Genetic evidence for natural selection in humans in the contemporary USA | Hacker News Reader

Genetic evidence for natural selection in humans in the contemporary USA

biorxiv.org

80 pointsgwern72 comments

I'm sorry was this an April Fool joke?

By definition a 'phenotype' is the expressed traits of a genotype. As far as I know there is no known gene or set of genes which express the 'phenotype' of 'Educational Attainment'.

I looked at the paper and the author provides no methods to replicate his study. For example, there is no discussion as to the SNP markers he was tracking to associate the 'phenotype' of EA.

What the author is implying here is that there is a gene for 'educational attainment' that is becoming more common in the population in general due to selection pressure.

I'm sorry but this is total hogwash. Anyone that correlates educational attainment to genetics is following a 'chopstick' gene.

Imagine that you correlate the ability to use chopsticks to genetic markers. What you'll find is that people with black hair and brown eyes have a much greater ability to use chopsticks then people with blond hair and blue eyes. You could even produce some pretty convincing graphs about this. However, as we all know the usage of chopsticks is cultural not genetic. There is nothing to prevent blond haired, blue eye people from using chopsticks genetically.

In fact you can correlate any two things that have no relationship to each other and find a correlation. For instance, the sun coming up in the morning and traffic congestion. They correlate, but have no causal relationship. i.e the sun does not cause traffic.

When the sun goes up, it's easier to do your work or job. So people get to work and cause traffic jams.

Correct me if I am wrong, but I though the chopstick gene example has more to do with population structure confounding than with the heritability of the trait?

For example one could make a similar example with malaria resistance and skin color.

>My results imply that natural selection has been slowly favoring lower EA in both females and males

lower educational attainment, not higher

see: GWAS of 126,559 Individuals Identifies Genetic Variants Associated with Educational Attainment (one of the authors is Beauchamp) http://science.sciencemag.org/content/340/6139/1467

>Three independent single-nucleotide polymorphisms (SNPs) are genome-wide significant (rs9320913, rs11584700, rs4851266), and all three replicate. Estimated effects sizes are small (coefficient of determination R2 ≈ 0.02%), approximately 1 month of schooling per allele. A linear polygenic score from all measured SNPs accounts for ≈2% of the variance in both educational attainment and cognitive function

So ... less educated people are getting more kids? (or the inverse: more educated people are getting less kids)

Yeah, this should be the case just based on the logistics involved. Higher education typically takes place during the years when a woman is able to have children. And to go a step further: generally people obtain higher education for a purpose: to further a career. So, then, after schooling is over, there will be a career to make, further reducing the number of years available for a woman to have children.

So this should be the case based on practical matters alone.

Sorry but these types of studies strongly smack of eugenics. This reminds of Stephen Jay Gould's book "The mismeasure of a man".

If the SNPs are inherited (and possibly racially so) and a lower educational attainment should be associated with a lack of native intelligence, can we then conclude from this study that certain groups have a lower IQ and therefore are genetically pre-disposed to have less educational attainment? Can we test for these people in the general population and prevent them from breeding?

Are these SNPs associated with criminal propensities or mental issues and therefore correlate with a lack of educational attainment? We could then identify them and drill holes in their heads: https://en.wikipedia.org/wiki/Lobotomy

Science is littered with shoddy peer accepted studies correlating IQ with genetics. This seems to be the 21st century version.

> What the author is implying here is that there is a gene for 'educational attainment' that is becoming more common in the population in general due to selection pressure.

No, actually, he finds that there was negative selection pressure on educational attainment.

Edit: You beat me, nabla9 ;)

The sun clearly does have a causal relationship with traffic, as traffic is exclusively the product of humans who are not sleeping, and is powered by the decayed products of photosynthesis. You may wish to choose a better example.

Also the study is apparently proposing that natural selection is favoring low educational achievement. This, not incidentally, is the fundamental premise of the movie Idiocracy, and certainly has a plausible logic to it, though I am wholly unqualified to judge the methodology of this study.

Robert Plomin has proved a genetic influence for UK A-Level (exam) results.

He uses the statistical method of comparing identical and fraternal twins to demonstrate that genetics accounts for twice the variation in exam achievement than different schools.

Plomin is keen to stress that individual genes effects are very very tiny, this is a 'whole genome' comparison.

Interviewed for BBC radio's Life Scientific : http://www.bbc.co.uk/programmes/b06j1qts

"In 1995 Plomin launched the biggest longitudinal twin study in the UK, the TED study of ten thousand pairs of twins which continues to this day. In this study and in his other work, he's shown consistently that genetic influences on intelligence are highly significant, much more so than what school you go to, your teachers or home environment. If only the genetic differences between children were fully acknowledged, he believes education could be transformed and parents might stop giving themselves such a hard time."

It's just funny that it is still called "natural selection" even when the environment that selects individuals is created by nothing but ourselves. I think it'd be much more accurate to call it "social selection" because for humans the environment is not nature, but society.

With that in your mind, those of you who are clever might wonder: what is the fundamental difference between "social selection" and eugenics, which many consider morally wrong? My answer is: they are essentially the same thing. It's because if we define eugenics as society deciding who to reproduce and who not to, that's exactly what is happening now.

I will not argue the common usage meanings that have come about for 'natural' and 'artificial' for things like plastic flowers or street lamps, but I look at 'ourselves', or humankind as part of nature, and not supernatural, or apart from it. I am not sure when this distinction or usage arose, but 'artificial' or using 'societal' here seems self-hating, and I don't think the term 'natural selection' is archaic or exclusive in this sense.

The same reason I find cell towers made to look like trees an eyesore, and I much prefer the 'natural' shape a truss tower takes with its wide supportive base, diagonals and ever-decreasing thickness to the spire. I personally find these more aesthetically appealing. Light from 'artificial' light bulbs is made of photos regardless of color temperature and the time of day it is turned on. I don't like plastic flowers, although plastic is also found in nature, but just not as structured, homogenized, or in the quantities we produce it.

I agree with you on the 'social selection' and eugenics bit. My Mom had bought me a hardcover 'Eugenics' textbook from an old bookshop originally published in England in the 20s. Scary content given what occurred 20 years later, but it was humorous to me when reading it in 1976, especially the chapter on 'The Honeymoon' and how to have the proper one with attendant drawings! I was twelve at the time.

There is a fundamental difference. Natural Selection works on an individual basis, everyone tries to maximize the amount of his or her DNA in the world. In eugenics (and to a lesser degree in social selection) someone else (or the group) decides that his or her DNA should not be inheritet. Basically capitalism vs. communism.

Humans are only "fundamentally different" in that we have alpha castes as well as alpha individuals.

It's not unusual for animal alphas to monopolise/deny mating opportunities. Eugenics is the same evolutionary invention taken up a level.

Ironically, this means proponents are acting just like animals.

In sexual selection there is also someone else (i.e. the group of females in a peacock population) that decide whose DNA should not be inherited. And sexual selection is considered a part of natural selection. So your argument does not hold.

Wikipedia: Sexual selection is a mode of natural selection where members of one biological sex choose mates of the other sex to mate with

https://en.wikipedia.org/wiki/Sexual_selection

I could even argue the same thing for natural selection itself. When a predator kills an animal, he decides that the DNA of that animal should not be spread further in the future.

Well, here there's selection against characteristics that eugenics typically favored.

CTRL-F "Idiocracy"

But seriously, can it really be said that this is related to natural selection rather than some other pressure? I guess it kind of depends on the definition, as one could argue that any result within a given environment is causal to that given environment. However, while I do anecdotally see that more highly educated people have less children (and also the opposite, which the author writes of), I still think that people generally try to select for the smartest and most capable people within their social sphere. In other words, I think people are still selecti mates which are smarter, but due to socioeconomic factors unrelated to breeding, the less educated end up having more children.

Another anecedotal (and probably wrong) idea related to this is the explanation of prevalence of ADHD in US relative to EU, for example.

The genetic selection idea goes something like this -- individuals who ended up migrating here, self selected for being bolder, ready to take on more risk, being less content staying in one place. On one side that is a postive trait, but on the other side manifests as ADHD.

There is a study from 2003 showing that it is not true and that it is due to how diagnosis is handled:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1525089/

Other places claim a strong genetic factor is there:

"ADHD often runs in families. Researchers have found that much of the risk of having ADHD has to do with genes"

http://www.nimh.nih.gov/health/publications/attention-defici...

Just to point out, there is no conflict between the claims (1) "ADHD is diagnosed more often in the US than in the EU because the US has looser diagnostic standards"; and (2) "ADHD is genetically determined".

Anecdotally, I have heard French people say that it is practically impossible to get a adult psychiatric diagnosis of ADD in France, regardless of how clear and debilitating your symptoms are. It may be that the apparent low rates in the EU are actually systemic diagnostic failures.

(Similarly, I have read that in [I think] Japan, it is actually illegal to begin a course of ADD medication after 18--if you were diagnosed as a child, you're fine, but if your parents never took you to a psych, you're screwed forever. Even in the US I've had a general-practice doctor tell me that they thought ADD just went away after a certain age. There's a lot of denial about adult ADD.)

Interesting. Wonder if it is cultural, and if there is any element of antagonism against US practices -- i.e. "we know US is pretty liberal with this diagnosis, we think it is rather wrong, so we'll just go the other way and deliberately shy away from diagnosing it as much as possible".

Even as a teen in France doctors basically brushed off the entire possibility of me having ADHD and just prescribed me a thing for drowsiness... despite having been treated for ADHD all through middle school with positive results.

I have yet to succeed getting medication that works for me in Japan (crappy laws mean no Adderall for me), but there are adult ADHD support groups that list doctors willing to work with adults despite no prehistory.

ADHD has been in my field of practice for quite a long time. It's true that many countries have greater restrictions on treatment, and medications available in the US are not approved elsewhere. Cultural biases against recognizing the condition are common, but is greater in some countries than others.

In this respect the US has been way ahead of most other places in terms of recognizing ADHD in adults and treating the condition. Even in the US, it still remains highly under-diagnosed and -treated, as its estimated only about 20% of cases ever come to clinical attention, and only a small proportion of those cases get proper and persistent treatment.

Over the last decade or so the stigma against diagnosis and treatment of adult ADHD has diminished, but remains a huge barrier for care of people with the difficulty. It's not a minor condition, in fact research shows ADHD has poor outcomes in all domains of functioning. IOW ADHD is frequently associated with lifetime disability which is usually reduced with appropriate treatment.

Been a couple of years since I've closely tracked the literature on some of these questions. Regarding diagnosis and prevalence of ADHD in children and adults, the population rates are similar internationally (Germany, Brazil, US) provided that the same diagnostic criteria are used. The new standard criteria given in DSM5 and ICD10 may reduce regional differences in the the criteria that are applied.

ADHD is one of the most studied conditions in terms of genetic predisposition. Across many studies the overall heritability is 0.75-0.80, denoting that 75% of the variability between affected and non-affected individuals is accounted for by genetic factors. IIRC the genetic risk of a child having ADHD if one parent has it is about 55-60%.

The migration theories don't hold up well to closer scrutiny, for one thing only some ADHD sufferers are prototypical restless risk-takers. At least an equal proportion are quite risk-avoidant, and would be less inclined to pick up and move.

Nonetheless ADHD may have evolutionary persistence because it confers reproductive advantage, though such advantages are hard to pin down. Perhaps energetic risky behavior is attractive or sexual impulsivity predisposes to having more children, but that's just speculation.

You would think that any bold/risky phenotypes would balance itself out: - On one side, maybe ADHD-types reproduce more or are better disposed to make sure their offsprings get to a reproductive stage (e.g. wealthy, due to a bold move which payed out). - On the other side, their are also more likely to have taken a risky move, which end up costing them a lot, including their life, so no offspring or no additional support to support kin-selection.

It's true that ADHD adults have higher mortality from auto accidents, more health problems, higher divorce rates, greater unemployment, lower educational achievement than non-ADHD peers.

Despite all their disadvantages, rate of ADHD occurrence in the population has been stable as far as can be determined. The inference is reproduction is not impeded, and may exceed the population mean in light of anticipated losses as you describe. I haven't specifically looked into this, not recently anyway, but data may well be out there.

Anecdotally it's a common pattern to see ADHD adults having had numerous partnerships which are not sustained beyond a few years. Not surprisingly, an ADHD parent can wind up with children from these several partners as a result, leading to the hypothesis that instances of high birthrate offset the fraction of ADHD adults with low birthrate.

Obviously, even a small positive offset in birthrate of all ADHD vs. controls would be sufficient to keep incidence of ADHD constant.

In modern societies, kids tend to get little quality time with their parents. In addition, they tend not to grow up in a relaxed carefree environment.

I see kids today with mobile phones, playing addictive games and being unable to focus on a single thing for more than a few moments.

It is tough to make multiyear controlled experiments on such things. I would discount heredity for now.

Economists dabbling in other fields have a uniformly appalling track record, but who knows, maybe this one is the exception.

No this isn't. He should have spent 1.5 more months thinking about his experimental design. He doesn't even try to make the distinction between causality and coincidence. We could achieve the same conclusions by talking about socio-economic changes rather than genotype-to-phenotype.

For a guy who was has is name in the big GWAS published in Science, he obviously didn't understand what his co-authors did or wrote. Sorry for the rant, I can't stand people who are getting a free-ride on a paper, and he just proved that is one of them.

Isn't it fairly known that less educated people have more children?

Yes. Education and economic prosperity have reduced population growth rates in many places. Too much, perhaps, in some places.

I don't understand. Why has educational attainment gone up and age at menarche gone down if there is selection pressure in the other direction?

Because these have non-genetic factors working in the opposite direction.

How is educational attainment a genetic trait? Intelligence is partially hereditary, but isn't educational attainment a behavior?

Intelligence is as much a behavior as education is. You observe it by what people do and how they perform. There is no physical or neurological correlate of intelligence like height or brain size or white matter integrity which measures intelligence as well as observing performance does.

As for how education can be genetically mediated, there's many ways. Through intelligence (high genetic correlation/overlap), but also through personality traits, ADHD-proneness, and other mental and physical characteristics which are undoubtedly under genetic control.

Still doesn't make much sense to use a secondary trait like educational attainment (58% heritability at highest) over a primary one like IQ (>80% heritability).

Most of these phenotypes strike me as pretty questionable. Maybe I missed something, but how well does BMI measured at age 45-50 correlate with BMI at reproductive age?

Wouldn't a far simpler approach be to measure allelic frequency at regions near GWAS QTLs in multiple generations?

> Still doesn't make much sense to use a secondary trait like educational attainment (58% heritability at highest) over a primary one like IQ (>80% heritability).

Educational attainment is phenotyped much more often than is IQ, so it's both available in this dataset and available in the much larger datasets of the GWAS in question. Aside from being a phenotype of interest in its own right and a case of looking under the lamppost, EA also proxies for intelligence on both the genetic and phenotype level. (In some cohorts, like the UK Biobank, the cognitive performance test used has such low test-retest reliability that anyone who is looking for intelligence variants is going to want to use the education variable anyway!)

> Maybe I missed something, but how well does BMI measured at age 45-50 correlate with BMI at reproductive age?

I'm sure the correlation is <1. But as long as it's not an inverse correlation, it's fine. That is just measurement error which will bias down to 0 the estimate and lead to underestimate of selection.

> Wouldn't a far simpler approach be to measure allelic frequency at regions near GWAS QTLs in multiple generations?

How do you aggregate those QTLs? You can't examine them individually, that would be horrifically underpowered. If you wind up summing them, don't you get a polygenic score back and do what OP did?

Considering that natural selection is generally considered to not operate on post-reproductive phenotype (prime case is antagonistic pleiotropy literature) I consider any conclusion based on these measures pretty suspect. Worse, specific heritability for these traits in this cohort, as far as I can tell, wasn't even estimated. Given that the supplemental says the mean in the literature is 40%, it could very well be sub 20%.

I suspect that selecting only known high effect QTLs would actually not impact your power all that much as it limits the number of tests performed. If you can't see selection at high effect alleles I'm not sure you'll be able to detect them in aggregate either, especially with very noisy phenotype data.

At the end of the day natural selection is a change in allele frequency driven by differences in phenotype. Sure, this maybe was an easy analysis to do given that the data was available, but I don't really see the value.

> Considering that natural selection is generally considered to not operate on post-reproductive phenotype (prime case is antagonistic pleiotropy literature) I consider any conclusion based on these measures pretty suspect.

Again, why? Is there any reason to think that late life BMI will inversely correlate with BMI such that the old fat people were actually the skinny young people? If it's just measurement error, then it's no worse than most variables which get used in health or sociology research.

> If you can't see selection at high effect alleles I'm not sure you'll be able to detect them in aggregate either, especially with very noisy phenotype data.

All of these are highly polygenic traits. A 'high effect allele' means explaining 1% or less of variance. Selection on such a trait is going to shift the frequency by a tiny amount. Grossly underpowered individually. You have to consider them jointly. Noisy phenotype makes that more, not less, true.

The antagonistic pleiotropy theory actually advocates exactly that inverse correlation, more investment into young years for increased reproductive fitness at the expense of worse later years.

Further, I suspect that number of kids or even just a binary having kids or not would have a direct effect on BMI in old age. More kids > less time to excercise > compounded over many years. There's also a known correlation between educational attainment and number of children in the literature which is generally hypothesized to be causal.

I have concerns about the validity exactly because of the expected tiny effects. Polygenic traits with extremely high heritability (height being the prime example) top out at 10% of variance explained even when considering all alleles. That just highlights the need for better phenotypes and multi generational typing. I think it's Botstein who I've heard advocates for more time points over more replication when given the choice.

> The antagonistic pleiotropy theory actually advocates exactly that inverse correlation, more investment into young years for increased reproductive fitness at the expense of worse later years.

Which would be great if you had any evidence that BMI is largely influenced by such antagonistic pleiotropy. But since BMI is not a disease like Alzheimer's, it doesn't even have prima facie plausibility. This shouldn't be hard: does low BMI in youth predict high BMI at middle or old age? I've never seen any result ever hinting at this, and no one believes that.

> I have concerns about the validity exactly because of the expected tiny effects.

Yes, it does make it harder, but fortunately, that's what we have polygenic scores for. Do you have any power analysis suggesting that the sample size here is inadequate?

> Polygenic traits with extremely high heritability (height being the prime example) top out at 10% of variance explained even when considering all alleles.

Totally wrong, where did you get this idea? The height polygenic score is at >>10% of variance (see Wood or http://ije.oxfordjournals.org/content/45/2/417.full ) and BMI polygenic scores are already at 10% and will go up since all the twin and GCTA estimates indicate larger heritability than that.

People who are highly educated like to feel like they are members of a higher species.

Wow... you ok?

I didn't mean to come off as being arrogant. I'm referring to the nature vs. nurture debate. Intelligence is determined by both nature and nurture, and not solely one or the other.

I'd argue that it's just as likely that our ability and desire to make teaching and learning more interesting (wider availability of college courses even without enrolling in a 4 year degree) could have had just as much impact on increasing the average EA.

The author looks at the relationship between fitness, educational attainment, and age at menarche. He measures fitness by number of grandchildren. So what he reports is that people with higher educational attainment, and women with later menarche, had more grandchildren on average.

That makes sense to me. It's pretty well established that educated people have fewer children. And early menarche might arguably be associated with long-term health problems.

The result of selection (lower education, etc.) will necessarily lag behind the selection pressure. Only the most extreme selection (immediate 100% death for non-selected variant) will show up immediately. Milder selection takes time. You can tell that it is happening if you look at millions of people, but it won't be easy to notice without some number crunching.

Suppose the difference is that one group averages 1.7 kids per woman, with typical generation-to-generation time of 35 years, while the other group has 2.1 kids per woman, with typical generation-to-generation time of 20 years. You probably can't notice, especially if those averages are blurred by lots of random variation.

It will take many generations to severely reduce the genetic predisposition toward learning. As it happens, it will be difficult to recognize. People learn less, but why? Many alternate explanations will be offered. There may in fact be multiple factors that each partially explain the result. Some of these factors may be taboo. We might explain away the situation by saying that the economy just doesn't support education, or that the culture has changed.

So, http://time.com/4327424/idiocracy/, basically.

I know who _I'm_ voting for: https://pbs.twimg.com/media/CV0dnFtWEAAwj_k.jpg

Side comment: the author is using first person in his paper, which is slightly odd. I know he is the only author, but would it be more suitable to use some other form?

I often find it more distracting when people avoid the first person than when they use it. If it makes sense, why not use it?

To be clear, "we" is first person, plural. I take it your comment and gp to mean "I" by "first person" and not "we".

Anyway, it's standard in scientific writing to use "we" instead of "I". https://en.wikipedia.org/wiki/We#The_author.27s_.22we.22

It is also a very subtle way to make it sound more credible, because it gives the reader a sense that multiple people did the work.

Or it was written by king :-)

https://en.wikipedia.org/wiki/Royal_we

On the other hand it can also some kind of proof of honesty and transparency. "Hey, I did that work alone, which is slightly odd, but hey at least I'm not hiding this fact."

I agree with haddr above. People should not be ashamed of using "I" instead of "we" when it's relevant.

For single author papers, I see many papers that switch from "we" to "I". We are probably observing a trend in the process.

I actually had a professor in graduate school who won multiple awards for his papers. His advice to us was:

* Write in first person

* Use the active rather than passive voice

* Structure papers in narrative form