18 May 2005
It is known that disease-associated genetic variants are more prevalent in some ethnic populations than others (for example, pathogenic mutations associated with cystic fibrosis in the CTFR gene are most common in Caucasians of northern European ancestry; beta-thalassaemia is most common in individuals of Mediterranean, Asian and Far Eastern ancestry). However, if such mutations are present in any individual, they almost invariably cause disease. The position with respect for common complex diseases (such as diabetes and cardiovascular disease) is much less clear; in this instance, multiple genetic variants may influence predisposition towards disease and also response to treatments. Identifying key genetic polymorphisms involved in complex diseases is a major research area. In the midst of this, there is on-going debate about the existence of racial differences in genetic associations with disease susceptibility and drug response; specifically, whether the effect of a given genetic variant is different when present in individuals of different racial backgrounds. This is an important question to address in the development of genetic medicine, since it potentially makes it much more difficult to assess disease risk and drug response.
A new study in Nature Genetics attempts to address this question by comparing the genetic effects of 43 validated gene-disease associations across different populations; these 43 meta-analyses were selected to satisfy certain criteria, including using data from at least two 'racial' or 'ethnic' groups and having significant results overall or for at least one ethnic group [Ioannidis JPA, Ntzani EE and Trikalinos TA (2004). Nature Genetics 36, 1312-1318]. The authors note the difficulty (and general inconsistency) in defining such groups, and opted to focus on major, distinct groups, primarily European, East Asian and African populations.
The frequencies of the genetic markers of interest in control populations varied between studies, as did the magnitude of observed effect of the markers. Racial group-specific frequencies of the genetic markers of interest also showed considerable variation between studies; significant heterogeneity was found in 25 of the 43 meta-analyses (58%), calculated using the I-measure, a statistical tool that gives an indication of whether observed heterogeneity is likely to be significant, or merely a chance finding. Next, the researchers looked at a statistical measure of the increased risk associated with a genetic variant, known as the odds ratio, for different racial groups in each study. They found that race-specific odds ratios were consistent in 32 of the 43 selected studies, in that variants that were associated with increased risk in one group had a similar effect in other groups. Although some estimated odds ratios in the other studies showed opposing directions (associated with increased risk in one group but decreased risk in another), in all of the 15 studies that showed significant risk association in two different racial groups, the effects were in the same direction. Significant heterogeneity in the effects of the genetic variants of interest was observed between different racial groups in only 6 of the 43 meta-analyses (14%).
The authors emphasise that, whilst racial group-specific frequencies of genetic variants showed significant heterogeneity in 58% of the studies, racial group-specific effects showed significant heterogeneity in only 14% of the studies, concluding that their data are