Every time I imply that nutrition is a relatively unimportant factor when it comes to height, brain size, or IQ differences within countries or generations, I get heckled in the comments by people who think they know what they’re talking about.

For starters I should say, that between generations and countries, the effects of nutrition are huge.  For example, the average white young American man was 5’7″ in WWI, but about 5’10.4″ by the early 21st century (an increase of about 1.3 standard deviations).  A parallel rise has probably happened to brain development, and this probably explains a big chunk of the Flynn effect as Richard Lynn brilliantly noted way back in the 1980s, though I think he overstated the case.  And btw, when Lynn talks about nutrition, he also includes stuff like disease, since that prevents the body from using nutrients.

And just as modern people score higher on IQ tests than Victorians, largely (though not mostly) because of nutrition, African Americans score higher on IQ tests than their West African cousins largely because of nutrition, as my fellow celebrity Steve Sailer brilliantly noted over a decade ago.

A study recruited peoples of West African ancestry from several places around the world.  The West Africans born in America (who had presumably lived in the developed world for centuries) had a mean height of 1.765 m (SD = 0.073) for men (see table table 1 of this document) and 1.634 m (SD = 0.064 for women).  By contrast, in the exact same study, men living in sub-Saharan countries had a mean of 1.684 m (Nigeria) and 1.701 m (Cameroon), so about 1.693 m overall; women in sub-Saharan countries had a mean of 1.583 m (Nigeria) and 1.607 m (Cameroon), so about 1.595 m overall.

In other words, Third World West African men were 0.99 SD shorter than African American men, and Third World West African women were 0.61 SD shorter than African American women; averaging across both genders, it seems being born in Africa stunts height by about 0.8 SD, relative to those born in the First World.

The problem with invoking nutrition to explain inter-generational and inter-national differences is people think it only applies to the poor.  They can’t wrap their brains around the fact that even elites in past generations and poor countries were also be malnourished.


The above is data from the famous Minnesota study of twins reared apart by Thomas Bouchard.  Now assuming his sample was roughly representative of American environments, the height correlation between identical twins raised apart is an astonishing 0.86, suggesting 86% of the variation in American height is genetic.  Taking the square root of 0.86 tells us that height in America correlates a stratospheric 0.93 with genetic height, AND THAT’S A CONSERVATIVE ESTIMATE, BECAUSE UNLIKE THIS STUDY WHERE PEOPLE WERE ADOPTED, IN MOST CASES TALL PEOPLE ACTUALLY GROW UP IN BETTER HOMES, SO THE TRUE CORRELATION IS FURTHER STRENGTHENED BY GENE-ENVIRONMENT CO-VARIANCE!!!!!!!!!!!!!!!!!!!!!!!!!!!

Now if we subtract the correlation between identical twins raised apart (0.86) from the correlation between identical twins raised together (0.93), we learn that adding identical family environment to identical genes, only increases the height correlation by 0.07, which means that only 7% of the variation in U.S. height can be explained by the type of home you grew up in.

Now if we take the  square root of 0.07, we get a 0.26 correlation between height and family environment.  Assuming the same correlation exists within sub-Saharan African countries,  where nutrition stunts the average black by 0.8 SD compared to blacks in the U.S., only one in a thousand black Africans should have a family environment nutrition enough to match the average American black.

We can estimate this because the one in a thousand level is about 3.06 standard deviations above the mean, and assuming a 0.26 causal correlation between height and family environment, the slope of the standardized regression line predicting height from family environment (independent of genetic effects since the figure was obtained from an adoption study) you have to be that far above the mean before the 0.8 SD African nutrition deficit is negated:  3.06 standard deviations above the mean multiplied by a 0.26 independent correlation between height and family environment = 0.8 SD nutrition gap between American blacks and their West African cousins.

The same calculations imply that if you’re an American born after 1980, then you’d need to have been born among the most elite one in FIVE MILLION late 19th century Americans to achieve the same height you’d have being born in a completely average modern American home.

So yes, nutrition matters for height (and by extension, IQ and brain size) within countries and generations, but the effect is so completely dwarfed by between country and between generation effects that I’m inclined to ignore it.

Now all these calculations assume the 0.26 correlation between family environment and height is the same in all countries and generations which is surely wrong to some degree, but perhaps not wrong in the direction people think.  I was recently sent a paper claiming the effects of family environment on height in the developed world is actually increasing, which means as countries get richer, family environment matters more to height, not less, which means that the effect would be even smaller in poor countries and past generations, WHICH MEANS THE CALCULATIONS I MADE UNDERSTATE MY CASE!!!!!!!!!!!!!!!!!!!!!!!