Meta-analysis of the heritability of human traits based on years of twin studies | Hacker News Reader

Meta-analysis of the heritability of human traits based on years of twin studies

nature.com

20 pointsjonifico18 comments

Where does the "all" come from in this headline. I only read the abstract but I read:

>For a majority (69%) of traits, the observed twin correlations are consistent with a simple and parsimonious model where twin resemblance is solely due to additive genetic variation

69% is not close to all. Especially when (I believe) trait is defined fairly strictly in the study to be something easily measured.

The complete sentence from the study is: "Our results provide compelling evidence that all human traits are heritable: not one trait had a weighted heritability estimate of zero."

So, all traits are at least partially heritable.

Not one? That seems highly suspect.

If I'm understanding the text correct, there are some correlation coefficients of 0.09, 0.1, 0.12 etc. I didn't see in my brief look at it if the sample was large that these would still be statistically significant. It seems misleading title still.

Like it has "social values" and "activities" in there but as far as I can tell it includes samples from various different countries. As long as twins are slightly more likely to end up living in the same country even when adopted - then its hard to say a score of like 0.1 has anything to do with genetics. I'm sure they tried to adjust for it somehow, but I'm pretty skeptical the headline result is capturing anything meaningful.

The next sentence: “The data are inconsistent with substantial influences from shared environment or non-additive genetic variation.”

Full text here for anyone that wants it: https://www.researchgate.net/publication/276922271_Meta-anal...

The average heritability is 49%. This would imply heritability and environmental influences would be about evenly split, no? So, all human traits are influenced by heritability, but they're also influenced by the environment. This finding seems less ground breaking then the title initially suggested to me.

Coincidentally, I'm in the middle of reading "The Neuroscience of You" by Chantel Prat. The author's comments on twin studies:

> some characteristics that have genetic influences, like extraversion, also influence the kinds of environments and experiences people will seek out. And other genetic factors, likes how tall or attractive you are, can influence your experiences by shaping the way others treat you. To further muddy the nature-versus-nurture waters, the rapidly evolving field of epigenetics is showing that environmental experiences can create chemical changes in our DNA!"

> To figure out how surprising the similarities between long-lost twins (or triplets) might be, we've got to ask the question, "How likely is it that any two random strangers who meet on the street would also be similar in this way?"

This was published seven years ago. Its arguments are very sketchy - as if settling nurture versus nature for humans were not social science, as opposed to science.

If identical twins are raised in a wealthy household versus a poor household, how can one determine traits coming from inheritance versus environment? This analysis does not really answer it.

The data of identical twins separated at birth is very light, as one can imagine. A look at the data of the major studies shows there really is not a lot of data - twins are separated at the age of 9 or so, or sometimes are separated at birth and then reunited.

Many of these twin studies are funded by the strange but well-funded Pioneer Fund. You can check out what they were up to in the 1930s if you're interested.

https://www.pioneerfund.vc/ ??? !!

( A fund of 300+ Y Combinator alumni investing in the top companies of YC and our Venture Partner network. )

Interesting coincidence, but I'm guessing you were referring to

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

> an American non-profit foundation established in 1937 "to advance the scientific study of heredity and human differences".

> The organization has been described as racist and white supremacist in nature.

> One of its first projects was to fund the distribution in US churches and schools of Erbkrank, a Nazi propaganda film about eugenics.

I'm assuming it's fair to say that there weren't many people left in 2022 who believed in a total blank slate? Or was that part still contentious somewhere?

But there’s a whole spectrum of heritability between “all traits are heritable” and blank slate.

Correct, which is why I was curious about the "total" version of it.

Re: the "in between", I remember someone on HN posting a meta-analysis suggesting that the split is about 50/50.

PDF is available here: https://www.researchgate.net/publication/276922271_Meta-anal...

All human traits? How about the trait of being an only child?

Having fewer children is certainly heritable.

The question I was actually raising was about the reliability of claims about a population that was explicitly excluded from the data.

Your implication of the heritability of childlessness is similarly troubling, but I imagine you might be thinking of large families where nearly all siblings are childless.

(2015)

Reply on news.ycombinator.com