The Amish heart-protecting triglyceride-busting null mutation

Toni Pollin and colleagues (2008) report one of the simplest medical research studies you’ll ever see:

Apolipoprotein C-III (apoC-III) inhibits triglyceride hydrolysis and has been implicated in coronary artery disease. Through a genome-wide association study, we have found that about 5% of the Lancaster Amish are heterozygous carriers of a null mutation (R19X) in the gene encoding apoC-III (APOC3) and, as a result, express half the amount of apoC-III present in noncarriers. Mutation carriers compared with noncarriers had lower fasting and postprandial serum triglycerides, higher levels of HDL-cholesterol and lower levels of LDL-cholesterol. Subclinical atherosclerosis, as measured by coronary artery calcification, was less common in carriers than noncarriers, which suggests that lifelong deficiency of apoC-III has a cardioprotective effect.

Gina Kolata covers the story in the NY Times:

For the sake of heart disease research, 809 members of the Old Order Amish community agreed to go to a clinic in Lancaster, Pa., near their homes, and drink a rich milkshake that was made mostly of heavy cream. Over the next six hours, a group of investigators took samples of their blood, determining how much fat was churning through their bloodstreams.
Most of the study participants responded as expected their levels of triglycerides, a common form of fat in the blood, rose steadily for three to four hours and then declined. But about 5 percent had an extraordinary reaction: their triglyceride levels started out low and hardly budged.

I’m generally interested in novel protective mutations, and this is clearly one – and far from the only one. Its current frequency is 5 percent in the Old Order Amish. Neither the article nor the paper report on its frequency in the general population; although there is the intimation that it is rare. The Amish individuals carrying the mutation all share a common haplotype, apparently (based on pedigree and LD) from a single 18th-century founder.

It remains an open question whether homozygotes for the null allele are better or worse off than normal APOC3 homozygotes. With a frequency of 5%, the allele is rare enough that homozygotes are as few as one in 400 people. They were not included in the present study. I can’t find any indication that homozygote nulls for APOC3 are a known Mendelian disorder.

I wonder to what extent the allele frequency in the Amish is due to selection.

The Amish have high frequencies of certain otherwise rare mutations. This is one of the textbook examples of founder effects – extreme genetic drift due to sampling a small number of founders from a much larger population. Today’s Old Order Amish in the United States trace most of their ancestry to an initial population of approximately 200 people in the eighteenth century. That means that any of the alleles carried by those 200 people, even if it was vanishingly rare in the European population, has a good chance of being half a percent or higher in today’s Amish.

But founder effect is only part of the story – there is also subsequent population growth. Those initial 200 people have more than 200,000 descendants today within the Old Order Amish. This number doesn’t count descendants who may belong to other sects that splintered during the nineteenth-century (like the Mennonites [see update below]), or descendants of people who left the church. These values suggest that the Amish population has increased by some 2.3% annually during the last 300 years; it’s current rate of growth is estimated at 4%.

This is very rapid population growth on an evolutionary time scale, equalling roughly 46% per generation. With this kind of population growth, strongly deleterious alleles may come to occur in a large number of individuals, even as they decline in frequency in the population. The susceptible population grows faster than selection can remove alleles. Hence, we find a number of rare genetic disorders within the Old Order Amish as a consequence not only of founder effect but also subsequent population growth.

The APOC3 mutation in this study was evidently not deleterious. Its current frequency of 5% suggests it may have been advantageous.

It’s not too hard to hypothesize why a mutation that decreases the risk of heart disease might have conferred a benefit in an agrarian religious sect over the last 300 years. To the extent that heart disease affects men in their 30’s and older, these are still active reproductive years for men who may have family sizes of eight children or more. Further, this is a time when men may come into property from their aging parents, may become leaders of new settlements, or may begin to affect the marriages of their children – a time when young people formally join the church. Being alive would seem like a significant fitness advantage for men in this society. Or perhaps other effects of the gene determined its success.

The question is just how strong such an effect might be. If the mutation began with a single copy in a population of 200 founders, its initial frequency would be 0.5 percent, or 0.005. Its present frequency in the Amish is ten times that, or 0.05. If we assume that 15 generations have passed, that growth would be consistent with a fitness advantage of around 15 percent for carriers of the null mutation. In other words, the Amish population grew around 46% per generation over the last 300 years; this mutation grew around 60% per generation.

That kind of differential increase is unlikely to have been driven by genetic drift. Considering the rarity of the mutation in the non-Amish population today, it is unlikely to have been carried by more than a single founder, although we can’t exclude the hypothesis that some number of founders were relatives who carried it. That hypothesis is the most likely way for an otherwise rare mutation to hit 5% by founder effect alone. Later, after the Amish population numbered more than a thousand or so, strong differential growth of a rare mutation by chance alone would be impossible. Still, we might imagine that in the initial few generations, one or two founders might have had a predominant effect on the subsequent Amish gene pool. We would need to suppose that the genes of such fecund founders now account for more than 10% of the present Amish gene pool. That’s a testable hypothesis. Selection is simpler – mainly because its effect can be spread across many more generations.

The interesting thing about selection in the Amish is that their population growth greatly affects the fixation rate of new advantageous mutations. In a constant-sized population, the fixation probability of a new advantageous mutation is roughly twice the heterozygote fitness advantage, denoted as 2s. But in a growing population, the fixation probability is 2(s + r) – when s and r are both small. If we assume a growth rate of 46% and a heterozygote fitness advantage of 15% for this null allele, it should be obvious that we’ve entered the territory where our small-value approximation no longer holds. New adaptive mutations are unlikely to exit the Amish population by genetic drift.

The subject of positive selection in founder populations is under-explored, from a theoretical perspective. Especially considering the very rapid growth of some human founder populations – measured against the generational time scale – there is a good chance that we’ll find many new adaptive mutations in such populations.

UPDATE (2008-12-15): A reader writes:

It is a common mistake to think that the Mennonites as a group broke off from the Amish. It is actually the other way around with the split occurring in Europe before both groups came to the Americas.

He kindly provided a couple of sites with more information (here and here). I appreciate the correction!


Pollin TI and 13 others. 2008. A null mutation in human APOC3 confers a favorable plasma lipid profile and apparent cardioprotection. Science 322:1702-1705. doi:10.1126/science.1161524