DNA testing is qualitatively different than family history

Jul 19th, 2011 | By | Category: Health, Learning

I am an advocate of DTC DNA testing. I believe the short and long term benefits of DTC and ubiquitous DNA testing far outweigh the cons. I’m in that camp with many other scientists, writers and bloggers.

But I am also not blind to the pitfalls of ubiquitous DNA testing from invasions of privacy to information that was once hidden that now is openly accessed. Many advocates of DTC seem to minimize these pitfalls, even to the point of suggesting that DNA tests are little different than what we already have.

Razib Khan’s recent post, is a case in point :

Here’s the problem, as I’ve noted before the government already has an excellent predictor of your predisposition to violence: your birth certificate. Genetic information only adds value on the margin, many of the things we can predict about people, from height to I.Q. to personality, are reasonably predicted by easy to access genealogical and sociological information.

This seems to me self-evidently incorrect.

This is akin to saying “Google search just adds value on the margin, many of the things we can find about the world are in library card catalogs.”

Agreed, family history information has strong predictive value . But that is a painstaking, labor intensive proposition. Not “easy-to-access.” It’s not in the ‘birth certificate,’ but rather in an exhaustive search of family records and data. Even if there are no adoptions, paternity questions, missing or misleading information, the effort needed to create an exhaustive and predictive family history of all known propensities is daunting to say the least. It’s not that it’s not doable, but it is labor intensive.  I have a very well-documented family history and I’ve spent lots of time building a health and sociological profile of that history. Gathering the health history of both parents and 4 grandparents alone was time-consuming, and there is no one better placed than I to gather that information.  Yet, faulty memories, missing documentation and other issues made the creation of a family health history (that captured as much useful information as necessary) both time-consuming and incomplete. Often even recent relatives are reported of dying of “cancer” or “natural causes.” In building a family health history, this sketchy information can be less than useful for a health professional or an individual. This says nothing of a ‘sociological’ history of behaviors and environments. There is also, of course, the other end of the spectrum, our adopted daughters for whom we have little or no family health and sociological information.

The point is, though a theoretical family health history might be, right now (and this is an operative phrase)  as good or better a predictor of health information than DNA tests, it is labor intensive and deeply flawed method for gathering information for healthcare provider or individuals, doubly so for governments.

In contrast, DNA testing costs a hundred or so, now takes little real effort and provides the individual (healthcare provider or government) with easily accessible extensive data, an ease of access and to an extent a family history rarely, if ever, provides. Additionally, as more research is done, more health information will be gleaned from that data that could never be gathered from a family history. Sociological information is an entirely different animal. Sociological information, from propensities toward violence to environment raised it, has the same issues when it comes gathering information for a specific individual.

And, importantly, DNA tests reveal information that a family history often never can.. specifically related to family histories themselves. From adoption, to paternity issues, to ethnic issues,  DNA tests reveal information with real-world consequences that a family history and the ‘birth certificate’ never will.

DNA testing is qualitatively different than family history

Share
Tags: , , ,

10 Comments to “DNA testing is qualitatively different than family history”

  1. MaryNo Gravatar says:

    Yeah–family history is sketchy at best. Who knows if the diagnosis they had was even close to correct? And all the new diagnoses–they didn’t have names for Alzheimer’s and autism in the past.

    And one side of my family all died of cancer before they were 55 or so. But maybe they were heavy smokers, or exposed to some horrifying things at work that we aren’t anymore. They worked in textile mills, and had really rough lives. It’s really nothing like how I live today. Also because of their shorter lifespans: we had no idea if they were going to get Alzheimer’s.

    Family history isn’t all bad–I’m just saying it has some gaps.

  2. razibNo Gravatar says:

    “Google search just adds value on the margin, many of the things we can find about the world are in library card catalogs.”

    i think for most people this is probably true. not for “power users” and nerds.

    also, i was responding specifically to the issue of criminality. that’s pretty easy to find so long as the criminal act was committed as an adult.

  3. Although the DNA sequence is obviously very important, how the genome interacts with genetic and non-genetic factors (environment, diet, pollution, stress, etc.) is what is ultimately responsible for medically relevant health outcomes. Considering the importance of non-genetic factors for the biology of health and disease, it is very apparent that one needs to put the genome in its environmental context in order to actually apply genomic information to identify health risks and to improve health outcomes. Family history is a robust approach to do this.

    From a practical point of view, family history really still is the best “genomic test” that we have to predict health conditions (exception- some single gene disorders and somatic mutations in e.g. cancer). As a genetic counselor who has taken literally thousands of family histories, I disagree that a relevant family history is so onerous to collect or interpret as you suggest. Multiple studies have also shown that the accuracy of family history information for major diseases is actually quite accurate (in most cases, in the 80 to 95% range) when compared to medical records, especially in close relatives. For multifactorial diseases, the family health history of close relatives is the most relevant anyway. And, as the awareness of the usefulness of family health history has increased (e.g. the Surgeon General’s Family History Campaign), it is reasonable to posit that the accuracy rate for this info will improve over time.

    Even if incomplete and not 100% accurate (the worst case scenario), family history is still so useful in practice because it 1) captures actual health outcomes; 2) is extremely accessible for most people (both in that it is intuitively obvious and that good info is usually available for at least for close relatives); 3) is extremely flexible; 4) it takes into account both genetic and non-genetic factors that impact health (e.g. environmental factors, shared behaviors, diet, culture, etc); and 5) is very predictive for relevant health states. In addition, precisely because it can capture non-genetic factors, it also can be useful in suggesting or shaping interventions to prevent or treat disease (e.g. Utah’s Health Family Tree).

    Having said all that, I am also fascinated by the newer, more hi-tech ways of identifying genomic factors that contribute to states of health such as whole genome scans and sequencing. I also believe that epigenetics may actually be more useful than the genomic sequence in understanding and addressing the biology of human health. For me, the more interesting questions are figuring out how best to use these tools together (along with biomarkers) to improve our ability to understand the underlying biology of health states and to identify the opportunities for improving health outcomes at the individual, family, community and population levels. I think these different approaches can inform the others and make the sum greater than the parts. Using them together helps to choose the right questions to ask and what information to concentrate on to identify and address health problems that already exist (secondary and tertiary prevention) and for primary approaches to prevent these diseases and to improve health.

    To me, one interesting insight of these types of studies has been to implicate common pathways that lead to quite different health outcomes. Although there have been some early attempts to apply this type of integration in practice in the literature, we still have a ways to go before an integrated approach is really ready for routine use in the clinic. But, these early attempts have been promising and I believe that we will get there before too long.

  4. TreyNo Gravatar says:

    @Kristen admitted, there is a lot more to complex diseases than genetics. Environmental factors, interactions, epigenetics (though I don’t put quite as much stock in the usefulness of epigenetics in comparison with genomic sequence are you do). I’m not questioning the predictive power of a family history concerning relevant health states, but I’d respond that it’s a single relevant health state you are speaking about. Doing a health history for all possible health issues is not trivial. I am not sure that your second point “2) is extremely accessible for most people (both in that it is intuitively obvious and that good info is usually available for at least for close relatives);” is exactly true. Proper, complete and comprehensive health history is indeed more work than most individuals can and do undertake. It’s possible (I’ve done mine, and yes, it’s been very helpful), but not trivial. A genetic counselor (and we need more of them and better paid :D ) who has done, as you have, hundreds and thousands of them, can indeed do them easily and interpret them well, but most individuals have little access to counselors.

    That said, I wholeheartedly agree with you here:

    “I think these different approaches can inform the others and make the sum greater than the parts. Using them together helps to choose the right questions to ask and what information to concentrate on to identify and address health problems that already exist (secondary and tertiary prevention) and for primary approaches to prevent these diseases and to improve health.”

    my own experience has been that my family health history in conjunction with my genetic test has been a huge boon that either individually would not have been.

    I learned from my family health history that I have a great propensity toward prostate cancer (which the genetic test shows) and hypertension (which the genetic test did not), and from my genetic test I found other issues I never could have known from my family history.

    And, though a minority, there is a large plurality of people who family health histories are unattainable (many adoptions) or difficult.

  5. TreyNo Gravatar says:

    @Razib

    Yeah, powers users and nerds won’t find that true (and I’ll put myself in the nerd category :) .

    Even with criminality though, I’m not sure it’s true. Could be, I’d like to see data on that though (sincerely). Is family history really a good predictor of criminality? Intuitively that seems true perhaps (though I’ll be it’s more sociological and environmental than genetic, so makes sense that a family history would be predictive). I think my response is more toward the second part of that quote “many of the things we can predict about people, from height to I.Q. to personality, are reasonably predicted by easy to access genealogical and sociological information.” and to the general idea (which I need to find as I’ve read it before), that family health histories are easily accessed by government.

    Which brings me to the underlying point (and this would be another part of my response to Kirsten) of the post: PRIVACY.

    For a government, corporation or organization (or unrelated individual) to get my family health history is not at all trivial. It would take a lot of research on their part to tease that all out.

    To get my genetic data takes one thing: a file.

    • razibNo Gravatar says:

      good point.

      and yes, adoption studies do show that criminal tendencies seem rather heritable. i’ll look it up. probably an intersection of intelligence and personality traits.

  6. MaryNo Gravatar says:

    There was a hilarious story on The Moth radio show once. Paul Nurse (Nobel winner, current head of the Royal Society) discovered while applying for his US Green card at age 58 just a few years ago that his birth certificate was _not_ indicating the parents he thought were his parents. And earlier in the story he had discovered that 2 of his grandparents weren’t who he though either (and that was *before* he found out his parentage was incorrect).

    There used to be a copy of it out there, but I can’t find the story now. I scoured The Moth site looking for it, and it’s not on iTunes. But if you ever have a chance you ought to listen to “Family Trees are Dangerous”. There are traces of it around (see this Science Blogs post: http://scienceblogs.com/grrlscientist/2009/06/the_science_of_storytelling_pa.php). I’ve been dying to do a post on it, but need the source material.

    There’s a text version of the story here: http://www.telegraph.co.uk/science/7607690/Sir-Paul-Nurse-Geneticist-inherits-a-mystery.html But he’s an awesome storyteller and you should hear The Moth version if you can ever find it.

  7. [...] very accurate in predicting your own health concerns. Recently, I wrote a post that claimed that a DNA test is qualitatively different than a family health history. The point was that for governments, corporations or others, obtaining your family health history [...]

  8. [...] is overstated, if not completely incorrect, as I’ve argued before.  Families of adoption have no such history to look at, Most families have poor family history [...]

  9. [...] This brings me to the second reason, simplicity. I’ve discussed this before, so I won’t go into it here much more.   But, to summarize, obtaining genomic data is very [...]

Leave a Comment