An attempt to find a causal relationship between neuroticism and the genetic expression of a particular neurotransmitter does not stand up to scrutiny, according to researchers in Jonathan Flint's group at the WTCHG.
A US-based collaboration led by Zhifeng Zhou at the National Institutes of Health had reported in Nature that variants of the gene for neuropeptide Y (NPY) showed small but significant associations with some measurements related to anxiety, such as regional brain responses to emotional stimuli and pain, and scores on the harm avoidance subscales of a personality questionnaire. They concluded that 'haplotype-driven NPY expression... inversely correlates with trait anxiety [neuroticism].'
Colleen Cotton, Thomas Campbell and Jonathan Flint tested this claim by looking for the same variants in 582 people who scored either very high or very low on the neuroticism measure of the Eysenck Personality Questionnaire. As reported in Nature in April, they found no difference between the high and low scorers in NPY expression as predicted by the combination of genetic variants they carried at the NPY locus. They conclude that if NPY does affect neuroticism, then its contribution to the heritability of the trait can be no more than 1.25 per cent.
The WTCHG researchers argue that genome-wide association studies (GWAs), which aim to discover patterns of association between a region of the genome and susceptibility to common conditions, will find causal variants more reliably than the 'candidate-driven' approach taken by Zhou et al. As in other conditions, however, the task is proving more difficult than first thought. So far the associations that have emerged account for only a small percentage of the risk, and some speculate that combinations of many rare genes may underlie complex traits.
The team sees grounds for optimism in the success of whole-genome approaches to other complex conditions, such as diabetes. 'I am confident that GWAs are the best way to find the genetic basis of complex traits such as neuroticism,' says Thomas Campbell. 'However, such studies will require an extremely large number of samples because of the complicated nature of human personality, and the small sizes of the effects of genetic variants associated with them.'
Colleen H. Cotton, Jonathan Flint and Thomas G. Campbell Nature 458, E6 (2009)
For more information on Prof Flint's research, click here.