Home genetic tests: How much do you really want to know?

Home genetic tests now claim to reveal if you’re at risk of cancer, or diabetes, or mental illness. But how accurate are they? And how much do you really want to know?

(Illustration by Paul Blow)

A round 2 per cent of the population will suffer from bipolar disorder at some point in their lives. That doesn't sound much. But what if there were a test which would tell you that you were likely to be one of those who will go on to develop it. Would you take it?

It may sound like the far realms of science fiction but a company in the United States called Psynomics Incorporated claims to have devised such a test, which involves you doing no more than spitting into a container and mailing it to San Diego in California where the saliva will be used for a DNA test. What's more you can order it on the internet for $399 – or $750 if you want also to find out what your genes reveal about your likely reaction to the serotonin-based drugs that are commonly prescribed to treat depression nowadays. There's talk of a spit-test to find out if you are genetically susceptible to schizophrenia next too.

Once genetic testing was the province of medical professionals. But now you can do it in the anonymity of your own home. And depression is only the start of it. Since the cracking of the human genome, genetic testing has been mushrooming all over the internet. There are now some 1,300 things you can be genetically tested for – including a genetic susceptibility to cystic fibrosis, muscular dystrophy, haemophilia, Huntington's chorea, sickle cell disease, multiple sclerosis, Crohn's disease, breast cancer, diabetes, heart disease and Alzheimer's.

Mostly the companies are based in the United States but they have begun springing up in other countries including the UK. They have names like DNA Direct, DeCodeMe and 23andme (a reference to the 23 pairs of chromosomes in human DNA). The Genetic Health clinic in Harley Street, London, offers DNA testing for cardiovascular disease, breast and ovarian cancer, prostate cancer, thrombosis, hypertension, osteoporosis, obesity, hormone replacement problems and Alzheimer's. Take the test, it proclaims, and "Take control of your life and your health". Knowledge will enable you to take preventative action in your most vulnerable areas and live healthily to a ripe old age. A few companies even offer to sell dietary supplements that will, helpfully, address the problems their tests have highlighted.

You can even have the genetic make-up of your dog tested, by BioPet Vet Lab in Knoxville, Tennessee, to make sure that you weren't misled when you were told it was a true pedigree. The doggie DNA test costs just $59.95 but, make no mistake, direct-to-consumer genetic testing, as it is called in the trade, is big money. Knome, a personal genomics company based in Cambridge, Massachusetts – where it "works alongside" leading geneticists, clinicians and bioinformaticians from Harvard and MIT – will sequence and analyse the whole of your genome for $350,000. For a more modest $2,500 you can get the California firm Navigenics to screen nearly two million genetic markers.

It all adds up. Myriad Genetics, the mother company of DNA Direct, saw its revenues grow from under $9m in 2000 to $43m in a five-year period. Most of the money came from breast cancer tests.

But a backlash has now begun. Critics have begun to say that many of the tests being sold to consumers are not backed by adequate scientific studies. Last month the California Department of Public Health launched a crackdown. It sent "cease and desist" letters to 13 genetic testing companies telling them they could no longer solicit business from state residents – unless they can prove that their laboratories are state certified and that customers can ask for a test only on doctors' orders. New York has already taken similar action against 31 companies. The US Federal Trade Commission has started investigations into possibly deceptive advertising or marketing of genetic tests.

How pervasive all this is in the UK remains unclear. "There's absolutely no data on how many of these services are used by people in Britain," says Stuart Hogarth, a sociologist at Loughborough University, who has been researching home testing for some years. "It's now very much an international phenomenon."

But it's now definitely a problem here too. So much so that the authorities have become concerned. Just two weeks ago the Human Genetics Commission, the government's new advisory body on the issue, held a meeting with gene testing companies from Europe, Iceland and North America operating in the UK, along with doctors, academics and government representatives. They agreed the need for a voluntary code of practice for tests sold directly to the public, which are currently marketed without any regulatory scrutiny.

They have some way to go to reach agreement, according to Dr Frances Flinter of the commission, who is also director of the Genetics Centre at Guy's and St Thomas's Hospitals in London. "Among the issues to be resolved are: have the tests on offer been properly evaluated? Do they measure what they say they do? If they do, is there anything that can be done about what they reveal? Who would develop the code and maintain it? And who should police it?"

The meeting ended with little but broad agreement, and no timetable for further action. "The companies are keen because they see it will give them a kind of kitemark seal of approval," said Dr Flinter.

"But," said one of the participants, Dr Ron Zimmern, "the devil will be in the detail." Dr Zimmern is director of the Foundation for Genomics and Population Health at Cambridge, and knows that a code means little if it is not a rigorous one. Dr Helen Wallace is director of the lobby group GeneWatch UK, which monitors genetic technologies from a public interest perspective. She has just given evidence to the House of Lords Science and Technology Committee on genetic testing – and she is far more sceptical. "For the public to be properly protected is going to require legislation," she says, "yet in the meeting the buck kept being passed from one government authority to another."

So is this home-brewed genetic testing a good or a bad thing on balance? The case for is trumpeted on the Genetic Health website under the headline: THE TEST SHOCKED ME INTO EATING BETTER. After taking the company's £825 Premium Female gene test, Victoria Hanlon, a 27-year-old account executive, said: "I was worried about stroke as it runs in my family, but it seems my genes are OK for that. However, I have a 140-fold increased risk of cancer as I don't clear pollutants very well – I could lower this risk by eating more fruit and vegetables, particularly cruciferous ones like broccoli and cabbage."

Potential benefits, say home testing's advocates, allow identification of those at risk, allowing preventative measures and earlier, more effective treatment. It will promote greater awareness among individuals of their own health and more control of diseases in the general population. Identification of carriers of genetic mutations will allow for more informed family planning.

The downside, say the critics, is that unless tests are adequately regulated, large numbers of people are likely to be either unnecessarily worried or falsely reassured. Healthy people could end up taking unnecessary medication. Unhealthy people could decide that a "good" test was a licence to binge. Public health messages on healthy eating, exercise or quitting smoking might be undermined, and busy family doctors further burdened by patients who have been internet tested and then seek either reassurance or medication.

There are wider ethical implications. A genetic test for one person may yield information that has implications for family members. People might send in someone else's saliva for DNA testing – as they already do with paternity tests where women send in their partner's chewed gum or samples from their hair brushes to have them tested without their knowledge – which can sour relations between parents and children for decades. Insurance companies, which at present have a voluntary moratorium on the use of genetic information to fix insurance premiums, could decide to start requiring it. Not to mention the whole minefield of tests that produce false positives and false negatives.

How is this to be resolved? There are three kinds of diseases currently being tested.

The first are diseases that can offer black-and-white results. If you are in a family which has the deadly Huntington's chorea you have a 50 per cent chance of getting the disease; if the test shows you are carrying the gene you can be certain you'll get it – though you don't know when, how severely or swiftly, or whether you might die of something else first. Tests for diseases with firm links to some individual genes, like cystic fibrosis, are readily available on the NHS. But when such conditions are tested privately that throws up ethical questions. Are the tests being accurately carried out? Who is regulating and policing that? Are people being adequately counselled before and after the tests?

Then there are the social implications. The negative side is illustrated by the 36-year-old German teacher who refused to have a test for Huntington's after her father died of it and lost her job as a result, though the courts eventually overturned the state's decision. The positive side is the testing in the United States of the Ashkenazi Jewish population where one in 30 carried the Tay-Sachs gene and where testing resulted in a 90 per cent reduction in births of children carrying the disease.

The second group of diseases are in a more complex area where a particular gene seems to contribute to increased risk but scientists are unsure why or how or whether it only does so when it interacts with a lot of other genes – there are 20,000 genes in the human body and many common variants – or some environmental factor. An example of this is breast cancer.

It is now possible to test for the BRCA1 and BRCA2 genes associated with the virulent form of breast cancer in younger women. This is one of the tests that the Genetic Health clinic in Harley Street offers. "I think we do have the evidence," its medical director Dr Paul Jenkins, a consultant at Barts, has said. "All of the genes we analyse have been published in very large-scale studies in the most eminent medical studies and show a clear association between those polymorphisms that individuals possess and their risk of having a disease... Those who discover they are at greater risk might think carefully about taking oestrogen or hormone-replacement therapy and having regular screening."

Others are less convinced. "These diseases are known to arise from the interplay of many genes and environmental factors, and often we understand only one or two of these influences," insists Dr Wallace, who has asked the Medicines and Healthcare Regulatory Agency to investigate Genetic Health's tests, which she claims are based on inconclusive evidence. All breast cancer has both biological and environmental causes, she says. Moreover, rare mutations in the BRCA genes account for about 5 per cent of breast cancer cases, and even women with these mutations are not certain to get it, though their risk is between 40 and 80 per cent. "But this offers no scientifically agreed basis for tailoring recommendations on HRT to a woman's genetic make-up. Small-scale studies have given conflicting results and don't give meaningful results."

"If you have the BRCA gene and no family history of breast cancer there's absolutely no increased risk," says Stuart Hogath. "So the risk is to only a small subset of the population. Are these companies trying to scare the whole female population into taking their test?"

These are deep waters. You can test positive for these breast cancer genes and not get the disease – since other factors like alcohol, smoking, lifestyle, menopausal status or other unidentified genetic risk factors may increase your risk. And you can test negative and get it.

And the implications can be grave. In addition to false reassurance or unfounded anxiety some women opt for unnecessary preventive surgery – having both breasts removed early in life to forestall the disease. "In Europe genetic testing is classed as a low-risk procedure," says Hogarth. "We need to change that. Canada, US and Australia all treat genetic tests as moderate or high risk. There are significant indirect risks – like a woman deciding to have a prophylactic mastectomy."

Most controversial is the third group of diseases, like heart attacks, diabetes, osteoporosis, multiple sclerosis and mental illnesses. Such conditions are highly polygenic: there may be 10, 20 or 30 or even 1,000 genes all increasing the risk in some way. The nine genes which private testers associate with Type 2 diabetes explain only a very small proportion of the condition in families – and they offer far less accurate risk predictors than existing risk measures like body mass index and fasting plasma glucose concentration. One test offered for multiple sclerosis looks at the IL7RA gene which accounts for less than 0.2 per cent of the risk of developing the disease. No region of the human genome has a uniformly large impact on hypertension. On coronary heart disease a recent study of nine common genetic variants associated with cholesterol levels found that use of the genotype did not improve clinical risk prediction in 5,000 subjects. And though the company offering the test for bipolar disorder claims that genes explain about 50-70 percent of what causes the disease most psychiatrists think stress, injury, environment and other factors play a bigger role.

Scientists from the Netherlands and from the US National Institutes of Health recently published a critique of the scientific basis of commercial genomic tests. The review found that fewer than half the 56 tests they examined were on genes with significant associations with disease risk.

"Scientists are becoming increasingly sceptical about the extent to which we'll be able to predict diseases," says Dr Wallace. "It's harder than predicting what the weather will be and they can only successfully do that five days ahead."

Which? magazine put it more bluntly. Two weeks ago it did its own survey submitting the same two DNA samples to four web-based testing companies. Two companies, DNA Solutions and Oxford Ancestors, examined the unique DNA blueprint of one volunteer and came up with worryingly different results. One company, 23andMe, hedged its bets saying that the DNA sample came from somebody of Polish, Arab or Irish decent. Expensive consumer gene-testing services it concluded "are a waste of money".

It is clearly a bit more complicated than that. "You can see that the companies themselves are making an effort to a sound a note of caution," says Hogarth. "But there's an inevitable tension between that caution and establishing themselves as leaders in a growing market."

"We have to be realistic," says Dr Zimmern. "This technology is out there and people are going to use it. We just have to make sure that it is accurate and used properly and that where it isn't, these tests are treated as a kind of fraud. After that it's up to the punter and his medical adviser to decide."

Don't say you haven't been warned. A little knowledge, said the poet, is a dangerous thing. Caveat emptor.