2600 words

What we eat is important. What we eat can increase or decrease our lifespan. But do different races digest and metabolize different macro and micronutrients differently? On a racial level in terms of individual diet, would one individual benefit from adopting the diet of their ancestors over another diet? Many claims have been made like this in the past few years, such as Europeans evolving to eat plants and grains. This, some people would presume, implies that if you have a certain ancestry then you must eat a certain diet or take different steps in regard to nutrition. I will show this is wrong and that, at least in regard to health and nutrition, individual variation matters more than racial variation (don’t call Lewontin’s fallacy on me. This is not a fallacy).

Different genetically isolated breeding populations evolved eating different diets based on what they had in their environment. Over time, humans eventually developed agriculture and then changed the course of human evolution forever (Cochran and Harpending, 2009). This then leads to large changes in how our genes are expressed and how our microbiome metabolizes nutrients and food we ingest. The advent of farming was, obviously, pivotal to human evolution (Cochran and Harpending, 2009). This then lead to heritable changes in the genome brought on by new foods the farmers ate. This also started the environmental mismatches we now have in our modern world, which is the cause for rising obesity rates and a large part of the cause of so-called diseases of civilization (for a discussion of these matters, see Taubes, 2008, chapter 5; see also page 8 in this summary of his book on diseases of civilization and also see Burkitt, 1973; Cordain, Eades, and Eades, 2003; Sharma and Majumdar, 2009; Sikter, Rihmer, and Guevara, 2017. For an outstanding review on the subject, read Daniel Lieberman’s 2013 book The Story of the Human Body: Evolution, Health, and Disease for in-depth discussions on this point and more in regard to nutrition and our evolutionary history).

Studies come out all the time saying that X population evolved eating Y food therefore Z. Then, people not privy to nutrition science, jump to large sweeping conclusions (mostly laymen and journalists, who are also laymen). These assumptions imply that people’s metabolic systems aren’t, first and foremost, based on an individual level with individual variation in physiologic and metabolic traits. This, I will show, is the reason why these studies don’t mean you should change your diet to what your ancestors supposedly ate based on these studies (though as I have argued in the past, high consumption of processed foods lead to obesity, insulin resistance, diabetes etc which is the cause of a lot of the modern-day maladies currently present in our population today). This assumption is wrong on numerous levels.

Buckley et al (2017), using data from the 1000 Genomes Project (see also Via, Gignoux, and Burchard, 2010), identified novel potential selections in the FADs region. The 1000 Genomes Project tested the genomes of 101 Bronze Age Europeans. They show that SNPs which are associated with arachidonic acid and eicosapentaenoic acid has been favored in Europeans since the Bronze Age (the selection for arachidonic acid being due to milk consumption which is a form of niche construction; see Laland, Odling-Smee, and Feldman, 1999; Laland, Odling-Smee, and Feldman, 2001; Laland and Brown, 2006; Rendell, Fogarty, and Laland, 2011; Laland, et al, 2016; but see Gupta et al, 2017 for a different view which will be covered in the future). They also hypothesize that differences in the selection of these regions is different in different population, implying different epigenetic changes brought on by diet (more on this later).

The FADS1 gene codes for an enzyme called fatty acid desaturase 1 which desaturates n3 and n6 which then catalyzes eicosapentaenoic and Arachidonic acid (Park et al, 2009). These genes code for enzymes that then aid in the breakdown of fatty acids. So, by testing Bronze Age Europeans and comparing their genomes with modern-day Europeans, researchers can see how the expression of genes changed and then work backward and hypothesize how and why the differing gene expression occurred.

The regions selected for are involved in processing n3 and n6 fatty acids. We need a certain ratio of them, and if either is thrown out of whack then deleterious effects occur. This, of course, can be seen by comparing our ratio of n3 to n6 fatty acid consumption with our ancestors’, who ate a 1:1 ratio of n3 to n6 (Kris-Etherson et al, 2000) which you can then compare to our n3 to n6 ratio, which is 14 to 25 times higher than it should be. The authors state that n6 is important, but it’s only important to have the correct ratio, having too much n6 is not a good thing (as I have covered here).

Twenty percent of the dry weight of the brain is made up of long-chain polyunsaturated fatty acids (Lassek and Gaulin, 2009). Therefore it is pivotal we get the correct amount of n3 fatty acids for brain development both in vitro and during infancy, the best bet being to breastfeed the babe as the mother packs on fat during pregnancy so the babe can have PUfAs during its time on the womb as well as during infancy through breastfeeding.

About 85kya selective sweeping occurred in Africa on the FADs genes. Buckley et al (2017) write: “Humans migrating out of Africa putatively carried mostly the ancestral haplotype, which remained in high frequency in non-African populations, while the derived haplotype came close to fixation in Africa. It is unclear why positive selection for the derived haplotype appears to be restricted to Africa. Mathias et al. (2012) suggested that the emergence of regular hunting of large animals, dated to ∼50 kya, might have diminished the pressure for humans to endogenously synthesize LC-PUFAs.” This is true. There is a wealth of important fatty acids in the fatty and muscle tissue of animals, which we need for proper brain functioning and development.

They also write about a study on the Inuit that proves that certain alleles have been selected for that have to do with fatty acid metabolism, which I have also covered in the past in a response to Steve Sailer. Nevertheless, on a population level, this is worth it, but individual variation in metabolism matters more than population. In the article, Sailer implied—with a quote from  New York Times science editor Carl Zimmer—that the Inuit have certain gene variants that influence fatty acid metabolism in that population. Sailer goes on to write “So maybe you should try different diets and see if one works better for you.” Of course, you should. However individual variation is more important than racial variation. (It’s also interesting to note that these genes that are expressed on the Inuit are also related to height.)

Nevertheless, it is true that selection occurred on these parts of the genome in these populations studied by Buckley et al (2017), but to claim that all populations wouldn’t benefit from a low carb, high fat diet is not true. I do agree with Sailer on, in the future, the scanning of individual genomes to see which diet would have a better effect. Though I would insist that most, if not all, humans should eat a higher fat lower carb diet.

Buckley et al (2017) cite a study (Mathieson et al, 2015) which “provides strong evidence of selection in the FADSregion in Europe over the past 4,000 years, in addition to the patterns of selection already reported in Africans, South Asians, and the Inuit.” Buckley et al (2017) also cite a study (Pan et al, 2017) which shows an SNP, rs174557, regulates FADS1.

In their analysis, they showed that “this variation is largely attributable to high differentiation between two haplotype clusters: a cluster widespread in Africa, largely containing derived alleles and possibly subject to a selective sweep (Mathias et al. 2011,, 2012), and an ancestral cluster, which is present across Eurasia.” They also showed that Neanderthal genomes cluster with the derived cluster, which is present in Africans, while Denisovans cluster with the ancestral cluster, which Eurasians also have.

Buckley et al (2017) write: “Thus the derived alleles appear to promote expression of FADS1 while simultaneously abating the expression of FADS2.” This is important to keep in mind for the end of this article when I talk about nutrition and how it affects the epigenome which can then become heritable in a certain population.

Buckley et al (2017) also confirm the results of the European sample using the Nurses Health Study and the Health Professionals follow-up study GWASs: “These results reinforce the associations with cholesterol from the GLGC GWAS. This confirms the hypothesized phenotypic effect of the selected variants in terms of increased EPA and ARA levels of the putatively positively selected variants in the European population.”

Selective (dietary) pressures on the three populations tested (Africans, Europeans and South Asians) have “have driven allele frequency changes in different FADS SNPs that are only in weak LD with each other [LD is linkage disequilibrium which is the nonrandom associations of alleles at different loci in a given population]” (Buckley et al, 2017). Further, the alleles (FADS1 and FADS2) that were under selection in Europeans were strongly associated with lipid metabolism, specifically reduced linoleic acid levels. An opposite pattern was noticed in the Inuit, where selection acted to “decrease conversion of SC-PUFAs to LC-PUFAs to compensate for the relative high dietary intake of LC-PUFAs.” The allele under selection was associated with a decrease in linoleic acid levels and an increase in eicosapentaenoic acid, which may possibly be due to improved metabolism in converting LC-PUFAs from SC-PUFAs.

Buckley et al (2017) hypothesize that the cause is eating a more plant-based diet which is rich in fatty acids (n6 and n3) while a subsequent decrease in fatty animal meats occurred. Of course, relative to hunter-gatherer populations, the increased plant consumption brought on by agriculture caused different methylation on the genome which then eventually became part of the heritable variation. So, of course, farmers would have eaten more plants and the like, which one then select for the production of SC-PUFAs to LC-PUFAs. This of course began at the dawn of agriculture (Cochran and Harpending, 2009).

Of course, this can help guide individual diets as we better map the human genome. These studies, for instance, can be used as guides for individual diets based on ancestral evolution. More studies, of course, are needed.

Also, in an email with correspondence with Arstechnica, one of the authors, Nelson Rasmussen, stated: “Of course, within the last century there have been drastic changes in the diets in many areas of Europe. Diets have typically become more caloric with a higher intake of simple sugars, and perhaps also more rich in proteins and fat from animals.  So selection is unlikely to be working in exactly the same way now.”

Though the strong claim from Arstechnica that “This is another nail in the coffin for the scientific validity of paleo diets” is a strong claim which needs much more evidence because low carb high-fat diets are mostly best for people since their insulin levels aren’t spiked too much which then leads to obesity, diabetes and along with it hyperinsulinemia.

Now I need to talk about how epigenetics is involved here. Nutrition can alter the genome and epigenome (Niculescu and Lupu, 2011; Niculescu, 2012; Anderson, Sant, and Dolinoy, 2012) and cause permanent heritable variation in a population if a certain allele reaches fixation, since there is evidence that maternal and paternal dietary changes possibly affecting multiple generations (Rosenfeld, 2017; though see Burggren, 2016 for the view that there is no evidence for heritable epigenetic phenotype in the genome. I will return to this in the future; see also the Dutch Famine Study showing heritable epigenetic change from famine; Lumey et al, 1993; Heijmans, 2008; Stein et al, 2009; Tobi et al, 2009; Schulz, 2010; Lumey, Stein, and Susser, 2011; Hajj et al, 2014; Jang and Serra, 2014; Tobi et al, 2014). Of course, based on what a population eats (or does not eat), epigenetic changes can and will occur. This not only affects the expression of genes in the body, but also the trillions of gut microbiota in our microbiome that partly drive our metabolic functions. Diet can change the composition of the microbiome, diet can change the epigenome and gene expression, and the microbiome can also up- and down-regulate genes (Hullar and Fu, 2014) Lipid metabolism is also related to developmental epigenetic programming (Marchlewicz et al, 2016). They showed that circulating fatty lipids in the mother during pregnancy are associated with DNA methylation in the genomes of the child. This can also, of course, contribute to health and disease risk in the future for the affected infant. FADS1 is also involved here.

Nutritional factors also come into play in regards to epigenetic inheritance (Alam et al, 2015). The n3 PUFAs also affect gene expression and DNA methylation (Hussey, Lindley, and Mastana, 2017). Further, DNA methylation is also associated with FADS1 and, to a lesser extent, FADS2 (Howard et al, 2014). This is strong evidence that, of course, that what was reviewed above in regards to selection for certain alleles for fatty acid metabolism in certain populations was strongly driven by the consumption of certain foods. Epigenetic changes that occur both in the womb and previous generations like the grandparents’, for instance, also have an effect in regard to which genes are expressed in the baby in vitro as well as consequences for future generations. The study of epigenetics, along with transgenerational epigenetic inheritance, of course, will be very important for our future understanding of both the evolution of humans and the evolution of the human diet.

Finally, I need to touch on why this doesn’t really matter in terms of individual diet choice. The fact of the matter is, anatomic, physiologic, and metabolic variation within race trumps variation between it. Two different randomly selected individuals will have different anatomy, along with different organs missing (Saladin, 2010). This implies that the individual differences in these traits trump whatever racial selection occurred since the dawn of agriculture 10kya. This is why, in my opinion, one should not look to just their ancestry when choosing a diet and should always choose a diet based that’s good for them, individually. Now, I’m not saying that this research is useless in regards to healthy diets, however, increased consumption of processed foods is the cause of obesity since processed foods (high in carbs) spike insulin which lead to obesity and diabetes (insulin causes weight gain). So, obviously, full-on plant-based diets will lead to these maladies. Contrary to the Alternative Hypothesis’ thesis on race and nutrition, this doesn’t really matter, not at the individual level, anyway. This could have small implications in regard to the population as a whole, but as an effect on the diet of individuals? No. Individual variation in traits matters much more than racial variation here (again, don’t call Lewontin’s fallacy because I’ve explained my reasoning which is logically sound).

In sum, the SNPs associated with the increased expression of FADs1 and increased the production of eicosapentaenoic and Arachidonic acid in Europeans occurred around 5kya. These studies are interesting to see how diet and how we construct our niches leads to changes in the genome based on those changes that we enact ourselves. However, laypersons who read these popular science articles on the evolution of diet in human populations will then assume that since they have X ancestry then they must eat how their immediate ancestors ate. The Arstechnica article makes some strong claims that Buckley et al (2017) prove that the paleo diet is not a viable solution for diseases of civilization. Do not make sweeping claims about eating X and Y because your ancestors evolved in Z environment, because individual variation in metabolic and physiologic functioning is greater and matters way more than racial variation

[Note: Diet changes under Doctor’s supervision only.]