In April 2002, I attended a meeting of genetic epidemiologists in London. The topic for discussion was UK Biobank: a planned study of the genetic and environmental factors in disease which would involve the collection of DNA samples from half a million people aged 40 to 69. Britain's National Health Service (NHS) planned to link genetic data from these stored samples with lifestyle questionnaires and with health outcomes recorded in electronic medical records.
At the meeting, I intended to ask some questions about the scientific basis of the study, about which I had some serious concerns. Instead, I became a silent witness to a ferocious argument. There were two sides to the debate: those who thought that the project was statistically meaningless and should be abandoned; and those who argued that it was a politically done deal and that they should accept the money and try to get some extra funding to add some scientifically useful studies onto the planned cohort. Ultimately, the latter faction won the day, but only after a bitter battle, conducted largely out of view of the potential participants.1,2 During the course of its development, UK Biobank was attacked as a politically driven project based on unsubstantiated claims.3 Particularly damaging were calculations which showed that its design could not deliver on its promise of characterizing gene-environment interactions and quantifying so-called "genetic susceptibility" to common diseases in the British population.4
Later research revealed the extent to which the British Labour government, led by Tony Blair, had been committed to funding UK Biobank as a pilot study for a DNA database of the whole population, linked to medical records in the NHS.5 Lobbyists for the project- including the Wellcome Trust, GlaxoSmithKline and the funders of the Labour Party known as the 'biotech barons'-argued that screening everybody's genome would allow common diseases, such as heart disease, cancer and type 2 diabetes, to be predicted and prevented, along the lines popularized by the former U.S. head of the Human Genome Project Francis Collins.6 The Government believed that transforming the NHS in this way would allow Britain to win the race to commercialize the human genome and build a new biotech economy. In addition to Blair's commitment to building UK Biobank, jointly with the Wellcome Trust, this led to a £12 billion commitment to build a centralised database of electronic medical records known as the 'Spine' (now partially scaled back); a proposal to sequence the genome of every baby at birth (subsequently abandoned); and a last ditch attempt by Blair's successor Gordon Brown to introduce a data-sharing law (proposed by Mark Walport of the Wellcome Trust) that would have allowed the government to share any medical, genomic or other data with private companies, police or governments without people's knowledge or consent. The latter proposal (hidden in clause 152 of a bill that was largely about the powers of coroners) was dropped within two weeks following massive public outcry.7
A few enthusiasts remain for the idea that one day everyone in Britain will have their genome sequenced; but as more data has emerged, long-term skeptics are increasingly being vindicated by findings that gene sequences have poor predictive value for most diseases in most people.8 This does not detract from the potential of genomics to shed light on disease mechanisms; to diagnose genetic disorders or rare inherited forms of common disorders; or to aid the development of new cancer treatments by studying mutations and gene expression patterns that emerge in cancer tumours. However, it does imply a growing recognition that there are likely to be fundamental limits to disease prediction based on genetic make-up, and that the heritability of common diseases is likely to have been exaggerated.9,10 This implies a need to recognize that tests for common genetic variants are unlikely to have clinically useful predictive ability for complex diseases at an individual level, and to focus on implementation of existing tests for certain rare genetic variants (associated with single gene disorders and single gene subsets of common diseases) that have demonstrable health benefits.11
Half a million people have now signed up to UK Biobank and their blood samples are being stored in a state-of-the art facility near Manchester. But the rationale for why the project is being undertaken is now unclear, as are the ethical rules about who will access the data and on what terms.
On the plus side, the idea that the project should focus on genetic prediction of disease has largely been abandoned. Some scientists remain convinced that the project should never have been funded, but others have been encouraged by its shift in emphasis. They argue that it is unlikely that such a large cohort will deliver nothing useful, even though its original design was based on deeply flawed assumptions. It is possible that the shift of emphasis from genetic to nongenetic factors means that new biomarkers (biological markers of risk such as LDL cholesterol levels) might be identified that lead to new preventive drugs.
On the down side, the UK Biobank resource remains constrained by poor environmental data and limited phenotyping; a cohort that (by starting at age 40) cannot investigate early social and developmental factors; and an expensive fixation on storing samples to test biological factors that are unlikely to provide the underlying reasons why most people develop most diseases. Future funding is also less than certain: the £65 million initial funding was allocated to setting up the resource for access by researchers in academia and industry. Actually generating any findings will require new public and private money.
The most likely outcome may be that UK Biobank will find little of interest from a mix of genetic and biomarker studies and gradually lose favour with researchers and with funders. This does not mean that it will prove to be entirely useless, but its role will be overshadowed by betterdesigned existing and new cohort studies which have a different emphasis, such as the famous Whitehall study of civil servants; the 1946 British birthcohort study; and a new study planned to look at babies' health; all of which include a much greater focus on social inequalities.
It is too late to stop the waste of money involved in recruiting half a million people to a study that was never properly designed. But there are important lessons for the future and for elsewhere in the world. Foremost among these must be the recognition that powerful funders (such as the Wellcome Trust and the Medical Research Council in Britain; the European Commission's research funding programmes; and the National Institutes of Health in the US) should not be allowed to decide research priorities with such a lack of transparency and so little consultation. As British geneticist Professor Steve Jones has pointed out it is hard for scientists to "bite the hand that feeds them" and criticise the limited value of some genetic studies to powerful funders like the Wellcome Trust.12,13 Such funding decisions remain in the hands of a tiny clique of powerful individuals who are inevitably slow to recognise the flaws in the underlying theories on which they have built their fame and their careers. This problem is exacerbated by conflicts of interest, as the history of promotion of the concept of genetic susceptibility clearly shows, with key roles played by the tobacco, food, nuclear and pharmaceutical industries: all keen to convince members of the public that cancer, hypertension, type 2 diabetes and obesity lay in their biology, not in these companies' products or pollution, and was preventable by a massive expansion of the drug market to treat genetic risk factors in healthy people. 14,15,16
A focus on the wrong research priorities has serious and harmful implications for both health and privacy. It is not inevitable that we all have our genomes sequenced: it is part of a marketing strategy, not a credible strategy for health.17 We need to find a better way to decide research priorities if we are to solve the problems that we face.18
Helen Wallace is the Director of GeneWatch UK.
1. Barbour V (2003) UK Biobank: A project insearch of a protocol? Lancet, 361, 1734-1738.
2. Wallace HM (2005) The development of UKBiobank: Excluding scientific controversyfrom ethical debate. Critical Public Health,15(4): 323-333.
3. House of Commons Science and TechnologyCommittee (2003). The Work of the MedicalResearch Council,
4. Third Report of Session 2002-03, HC 132.London: Stationery Office.
5. Clayton, D., & McKeigue, P. M. (2001). Epidemiologicalmethods for studying genes andenvironmental factors in complex diseases.Lancet, 358, 1356-1360.
6. GeneWatch UK (2009) The history of UKBiobank, electronic medical records in theNHS, and the proposal for data-sharing withoutconsent. January 2009.http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/UK_Biobank_fin_1.pdf
7. Collins, F. S. (1999). Shattuck Lecture-medicaland societal consequences of the HumanGenome Project. New England Journal ofMedicine, 341, 28-37. For more informationsee: http://www.genewatch.org/sub-563487
8. Wallace HM (2009) Genetic screening forsusceptibility to disease. In: Encyclopedia ofLife Sciences. John Wiley & Sons Ltd., Chichester.http://www.els.net/[Doi:10.1002/9780470015902.a0021790] September2009.
9. Lander ES (2011) Initial impact of thesequencing of the human genome. Nature,470, 187-197.
10. WallaceHM(2006) A model of gene-geneand gene-environment interactions and itsimplications for targeting environmentalinterventions by genotype. Theoretical Biologyand Medical Modelling 3(35): doi:10.1186/1742-4682-3-35. http://www.tbiomed.com/content/3/1/35.
11. Genomic Medicine: An IndependentResponse to the House of Lords Science andTechnology Committee Report. PHG Foundation(2010). ISBN 978-1-907198-04-5.http://www.phgfoundation.org/file/5441/
12. Jones S (2009) One gene will not reveal alllife's secrets. The Telegraph, 20th April 2009.http://www.telegraph.co.uk/scienceandtechnology/science/stevejones_viewfromthelab/5189941/One-gene-will-not-reveal-all-lifessecrets.html
13. Alleyne R, Devlin K (2009) Genetic researchin a "blind alley" in search for cures for commondiseases. The Telegraph, 20th April2009.http://www.telegraph.co.uk/health/healthnews/5189873/Genetic-research-in-a-blind-alley-in-search-for-cures-for-common-diseases.html
14. GeneWatch UK (2010) History of theHuman Genome.http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/HGPhistory_2.pdf
15. Wallace HM (2009) Big Tobacco and thehuman genome: driving the scientific bandwagon?Genomics, Society and Policy, 5(1),80-133. www.gspjournal.com
16. Gundle KR, Dingel MJ, Koenig BA (2010)'To prove this is the industry's best hope': bigtobacco's support of research on the geneticsof nicotine addiction. Addiction, 105,974-983
17. GeneWatch UK (2009) Is 'early health' goodhealth?http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/Data_mining_brief_fin_3.doc
18. GeneWatch UK (2010) Bioscience for Life?Who decides what research is done in healthand agriculture?http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/Bioscience_for_life.pdf