THE quality of men鈥檚 sperm declined steadily in the early years of the 21st century until hardly anyone could reproduce in the normal way. Meanwhile, the countryside was virtually emptied of animals as natural populations crashed. The first signs of the impending catastrophe were noted in the 1990s, but few people then believed that a cocktail of pollutants that mimic human hormones could have such profound effects. And so nothing was done 鈥� until it was too late.
This chilling scenario has already inspired a best-selling thriller by P. D. James, The Children of Men. But could any of it be true? Are pollutants altering the reproductive health of humans and wildlife? 鈥淓verybody in the environmental movement is drawing their breath and saying my God, what if,鈥� says Gwynne Lyons, pollutant consultant to the World Wide Fund for Nature. Her organisation, along with Friends of the Earth and Greenpeace, has recently campaigned against hormone-like pollutants 鈥� or 鈥渆ndocrine disrupters鈥� 鈥� in the environment. And she believes we need urgent action to reduce the exposure of humans and wildlife to these chemicals in water, food and air.
Flawed statistics
Advertisement
Yet last month, a report for the British government, by the Institute for Environment and Health at Leicester, found no direct evidence of a causal link between these environmental chemicals and reproductive problems in people. The report concluded that 鈥渋t is currently very difficult to assess鈥� whether environmental oestrogens might be responsible for a fall in sperm counts or rising rates of testicular cancer. Lewis Smith, director of the institute, added that a ban on suspected chemicals would be premature.
Others are downright sceptical. 鈥淚 am absolutely not worried,鈥� says Jack Cohen, a developmental biologist at the University of Warwick, who believes that the alleged fall in sperm counts is based on flawed statistics.
So who is right?
Most researchers would agree that an overdose of oestrogenic chemicals could damage a fetus: sex hormones such as oestrogen influence the development of the reproductive system in the developing fetus, including the Sertoli cells that control future sperm production. And since sex hormones can also stimulate the proliferation of cancer cells, in theory environmental chemicals that mimic them could be responsible for the rise in testicular cancer. In theory, too, such chemicals could damage the reproductive health of women. Indeed, medical researchers have already seen what can happen later in life to fetuses exposed to an excess of hormones during pregnancy thanks to a famous 鈥渕edical mistake鈥�. Between the 1940s and the late 1970s, doctors prescribed a synthetic oestrogen drug, DES or diethylstilboestrol, to an estimated 2.3 million pregnant women worldwide in the belief that it would prevent miscarriage. The adult children of the women given DES are now acknowledged to be at greater risk of developing various cancers of the reproductive system. They also suffer a higher incidence of infertility and other reproductive problems.
But the fact that certain substances could damage reproductive health doesn鈥檛 necessarily mean that the chemicals people are currently exposed to in their everyday lives are actually doing so. Establishing such a link scientifically is no easy task, and most of the evidence accumulated so far is dogged by uncertainty.
The strongest evidence that something is amiss comes from surveys of men鈥檚 reproductive health, and in particular from semen samples analysed at sperm banks or fertility centres. According to studies in Denmark, Scotland and France over the past three years, the number of sperm generated by the average 鈥渘ormal鈥� man in an ejaculation has fallen substantially 鈥� by almost half 鈥� compared with samples taken 50 years ago. Over the same timescale, the rate of testicular cancer in adult men has doubled, and there has also apparently been a rise in the number of boys born with misshapen penises or testicles.
Two years ago, an Anglo-Danish research team attempted to link all these observations. Richard Sharpe, of the MRC鈥檚 Reproductive Biology Unit in Edinburgh, and Niels Skakkebaek, of the University of Copenhagen, argued that oestrogen-mimicking chemicals in the environment might be damaging male fetuses in the womb. 鈥淚t is only a hypothesis,鈥� Sharpe emphasises, 鈥渂ut all the facts fit.鈥�
The problem is that the central 鈥渇act鈥� of the case 鈥� that sperm quantity and quality has declined 鈥� is strongly disputed. A comprehensive American study published in 1979 found no evidence for any decline from 1951 to 1977, and more recent studies have also been attacked by researchers in Britain and the US. Last year, for example, Cohen, with Ian Stewart, a mathematician at Warwick, published a paper in the British Medical Journal which claimed to find statistical flaws in the Danish study that first reported the decline. The study is flawed, says Cohen, because the WHO 鈥渃hanged the goalposts鈥�, raising the lower limits of a 鈥渘ormal鈥� sperm count, in the mid-1960s. 鈥淭he change in reference value accounts for almost all the decline, and what鈥檚 left is probably just statistical noise,鈥� comments Stewart.
Earlier this year, a group of American epidemiologists from the Dow Chemical Company, Shell and Baylor College of Medicine in Houston reanalysed the data from the Danish study using three different statistical models, in a report published in Fertility and Sterility. The researchers point out that the linear regression model used in the Danish study, if extrapolated forward in time, 鈥減ortends the collapse of traditional means of procreation by the middle of the next century鈥�. Yet alternative models which fit the data just as well suggest that average sperm counts have been increasing since the late 1970s, while a third approach, the 鈥渟tairstep鈥� model, indicates that a large drop in sperm counts happened in the mid-1960s with counts constant ever since.
The most recent study, of Parisian sperm donors, has stood up better to statistical scrutiny. Nevertheless, reviewing all the published studies, the Institute for Environment and Health report concludes that the evidence for falling sperm counts is 鈥渘ot entirely convincing鈥�.
If proving that sperm counts are falling is difficult, establishing a direct link with specific environmental chemicals 鈥� and ruling out the effects of changes in people鈥檚 lifestyles and diets 鈥� is harder still. Certainly, the list of chemicals alleged to mimic or interfere with sex hormones is now vast. It includes ubiquitous industrial chemicals and products of petrol combustion such as polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs) and dioxins; phthalates which are added as plasticisers in plastics and used as ingredients in paints, inks and adhesives; alkyl phenolic substances (such as octyl and nonyl phenol), which are breakdown products of alkylphenol polyethoxylates (APEs) used as surfactants in industrial detergents and also found in paints, herbicides and some plastics; and organochlorine pesticides such as DDT, aldrin and dieldrin.
Plastic tubing
However, only a tiny number of these chemicals have actually been shown in lab tests to behave like a sex hormone. The research just hasn鈥檛 been done, partly because there are no quick and easy ways to screen chemicals for oestrogenic effects. Nor is the degree of our exposure clear. What is known is more by accident than design.
For instance, in 1991 Ana Soto at Tufts University in Boston stumbled on the truth about one chemical, a common additive in plastics, when it leached from plastic tubing into the cells growing in the laboratory. She noticed that something was encouraging the growth of the human breast cancer cell line. The culprit, nonylphenol, turns out to act like an oestrogen hormone, and stimulates breast cells to divide. 鈥淚ndustry uses about 300 000 tonnes of APEs every year, 60 per cent of which ends up in water,鈥� says Charles Tyler, a researcher studying oestrogens in the environment with fish endocrinologist John Sumpter at Brunel University. A new paper by Sharpe and Sumpter, to be published in Environmental Health Perspectives, reports an experiment in which pregnant lab rats were fed nonylphenols at doses comparable to what a human might be exposed to. At maturity, the male offspring had significantly smaller testicles and lower sperm counts.
Sumpter began his investigations when anglers began turning up odd-looking fish with ambiguous genitalia in the mid-1980s. With government funding, he placed caged male fish (typically rainbow trout) near sewage outflows into rivers to act as sensitive 鈥渂iomonitors鈥� of hormonal pollution. If the male fish detects oestrogen, says Sumpter, it produces an egg yolk protein, vitellogenin, usually produced in quantity only by females. The more vitellogenin in his blood, the more oestrogen in the river, Sumpter surmised. He found the male fish placed in some 30 different sites in British rivers made vitellogenin in very large amounts 鈥� 鈥減hysiologically, they were behaving like females,鈥� he says.
The pesticide DDT has also been identified as an endocrine disrupter in wildlife, not least because of an outbreak of male alligators with abnormally small penises in Lake Apopka in Florida 鈥� which had been accidentally polluted by a massive pesticide spill in 1980. Most of the male reptiles were sterile 鈥� the result, it was thought, of 鈥渇eminising鈥� exposure to oestrogen-like DDT. But there was a new twist to the story in June this year when William Kelce of the US Environmental Protection Agency and his colleagues reported in Nature that DDE, DDT鈥檚 main breakdown product, is an 鈥渁ntiandrogen鈥�: rather than acting like an oestrogen, it blocks the action of 鈥渕ale鈥� androgen hormones instead.
Like a wrong key that fits into a lock but won鈥檛 quite open it, all these endocrine-disrupting compounds can bind to hormone receptors on the surfaces of human cells. So DDE binds to the androgen receptor, for example, while nonylphenol binds to the oestrogen receptor. Once securely plugged into a receptor, these artificial hormones then behave abnormally. They set off only some of the biochemical reactions normally stimulated by bona fide oestrogen or androgen. What鈥檚 more, their very presence on the receptors blocks the normal lines of communication between the natural hormone and a cell.
Popular headlines call these substances 鈥渢he emasculators鈥� or talk about 鈥渢he chemical attack on the male sex鈥�. But female wildlife and women may also be at risk. For instance, an epidemic of infertile eggs laid by western gulls and common and roseate terns on the Pacific coast of the US has recently been attributed to oestrogenic effects of DDT and PCBs.
Some believe that oestrogen-like compounds in the environment may be contributing to the steady rise in the rate of breast cancer.
But epidemiologists in Britain are by and large sceptical of such claims. 鈥淭here has been no sudden upturn in deaths from breast cancer since pesticides were introduced,鈥� says Valerie Beral, director of the Imperial Cancer Research Fund鈥檚 Cancer Epidemiology Unit in Oxford. The increasing trend for women to have fewer children later in life is probably largely to blame, she suspects. Similarly, the steady rise in testicular cancer rates could also be linked to marked changes in lifestyle since the Second World War.
Attempts to blame human health problems on oestrogenic chemicals must also contend with the fact that fetuses are inevitably exposed to their mothers鈥� bloodstream, which is awash with oestrogen during pregnancy. But Sharpe and his like-minded colleagues argue that while fetuses may be somehow protected against natural oestrogen, they are not safe from synthetic variants.
One sign that natural oestrogenic chemicals found in plants are benign is the longevity and good health of the Japanese. The Japanese diet is extraordinarily rich in oestrogen-mimics produced by plants such as soya beans. Yet there is no sign of increased infertility in Japan, and the nation鈥檚 cancer rates are among the lowest in the world. Indeed, nutritionists at the MRC Dunn Nutrition Unit in Cambridge and elsewhere suspect a diet high in plant oestrogens may actively lower a woman鈥檚 risk of developing breast cancer.
Polluted fat
Plant oestrogens may be something we鈥檝e evolved to cope with, suggests Lyons at WWF. 鈥淲hat is new is manmade chemicals mimicking hormones.鈥� Certainly, the ability of synthetic chemicals such as DDT, phthalates and APEs to accumulate in fats in the body sets them apart from natural plant oestrogens and may make them more dangerous. These chemicals can be taken up by algae and fish so that trace amounts in rivers, lakes and the sea become multiplied thousands of times in food. The chemicals can then be stored in human fat; if this fat is broken down, as often happens early in pregnancy, concentrated pollutants may flood the system.
The chemicals may also have additive effects, as researchers at Brunel have discovered. When fish were exposed to a mixture of oestrogenic compounds, the vitellogenin response was larger than if either chemical had been added alone, says Tyler. The message here is that even if chemical oestrogens are much less powerful than natural steroids 鈥� and some PCB-type compounds are known to be a million times less potent 鈥� the combined effects of a multitude of different chemicals could still be noticeable. 鈥淲e are particularly concerned that with the tens or hundreds of different substances acting on the same receptors,鈥� says Lyons, 鈥渆ven small amounts of pollutants in the environment could have quite startling effects.鈥�
A report produced for the Danish government and published earlier this year called for urgent research, particularly into the exposure of women of child-bearing age to oestrogenic pollutants. It also warned that existing methods of testing for toxicity may not pick up the harmful effects of oestrogenic pollutants. In the US, environmental campaigners have called for new chemicals to be screened for their ability to mimic hormones before they are released into the environment. Sumpter and his colleagues at Brunel are working to develop a genetically engineered strain of yeast that contains the human oestrogen receptor. This could streamline the screening of thousands of chemicals and become a standard test for oestrogenic effects.
The Institute for Environment and Health report, accepted by the government, argues that more research is needed before any chemicals are banned. 鈥淚t鈥檚 a major nightmare trying to unravel what is going on, to find out how much oestrogen is present and what its consequences are 鈥� 50 years of research is needed,鈥� says Sumpter. But Lyons and fellow campaigners say we can鈥檛 wait for scientific proof of cause and effect. 鈥淲ith widespread and multiple exposures to chemicals, it will be more or less impossible to get that proof,鈥� she argues. 鈥淲e must act now on the weight of evidence to reduce the release of these substances.鈥�