Evolution, pregnancy, and food

The populations at lowest risk for developing gestational diabetes — namely, ladies of European decent — come from cultures that eat (and have eaten, for thousands of years) dairy and wheat-heavy diets that would, normally, increase your risk. Meanwhile, writes Carl Zimmer at The Loom, Bangladeshi women, who have one of the highest risks for gestational diabetes, come from a culture that traditionally ate a low-carb, low-sugar diet. What's going on here? The answer might lie in evolution. It's a particularly interesting read given the ongoing pop-culture debate about whether 10,000 years is enough time for humans to adapt to eating certain foods. This data on pregnant ladies would suggest the answer is, at least in some respects, yes.


  1. I’d be interested to know how much of the evolving has happened on the ‘human’ side and how much has been handled by changes in intestinal flora. There are certainly plenty of microbes involved, and we know how fast those little guys can evolve…

    1. Ditto. My hunch would be that, even if there is a lot of purely human evolution happening in response to diets, it’s accompanied by gut bacteria evolution. I’m not really certain you could totally separate the two. 

    2. Those us who are participating in the American Gut Project were surveyed and asked to keep a food diary, so we’d have some idea of the correlation between what we’re eating and the dominant populations of flora in our guts, when we submitted our samples for sequencing.

      We were also asked questions like were we born caesarian or vaginally, and something of our histories with antibiotics.  What I want to out of this, is to understand how they’ll draw their hypotheses about what represents a ‘healthy gut’, that skirts most of diseases we associate with a Western diet.  We are a culture of immigrants.

    3. At the same time I think people underestimate how quickly evolution can take place because they assume it has to happen through random mutation.  Random mutation doesn’t explain the vast variety of different kinds of dogs (all the same species) or the fact that dairy cows produce so much more milk than is needed for their young, or the fact that we can breed flies that live four times as long as normal flies just by delaying their ability to first reproduce.  The amount of variation possible within species through genes that are already there is very large, and we’ve used selective breeding to demonstrate that huge variation can occur in just a few generations, let alone 10,000 years.

      1. Random mutation doesn’t explain the vast variety of different kinds of dogs

        Yes, it does. Every dog breed is a result of random mutations, which were then artificially selected.

  2. Lactase, amylase, alcohol dehydrogenase – pathways for all of these show signs of selective pressures in some human populations over the past 10K years. Check out the references in this paper: http://www.hindawi.com/journals/ijeb/2011/484769/#B52

    There are other examples of adaptation which may have take place in historic times. When the Romans introduce wine to Gaul, the locals were prone to alcoholism in much the same way as the natives of North America were when the Europeans came. The modern French don’t seem to have this problem. To be fair, though, I know of no smoking gun in the genetics to account for this…

  3. What about these changes reflecting epigenetic changes, eg the expressivity of a gene can be modified by the environment and diet of the mother, and not true mutations of the DNA? Changes to the phenotype of an organism does not always imply that evolution is happening.  

    1. http://evolution.berkeley.edu/evolibrary/news/070401_lactose we domesticated cattle some  8-9000 years ago, but attempts to find a lactose gene in 7000 year old DNA failed. If the gene were present but unexpressed at the 90% rate that we see today we should have still found the gene for adult lactase; it simply wouldn’t have been expressed in the individual. If it wasn’t present as often as it is today then there was change in the genetics of a population over time. It’s a given that the gene for adult lactase existed before domestication, but the large percentage of specific groups that currently express it speaks very strongly of an evolutionary shift.

      Edited because I found information on a specific search for lactose genes.

  4. The idea that there is a single diet appropriate for every human being has always been rather obviously ridiculous; but, somehow, it just won’t die.

    I always tell people to pay attention to what their direct ancestors ate and what their siblings eat now, and develop their diets based on observing the health and morphology of those people.  You can still go wrong – my dad’s problems are due to on-the-job chemical exposure, not his diet – but you’ll probably get better guidance than you can get from any doctor, nutritionist or textbook.

  5. Phenotypic changes can manifest in less than 10 generations if the selective pressure is strong enough. 

  6. Turns out there’s a high correlation between low birth weight and adult diabetes…


    Low birth weight has been associated with an increased risk of type 2 diabetes in adulthood. Poor fetal nutrition has been suggested to explain this association. Our objective was to determine whether genetic factors contribute to the association between low birth weight and subsequent risk of type 2 diabetes.

    We retrieved information from original birth records on same-sex Swedish twins with known zygosity, born from 1926 to 1958. We used regression models to investigate whether birth weight was associated with risk of type 2 diabetes in the cohort of twins overall, and in case-control analyses within disease-discordant dizygotic and monozygotic twin pairs.

    Of 18,230 twins, 592 (3.2%) had type 2 diabetes. The rate of type 2 diabetes consistently increased with decreasing birth weight, from 2.4% among twins with birth weights of 3500 g or more to 5.3% among those with birth weights less than 2000 g. In the cohort analysis, in which twins are analyzed as independent individuals, the adjusted odds ratio (95% confidence interval) of type 2 diabetes per 500-g decrease in birth weight was 1.44 (1.28-1.63). When we compared the diseased twin with the healthy cotwin, the corresponding odds ratios were 1.38 (1.02-1.85), among dizygotic twins, and 1.02 (0.63-1.64), among monozygotic twins.

    Low birth weight is associated with type 2 diabetes in adulthood. The difference in this association between monozygotic and dizygotic twin pairs suggests that genetic mechanisms play an important role in this association.


  7. What I’m curious about is the broader demographic impact.  That is, American food, while pulling from tons of disparate sources, is still largely pan-European in the broad strokes, right?  I’m not a food historian, but that is my guess.  Which…this pan-European based diet is being eaten by people who AREN’T of European descent– & I wonder how it ties in with health & fitness for them.  I haven’t read the article yet but very interesting.

Comments are closed.