Reader Comments

Post a new comment on this article

Referee Comments: Referee 2

Posted by PLOS_ONE_Group on 10 Apr 2008 at 18:53 GMT

Referee 2's Review:

This paper reports the results of 2 genetic association studies between 4 SNPs of the GLUT9 gene and 1. gout in a study of 665 cases and 665 controls and 2. CAD/MI in a study of 1473 cases and 1241 controls.

The rational for initiating this study was that on one hand GLUT9 SNPs had previously been reported to be associated with uricemia, an important causal factor of gout, and on the other hand an independent association has been reported between hyperuricemia and CAD in some studies but not all.

The results are very interesting in showing a strong association between GLUT9 SNPs and gout but no association with CAD/MI. Both results are important.

• The reduced risk of gout ( OR= 0.62 ) associated with the minor alleles of the GLU9 SNPs is consistent with their previously reported effect on Uric Acid (UA) levels. The fact that none of the SNPs studied or previously reported appears to be causally related to UA production may be a matter of concern, the direct implication of the GLUT9 gene may even be questioned. The authors discuss this point carefully p12-13. Clarification of the mechanism underlying the association will be an important next step in this research as it may have implications in the treatment of gout.

• With regard to CAD/MI, the results are also important as they suggest a non-causal relationship between elevated UA levels and CAD. As for many studies reporting phenotype-disease associations, the association between UA levels and CAD in epidemiological studies may be secondary and not etiological.
The authors discuss some limitations of their study, especially the fact that the diagnostic of gout was established without knowledge of UA levels. There are 2 other aspects however which might be of concern and should be discussed further to ensure that they might not bias the results: 1. the gout case/control study was nested within a CAD study, which explains the very high prevalence of CAD in both gout cases and controls; 2. The cases and controls for the CAD study are very different in term of gender distribution. This is not the kind of matching that we would expect in a case control study of a disease that is strongly associated with gender. This imbalance may be related to the particular mode of selection of the controls in this study (relatives of cases for a disease having a much higher prevalence in males). The lack of gender balance might have consequences if interaction existed between disease status, gender and the genetic factor. This should be discussed and the possibility of a bias eliminated.

The statistical and genetic analyses, including haplotype analyses and careful studies of the genetic context of the GLUT9 gene using HAPMAP and recent results of GWAS are appropriate and add to the interest of the paper.

Overall, this is a very interesting paper which in addition is clearly written.