogenic micro‐organisms, including viruses and bacteria from the human intestinal tract. We should also remember that according to the WHO more than 1.5 billion people around the globe still do not have access to safe drinking water, and diseases associated with dirty water kill more than 25000 people per day – more than 9 million each year – around the world.
The examples given above serve as a warning of how well‐meant action and policy have had, and could have, detrimental consequences for public health. The endocrine disruptor hypothesis has all the ingredients for an environmental health scare that could lead to negative public health. We hope that scientists will carry out research mindful of the limitations of scientific methods and of the scrutiny of media, the public and politicians. Even the posing of a study question has a potential impact on public health. Resulting evidence must be weighed up and interpreted from a balanced perspective within a wider, integrated cost–benefit assessment in terms of the health of the public and society as a whole. The recent Committee on Toxicity's report on the human male reproductive system and potential chemical‐induced effects9 offered an opportunity to embrace this wider perspective. However, we find it advocates a “more of the same” approach to investigating the endocrine disruptor hypothesis. It calls for more toxicology and epidemiology, whilst acknowledging their limitations in resolving the question, and the fact that “…a clear link between experimental data and epidemiology is still missing”.9 We, as a community of researchers, must ask ourselves, are we moving forward this body of research or do we simply resemble the “Ascending and Descending” figures in Escher's famous lithograph? (fig 11).
Competing interests: None.