Dr Pound looked at the relationship between facial asymmetry and illness in more than 4,700 15-year-olds. He drew on data collected as part of a project called the Avon Longitudinal Study of Parents and Children (ALSPAC). This has accumulated detailed records of the childhood health of participants by sending questionnaires to those children’s parents once a year. In addition, 2,506 of the girls involved and 2,226 of the boys agreed, for a previous study conducted when they were 15 (and carried out by one of Dr Pound’s co-authors, Arshed Toma), to have their faces scanned to create three-dimensional images. Dr Pound used these images to assess participants’ facial asymmetry, and then looked for correlations with rates of childhood illness, as recorded in the questionnaires. (via Asymmetry and attractiveness: Facing the facts | The Economist)
- There were none. He examined the number of years in which each child had been reported to have suffered any illness at all; the rate, each year, of symptoms such as diarrhoea, vomiting and coughing; and also a child’s total infection load, defined as the number of illnesses from a list of 16 (including measles, chicken pox, mumps, influenza and glandular fever) from which he or she had ever suffered. In each case, facial asymmetry was uncorrelated. As far as susceptibility to infection is concerned, then, asymmetry is a useless indicator.
- Dr Pound and his colleagues did, though, turn up some evidence for a second hypothesis: that symmetry is correlated with intelligence. They found an inverse relation between a child’s facial asymmetry at 15 and the results of an IQ test given to ALSPAC’s participants when they were eight.
- The effect was slight—less than 1% of total observed variation in those participants’ IQs. But previous studies of facial asymmetry, with smaller sample sizes, have suggested a similar effect. Sceptics often ascribe these earlier results to publication bias (the tendency of both researchers and journal editors to prefer to publish studies that show correlations, rather than ones that do not). But Dr Pound’s research, whose main conclusion is just such an absence of correlation, can scarcely be accused of suffering from that.