Interesting. Identical twins raised separately have a correlation of .76, while fraternal twins raised together have a correlation of just .55 and non-twin siblings raised together have a correlation of .47[0]. To me that suggests a strong genetic (though not necessarily heritable) component. And perhaps the .08 difference between fraternal twins and non-twin siblings is explained by differences in the mother's nutrition.
How do you measure the effect of nutrition? What studies are you looking at?
[0]: http://en.wikipedia.org/wiki/Heritability_of_IQ#Correlations...