Social barriers to Indian public health & non-Aryan “invasions”

By Razib Khan | February 5, 2013 1:16 pm

A reader points me to a talk given by David Reich at the Center for Human Genetic Research 2013 Retreat. One of the issues Reich brought up is old, but perhaps worth reemphasizing: due to endogamy many South Asians carry a higher load of recessive ailments. This is not due to recent inbreeding (which is barred by custom in many South Asian groups, which enforce kin-level exogamy), but long term genetic isolation. Over time even a moderate sized population can be affected by drift. This was one of the major points in the 2009 paper Reconstructing Indian History, but not one particularly emphasized in the press follow up. A major implication is that a relatively simple public health measure for South Asians would be to marry outside of their jati. The social or genetic distance need not be great. But one generation of outbreeding should “mask” many of the deleterious alleles. If this model is correct one should be able to track decreases in morbidity within the American South Asian population, where there are many inter-caste and inter-regional marriages (yes, this is between people of putative high status, but this doesn’t matter).

Second, Reich agrees that the ANI (West Eurasian, “Ancestral North India”) admixture into the India population exhibits at least two admixture events. There were hints of this in the original 2009 paper, and looking more closely at the South Asian data others have suggested this more explicitly. This seems the best explanation for why non-Brahmin upper castes in South India do exhibit distance on the ANI-ASI cline from lower castes, but without clear connection to many ancestral components with a “northern” affinity present at non-trivial levels in Indo-European speaking groups and South Indian Brahmins (or those groups which have admixed with Brahmins, such as Nairs).

The hypothesis I prefer is that there was an initial wave of West Asian agriculturalists who arrived in the Indian subcontinent <10,000 years B.P., and admixed with the ASI ("Ancestral South Indian") substrate. Then, there was at least one further substantial demographic wave of West Eurasians, probably bringing the Indo-European languages. This population had more northern affinities (though not exclusively; the Basque vs. non-Basque difference in European seems to be a West Asian element), which explains the subsidiary minor explicitly European-like element found in many upper caste populations, and to a lesser extent Indo-European speaking South Asians generally. Finally, I do suspect that some groups in the Northwest, such as Jatts, were shaped by later migrations.


Comments (11)

  1. ohwilleke

    * Your hypothesis seems very solid. I would sometimes think in terms of (1) at least some underlying N-S cline within ASI prior to other waves, upon which were added to form ANI, at least, (2) an early Neolithic Indus Valley Civiilzation (with a big demic impact) ca. 9000 years BP, and (3) a later Indo-Aryan layer (with a moderate demic impact in North India) ca. 4000-3500 years BP are imposed before ANI expands in the rest of South Asia – with (4) a possible less heavily demic component associated with Munda admixture in ANI as well (particularly in light of recently released evidence of significant rice consumption all of the way back to the beginning of the Mature Harappan phase ca. 4600-4500 years BP).
    * One would expect (1) to be completely gender balanced, (2) to be mostly gender balanced, (3) to be heavily male biased, and (4) to be male biased but perhaps not so much as (3) but more than (2).
    * If one wants to look at the demographics of the South Asian Paleolithic, one would want to confine oneself pretty much to the ASI layer and older Y-DNA and mtDNA haplogroups with maximal basal diversity in South Asia (e.g. Y-DNA haplogroup H) and remove the clutter of the other layers of the palmpiset first. For example, my inclination would be to attribute Y-DNA haplogroups L and R2 to the Neolithic Indus Valley Civilization wave, R1a to the Indo-Aryan layer, and O to the Austro-Asiatic layer (apart from more recently arriving Tibeto-Burman populations that are fairly distinct from other South Asians genetically). (The most exciting and mysterious of all in South Asia is the distribution of Y-DNA haplogroup T for which there isn’t a clear answer yet, in part, because of insufficient subtype detail in the available data – details studies of Y-DNA haplogroup T are available for Africa and Europe and some of West Asia, but not yet for India, as far as I know).
    * Of course, just because we don’t have good historical candidates for significant waves of Holocene era admixture in ASI, this doesn’t necessarily mean that there aren’t any, although the association of ASI with Andamanese as the closest modern reference population does seem to suggest this possibility and Reich doesn’t seem to mention multiple wave theories for ASI in the Holocene at least.
    * Any demic Munda influence on ANI may be too close in time to the Indo-Aryan impact to distinguish statistically in autosomal genetics as separate waves of admixture temporally (given the limited resolution of the already imprecise genetic admixture dating methods used) and might be possible to disaggregate at all within the overall ANI mix only through good reference populations (ideally from ancient DNA at least for the Indo-European part) to stand in for Bronze Age contributions from IE and Austroasiatic peoples respectively.
    * It is also unclear to me if a model of Harappan-Indo-Aryan synthesis into ANI followed by expansion elsewhere in that sequence is the predominant story (i.e. two back to back arrows), or if there is a meaningful amount of unadmixture contributions separately to populations with high ASI proportions (i.e. two back to back arrows, plus a IVC to South India arrow and an Indo-Aryan to South India arrow, with the latter two signals muddying each other a bit into a secondary Harappan-Indo-Aryan admixture phenomena within South India in addition to the “pre-mixed” ANi population migration to South India.
    * Of course, there are also “modern” (i.e. post-Indo-Aryan historic era) layers in some places (most notably, a historic era layer of African immigration that shows up in genetic studies in one ethnic group in one region of South Asia – the name escapes me). These should disproportionately impact ANI relative to ASI as well, for the most part.
    * The very great cultural impact of the British colonial era, but very minimal demic impact, right at the dawn of the modern era (i.e. the 17th century) is also a good counterexample to the general proposition that all major cultural shifts are demic.
    * I don’t have a good sense of how much demic impact the current era invasions from West Asian (first Persian and then Islamic) had on South Asia that could be reflected in ANI – other than that it is probably non-zero and probably discernable in at least some substantial swaths of Musliim South Asia (I don’t know the history well enough to know which parts within Pakistan and Bangladesh and perhaps parts of Northern India), and that it is probably a fairly thin demic contribution (less than 5%?)
    *[edited to use asterix to mark paragraph breaks]

    • razibkhan

      agree with lots. one issue that i think i would demur: if you look at ADMIXTURE it looks like munda are like dravidian ppl + east asian. the residual “european” element found in IE/upper caste populations is lacking in them.

  2. Dmitry Pruss

    So Razib, does this degree of endogamy mean that for genetic association studies in S Asians, one should always draw jati-matched controls? (it’s hard to imagine how that might ne possible)
    … Even for South Asian *Americans* (since most would be 1st gen and/or children of marriages from the old country)? Are there ancestries which are less endogamous (does the traditional endogamy affect Muslims at the same degree? Are there easily identifiable jatis of larger population sizes, which would be less likely to have AF%s affected by drift? If I get it right, Indian Americans tend to be, disproportionally, from Gujarat?

    • razibkhan

      So Razib, does this degree of endogamy mean that for genetic association studies in S Asians, one should always draw jati-matched controls?

      probably too extreme. there are south asian specific markers out there that have been detected. in general the GWAS which have picked up at things have surprised me in being portable across populations. but that’s not the low hanging fruit at this point. so yeah, in the future you need to control more finely for genetic background. you don’t necessarily need to limit to one jati. but i’m just saying here that this might be a quick & dirty social policy to reduce morbidity (just like getting mid eastern & south asian muslims to stop engaging in inter-generational cousin marriage).

      • Sereno Barr-Kumar

        I think a good contrast would be to asses recessive ailments in Sri Lankan populations where cross cousin marraige has been the norm, but has had genetic admixture by being in the middle of Indian Ocean trade for at least 2000 years (Sinbad stories are based on Sri Lanka). Add on to that another 500 years of European colonial occupation, Javan populations seeking refuge in Sri Lanka and brought in as soldiers by the Dutch it is one big mix.

        I would expect this of Kerala and Goan poulations too, but maybe to a lesser extent

  3. Matt

    Then, there was at least one further substantial demographic wave of West Eurasians, probably bringing the Indo-European languages. This population had more northern affinities (though not exclusively; the Basque vs. non-Basque difference in European seems to be a West Asian element)

    The West Asian components found by Dienekes’ Dodecad project as the extra-South Asian West Eurasian modal components contributing to Indian populations, tend to be most similar to (if still quite distant to, compared to between country fsts in West Eurasia / Europe), the North European / Atlantic-Baltic modal components he finds (most prominent in Lithuanians).

    Of course these components probably (?) don’t represent real populations, perhaps they do represent some similarity in population dynamics and descent.

  4. stevesailer

    There doesn’t seem to be that much regression toward the mean among second generation Indian-Americans relative to their often highly selected parents who came to the U.S. for academic purposes. Better health and nutrition is no doubt a reason, but this suggests that less inbreeding could be helping too.

  5. stevesailer

    About 30 years ago, Daniel Seligman speculated that Italian-Americans born after WWII were substantially taller because they were more outbred, with their parents marrying not just within their transplanted village. (Seligman was Jewish-Irish, by the way.)

  6. Its amazing (to me, anyway) how the The Cambridge history of India *1922* pretty much says these same things regarding population movements etc. Respect!

  7. S. V.

    I have already tried debating this topic, with different people from my “caste” (although I don’t follow the same). The biggest stumbling block they claim, are dietary customs, and a perceived sense of “high culture”, that other groups (including non-Indians) apparently lack. These days, it seems that the diaspora is holding on to such values more tightly than those back home in the subcontinent.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar