Genomics Confidential: Iceland Not So Wonderful

Many people think that personal genomics will change medicine. Doctors will choose treatments based on your genome, learning your genome will tell you what diseases you are at high risk of so you can take precautions, and so on. One person who believes this is Eric Topol. In his new book, The Creative Destruction of Medicine, he writes:

The biggest leap came in the first decade of the twenty-first century. The six billion bases of the human genome were sequenced, and this led to the discovery of the underpinnings of over one hundred common diseases, including most cancers, heart disease, diabetes, autoimmune disorders, and neurologic conditions.

Here is the founder of a company that makes sequencers: ““I believe that the impact on the medical community of whole human genome sequencing at a cost comparable to a comprehensive blood test will be profound.”

I disagree. I have seen nothing that suggests genes make a big difference in any common disease and plenty that suggests environment makes a big difference. My self-experimentation led me to one powerful environmental factor after another, for example. Biologists have invested heavily in the study of genes for reasons that have nothing to do with practical applications, as Thorstein Veblen would be the first to point out.

In 1999, New Yorker staff writer Michael Specter wrote an admiring article about a neurology professor named Kari Stefansson. Stefansson had returned to his native Iceland to take advantage of Iceland’s genetic homogeneity to find genes for common diseases. “In the past, drugs were discovered almost by chance,” Specter wrote, as if this would soon change. The wishful thinking involved is indicated by passages like this:

[Stefansson] and Gulcher selected the five per cent of Icelanders among the hundreds of thousands in their genealogical database who had lived the longest— most of them over ninety. The database allowed the two scientists to seek an answer to a simple question: Are these people who live so long related to each other more often than the average in Iceland? The answer quickly became apparent. People over ninety are much more closely related to each other than people in the general population are, and their children are more likely to live longer than the children of others. That provides strong evidence that the trait is inherited.

“Strong” evidence? The “people over ninety” observation is strong evidence that longevity is inherited only if relatives share nothing but genes. The “their children are more likely” observation is strong evidence of genetic control only if parents pass on to their children only genes. Both assumptions are highly unlikely. For example, surely an Icelandic person lives closer to his relatives than to randomly selected Icelanders.

The article quotes no one with my view (geneticists are overstating the practical value of their work), but it does say that “Stefansson set out to raise capital at a time [1996] when investors had become skeptical about the many unfulfilled promises made by companies claiming that genetic research would solve the ills of humanity.”

Will reality overtake hype? Here is an indication this is happening:

Kari [Stefansson], a neurologist, was a Harvard professor when he co-founded deCODE in 1996. Two years later, Iceland’s parliament gave deCODE access to one of the country’s unique resources—health records of the genetically homogenous population. DeCODE debuted on the NASDAQ stock exchange in 2000, and it made dramatic discoveries of genetic factors associated with cancer, heart disease and other conditions. But the company never turned a profit and filed for bankruptcy protection in 2009.

 

7 Responses to “Genomics Confidential: Iceland Not So Wonderful”

  1. Albert Says:

    Michael Rose found that he could breed longer living fruit flies by natural selection (http://www.nytimes.com/2005/12/06/science/06conv.html?pagewanted=all), so it is not that far-fetched to think there could be something in the genome to account for longevity.

    Seth: No doubt genes influence longevity. But how much — the size of the effect — is important. Is the genetic influence on longevity like pieces of gold laying on the sidewalk (high benefit/cost ratio) or like the gold found in seawater (low benefit/cost of extraction ratio)? In other words, how does studying genetic effects on longevity compare to studying other effects (e.g., nutritional effects) on longevity?

  2. Alex Chernavsky Says:

    See also this:

    Personality Without Genes?

  3. mike Says:

    Parents increase their child’s risk of coronary heart disease through their genes and not through the family’s diet or lifestyle

    a study of more than 80,000 men and women who were adopted as children showed that susceptibility to the disease is transmitted in the womb and not in the home.

    They found that adoptees who had at least one biological parent with CHD had up to 60 per cent more chance of suffering the disease themselves, compared with a control group.
    In contrast, growing up in a home with adoptive parents who suffered from CHD resulted in no additional risk for the child, even if both parents had the disease.

    http://www.telegraph.co.uk/health/healthnews/8725739/Heart-disease-risk-inherited-through-genes-not-behaviour.html

  4. JRM Says:

    Maybe it won’t change medicine, but it is extremely interesting nonetheless.
    I just got my 23andme results back. One of the things they look at is warfarin sensitivity. I have a gene that reduces the efficiency of vitamin K recycling. Guess what? Liquid vitamin K2 drops is one of the few supplements where I notice the effects. Before, I always had chapped lips. Personal genomics has the potential to indicate which supplements would be most useful.
    What else? I apparently have some gene variants for increased intelligence. Over time, personal genomics will help break the taboo on differences in IQ.
    I am a carrier for cystic fibrosis. The gene affects transport of chloride ions. Personally, I am neurotic about salt intake. It has to be just right. If I eat fruit, I need to balance it with salt because the potassium in fruit throws off the electrolyte balance. CF sufferers produce mucus in excess. So, I imagine that carriers still produce abnormal amounts of mucus. This would provide an explanation of why I suffered chronic ear infections as a child. Mucus would block the Eustachian tube creating a breeding ground for bacteria. Of course, it was’t the sole cause. Mucus producing foods such as dairy would also be a cause. Since going paleo, it is not a problem anymore.
    What else? One of the reasons I got tested was to see if I had genes for hemochromatosis. I don’t. So I can keep on eating my 1.5 pounds of meat a day.

    Seth: You sound like an extremely sophisticated consumer of genetics information. Yet even for you the only practical benefit seems to be what you say in the last three sentences: “One of the reasons I got tested was to see if I had genes for hemochromatosis. I don’t. So I can keep on eating my 1.5 pounds of meat a day.” I wonder why you think we know everything about “genes for hemochromatosis”. How much variance do the hemochromatosis genes we know about explain? what if they explain only 10% of the variance? If I were worried about too much iron in my blood I would simply have it measured. So even your one practical example doesn’t make sense. I’m not saying the info isn’t interesting…but the defenders of its practical value seem to have a hard time being convincing.

  5. threepipeproblem Says:

    This guy has a bunch of anecdotes of the kind you said you haven’t seen.

    http://www.ted.com/talks/richard_resnick_welcome_to_the_genomic_revolution.html

  6. Pauline Says:

    This is a nice article reminding us of the limitations of the scientific method.

    http://everythingyouknow-iswrong.blogspot.co.uk/search?updated-max=2011-02-25T04:37:00-08:00&max-results=2&start=18&by-date=false

  7. Sadie Says:

    “I have seen nothing that suggests genes make a big difference in any common disease and plenty that suggests environment makes a big difference.”

    You are dead right about this. The more we learn about epigenetics, the less our previous model of genetic determinism makes sense. We know beyond a doubt that gene expression can be turned off and on (it’s an essential part of how we determine a gene’s function in the first place). Any given gene’s function is affected by other genes and by the cellular environment.

    What we learn from genetic studies in the future may help us understand the mechanisms of how gene expression can be modified by behavior, but it’s missing the big picture– that behavior can change gene expression, and thus every bodily process!

    I’m interested in JRM’s example: “I have a gene that reduces the efficiency of vitamin K recycling. Guess what? Liquid vitamin K2 drops is one of the few supplements where I notice the effects.” Did you experiment with vitamin K2 drops before you received your genetic analysis, or after?