Genetic contribution to neuroticism associated with affluence, health, longevity | Hacker News Reader

Genetic contribution to neuroticism associated with affluence, health, longevity

biorxiv.org

67 pointsgwern18 comments

This seems intuitively correct. Neutrotisicm is associated with risk-averse behavior and thus less likely to engage in risky life-threatening activity. A highly neurotic personal can be paralyzed by enumeration of every conceivable outcome, and as such not ever do the dangerous thing. There seems to be a positive correlation of intelligence, so the individual may make long term conservative bets that further increase longevity.

Be careful: this post just reeks of confirmation bias. Correlation (association) does not imply causation and it’s tantalizingly attractive to want to draw conclusions that confirm the way we think about the world.

I, personally (for example), would not generally assume neuroticism is associated with risk-aversion. In my experience risk aversion (or tolerance) tends to be completely decoupled from other traits.

I’m not saying you’re wrong, and it would equally be a fallacy to believe my anecdotal experience is anything more than an anecdote, but just be careful jumping to conclusions.

Neroticism [1] also correlates with being female. And women, on average, have longer lifespans.

[1] The big 5 personality trait, not to be confused with neuroses.

A difference in metabolism, risk taking, or perhaps both?

The article is looking at genetic differences, not environmental ones. The small gender difference in neuroticism is actually overexplained by the difference in rates of childhood sexual assault alone, meaning that if anything females are genetically predisposed to be less neurotic.

I'd be really curious to see data on this, do you have a link?

> I, personally (for example), would not generally assume neuroticism is associated with risk-aversion. In my experience risk aversion (or tolerance) tends to be completely decoupled from other traits.

I wouldn't even say that risk aversion is a one-dimensional thing.

Anecdata: me and my SO have complementary risk tolerance patterns. She avoids financial/material risks but is happy to put herself in, let's say, interesting social situations, while I, conversely, am able to spend money to the point where I'm "running on fumes"(if I believe there's a good reason to do that), but avoid any situations where I would risk making somebody upset, even if they should be confronted.

A cursory search finds that neuroticism does seem to be correlated with risk aversion, which suggests that your claim of confirmation bias, and your admonition to be careful about jumping to conclusions, both based on your personal experience, are rather ironic.

... seriously? You didn’t read the last paragraph or what?

Indeed I did - in fact, you might have noticed that I quoted part of it. The first and last sentences of your post admonish the person you are replying to for jumping to conclusions and falling for confirmation bias, yet the closest you get to stating your grounds for believing they have ignored contrary evidence is your own personal opinion about neuroticism. Confirmation bias, of course, is the selection of data that supports one's bias, but here you are not even attempting to offer data in support of your contrary bias. A valid claim of confirmation bias needs more support than the fact that you don't agree.

And there definitely is different categories for risk. People probably don't have equal risk tolerance for eating questionable food and gambling.

Yet the longest living cultures (Sicilian, Okinawan) are renowned to have devil-may-care attitudes towards life in general.

“Longest living cultures...”

Somewhere an aboriginal Australian is sighing at you. A bunch of Israelis and Greeke are pretty annoyed too. China and India won’t be thrilled at your comment either.

from the context of OP: longevity of members of the culture, not longevity of the culture.

Oooooops. You see, this is one reason why I keep trying (and failing sometimes) to be less snarky. Thanks for the correction, and I apologize.

no problem

They take a 12 question test, so 12 bits of entropy (asfar as I could see these are indeed yes/o questions) and were able to correlate that to gene code, which has orders of magnitudes more entropy (I assume)?

> For the general factor of neuroticism we identified 1,436 SNPs that were genome wide significant and formed 11 independent loci. [...] Again, these findings were comparable to those from the original study by Smith et al.

> We include them here in order to compare them with the first GWASs of neuroticism factors, which we report next. Four SNPs, all on chromosome 12, were genome-wide significant for the worry/vulnerability phenotype. These SNPs were located in one locus, spanning 219kb. This region contains the gene PPFIA2, which is known to be part of the postsynaptic density in humans [14,15].

And further

> The largest difference in the pattern of enrichment found was identified when examining which tissues showed enrichment. For each of the three neuroticism phenotypes, significant enrichment was found for the tissues of the central nervous system (general factor fold enrichment = 2.76, P = 1.35 × 10−4, anxiety/tension fold enrichment = 3.13, P = 1.90 × 10−4, worry/vulnerability fold enrichment = 3.57, P = 2.79 × 10−4);

> however, for the anxiety/tension factor significant enrichment was also found for the adrenal/pancreas (fold enrichment = 4.57, P = 6.52 × 10−4), cardiovascular (fold enrichment = 3.76, P = 0.004), and skeletal/muscle tissues

They actually had data quantifying the nervous tissue. Wow!

> The genetic variants associated with an increase in the general factor of neuroticism were also associated with a genetic risk for a lower household income (rg = -0.39, P = 2.67 × 10−16), and living in an area with a higher level of social deprivation (rg = 0.24, P = 6.95 × 10−5). However, both the anxiety/tension, and the worry/vulnerability factors showed significant genetic correlations in the opposite directions to the general factor of neuroticism for both household income (anxiety/tension rg = 0.25, P = 7.64 × 10−4, worry/vulnerability rg = 0.24, P = 3.57 × 10−4), and living in an area with a higher level of social deprivation (anxiety/tension rg = -0.31, P = 3.87 × 10−5, worry/vulnerability rg = -0.31, P = 5.02 × 10−5).

Amazing, but association is not cause. I can imagine living in a terrible neighbourhood can leave a person, well, terrified, not to say neurotic. One question of an older test they used was removed from the revised test ("do you lock the door at night"), because it didn't fit their P values. Whatever that means, the test seems like a very rough measure to arrive at any statement at all.

And the fact that two traits are inversely related means to me without giving it further due thought, that headline is largely misleading.

That should be "One question of an older version of the test they used".

Reply on news.ycombinator.com