Risky business New Scientist Jan 3 98
A CALL for a two-year moratorium in Britain on using genetic tests to fix insurance premiums sets the scene for an indefinite ban, says one of the country's leading bioethicists. The Human Genetics Advisory Commission, which recommended the moratorium before Christmas, has said that the ban should remain in place until the insurance industry justifies-with scientific evidence on a case-by-case basiswhy certain genetic tests should be used. The industry would have to show that specific test results could predict events that are relevant to insurance risk. "It's too early and unsafe to make predictive statements from genetic tests," says Colin Campbell, vice-chancellor of the University of Nottingham and the commission's chairman. David Shapiro, who recently retired as director of the Nuffield Council on Bioethics, predicts that not many genetic tests will be shown to be useful for insurance purposes. What may result is a wide-ranging ban, he says. The commission has no statutory powers. But its report has been welcomed by the government, which will make its views difficult for the industry to ignore. Nevertheless, on the same day the commission published its report, the Association of British Insurers produced its official code of practice. This identifies eight genetic tests that association thinks should now be accepted as predictive, among them tests for Huntington's disease, the breast cancer genes BRCAI and BRCA2, and a test for Alzheimer's disease. Alison Motiuk
Your Money or Your Life New Scientist 18 Jan 97 - p12.
WHEN Demetra Pappas was in her teens, her father started to die from Huntington's disease. He was in the American foreign service. But in the prime of life, he was reduced to a physical and mental invalid. Not only did Pappas have to cope with the trauma of prematurely losing a parent, but she knew that she had a 50:50 chance of inheriting the fatal disease. Pappas says she had a "compelling need" to know whether she would inherit the disease. So when a genetic test was being developed in the early 1990s, she volunteered to take part in the trials. For Pappas, the test has been liberating. She does not have the gene for Huntington's disease. But genetic testing is fraught with problems. The British government's Human Genetics Advisory Commission is due to meet next month. One of the issues it will consider is the impact of genetic tests on the insurance industry. After being tested, some people may become uninsurable. Before the test for Huntington's disease was developed, anyone who had a parent with the disease automatically had to pay about 3-5 times the standard rate for life insurance. And that was if they were luckyoften they could not get insurance. Soon everyone unlucky enough to have inherited the disease gene will have been identified. They will almost certainly find it impossible to insure their lives.
Researchers are making great strides in identifying the genetic basis, of many crippling diseases. Asthma, diabetes, rheumatoid arthritis and hypertension are all underpinned by genes. Will insurance companies start to exact higher premiums from people who have the genes that predispose people to such diseases? While genetic tests will tell us a lot about our potential state of health, and may even hint at our longevity, they are unlikely to predict illness accurately enough to undermine the whole basis of insurance-uncertainty. As a result, many people both inside and outside the industry are coming to the same conclusion: that better genetic information may not make a lot of difference. "At the moment," says one actuary working for a big City insurance company, who wishes to remain anonymous, "there is nothing that genetic tests can tell us that family history doesn't tell us already." Paul Grace, president of the Faculty of Actuaries, agrees. Family history, he says, "is already being used, and it's undoubtedly genetic information". And even if genetic tests could add to family history, they run the risk of defeating the whole purpose of insurance. "If the incidence of Alzheimer's and other dementias were predictable," says Onora O'Neill, who chairs the Nuffield Council on Bioethics, "long-term care policies would become unaffordable for those who really needed them, and unneeded by those who could really afford them." Peter Harper, a medical geneticist at the University of Wales at Cardiff, argues that insurance companies should keep the uncertainty that underpins insurance. "The insurance industry has nothing to lose and a lot to gain from not asking for genetic information." Insurance companies offer a wide range of sickness-related policies, apart from life insurance. There is health insurance, which pays the costs of private medical care. There are income protection policies, which pay out if you suffer a sudden loss of income. There are also a variety of policies that will pay your mortgage or other credit bills should you fall ill.
Sharing the risk
Health and life policies are the most likely to be affected by better genetic information. But more so in some countries than others. In the US, health insurance is private and follows the principle of "mutuality", or sharing a risk as close as possible to your own with other clients. Some 400 different providers of healthcare are competing vigorously, says O'Neill, and they have every reason to want only low-risk clients on their books. As a result, those most likely to need healthcare are excluded or face huge premiums. "It's the best argument we've got for a robust public health insurance system," she says. Under a state-sponsored health scheme, on the other hand, such as Britain's National Health Service, all risks are pooled together. This strategy is known as "solidarity". The idea is that there will always be enough healthy people paying into the system to compensate for the few who fall ill and require expensive care. Walter Bodmer, the former director-general of the Imperial Cancer Research Fund, who is now at the University of Oxford, says: "The more we know about genetics the more important it is to count on solidarity." In Britain, therefore, life insurance is a more important issue than health insurance. "Everyone will die once, and only once," says Grace. "But not everyone will make a claim under a sickness policy.11 Insuring your life is a little like playing the lottery-except that you do not want to win the jackpot. Insurance companies are gambling that you will not die prematurely. Even if you do, they have calculated that most of their other clients will stay alive and keep paying out premiums. But fatal diseases caused by a single gene are extremely rare. Huntington's disease is one example, and most importantly, it strikes in mid-life-leaving insurance companies to foot the bill. Many other diseases caused by a single gene strike early on, such as cystic fibrosis, or at the end of life, and so are less important to life insurers. The majority of inherited fatal diseases, however, are vastly more complicated. For though great advances have been made, and new genes are being found all the time, just how they interact with each other and with the environment to produce an outcome is extremely difficult to assess. For example, scientists know that BRCA1, a breast cancer gene, accounts for fewer than 5 per cent of all breast cancers. Among those women who carry the gene, 70 per cent will die of the disease before they reach their seventies. But 30 per cent will never get breast cancer. "What is different about that 30 per cent?" asks O'Neill. "We don't know." Should these women be grouped along with the unfortunate ones, simply because we have not yet discovered what protects them? "These tests require expertise to interpret," says Martin Bobrow ' professor of medical genetics at the University of Cambridge and a member of the Human Genetics Advisory Commission. "It makes me nervous that they are trying to set prerniums on the basis of things neither they nor we understand." The same argument can be made about crippling inherited autoimmune diseases, such as childhood diabetes and rheumatoid arthritis, says Bodmer. Only a fraction of people who have the genes end up getting the disease, he says. And even nasty genes that end up expressing themselves are not always a very good indicator of life expectancy, he adds. The most common cancer, apart from the tobacco-induced varieties, is bowel cancer, and it is known t to be inherited. "But if you remove the colon," says Bodmer, "the person lives a normal life." In the Western world, some 95 per cent of all life policy holders pay what is known as a "standard" rate. Within that catchall category, a few simple distinctions are t usually made. The old pay more than the young, men pay a bit more than women, and smokers pay more than nonsmokers. If other details demanded by the policy proposal form, such as height, weight and medical history, suggest nothing untoward, insurers leave it at that. About 5 per cent of clients are deemed "substandard", however, and have to pay extra. A 45-year old obese man with high blood pressure who wants 10 years of life cover may be asked to pay up to twice the going rate. if his blood pressure is really high, he d may be turned down. Under Britain's Disability Disncer crimination Act 1995, insurers are not supposed to discriminate unless they can demonstrate that it is justified. it will take years before the relevance of single genes to multifactorial diseases becomes clear, says Bodmer, and it is doubtful that the risks will increase enough to justify a special premium. Many insurers say they are committed to keeping a simple standard rate. "No one expects it to change much. It hasn't changed much even across a time when medical science has advanced rapidly," says Malcolm Tarting, a spokesman for the Association of British Insurers.
Not worth asking
"It simply wouldn't be cost-effective to ask for all that information," adds Grace. Measuring blood pressure and testing a urine sample could also reveal a lot about a person's health, he says, but most insurers do not ask for these tests as a matter of course. "The volume of paperwork. . ." he groans. "I don't think there are many genetic tests that would. add anything to the underwriting process,' says the City actuary. Could insurers start offering cut-rate policies to the "superfit" to encourage low-risk customers to have genetic tests and identify themselves? These customers would in effect be able to collect their noclaims bonuses in advance. The strategy would make it easier to charge higher premiums for everyone else without demanding that they be tested. "It could be quite attractive financially," says O'Neill. But insurers wanting to sell these policies, with heavily discounted premiums, may have a hard time pinpointing what "superfit" really is. At a meeting organised last autumn in London by the Royal Society, the Institute of Actuaries ahd the Faculty of Actuaries, one actuary offered a definition of "superfit" as "no predisposition to any of the major diseases". Nevertheless, fears remain that perfectly healthy people might soon be paying steep penalties for genes that predispose them to one thing or another, but which will never affect their health. In the absence of any guidance, says O'Neill, people susceptible to cancers and early heart attacks may soon find themselves paying over the odds. Fred Pearce