Reader Comments

Post a new comment on this article

Advancing Issues in Understanding Health Disparities

Posted by Lillian on 17 Sep 2009 at 15:53 GMT

Gravlee and colleagues have contributed approaches that will increase understanding of ethnic disparities in health. The use of 2 measures of skin color rating is important: admixture analysis allows us to know proportions of genes that derive from ethnic groups with large gene frequency differences at 100 different polymorphic sites. This is more precise than reflectometry since the genes and their frequencies are known and can lead to greater insights. But, more important, Gravlee and colleagues also used socially ascribed skin color. This method is valuable inasmuch as people react to others by the way they look, and not by their percent admixture. Thus, Gravlee notes that "Puerto Ricans perceived as negro, as compared to trigueno or blanco may encounter more frequent frustrating interactions in high-SES settings due to institutional and interpersonal discrimination..." which can lead to "elevated blood pressure" (see references 38,49-51 in Gravlee et al article, PLoS ONE, 4(9)e6821 for more details). This interpretation could be an example of Dressler's concept of cultural consonance in lifestyle. By the way, Ernest Harburg was much criticized for his 4-level skin color variable, but it proved quite successful in numerous studies by numerous scientists: I believe the main reason for its success was that people interacted with others according to the way they looked and not according to their genetic ancestry).

The addition of a polymorphic gene to the multiple regression is problematic. It is a reasonable choice of a SNP for cardiovascular experiments. However, it seems to appear arbitrary. It certainly would need to be replicated to be accepted. I would like to know the proportion of homozygotes in each color group.

Finally, I am not fond of the organizational style of the PLoS ONE articles. The idea of placing the methodology after the results and conclusions make no sense; it was confusing for me.

In sum, Gravlee and colleagues have made excellent contributions which need more elaboration. A good piece of work!

No competing interests declared.

RE: Advancing Issues in Understanding Health Disparities

cmulligan replied to Lillian on 25 Sep 2009 at 15:45 GMT

Dear Lillian,

Thank you for your comments on our paper. I’ll be happy to respond to your questions about our choice of candidate genes and the number of homozygotes per color/SES category.

Our choice of candidate genes was based on prior identification of these genes as being associated with hypertension phenotypes as well as availability of inexpensive genotyping platforms at the University of Florida. We were limited in the number of genes and genetic variants we could assay since the study was unfunded so we tried to pick top candidate genes. It is important to emphasize that the point of our study was not to identify, or confirm, candidate genes for hypertension. Instead, our goal was to investigate whether genetic data (e.g. genetic ancestry), biological (e.g. skin color) and/or socio-cultural variables (e.g. color) could modify particular candidate gene associations with hypertension. That is the case we found with α2CDel322-325. The association between α2CDel322-325 homozygotes and variation in blood pressure achieved significance only when the socio-cultural variable, color x SES, was included. When we looked at the group of α2CDel322-325 homozygotes, it became clear that there were three categories of color/SES individuals; negro/low SES, negro/high SES and trigueño + blanco/low SES. The sample sizes within these categories are quite small since we partitioned the dataset in three ways with α2CDel322-325, color and SES. Thus, the number of homozygotes/total individuals are 2/14 for negro/low SES, 1/10 for negro/high SES, and 1/21 for trigueño + blanco/low SES. Within each group, the α2CDel322-325 homozygotes have significantly lower blood pressure relative to the other members of their color x SES group (an average reduction in SBP of 14 mmHg).

Connie J. Mulligan

No competing interests declared.