Do our genes hold the secret to health and disease risk?
Genetic tests have received a lot of media attention. The public is being led to believe that genetic screening can identify people at risk for diseases, such as heart disease, cancer and diabetes, and that genes hold the promise of personalized medicine and nutrition. It all sounds so scientific.
While the marketing of genetic screening tests gets lots of buzz, the information from genetic research scientists, medical professionals, health departments, federal regulatory agencies and consumer protection groups rarely does. For years, they’ve been trying to warn the public about the lack of scientific validity and commercial interests behind genetic screenings, and the serious potentials for misuse of private genetic information. Meanwhile, just as health agencies across the country crack down on commercial genetic testing, the companies find another way to skirt regulatory censures.
Helen Wallace, Ph.D., Deputy Director of GeneWatch UK has equated tailoring your diet to your genetic make-up as being as scientific as tailoring your diet to your star sign. “With only a few exceptions, what the genomics companies are doing right now is recreational genomics,” said David B. Goldstein, Ph.D., director of the Center for Human Genome Variation, Institute for Genome Sciences and Policy at Duke University in Durham, NC. “The information has little or in many cases no clinical relevance.”
As Dr. Kenneth Offit, M.D., MPH, with the Department of Medicine at Memorial Sloan-Kettering Cancer Center in New York, has explained, personalized medicine is not scientifically supportable and genetic markers have not been shown to be valid in predicting disease, a caution also issued by the CDC, FTC and FDA.
Genetic markers are little more than a vast collection of hundreds of thousands of correlations gleaned from data dredges on vast populations, comparing genomes of patients and healthy people. They come from what are called genome-wide association studies, such as the Human Genome Project, that seek to identify inherited genetic variants associated with risks for diseases. Their significance, however, is far different from what the public is led to believe. The emerging consensus of the scientific community is that these correlations are not true disease markers but part of a blizzard of false positives that can be found anytime hundreds of thousands of markers are probed at the same time.
The scientific basis of genomic profiles for assessing health risks and personalize health interventions was recently reviewed by international genetic experts, led by A. Cecile J. W. Janssens from the Department of Public Health at Erasmus University Medical Center in Rotterdam, The Netherlands. Their review of the scientific evidence published from 2000, appeared in the American Journal of Human Genetics. Of the 260 studies in their analysis, examining genetic screening tests for 160 unique polymorphism disease associations, nearly two-thirds of the associations (62%) were found to not even be statistically significant. Of the significant ones, the odds ratios were modest, ranging from 0.54 to 0.88 for protective variants and 1.04-3.2 for risk variants.
As other experts have reported with other genetic links, for example, genes said to be associated with cardiovascular diseases were more frequently associated with noncardiovascular diseases than cardiovascular diseases! Dr. Janssens and colleagues concluded that there is no scientific evidence to support genomic profiles as assessing genetic risks for common diseases such as cancers or heart disease, or in developing personalized diet and lifestyle recommendations for disease prevention.
Concerns surrounding this issue go far beyond the recreational use of online genetic tests by consumers being sold ‘healthy’ diet and lifestyle products and interventions. In the UK, the Nuffield Council on Bioethics has just launched an investigation into personalized genomics and medical profiling, electronic medicine with its strong focus on screening and prevention, and genetic testing being claimed to predict a person’s risk of developing a range of diseases. As the NHS explains, these tests are different from proven DNA tests for single gene disorders, such as cystic fibrosis. Genetic screening tests are being criticized for delivering potentially misleading, unreliable and inconsistent results with no evidence for any real health benefits. There are numerous examples of investigations into different genetic screenings where healthy people are given vastly different reports of their risks for developing diseases.
Genomics: association studies
The current issue of the New England Journal of Medicine features articles discussing genomics from a variety of genetic researchers. The overriding take-home message is that genome-wide association studies are the world’s largest data dredges and challenge us to remember that: health risk factors are correlations, correlations do not make causation, relative risks are different from absolute risks, and statistical significance is not the same as being tenable or biologically relevant.
While the more than one hundred genome-wide association studies have mined the genome for countless associations, in reality, they have little real meaning, Dr. Goldstein said. “Most common gene variants that are implicated by such studies are responsible for only a small fraction of the genetic variation that we know exists,” he wrote. Among the facts universally known about human variation, said Richard C. Lewontin, Alexander Agassiz Professor Emeritus of Zoology at Harvard University, is that “any two unrelated human beings differ by about 3 million distinct DNA variants.”
Genome-wide association studies identify common genomic markers, single-nucleotide polymorphisms (SNPs), and although the strongest among these common variants have been identified, scientists are finding that they’re packing much less of a punch than had been anticipated, said Dr. Goldstein.
An unreasonably large number of such variants appear to account for the genetic component in risks for common diseases, such as diabetes and cancers, as well as for inherited traits, such as height. The estimated number of SNPs required to explain 80% of the population variation in height (the most common estimate of height's heritability), for example, is 93,000 SNPs. If it takes “a large chunk of the genome to explain the genetic component of a disorder,” he said, genetics provide relatively little guidance about the biology of these conditions because most genes are “height genes” or “type 2 diabetes” genes.
“In pointing at everything, genetics would point at nothing,” he said.
Genetic screening tests, which focus on just a few variants, are unlikely to credibly identify people at highest risk for common diseases, and instead, are more likely to overestimate the significance of genetic results to the public and, especially, to be misused.
“Most variants that have been identified to date are markers, not causal variants,” he wrote, “and are generally assumed to reflect the effects of some other, as-yet-unidentified common variant.” Or, some of these associations may also be multiple rare variants that occur, by chance, more frequently in association with one allele at a common SNP than with the other.
“It seems much more likely,” he said, “that most genetic control is due to rarer variants, either single-site or structural.” These aren’t going to be found in genome-wide association studies, he said. “It’s hard to have any enthusiasm for conducting genome scans with the use of ever larger cohorts.”
As epidemiologists Peter Kraft and David Hunter of Harvard School of Public Health in Boston explained, the vast majority of these risk-marker alleles identified in genome-wide association studies also confer very small relative risks. Most range from 1.1 to 1.5, they said, far below the relative risk of 3.0 that researchers would consider tenable — above random chance for such data dredges, for instance — even though they meet mathematical criteria of statistical significance. Even when combining these alleles, these genetic markers have a poor ability to identify or predict those people who are susceptible to disease, they said.
A striking fact about genome association studies, they wrote, is that even collectively, genetic markers explain only a tiny fraction of the genetic role for most diseases. For example, the approximate number of risk alleles that would be needed to statistically increase a sibling’s risk for getting diseases such as diabetes, heart disease and many cancers go far beyond a single marker. It would take 550 to 1,374 alleles, each associated with a 10% relative risk of a disease, combined to reach a tenable relative risk of 3.0.
But this still fails to show that these correlations are clinically meaningful or have a causal role. “The clinical value of a genetic test also depends on its sensitivity, specificity, and positive and negative predictive values; the costs and benefits of interventions; and the availability of data linking specific variants to improved clinical outcomes,” wrote Drs. Kraft and Hunter. The fact is that even when we can detect subtle differences in risk as more allele are identified, most people will fall around the medium level of risk. For less common diseases, the predictive value of a genetic test “will almost always be low,” they said.
So, while genetic tests claim to identify someone with genetic markers for a disease, wrote Drs. Kraft and Hunter, “the identified variants do not contribute more than a small fraction of the inherited predisposition.”
Genomics is not only considerably more complicated than most consumers hear, it appears to be little more than a fad — one with tremendous potential for financial and political misuse.
As neurology professors John Hardy and Andrew Singleton wrote in this NEJM issue, genome-wide association studies have identified large numbers of associations and made it routine to identify common, low-risk variants. But “the initial contention surrounding the viability of genome-wide association studies has largely subsided” with what many now see as diminishing returns. “Genome-wide association studies identify loci and not genes per se and cannot easily identify loci at which there are many rare risk alleles in any given population.” This approach had been created to fit the “common disease–common variant hypothesis of human disease,” they wrote.
However, the realization “that has taken many observers by surprise is that most loci that have been discovered through genome-wide association analysis do not map to amino acid changes in proteins. Indeed, many of the loci do not even map to recognizable protein open reading frames.” In other words, they might affect gene expression, but such effects are extremely varied and complex and can occur at numerous stages (transcription, messenger RNA stability, and splicing or translation efficiency). Ever sorting it all out is an extremely unlikely probability.
Then, there’s the widespread belief in the role of environmental and behavioral influences on disease. “To state that most complex diseases are caused by an interaction between genome and environment is a cliché,” they wrote. “Such interactions, while likely, have for the most part not been demonstrated, and we should be cautious about universally subscribing to this belief without evidence.” Yet, they pointed out, “this is one of the goals of the recently announced Genes, Environment, and Health Initiative of the National Institutes of Health.”
The most disturbing re-emerging trend is the use of genetic markers to “identify” biological weaknesses and disease risks associated with race and ethnicity. Genomics is stimulating a new field of race-based public health, government-funded research and health policy, euphemistically couched as addressing “health disparities.” If you missed the coverage of the science of genomics, race and genetics, and health disparities, please read and think about the Council for Responsible Genetics’ review, “History of Race and Science,” and the Social Science Research Council’s “Is Race Real?” series, and follow it to its historical, logical and troubling conclusions.
The great and terrible war which has now ended was a war made possible by the denial of the democratic principles of the dignity, equality and mutual respect of men, and by the propagation, in their place, through ignorance and prejudice, of the doctrine of the inequality of men and races…Biological studies lend support to the ethic of universal brotherhood… For every man is a piece of the continent, a part of the main, because he is involved in mankind. — United Nations Educational, Scientific and Cultural Organization, The Race Question, Paris, 1950.
The risky gene
In light of the growing popularity of genomics in public health, an article in Patterns of Prejudice offered another thought-provoking perspective. “The risky gene: epidemiology and the evolution of race,” was written by epidemiologist Philip Alcabes, Ph.D., associate professor of Urban Public Health at Hunter College and the City University of New York.
Did you know that the United States is the only industrialized nation in the world that does not collect health data according to economic class? “Instead,” he began, “we report almost every morbidity and mortality statistic by race.”
Today is the heyday of behavioural epidemiology. Eat too much and you can become obese. Smoke and you raise your chances of dying of cancer. Have sex without a condom and you might die of AIDS. Life seems thick with risk, some of it still waiting to be detected by a passing epidemiologist. What did you eat for breakfast? Do you buckle up?
Epidemiology has departed from systematically, scientifically generating knowledge for the purpose of better understanding the social inequities that result in health problems among disadvantaged. “Now, the people who generate such knowledge are the activists and advocates, spokespersons for one or another identity group,” he said.
Is American epidemiology turning away from its original raison d’eˆtre in order to join in the pursuit of a new American order in which those inequities, like poverty and housing discrimination, that are easily seen to be causes of health problems, are to be ignored? …
No epidemiologist can tell you what your risk of disease is based on your behaviour… Which is exactly why you hear so many conflicting stories about whether or not you are supposed to be eating broccoli and Grape Nuts. Still, the thrust of epidemiology lately goes towards calculating the risk allegedly associated with particular activities.
Research articles on behavioral disease risk published in professional journals now far outnumber epidemiological studies examining the greatest health risk, socioeconomic status. “What my colleagues do not realize is that their work also supports the delivery of health-risk-avoidance messages to the public in terms that are heavily moralistic,” he said.
Their work contributes, albeit indirectly, to a new set of pariahs in our society: smokers, women who drink coffee while pregnant, unhelmeted motorcyclists, people who have ‘unprotected’ sex… I must emphasize that epidemiology has very little to say about whether you will be better off if you engage in ‘safe’ behaviors…
The leading killer throughout the nineteenth century was tuberculosis (TB). Here, too, the effects of social structure were displaced on to behaviour, and behaviour was used to define, and then indict, a race. TB is transmitted by the only unavoidable and unmodifiable activity of life, breathing. It is the classic example of a disease whose distribution is determined purely by social conditions… Unsurprisingly, it was the leading killer among tenement dwellers in the industrial revolution era, and it remains dangerous to the poor in densely populated places everywhere today. In the United States before the Civil War, TB was responsible for 20 percent or more of all deaths, and was more pernicious in the industrial cities of the North than in the still agrarian South… Proponents of slavery seized on the North/South difference in TB…in particular, cited the lower rate among slaves compared to Northern Blacks as evidence that black people were a race apart: the predisposition of the African American towards TB could be contained only through the benign paternalism of slavery….
In our own time, higher rates of both AIDS and heart disease among black Americans have been attributed to habits… Heart disease is supposedly a function of diet, or an eating ‘culture’ that disdains low-fat foods, or hip-hop advertising campaigns by McDonalds. Here, too, culpability is fixed by holding African Americans responsible for mythic behaviour: sexual abandon, sexual deviance, primitive eating and dancing rituals… If only they could control themselves…
In each of these cases, the victims of disease were seen as having brought it on themselves, through some activity rumored to be characteristic of their group, he said, much like the “obese.”
Aspersions cast on behaviour are magnified by linking the behaviour to a fearsome disease. And the disease thereby helps to identify people as members of the reprobate group.
Identifying behavioural causes of disease produces more investment in yet more campaigns to get people to alter their behaviour and live the healthy life. And that allows us, as a society, to continue to ignore the social problems, economic policies and other forms of social stratification that underlie differences in disease susceptibility... The distance between ‘us’ and ‘them’ is highlighted by their supposed indiscretions. It makes them a group apart, a race. And while it incriminates them, it exonerates us.
I believe that implicating people in the harming of society, through the allegation that their behaviour causes widespread disease, and imagining that people’s behaviour expresses their place in a purported hierarchy of races are two harmful forces in public health… In the United States, ‘race’ is often the name we give to such discontinuities when they really come from class. In general, the race allegation refers to deficits in wealth or access to power. But it also, nowadays, resonates with assumptions about fitness. You are part of a race if you are identifiable as a member of the lesser classes, not just of the lower strata economically, but of the groups whose disease rates are higher.
We can say that the differences we observe in disease rates arise from genetics, or we can say that the differences arise from engaging in improper behaviour. But in either case what we mean is that “those people” — the ones who are ‘at higher risk’, the ones who are ‘hereditarily’ susceptible — are the loci of disease. They carry risk in their genes. They are socially suspect. They are implicated. And since we are innocent, we do not have to solve, or necessarily even examine, the vexing problems of modern society that truly make some people sick while others are well.
Government research funding
These observations from the scientific community — on the science of genomics and its role in the pursuit to identify those at risk of disease and find treatments — take on added relevance when we compare them to the research and health policies our government is supporting.
Tomorrow is the deadline to cash in on the $200 million allocated to the NIH for comparative effectiveness research as part of the American Recovery & Reinvestment Act of 2009 (Stimulus Bill). What isn’t often publicized is that federal research grants only go to projects that specifically support the issues and agendas outlined by the government.
Here is a sampling of research topics the NIH has designated for its new initiative called the Challenge Grants in Health and Science Research:
● The role of health behaviors in cancer prevention (behavioral change studies are encouraged, including diet, physical activity and adherence to cancer screening)
● Role of nutrition in cancer (“application of multiscale modeling to linking effects of nutrition from the molecular to cellular to organism to population studies”)
● New genomic technologies (“physician utilization and/or patient acceptability”)
● Unified informed consent for biobanking and subsequent analysis of human biospecimens (informed consent “remains a challenge that impedes efforts related to biobanking” and the later use of genetic material for research)
● Health disparities and participation in research (identify barriers to minority participation in biomedical and clinical research)
● Development of an electronic health information infrastructure and the sharing of health information
● Translating genetics (“genetics and genomics have great promise for the development of personalized medicine…examples of studies include those related to board sharing and use of new genetic information and technologies”)
● Fingerprints (genetic markers) for the early detection and treatment of cancer
● Projects to “support a large scale bioinformatics effort to identify biomarkers…”
● Biomarker discovery and validation among racial and ethnic minority populations
● Policy for Challenge Grants (“incorporation of analysis of race/ethnicity differences into Challenge Grants”)
● Oversampling minority populations in clinical research (“financial incentives/supplements to NCI-supported clinical trials where oversampling for minority populations is feasible”)
● “Designing clinical research studies based on unique health characteristics of a minority community cohort with regard to co-morbidities” (including the development of clinical guidelines, upgrading surveillance and tracking systems in these communities)
● Health informatics (integrated electronic medical records) to increase cancer prevention compliance among primary care physicians
● Cancer Intervention and Surveillance Modeling Network (consortium focused on use of modeling to determine best strategies for prevention, screening and treatment strategies among the population)
● Enhancing electronic tools to assess patient-reported symptoms, functioning or health-related quality of life and/or health behaviors such as physical activity
● Implementing FDA regulation of tobacco products
● Quantum biology in cancer (an “emerging interdisciplinary field… biological interactions are modeled using mathematical computation and physical measurements in light of quantum mechanics effects”)
● “Augmenting Genome-Wide Association Studies” (high priority)
● “Role of gene-environment interactions in cancer health disparities research” (“Minority and underserved communities usually depict higher incidence and mortality rates for a number of different cancers”)
● Genomic analysis and development of a bioinformatic pipeline for rapid genomic analysis
● Molecular and population epidemiology and health disparities (“including the novel identification and validation of functional or ancestral biomarkers for risk susceptibility”)
● Genome-wide association studies in cancer prevention (“as well as the influence of environmental factors”)
● “Enhance genomic studies with social determinants of disparities”
● “Genomics research targeting minority populations”
● Genetic and environmental associations in cancer incidence
● Health disparities cancer research (“The role of biological factors in cancer health disparities is now a reality with studies that show that genetic risks to cancer varies by racial/ethnic groups.”)
Science or something else? Do enough Americans understand science and remember recent history to see?
© 2009 Sandy Szwarc