Arnold Kling  

Lifespan and Heritability

PRINT
Health Care and Cost-Effective... The Panel on Crisis of Abun...

This gets my vote for most important research finding of the year (my emphasis):


“How tall your parents are compared to the average height explains 80 to 90 percent of how tall you are compared to the average person,” Dr. Vaupel said. But “only 3 percent of how long you live compared to the average person can be explained by how long your parents lived.

...Matt McGue, a psychology professor at the University of Minnesota who studies twins, contrasts life spans with personality, which, he says, is about 50 percent heritable, or attention-deficit hyperactivity disorder, which is 70 to 80 percent heritable, or body weight, which is 70 percent heritable.

“I’ve been in this business for a long while, and life span is probably one of the most weakly heritable traits I’ve ever studied,” Dr. McGue said.


I challenge somebody to explain this finding.

I am in shock, because I have always read that characteristics that we know are associated with longevity, including race, IQ, and weight, are highly heritable. In fact, if weight is 70 percent heritable and longevity is only 3 percent heritable, how can anybody find a connection between weight and longevity? Is the connection statistically significant, but so small that it is swamped by random variation in longevity?

If I were allocating money for medical research, I'd throw a bit more at the oddballs who are studying bacteria in the gut as a source of disease and a bit less at the folks massaging the genome.


Comments and Sharing


CATEGORIES: Economic Methods



COMMENTS (10 to date)
Swimmy writes:

I'll say what I said at MR in a different way: Even if weight is heritable and threatens longevity, you can't ignore time. If I were overweight (I happen to be slightly underweight), I would now have access to more medicine, surgeries, or even healthier foods than were available to my parents at my age. And that's just a start; there are all kinds of changes that have increased my chances of living longer than my parents.

(Barring, of course, an unexpectedly short life. What do you suppose would happen if teenage car accidents alone were removed from the picture?)

Matt McIntosh writes:

Arnold: If you haven't read Plague Time by Paul Ewald, I suggest it.

Todd Flemming writes:

Arnold,

I agree the statistical colorations don’t' seem to make sense. Though, this is probably over simplified. There are so many variables that have likely not been considered, such as the impact of modern medicine. It only makes sense that if average longevity is increasing each year that a close genetic coloration would be harder to come by due to the introduction of an exponentially increasing number of variables.

jaimito writes:

Slowly, please. All insurance longevity tables take into account the age of death of the parents and grandparents, and those table actually make money for the insurance companies, so they must be right or wrong in a right way. 3% is ridiculous, or we are talking about different things.

Una golondrina no hace verano.
robin writes:

Is there a typo or something somewhere? I did a quick google scholar search on Longevity and Vaupel and got ,eg, this paper, where they got about 25% heritability for longevity. Note Vaupel is a coauthor on that paper. I assume it's the same person, but I don't know. Also on this one, which reports about a 33% heritability, and on which McGue is a coauthor. These are 90s papers, so maybe something changed a lot.

Anyway, I experimented for a few minutes, and it was pretty easy to construct correlation matrices corresponding to high heritability for weight, low heritability for longevity, and moderate negative correlations between weight and longevity, so a result like this wouldn't necessarily be internally inconsistent, correlation-wise.

Phil writes:

Is 3% the r, or the r-squared? If it's the r, I flat out don't believe it.

If it's the r-squared, which would make the r about 17%, then it makes a bit more sense.

Steve Sailer writes:

I don't know what the actual heritability is for longetivity, but there probably isn't much Darwinain selection pressure for whether you live to be 74 or 92. There's a huge amount of selection pressure to keep you alive through your reproductive and childrearing years, and then a fair amount to keep you alive through your grandparent years so you can help your kids raise your grandkids, but when you get into your great-grandparent years, you aren't contributing much to your descendents' survival and might be a net burden on the reproduction of those who have your genes.

So, evolving to burn out rather than to fade away can make sense from a Darwinian standpoint. It's probably pretty close to a toss-up what strategy involving longetivity works best from a selfish gene point of view, so you get a lot of variation in longetivity within a single family.

As an earlier commenter pointed out, the Cochran-Ewald New Germ Theory suggests that most of the major diseases that kill people before age 60 or so are not hereditary genetic diseases, but are caused by infections. This is starting to be taken more seriously by the medical establishment.

jaimito writes:
isn't much Darwinain selection pressure for whether you live to be 74 or 92.

That seems true, but in fact we dont know. One of the great discoveries in how humans survive is the role of the grandmothers. They consistently contribute to family survival, provide food to children and play a critical role during famine. Even nowadays families with grandmother living nearby have more children than families without grandmothers. Grandmothers are very useful as reliable and cheap babysitters, I assure you.

Old people tend to be useful in rural homes. When they feel they are not useful, they kill themselves.

Old leaders are very valuable in politics, they are the only ones able to ensure cooperation among younger subchiefs. Death of long senile kings and dictators always gives place to chaos and violence.

In fact, we do not know. Centenarians are not freaks, they are because they are useful.

Steve Sailer writes:

Dear jaimito:

Yes, those are good points you make. This whole question is worth studying more. I was too glib and jumpted too quickly to a conclusion.

MR writes:

There's a huge amount of selection pressure to keep you alive through your reproductive and childrearing years, and then a fair amount to keep you alive through your grandparent years so you can help your kids raise your grandkids, but when you get into your great-grandparent years, you aren't contributing much to your descendents' survival and might be a net burden on the reproduction of those who have your genes.

Comments for this entry have been closed
Return to top