In the news

Behind the recent plethora of genome-wide association studies identifying common risk variants for multifactorial diseases, is the assumption that an individual’s overall risk of developing a particular disease is related to the cumulative effect of multiple gene variants. This idea is often known as the common disease common variant hypothesis (CDCV). However, until recently, there has been limited evidence that the number of risk variants possessed by an individual correlates with their phenotype or risk of getting a disease.

 

Since multiple risk variants are now known for some specific complex diseases, it should be possible to test this idea, and recently two independent studies have shown that – at least in principle – the hypothesis is valid.

 

The first study, published in the New England Journal of Medicine, assessed the cumulative effect of 16 previously identified single nucleotide polymorphisms (SNPs) at five different chromosomal locations against the risk of developing prostate cancer [Zheng et al.(2008) NEJM 358(9):910]. Each of these SNPs individually have odds ratios of less than 1.65. However, the presence of high risk variants in all five regions combined with family history (with an odds ratio of 2.2) explains nearly half of all cases of prostate cancer, with an overall odds ratio of 9.5.

 

The second study, published in the same journal a month later, assessed the cumulative effect of 11 previously identified SNPs at nine different locations against the risk of cardiovascular disease [Kathiresan et al. (2008) NEJM 358(12): 1240]. The risk variants were compared with plasma levels of low-density lipoproteins (LDL) and high-density lipoprotein (HDL), which are associated with the risk of cardiovascular disease. They found that, as the number of risk variants increased, the level of LDL in the blood increased whilst the level of HDL decreased, in-line with a higher overall risk of cardiovascular disease. During the follow-up period, the number of risk variants also correlated with the incidence of cardiovascular events.

 

Comment: Although these studies have relatively small sample sizes and both use naïve additive models to combine the effect of multiple risk variants, they nonetheless represent the first real indication that risk stratification of the population, using common genetic susceptibility variants, might be possible. Whether or not genetic testing of this kind could ultimately be useful for improving public health is still under debate.