January 27, 2014

Is the ability to digest milk in Europeans caused by ancient social inequality?

I've got involved these days in a discussion at Dienekes' Anthropology Blog on the causes of lactase persistance (LP), i.e. the ability to digest milk as adults, in Europe. 

The discussion orbits around a recent pay-per-view study by O.O. Sverrisdóttir, which claims, with some soundness for what I can discern, that LP in Europeans must have gone through positive selection. 

Actually the study, as most of its kind, deals only with one LP marker, the well known SNP rs4988235, whose T variant allows adults (in dominant fashion) to digest milk, an ability often lost after weaning. 

As I discussed back in 2010, there must be other such SNPs because actual LP phenotype only partly corresponds with the known LP alleles. But for whatever is worth, this is the (2010) "known allele" LP map:

In Europe at least, it essentially corresponds with the T variant of rs49235, which concentrates in Scandinavia, Atlantic Islands and the Basque/SW French area. 

At first I boarded the discussion with perplexity, because, even if the positive selection argument seems sound, it seems hard to find a reason for it: milk is not such a "great" source of food, excepting the issue of calcium and a high content of protein and fat, and the occasional claims that it is related to vitamin D deficiency seem extremely feeble because this vitamin is present at extremely low frequencies in natural milk, being rickets (where milk's extra calcium could play some role) only a "less important" side effect of vitamin D deficiency, because its main harmful effect is to impair early brain development, a most serious problem for which calcium seems quite meaningless. So why would the ability to digest milk would have become such a matter of life or death to be actively selected for generation after generation until near-fixation?

A key piece of information is that not a single sequenced Neolithic farmer has ever been found to carry the relevant LP allele (being all CC) and only since Chalcolithic we begin to find some TT and CT individuals. These are found in Sweden and in the southern areas of the Basque Country (see here for a lengthier discussion):
  • In Götland (Pitted Ware culture) only 1/20 alleles was T (i.e. 1/10 persons had the CT combo, all the rest being CC and therefore likely lactose intolerants). 
  • In Longar (Navarre, dated to c. 4500 BP) 1/7 individuals was TT, while the other six were CC (intolerant). There were no CT cases.
  • In San Juan Ante Porta Latinam (SJAPL, Araba, dated to c. 5000 BP), 4/19 were TT, 2/19 were CT, while the remaining 13 were CC.

In the Basque cases we can appreciate that there must have already been two different populations regarding this SNP, because the CT cases are rare, implying that the two groups were only beginning to mix. It is worth mentioning that the Basque sites are odd in several aspects: on one side they seem to be military cemeteries (mostly males, arrow injuries and arrow points) and, on the other, they are rather exceptional in the Basque historical sequence of mtDNA pools (a lot more K and some other lineages than usual, less H and U).

But a key finding in this study is that a Neolithic sequence from Atapuerca (near Burgos city, historical Basque SW border) was again CC for the relevant SNP (and therefore likely lactose intolerant). So it is very possible that proto-Basques did not have the T allele in notable frequencies either (although I keep some reservations for lack of larger samples).

Whatever the case, if the T allele was selected positively as it seems, there must be a powerful reason for it. Was it cows, as some have claimed a bit too vehemently? I doubt it. 

Why? Because for all we know from the Middle Ages, a period very similar in many aspects to the Metal Ages, it were goats and not cows the main providers of milk. This makes total sense because the hardy goats are rather inexpensive to rear, while cows are more costly and were often reserved for traction jobs. In most cases, cow produce, be it milk or meat, was an expensive luxury apt only for the upper echelons of a society that was becoming more and more hierarchical and unequal since precisely the Chalcolithic period. 

Some oral accounts I have heard tell that not so long ago "acorn bread and goat milk" were often staple for the poor. In other areas maybe it was not acorn bread but, say, oat meal (or whatever else), but almost certainly the milk came almost invariably from goat udders, which very efficiently transform leaves and almost any vegetable, even thorny ones, into milk (and meat) for our consumption.

It is crucial to understand that only if milk was a key survival staple, LP would have become fixated. Otherwise people would have preferred alternative foods and survived in similar shape, so positive selection would never have happened at this locus (non-LP individuals would have survived easily, selection would never have happened or would have been mild enough to retain much greater diversity). 

It is also crucial to understand that, for all we know, this positive selection only happened since the Chalcolithic, i.e. when social stratification, inequality and private aristocratic property became common. Obviously the upper classes (or castes) had no problems accessing high quality foods, including meat, but the masses probably had growing problems in this aspect as the land and cattle became more and more concentrated in few hands. 

Even where a wide class of free peasants existed, as was probably the case in much of Atlantic Europe, these were surely often not well-off enough to afford dairy cows. Instead goats would have been available for almost everybody, even the poorest of farmers. And very likely they were the only steady supply of proteins and fat, mostly via milk.

Plausibly this need of extra nutrients of animal origin was more intense in the Atlantic areas of Europe because cereals do not perform so well in the prevalent humid conditions. Also before the medieval development of the heavy plough, the deep Atlantic soils were not at all as productive as they are now (and that's why NW Europe only got its economic prominence in the last millennium, being before a peripheral area to the much more productive Mediterranean climate). 

But climatic and agricultural issues aside, I strongly suspect that the main driver of LP positive selection, were goats, because these and their dairy produce were almost certainly available for almost everyone and, in the Metal Ages, the vast majority of people were farmers, often rather poor peasants who had to rely on their goats for survival, very especially in the bad times.

I really do not see any other explanation that fits the data.

PS- This social inequality & goats argument makes sense assuming that the positive selection theory is correct. However before I fully embrace it, I would need a half-decent sample of aDNA sequences from the Atlantic areas of Europe, notably Britain & Ireland, the Basque Country & SW France and mainland Scandinavia, where the T allele peaks. I say because what we find in some Chalcolithic sites, notably in the Basque Country, rather strongly suggests that there was already a TT population somewhere and we have not yet found it. So maybe some of the premises of the positive selection theory are not as sound as I said above - but we do not know yet.


  1. There's two arguments really
    1) Does Funnelbeaker represent a transition from forager -> shifting agriculture forager/herder which led to a population expansion among the native foragers along the Atlantic coast and Northwest

    2) Is that responsible/connected to the strong selection for LP in that region.

    The two arguments are separate but i have a bad habit of arguing them together.

    I agree the dietary mechanism is likely to be bowls of milk + grown/foraged grains, berries, nuts, acorns etc i.e. a porridge/muesli type staple in the diet but did Funnelbeaker use cow's milk or goat's?

    Regardless of LP if Funnelbeaker was a transition from forager/fisher -> shifting agriculture forager/herder then there ought to be a population expansion and bottleneck in I1 around 6000 BP.

    1. Funnelbeaker can only explain changes in Northern Europe senso stricto (Denmark and surroundings), having no influence further West (also further South it includes many different cultures of various origins: Danubian, early Kurgan, etc. - so I'm wary to identify it as a simple unity) Do you know if the ancient Megalithic farmers from SW Sweden (who were likely connected to Danish TRBK) had the T allele?

      "... but did Funnelbeaker use cow's milk or goat's?"

      Plausibly both. I see no reason why they would not have goats, being so versatile and productive, able even to forage in the forests. All cultures of Europe had all or nearly all kinds of food resources, my hypothesis goes rather in the line of how much people did each kind of resource feed, especially considering growing inequality and resource accumulation by the elite few?

      As the Mexican song says: "las penas y las vaquitas se van por la misma senda, las penas son de nosotros, las vaquitas son ajenas" ("the pains and the cows leave by the same road, the pains belong to us, the cows to others"). Cattle has "always" been a measure of wealth in some languages both terms get confused (for example in Basque abere=cattle or livestock, aberastasun=wealth, lit. "cattledom" or "cattleness", aberats=rich), so I'm guessing that only goats among milk-producing animals could feed the masses in order to cause such a massive sweep, especially if this happened when class stratification was already established, as it seems.

      I'm anyhow split between the two logical paths: (partial) replacement (assuming a very ancient Atlantic TT population) or positive selection, which almost necessarily must be because of a "cheap" dairy animal, not a costly one, or maybe both. Ancient DNA is not sufficiently informative yet on this issue.

    2. The song is not Mexican, it is from Argentina and its author is Atahualpa Yupanqui.
      I love this blog, I do not understand everything but it is really good for a dilettante in human evolution as me. And besides that it improves my English skills!

    3. Incidentally the song is Paisajes de Catamarca and it comes from Argentina, not Mexico.
      I love this blog, it is very good even for a mere dilettante in human evolutions as me. And,l moreover, it improves my English skills. Bye

    4. Oops, my bad. Now I know, thanks. It's a beautiful classical song in any case.

  2. "Funnelbeaker can only explain changes in Northern Europe senso stricto"

    Yes. I'm focused on TRB for proving a model - the possibility of a forager -> forager/herder transition at the edge of a full farmer region. If correct, then similar events might also be possible even if with a different mechanism.

    "I see no reason why they would not have goats"

    Yes, i got a bit fixated on the cows. The LP part doesn't have to be connected to the forager->herder transition part at all.

  3. "Because for all we know from the Middle Ages, a period very similar in many aspects to the Metal Ages, it were goats and not cows the main providers of milk."

    But that is not true for all of Europe. Much of Northern Europe's and North-Central Europe's medieval agriculture was clearly cattle-centric, with few goats, and a lot more sheep than goats. There is an ancient cattle path going from (now) Denmark south and west, where surplus cattle were driven to Frisia for fattening on the endless coastal grasslands. Also, the Angles, Saxon, and Frisians taking advantage of the Roman withdrawal on the Isles brought with them cattle, whereas the locals withdrew with sheep to the mountains or became your second class citizen but left little genetic legacy in the lowlands. And those things didn't change much during the second half of medieval times: even under the feudal/ manorialism system, locally, cattle were the most important and fully taken care of by the peasants (on the serf or completely independent on free peasant land).

    So, any explanation of the extremely wide spread of LP needs to take these huge geographical differences into consideration.

    "It is crucial to understand that only if milk was a key survival staple, LP would have become fixated. Otherwise people would have preferred alternative foods and survived in similar shape, so positive selection would never have happened at this locus"

    But there is one vector that does not require milk to be a main food staple, just a common one: that is, immunization/ protection from diseases due to antigens, bacteria, and viruses in the cows milk. The metal ages saw an unprecedented growth in population density, which, as documented, went parallel with numerous set-backs caused by waves of epidemics. Given the not-uncommon decimation of the population to <30% of previous levels, it would not have taken many of those for a selective sweep.

    Vitamin D, Calcium, and folic acid were likely still important contributing factors: a double-whammy of reproductive success for those drinking milk more than occasionally.

    1. Milk or meat, Eurologist. I'm founding my claim on pan-European Middle Ages history (Han Dhont's book), based especially on the Frankish Empire, which included France, Germany, the Benelux, Switzerland, Austria and parts of Italy.

      I know that sheep became very important as source of wool for the textile industry but this one took some time to fully develop (it's mostly a late Middle Ages to Early Modern thing). Simultaneously beef may have become growingly important as source of commercial meat (your reference talks of "fattening" not dairying) but probably not milk in most cases.

      Anyhow most of what you say refers to very specific medieval episodes, what does not directly weight on what happened in the Chalcolithic and Metal Ages, which is the real period of the genetic sweep (if it happened at all).

      I also have to strongly disagree on your claim of that ancient Britons "left little genetic legacy in the lowlands". That makes zero sense to me on light of what I know from British genetics, even from the one-sided Y-DNA aspect only. Only in Orkney probably the Scandinavians (Norwegians specifically) made a major impact. The mild relation of East England to Denmark-Saxony elsewhere can largely be attributed to Epipaleolithic and/or Neolithic flows across the North Sea, although of course these may have been slightly enhanced by Germanic migrations.

      But, well, this is a different debate, because it is obvious that Welsh, Scots and Irish also have high LP scores and they have no meaningful Germanic ancestry. Same with Basques or Bretons.

      I think you guys focus too much on Nordic Europe and take the part for the whole.

      "But there is one vector that does not require milk to be a main food staple, just a common one: that is, immunization/ protection from diseases due to antigens, bacteria, and viruses in the cows milk".

      This is a new concept to me. But anyhow: why cow milk and not goat milk? Why this obsession with cows?

      "Vitamin D, Calcium, and folic acid were likely still important contributing factors".

      Naturally vitamin D is nearly not present in milk. Only fish are a decent source of nutritional vitamin D. Modern enriched milk was not available before the 20th century: it is an artificial product. Someone should tell those geneticists that milk does not come with relevant amounts of vitamin D (unless artificially enriched).

      Folic acid can be obtained from a variety of food sources, notably veggies. Calcium is the only element that milk really helps with of all you mention.

      However milk is a good source of protein and fat, especially for those without regular access to meat or fish. And those without regular access to meat were indeed the poor peasant masses and for them, most likely cow milk was not an option either, instead goat milk would have been readily available in almost all circumstances.

      That's the bulk of my argument: poverty means no meat but still probable access to "cheap" (but nutritious) goat milk.

    2. @eurologist

      "But there is one vector that does not require milk to be a main food staple"

      Another might be if there was a dramatic differential effect on children.

    3. @Grey: all children, at least the very young ones can digest milk (logically). The issue of LP is essentially about grown-ups, although of course it may affect young children via their mothers (when pregnant or lactating) if they do not get enough nutrients.

  4. "This is a new concept to me. But anyhow: why cow milk and not goat milk? Why this obsession with cows?"

    Because cattle were omnipresent throughout Europe, whereas goats and sheep were local exceptions.

    See, e.g., the well-known protection cowpox deliver against smallpox.

    1. I don't think that was the case, Eurologist. Goats and sheep have always been there, I mean: since the arrival of Neolithic. And, considering the socio-economic changes associated to the Chalcolithic (class society), I suspect that goats were available for all, while cows only for some.

      I'm not familiar with the issue of cowpox nor if there is a "goatpox" which could do the same. Excuse my ignorance.

    2. My issue is that the rich, those with easy access to cow milk, also had access to meat and every kind of food. Also they were a minority. In order to explain such a broad genetic sweep, we must consider the impoverished masses of peasants.

      I guess that with cowpox, you are talking of the known case of Pasteur's milkmaid, right? Not sure if and how well applies.


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).