© 2008 American Thoracic Society
The Relative Importance of Unmeasured Covariates in Racial/Ethnic Disparities ResearchFrom the Authors:We thank Drs. Cooke, Frank, and Frank for their thoughtful letter on our article (1). We regret, though, that they did not take more care to quote us fully when reviewing our discussion. For example, in quoting our statement that genetic factors may play a role in explaining our results, they omitted that we prefaced our point with "for example." Furthermore, they suggest that we believe enthusiastically that the differences between whites and blacks in the rate of severe sepsis are genetic. Yet, we would contend, as stated in the article (1), that racial differences likely reflect a mixture of environmental and biologic differences. In contrast, it is Cooke and colleagues who appear to vigorously defend the idea that there can only be one realistic explanation for our results—residual confounding due to unmeasured environmental factors that correlate with race. In defending their position, they cite work by Cooper and coworkers, without noting that Cooper's opinions have been criticized for straying from "middle ground" (2) or that many of Cooper's concerns relate to the failure to adjust for a number of environmental factors which, in fact, were accounted for in our analysis. Furthermore, Cooper acknowledges that functional genetic allele frequencies that may influence disease vary significantly across different racial/ethnic groups (3). His concern is that these differences not be overinterpreted as being an "effect of" rather than "association with" race, lest oversimplistic assumptions be made about the link between race, a single gene, and disease in multiattribute non-Mendelian conditions. We could not agree more. There is very strong evidence of hereditary susceptibility to infection and sepsis (4), yet unraveling the exact genetic basis is a gargantuan task. Variation within different geographic and racial/ethnic populations provides important clues. For example, recent data suggest that TLR4 haplotypes that appear to increase the risk of sepsis are more prevalent in African populations in part because of evolutionary pressure from malaria, and not simply due to genetic drift (5). The point here is that there do appear to be important differences in the rates of severe sepsis between different self-identified racial/ethnic groups in the United States. That a large part of these differences is due to differences in environmental factors establishes a clear social and public health agenda. That a large difference persists despite efforts to adjust for environmental factors also leaves open the possibility that not all the difference is environmental, and that continued efforts to explore the potential causes may not only redress disparities in care but potentially offer insights into the pathogenesis of sepsis.
University of Pittsburgh School of Medicine
ZD Associates, LLC
University of Pittsburgh School of Medicine
Genentech, Inc. FOOTNOTES Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
