Science: When your make-up lets you down

You've Slogged over the application form, sparkled in the interview and been offered the job. Just a routine health check before it's all official. It's a step up the ladder, a feather in your cap. You're fit and healthy and full of enthusiasm. But when the official letter arrives, it says: "Unfortunately we are unable to offer you this position due to the findings of our genetic risk assessment carried out as part of your pre-employment medical examination."

Could it happen? Could your genetic make-up and the future health it predicts be judged more important than your CV? Could a company legally reject you on the basis of a disease you might - or might not - develop? And could you be gene tested without your consent? Right now, the answers are an unreassuring "possibly".

There are announcements of genetic discoveries almost weekly. Breast cancer, colon cancer, Alzheimer's disease, Huntingdon's disease, cystic fibrosis, schizophrenia, heart disease, even alcoholism and criminal behaviour have been linked to specific genetic profiles. Screening tests currently exist for just a few of these, but many more are in the pipeline including tests intended for mail order sale.

Using this novel form of fortune-telling as an employment tool may be rare in Britain but it is starting to catch on in the US, where company- funded health insurance means employers have a vested interest in their employees' demand for healthcare. A survey carried out last month by the American Management Association found that 6 per cent of the 1,000 participating companies had started using genetic tests in pre-employment assessments. While the common response in Britain is that NHS treatment means it won't happen here, healthcare is not the only cash call that illness makes on employers.

Weeding out employees who are shown to have a predisposition to an illness associated with particular industrial conditions, such as asthma, eczema or cancer, could have a significant impact on insurance, compensation and re-recruitment costs, while testing potential employees for predispositions to conditions such as heart disease, cancer, arthritis or mental illness could reduce sickness absence costs, identify potential poor performers, prevent the drain of early retirements on pension funds, and reduce premiums on corporate health insurances. Some go as far as suggesting that employers might even find it useful to know whether a potential recruit has a high risk of having a child with a serious genetic illness which could detract their attention, time and energy from work.

On the positive side, there is the argument that genetic testing could protect us by letting companies know which of their employees might be more vulnerable to exposure to certain chemicals, for example, and should take greater precautions. Or perhaps, you might decide yourself not to pursue a certain career if a test highlighted a high risk factor. But all this assumes that genetic tests offer accurate predictions. With the endless stream of fascinating information about genes for this and that, it is easy not to appreciate that there are limitations on the predictive ability of genetic testing.

Basically there are two kinds of tests: those with a high predictive value that makes it almost certain you will develop the disease if you have the gene, and those with lower predictive values which show whether you are at increased risk of developing a disease.

The first scenario which is relatively rare relates to diseases caused by single gene defects. For instance, if you have the defect for Huntington's chorea it is 95 per cent certain that you'll develop it between age 40 and 50. Note, though, that even here there is not 100 per cent certainty. Also, the vast majority of diseases that affect working life are not the result of a single testable gene mutation - but of a tangled, multifactorial bundle of DNA, environmental and lifestyle triggers.

Just how dodgy this "nature/nurture" balance can make genetic crystal ball-gazing is illustrated by another condition, familial hypercholesterolaemia. Caused by a gene mutation which disrupts cholesterol elimination, FH affects roughly one in 500 people. If you have this gene and have no treatment, it guarantees heart problems from as early an age as 20 or 30. But with treatment - such as a low cholesterol diet and/or cholesterol-lowering drugs - a person with FH has a lower risk of developing debilitating heart disease than a chip- devouring, smoking couch-potato with "good" genes. Which then makes the stronger guide to ill health: genetic predisposition or habits?

To complicate matters, most common diseases are not only multi-factorial, they also involve multiple genes. In breast cancer, faulty genes are thought to cause one in 20 cases and the much-feted genes for breast cancer, BrCa1 and BrCa2, account for 2 per cent of these So if an employer decided to screen out female recruits with a susceptibility to breast cancer using tests for BrCa1 and 2, they would miss 97 per cent of women who will develop breast cancer.

Then there's the problem that a positive gene test gives only a probability of developing a condition, not a certainty. For example, research shows 94 per cent of those who have the ApoE4 gene which predisposes to Alzheimer's never develop the disease.

On a personal level, test results could cause considerable misunderstanding and confusion. In employment terms they could become dynamite as employers, deluded about the predictive powers of genetic testing, start excluding healthy people whose genetic predisposition for an illness will never be realised. Even where there is strong predictability - such as with Huntington's disease - refusing someone employment 20 to 30 years before symptoms affect their working ability could represent an abuse of such genetic knowledge.

Richard Excell, disability policy officer of the Trades Union Council, says this kind of misunderstanding would create an underclass of unemployable people. "We are concerned that some employers might just take the easy route and discriminate against people with a genetic predisposition - particularly if that condition could be made worse by the work environment or could affect work performance. This could lead to a group of people being denied employment throughout their lives, yet never actually developing the condition for which they have an apparently higher than normal risk."

He rejects the argument that genetic testing could be used positively to protect the susceptible: not only would some people not be in a position to refuse a job regardless of its health risks but "unscrupulous companies could refer to their falling rate of industrial injuries as reason for reducing health and safety precautions," he says.

Professor Anna Kessling, director of North West Thames Regional Genetics Service, says that knowing an employee's genetic profile would offer little practical advantage to employers or employees. "One in four of us will die of cancer, one in five will die of heart disease. These are huge numbers. Unless you can predict precisely that someone is going to die before, say, 43 or whenever the pension kicks in, what good is it to employers to know that someone has an increased risk? Fairly soon, we will be able to test for predispositions to many diseases, then none of us might be employable if these tests are employment criteria because we will all have some sort of risk - though we may never develop a disease."

The scenario would also create huge potential for anxiety, uncertainty and distress, she says. "Most people go into any kind of testing expecting the answer to be that they are unaffected. Then overnight, they may face the prospect of a life-threatening disease. If employers are going to do this type of testing, they will have to be very careful about accurate interpretation of the results and set up extensive counselling services."

In fact, the practical limitations of trying to screen out high risk individuals from the workplace - even with good intentions - have already been thoroughly rehearsed with the use of "atopy" testing - a skin prick or blood test that shows if a person is prone to allergic diseases such as asthma or eczema. Atopy testing was widely used in the 1980s by companies involved in detergents, chemicals or laboratory animals, where there is a higher risk of both asthma and eczema. While the test doesn't actually involve looking for a gene, it offers a similar kind of predictive information.

These tests have now been abandoned as a pre-employment screen, says occupational medicine expert Dr Paul Cullinan, of the National Heart and Lung Institute in London. "They're such a crude measure. Although atopic individuals are at increased risk in some industries, most atopic people do not get asthma. Then, there's the problem that over 35 per cent of the population is atopic. So in the laboratory animals industry, for instance, we estimate you'd have to exclude 12 atopic people from employment to prevent one getting asthma. You also have to bear in mind that atopic people get sensitised because they are exposed to an allergen. If they are not exposed, they do not develop asthma. So the real issue is reducing exposure."

One company, minerals giant Johnson Matthey, which used testing to weed out atopic recruits at its platinum salts plant until last year, stopped when it made a discovery which spotlights the fundamental problem with genetic testing. "We found that someone who smokes 20 cigarettes a day has five to eight times the risk of developing occupational asthma through exposure to platinum salts than a non-smoker, regardless of whether they are either atopic or not," says JM's occupational health physician Dr Peter Linett. Ethics aside, he says, if employers want to discriminate on the basis of potential ill health, they would do better to forget gene testing. Smoking is a far better predictor of who will develop lung disease, not to mention heart disease, stroke, diabetes and osteoporosis.

Nevertheless, genetic tests are coming. And even if the company you work for, or want to work for, isn't keen on genetic spying, insurance companies, who see genetics as another way to load the risk dice, are likely to start applying pressure, increasing premiums or refusing compensation claims to employers they see as neglecting to minimise health risks.

Legally, too, there is nothing to stop discrimination based on genetic testing since an employer which decides to enforce pre-employment tests would not fall foul of either the Disabilities Discrimination Act or the Employment Rights Act. Under the DDA, job seekers are protected from discrimination if they already have a disabling condition, but healthy applicants can be legally rejected because of a genetic predisposition, while under the ERA, an employer can make testing a condition of employment, though they have no right to results of tests taken privately.

This means, says the TUC's Richard Excell, that employees have no legal right to refuse to take a genetic test, nor to claim discrimination or unfair dismissal as a result of refusing. And since taking a genetic test for any reason is now insurance-form declarable (the Association of British Insurers has agreed to ignore adverse test results, except on life cover of more than pounds 100,000 for the next two years only), the genetic test you have been forced to undergo in order to get a job could jeopardise your ability to secure life insurance, pension or a mortgage.

Of course genetic testing - like any medical procedure - can only be done with the patient's consent, but says Olga Aikin, of employment law experts Aikin Driver, exactly what giving consent to a sample allows employers to do is an enormous grey area. "Does it mean an employer can use it to test for anything they want? Does it allow a doctor to pass on the information to an employer, to an insurer."

Since genetic tests can be done on any living cells, the scope of unauthorised testing is vast, says Aikin, who warns potential job seekers at the very least to ask for a written list of tests that will be carried out before they even consent to giving a sample (see box above).

Given the enormous potential for abuse, it is just as well genetic CVs are not yet a reality in Britain. But we haven't got long, and current anti-discrimination laws give little reassurance. With the shadow of the double helix falling further and further across our lives, it's high time for a full review. !

UNAUTHORISED TESTING

GENETIC TESTS can be done on any living cells - urine, blood, tissue, hair - but employees should beware of employers assuming that permission to take a sample means they can investigate anything they want. Employees should make absolutely sure that they know exactly what tests they have consented to. Two court cases, one ruled on in the European Courts of Justice and one currently ongoing in the US, illustrate the potential pitfalls of loose consent for both sides.

In 1995 the European Court ruled on a case concerning a job applicant who had been refused a temporary position at the European Commission. The applicant had consented to a medical check including a blood test, but expressly refused to take an HIV test. Although no HIV test was done, the medical officer asked the laboratory for data from the blood sample from which he could assess whether the person had HIV. The applicant claimed breach of privacy because he said he had not given consent for his HIV status to be examined. The court initially rejected the claim, arguing that the applicant had only refused consent to an HIV test, not to any other test from which his HIV status could be inferred. On appeal, however, the court decided that the applicants' objection to HIV screening should have been observed for any method of testing. But it also warned employees in Europe to be aware that agreeing to give a sample without specifying exactly what could be done with it allowed employers to test for any condition at all.

In the currently running US landmark case, the federal appeal court in San Francisco could establish that employers must have explicit consent from staff for each test they want to run on a given sample. In the case, employees of the University of California's Lawrence Berkeley Laboratory claim their privacy was invaded when blood and urine samples, taken in routine pre-recruitment and annual medical checks, were tested for sickle cell trait, pregnancy and syphilis. Though there is no evidence that the results were used to inform decision-making, results of the tests were not passed on to the employees, who maintain they had no idea the samples would be tested in this way. Should the employees win, the case will be the first to rule that explicit consent is needed for the actual tests to be carried out as well as the sample to be given - and could cost the laboratory thousands of dollars in damages.

Employment law expert Olga Aikin says with genetic tests coming on to the market thick and fast, these cases highlight the dangers of loose consent. "Employees should ask what the test is for and say 'I will be tested for this and that, but I do not give consent for any other investigation'."