THE CHANCES OF A LIFETIME

SCIENCE; Which poses the greater risk to humans: cigarette smoking or nuclear power? Are you more likely to be harmed by toxic chemicals or by peanuts? Public perception differs radically from expert opinion

LIFE is full of risks. No sensible people would argue with that, but they are almost certain to disagree on how big any given risk is. For many people nuclear power represents a huge risk; some may even consider it to be the biggest risk in life. Others may believe the biggest threat comes from chemicals in the environment.

Whatever the disagreements, the gap between the public perception and the expert assessment of risk is certainly getting wider. Researchers who carry out surveys into what risks the public fears most conclude that people tend to overestimate the frequency of rare causes of death and to underestimate the risk of common ones. Accidents, for instance, are often put on a par with death from illnesses, whereas in reality diseases account for about 15 times as many deaths as accidents.

This difference between the public's idea of risk and the expert's is perfectly represented by fears about nuclear power and the chemicals industry. A survey of Americans put nuclear power top of the list of risks in life, even though it is impossible to prove that any member of the US public has died as a result of radiation from the civil nuclear industry. A similar survey, also in the US, showed that Americans believe murders are more frequent than diabetes and cancer of the stomach. They greatly overestimate the risk of death from botulism, tornadoes and pregnancy - along with animal bites, stings and fires. The only conceivable explanation for this gross distortion of reality is the media. The most reported stories involving fatalities corresponded fairly accurately to the most overestimated causes of death.

Sean Paling, a former professor of biology at Oxford, has tried to allay the public's disproportionate fear of what he sees as trivial environmental risks by constructing a "perspective" scale of risk. At one end of his scale is the sort of risk territory we all inhabit - relatively probable events that will happen to a good proportion of us, such as death from cancer. At the other end is the risk of a "one-off" event that happens to one person but which dominates the media, and consequently the public mind. Professor Paling, who now works as a freelance lecturer in the US, calls this the "Bobbitt zone", after the man who had his penis cut off by his irate wife.

"We begin by recognising that everything we do in life has some risk associated with it. For example, just staying in your own home for your 70-year lifetime holds 7,700 chances in a million of you incurring a fatal accident... and staying in bed every day would give you some calculable risk of developing bed sores," he says.

Life is indeed full of risks, but no matter how common they are many people seem to have an overwhelming desire to introduce more. For some it may amount to no more than buying a national lottery ticket, even though if you buy one on a Monday the probability of winning is about the same as being dead by the time the Saturday announcement comes around. Others, unsatisfied with the average risk of death may feel, like climber Alison Hargreaves, that it is "better to live one day as a tiger than a thousand years as a sheep". Should they opt for a few hours a week of solo rock- climbing, they would increase their chance of sudden death by about the same degree as working in the construction industry (one of the riskier occupations) for a whole year.

At the same time, we all try to reduce some of the obvious risks of life to a bare minimum. Some of us may believe that buying a water-filter can significantly reduce the risk of cancer, or that taking herbal remedies can stave off sickness and ill-health. We may switch to decaffeinated coffee and organically grown vegetables while cutting down on barbecued meat (polycyclic aromatic hydrocarbons in burnt meat may increase the risk of cancer).

The inconsistencies in what risks we worry about is illustrated by a conversation Professor Paling once had with a woman who asked him whether he thought it would be safer to use a water purifier. The irony was that at the time she was smoking a cigarette, probably the biggest avoidable health risk in the world today. On average one person on the planet dies every 10 seconds as a direct consequence of tobacco. In 25 years time, if present trends continue, that death rate will rise to one every three seconds.

Professor Richard Peto at the Imperial Cancer Research Fund in Oxford, who has spent much of his professional life studying the effects of smoking, says the public is generally confused about health risks. He is particularly concerned that young smokers seem unaware of or unconcerned about the long-term risks they deliberately choose to run. "The chief source of misunderstanding has been the very long delay between cause and effect," he says. "If cigarette smokers start young and don't stop, then eventually about half of them will be killed by tobacco. But these big risks come decades later, in middle or old age."

Experts like Professor Peto argue that the public can be easily sidetracked by the tiny risks that often dominate the media - such as contracting mad cow disease from eating burgers, for instance - and remain woefully ignorant about really big avoidable risks, such as smoking. To put this into perspective: the risk of a person dying in any one year from smoking 10 cigarettes a day is one in 200, compared to a risk of one in 850 of dying from all natural causes, one in 3,300 from any kind of violence or poisoning and one in 8,000 from a road accident. The smoking risk is even more stark when compared with the estimated risk of death in any one year by radiation from a nearby nuclear power station - one in 10 million.

Some people will argue, however, that smoking is a risk of individual choice and thus qualitatively different from the unavoidable risk posed to society at large from the nuclear industry. In any case, how is it possible to estimate the risk of nuclear accidents when they have such a potential for catastrophic damage?

Sam Harbison, Her Majesty's Chief Inspector of Nuclear Installations, helps to calculate these risks and assess the level that is tolerable or acceptable for nuclear workers and the public at large. The public, he says, measures risk in a totally different way from him and his colleagues at the Health and Safety Executive (HSE), the statutory agency responsible for monitoring nuclear power plants.

"Members of the public are more interested in the consequences of risk than they are in the probabilities," Dr Harbison says. "We can deal fairly well in our perception with risks that only affect single members of the public. But it's far more difficult explaining risks that we think may affect the whole of society, such as nuclear accidents, which people perceive will not only affect this generation but successive generations for thousands of years to come."

Following the public inquiry into the Sizewell nuclear power station in the mid-1980s, the HSE has formulated a system of assessing nuclear risk as scientifically as possible. It compares nuclear risks with known risks in other industries, such as construction, where one worker out of 10,000 dies each year in accidents. (This is about the same average risk we all face from being killed in a traffic accident.) Even more dangerous industries, such as mining or deep-sea fishing, show an average risk of about one death in 1,000 workers.

Using these risks as a benchmark, the HSE has proposed that the highest risks a worker can be exposed to should not exceed one in 1,000, and for members of the public the risk from any large plant in any industry should not exceed one death per year in 10,000 people. Because of the "special public concern" about the nuclear industry, the HSE has adopted a limit for new nuclear power stations that is 10 times smaller than for other industrial plants. They must operate to a maximum risk limit of one death in 100,000 members of the public per year.

The HSE says it places even more stringent impositions on the industry to reduce these risks even further, "as far as reasonably practicable". As a result, the average risk to members of the public in the vicinity of a nuclear plant will generally be no more than one in a million per year. "So if 10,000 people were to spend their entire lives near a nuclear plant, we would expect on average rather less than one of them to die from cancer as a result of operation of the plant," Dr Harbison says. "To put this in context, statistics show that about 2,500 people (one person in four) in the same population would be expected to die from a whole variety of cancers, many of whose origins are still unknown."

Chernobyl has demonstrated, however, that when major nuclear accidents occur they affect thousands of people over considerable distances. Aside from the relatively few initial deaths - 31 officially - scientists estimate the radiation expelled from the plant will result in several thousand cases of cancer and genetic damage for decades to come. What is the HSE view on this type of large-scale, long-term risk?

"The maximum risk that should be tolerated by society of a major accident (causing between 100 and 1,000 deaths) at any nuclear plant in the United Kingdom should be no more than one in 100,000 per year," the HSE concluded. "However, it is not enough for the nuclear companies just to meet these targets of tolerability; they are required by law to make the risks as low as is reasonably practicable."

What this means is that the HSE expects the nuclear industry to work to an actual risk of a major accident which is 10 times smaller than the maximum tolerable value. "In other words, for 25 such reactors operating together continuously over a period of 40 years, there would be odds of about 1,000 to one against any one of them ever suffering such an accident."

It is even more difficult to estimate the risks of the approximately 100,000 man-made chemicals on the market, whether they occur in shoe polish, sausages or slug pellets. Some may be known to cause cancer in laboratory animals in relatively high doses but for the vast majority there is little or no toxicology data.

"The scientific rigour of what we know about the safety of chemicals is quite a bit less than we know for radiation," Dr Harbison says.

Again, the public fear about chemicals tends to be dominated by accidents such as Bhopal in India, which result in mass poisonings involving high doses, rather than insidious pollution at very low levels. Naturally occurring chemicals, such as the toxins found in botulism and in the mould Aspergillus flavus, which grows on peanuts, are often far more dangerous weight for weight than anything produced by the chemicals industry. And many compounds known to cause cancer are found in plants such as comfrey, coltsfoot and sassafras, popular ingredients of herbal remedies.

What increases the public's perceptions of the dangers of chemicals is the growing ability to detect them in the environment at incredibly low concentrations. A century or more ago a level of purity of 99.5 per cent in a foodstuff would have been impressive. Today, this figure is interpreted as an impurity level of 5,000 parts per million, which for many contaminants would result in a food product being banned. Analytical techniques today routinely measure impurities to a few parts per million, and can even go down to parts per billion. In 1990, for instance, Perrier had to withdraw its mineral water when it was found to contain levels of benzene, a known carcinogen, at 15 parts per billion. Finding one in a billion is equivalent to being able to pick out a single grain of sugar in a hundred one-pound bags.

As analytical chemistry becomes increasingly sophisticated, with the ability to demonstrate the environmental ubiquity of a wide number of potentially toxic chemicals, educating the public about the reality of a risk may become even more difficult. It will be especially hard if the mass media is increasingly geared towards trivialising important risks - such as smoking - and exaggerating minor ones. Risk experts of the future may find their messages hardly ever get through - though in the risk business there is one certainty: you can never say never. !