Tuesday, June 17, 2008

Hmmm ...

... Healthy lifestyle triggers genetic changes: study.

If behavior brings about genetic changes, would those changes be heritable? And if so, would that not be the inheritance of acquired characteristics? And if those changes were brought about intentionally, they would, ipso facto, not be brought about randomly, right?
I mean, I thought evolution was random, purposeless, and that the inheritance of acquired characteristics was Lamarckism.

Update: Do read Jeff Mauvais's comments attached to this post. I am lucky to have such well-informed readers. Reuters needs someone like Jeff to edit a story like this.


  1. My sentiments exactly.

  2. Anonymous8:33 PM

    The research refers to changes in gene activity, not gene structure.
    Nothing novel here except for the specifics. Changes in gene activity in response to extra-genetic factors are ubiquitous and, indeed necessary in all living organisms. Only the gene's ability (or inability) to respond to external cues is inherited.

    For example, germinating plant seedlings are white (etiolated)until exposed to sunlight, which activates chlorophyll biosynthetic genes, turning the seedlings green. However, this doesn't mean that the next generation seedlings will be green before exposure to sunlight. Only the genetic response to sunlight is heritable and is encoded in the structure of the plant DNA. Structural changes to the DNA can, and do, abolish this capability, but are not heritable since non-photosynthetic seedlings die (natural selection).

    The research in the link simply demonstrates that human gene function responds to the internal human physiological environment, which in turn is affected by diet, exercise, cortisol levels (stress), etc. No surprise there. However, the work could lead to the identification of specific genes involved in the response, which could be useful medically.

  3. Thanks, Jeff, for explaining that. Now I understand (shows that a little knowledge remains a dangerous thing). BTW, are you familiar with Edward Steele's research into retrogenes and his book Lamarck's Signature?

  4. Anonymous9:11 PM

    I just re-read the Reuters article and found the probable source of confusion. It's Ornish's statement that "In just three months, I can change hundreds of my genes ...". What he meant, as the rest of the article makes clear, is "In just three months, I can change the activity of hundreds of my genes...". That is, by changing one's internal physiological environment, one can turn on disease-preventing genes and turn off disease-promoting genes. But only if one actually carries a given gene, and if it is responsive to environmental cues. Not every individual will carry all of the "good" or "bad" genes, which accounts for individual differences in susceptibility to various dieases.

    I'm not familiar with Steele's book, but know a bit about retrogenes. I'll check it out. Thanks for the suggestion.

  5. Anonymous9:27 PM

    I just read the description of Steele's book on Amazon -- sounds interesting. Viruses are certainly the wild cards in the game of evolution.

  6. HI Frank & Jeff,

    Here is the paper:

    PNAS: Changes in prostate gene expression in men undergoing an intensive nutrition and lifestyle intervention

    It is a study of events, so is necessarily inconclusive. It does not look like they had a control group, but instead, relied on this line of thinking: Epidemiological evidence (1, 2) and migrant studies (3) indicate that the incidence of clinically significant prostate cancer is much lower in parts of the world where people eat a predominantly low-fat, plant-based diet.

    What we don't know is whether the chicken came first. Did a change in genetic activity lead to behavior that improved lifestyle? Or, did lifestyle lead to the change in genetic activity? The study shows that each took place, but without a causal link reported.

    Here's the Ornish quote that has a lot to do with misleading readers:

    "It's an exciting finding because so often people say, 'Oh, it's all in my genes, what can I do?' Well, it turns out you may be able to do a lot," Ornish, who is also affiliated with the University of California, San Francisco, said in a telephone interview.

    "'In just three months, I can change hundreds of my genes simply by changing what I eat and how I live?' That's pretty exciting," Ornish said. "The implications of our study are not limited to men with prostate cancer."

    That looks like an advertisement for soy shakes and diet pills, because the study only looks like it supports the idea that what he is excited about is correct. Notice his qualifiers, "you may be able" and "The implications of our study." Nothing is firm.

    It's not that the study is unimportant. If the findings were different, then the hypothesis would have to be discarded or changed. Also, data was collected that can be used in the future.


  7. Anonymous12:28 AM


    Thanks for the link to the PNAS paper. This was a well-designed pilot study, with the limitations characteristic of this phase of research: small sample size, incomplete controls, grouped variables, etc. But it does what pilot studies are supposed to do: identify specific questions for future research. All the necessary caveats are provided in the Discussion section of the paper. We have to remember that science is a cathedral-building enterprise: one stone at a time.

    I would have preferred the inclusion of a placebo (no lifestyle modification) control, but it looks like they didn't have enough experimental subjects at this time. Interesting preliminary results like this usually generate significant grant funding which enables larger, more focused studies in the future. But the research was not completely without controls. Each of the 30 subjects supplied a prostate RNA sample prior to the commencement of lifestyle modification, and then another after three months of treatment. This provides internal control (or baseline) data.

    The results contain a nice mix of expected and unexpected data. One would expect down-regulation of lipid metabolism genes in response to a low-fat diet, but not necessarily a similar response with a group of oncogenes. The expected results inspire confidence in the reliability of the protocol, making the unexpected results meaningful.

    I only wish they could rapidly separate the effects of the grouped vaiables, so that I'd know whether to emphasize the jogging or the broccoli.

  8. Hi Jeff,

    You essentially elaborate on the problems I point to.

    I go a step further and call Ornish on his misleading exuberance. The study only showed that he may hold to his hypothesis for a little while longer, anyway. His "it turns out you may be able to do a lot" statement is in the Reuters article and was not part of his report's discussion, but it is how he is selling his work to the public. So, he sticks to the proper format when creating his report, and then puts the snake oil PR in afterward.

    I'm with Frank on this. It's okay if he found a writer to get the word out through Reuters, but this needed better editing.


  9. Anonymous2:15 AM


    Please excuse the piecemeal nature of these comments. I received my child's first college tuition bill today and I have been slipping in and out of consciousness since I opened the envelope.


    You are absolutely right about Ornish. My first two comments were an effort to explain and paraphrase the Reuters article. Once I read the PNAS paper, I realized that Ornish was engaged in some serious overreaching in his comments to the press, particularly in referring to disease-promoting and disease-preventing genes. Many of the genes currently have no known function.

    What did the research actually demonstrate? That one or more of a group of lifestyle changes were associated with changes in the expression of several hundred genes in prostate tissue taken from men with indolent prostate cancer. Among the many unanswered questions: Which lifestyle changes were associated with which changes in gene expression? What do these genes encode (some known, most not)? Are changes in the expression of specific genes associated with changes in prostate cell function? Are any of these changes associated with a change in disease state?

    Lots of unanswered questions, I agree, but none of them askable before this research was done. And all of them potentially answerable with further, more focused research.