Does national IQ depend on parasite infections? Er…

View Images

[I was originally going to avoid this, but decided to do it for the critical analysis, because I suspect it will be widely but badly covered, and because I also suspect that very little of this coverage will point out the publication record of these authors. Which is worth pointing out. Have fun in the comments!]

Why do different countries have different IQs? You’d first answer probably has something to do with education, but a trio of US scientists have put forward a radically different hypothesis – international variation in intelligence is related to the prevalence of parasites in a country. As is, according to them, pretty much ever y major aspect of human culture (but more on this later)…

Christopher Eppig, Corey Fincher and Randy Thornhill (yes, that one) from the University of New Mexico have suggested that fighting off parasitic infections during childhood takes up valuable energy that might otherwise go towards the development of the brain. More parasites mean less well developed brains and weaker mental abilities.

To test their controversial idea, the trio collected average IQ values for countries all over the world using three separate sets of data. They also used the World Health Organisation’s data on global “disability-adjusted life years” (DALYs), a measure of a country’s disease burden that looks at the number of years of ‘healthy’ life lost by an average citizen because of poor health. They found a strong correlation between these two figures, both across all nations and within each continent (except South America).

They claimed that the prevalence of infectious diseases is the “most powerful predictor of average national IQ”, even after they adjusted the results for other factors, like each country’s temperature, GDP, literacy rate, enrolment in secondary school and more. They also suggest that this could help to explain the mysterious Flynn effect, where IQ increases sharply as a nation develops.

The very obvious caveat to all of this is that old adage that correlation is not causation. In this case, a link between infections and IQ tells us nothing about whether infected people grow up to be less intelligent, or whether intelligent people are less likely to become infected. Intelligence, after all, could affect one’s understanding of what a disease is, how to avoid it, and how to seek help for an infection.  And perhaps a third factor is at work here – higher education could lead to both greater intelligence and the knowledge to avoid common infections. Readers may enjoy trying to come up with alternative explanations of their own.

These problems become particularly astute when you’re looking for correlations between statistics that represent entire nations. This broad-brush ‘ecological’ approach tells us nothing at the individual level. In a given country, do children who acquire early infections grow up to have lower IQs? We simply don’t know.

In fairness to Eppig, Fincher and Thornhill, they say, “We are not arguing that global variation in intelligence is only caused by parasite stress.” They also frame their paper as a way of introducing a hypothesis and suggest ways of testing it. Fair enough, but they have supported their hypothesis with data that are, at best, inconclusive. As such, I wonder what this study is doing in a Royal Society journal rather than, say, Medical Hypotheses.

Indeed, as I alluded to earlier, this new paper is the latest in a long line of hypothesis-generating publications from Fincher and Thornhill linking parasites and infections to pretty much any sweeping aspect of human life you can think of. Through similar studies based on correlations at the national level, Thornhill and Fincher have suggested that infections are linked to individualism and collectivism, religious diversity, linguistic diversity, armed conflicts and civil war, and democracy and liberal values.  Like any attempt to explain very complex patterns of human behaviour through a single cause, this ought to raise an eyebrow. I’m raising two.

Reference: Proc Roy Soc B http://dx.doi.org/10.1098/rspb.2010.0973