Fat is an intellectual issue

As a general rule, social scientists should be restricted to three factors, and allowed to deal with four factors only when they have achieved public renown. In my own case, I try to restrict myself to two, because holding two ideas in mind is hard for a psychologist.

It is clear, from both epidemiology and casual observation of the British street scene, that Britons are getting fatter. The great fatness that was visible in the US has now spread here. Clearly, this has a simple, one factor explanation: people are eating more. A cursory look at human metabolism shows that a small amount of food goes a long way, unless you are doing heavy manual labour, where a moderate amount of food goes a long way. Replacing the calories expended by exercise takes about one minute. I can be precise about that. Yesterday my 600 meter, 23 minute morning swim used up 175 calories (and my exercise watch is probably being generous). My usual yoghurt contains 154 calories. Croissants? Why should I touch those? I am told that an almond croissant contains 504 calories, but that need not concern us here. Assuming normal exercise during the rest of the day, the maximum I will need is 2,500 calories. Yes, no more than 5 almond croissants, but let us drop the subject.

Let us search for the second factor. Why is it that some people are fatter than others? At this stage the “multifactorial” gang make their entrance: glands, genetics, advertising, international food businesses, supermarket display shelves, television ads, leisure, lack of local sports facilities, and a profound crisis of capitalism/consumerism/spiritual values. I agree to all of the above, but what I am asking for is a predictive variable measured when children were young, and then evaluated against fatness in middle age.

Such a second factor has been identified by Kanazawa (2013). http://personal.lse.ac.uk/kanazawa/pdfs/O2013.pdf  He has used a famous population sample: all those babies (n= 17,419) born in England, Wales, and Scotland during 1 week (03-09 March 1958).  This is a proper sample: none of your convenience sample of n=30 sort of rubbish. Kanazawa had been surprised that other researchers had argued that, once you corrected for education, there was no link between intelligence and subsequent obesity. It turns out that this finding is based on partial use of the intelligence data. He found that if you include the full set of intelligence tests at 7, 11, and 16, and measure weight at 51 years of age (oh, the power of population samples) then there is a strong relationship, such that even intelligence at 7 predicts weight at 51.

In a very interesting twist, the effect of intelligence on weight only becomes apparent when children leave home, and can control what they eat. It is the opposite of all the “you are conditioned by your family environment” theories. Freedom of choice leads to different life styles, mediated by intelligence.

At this stage you may wish to stop reading, because I have just shown you evidence that intelligence at age 7 predicts obesity at age 51, QED. Stay with me a while. Parental BMI also has an influence. Fatter parents have fatter children. How does one explain that? Is it that children ape their parents through some cultural modelling? Probably not, because as we have just learned, the effect happens long after children leave home. It may be that parents with low IQ are themselves more likely to grow up to be fat, and that they have passed on the genes for low intelligence to their children. Obesity is just a side effect. It’s that pesky problem of collinearity again, and again. Are children inheriting genes for intelligence or genes for obesity, or both? A proper genetic study (of the sort which has not really been possible until now) might elucidate the causal pathway. Furthermore, although the picture is clear, the IQ/obesity link is not all that strong, and other factors also have an influence. Perhaps, after a genome wide study, we can identify the three or perhaps even four factors which account for most of the variance.

In fact, the whole issue is almost as much of an intellectual challenge as to whether one should eat an almond croissant.