An
interesting new study on the heritability of childhood obesity has been widely
publicised.
The paper, published in the American Journal of Clinical Nutrition,
found only a modest effect of shared environment on body mass index. The study
used the common technique of comparing monozygotic and dizygotic twins; that
is, twins who share all or half of their segregating genes. If a significant
group level effect is found between such groups, then the effect is very
probably due to genes.
Here is how The Times
headlined the story
Genes not poor diet blamed for most cases of childhood obesity
But the
study does not license any such conclusion.
thing, the study did not examine childrens’ diets at all. It is consistent with these results that all the children
had poor diets, and that genes played a protective role in keeping some
children slimmer than others. But the more important point is that showing that
there is a genetic influence on a trait, in the absence of any knowledge
regarding the causal route, tells us nothing the role that the environment
might be playing. Environmental can be crucial mediators of effects that are
‘due’ to the genes, in ways that make it more natural to select the environment
as the cause then the genes.
for instance, IQ is correlated with skin color, and skin color with genes. That
might indicate that genes play a direct causal role in IQ. But equally it might
indicate that racism plays a direct causal role in IQ: it is the responses of
people to a trait that is (partly) ‘due’ to the genes that plays the
explanatory role. How might environments mediate the effects of genes to
produce obesity? We don’t know (that’s the point). It’s likely that the
influence is on dispositions or preferences, and thereby on BMI. Perhaps genes
produce a sweet tooth. Or perhaps they produce more persistent or persuasive
naggers. Without knowing how the genes produce the effect, though, we shouldn’t
be drawing any conclusions about the significance of environmental factors like
poor diets.