Personalized, not “racialized” medicine: Substituting racial categories with genetic markers in clinical research

The promise of personalized medicine to prevent adverse drug reactions depends on accurate methods of identifying individuals who require genetic testing. A recent FDA drug label for carbamazepine uses ancestry to accomplish this goal. The label uses Asian ancestry as a means of identifying who is most likely to need a genetic test. This approach creates a number of bioethical/policy problems. This research project assesses a possible solution to these problems: utilizing genetic markers to characterize the diversity of individuals. Ten markers, identified in an extensive literature review, were genotyped in self-identified African American children participating in an asthma gene-environment study. Simpson's index of diversity (D) was calculated for all markers and chi-square tests were performed to determine if markers linked to African populations were able to characterize all self-identified African Americans. The results demonstrate that there are ancestral similarities and differences among research participants which are not captured by their racial category. For cases, D ranged from .39 to .92 and for controls, D ranged from .35 to .75 indicating diversity in both groups for most markers. Chi square tests demonstrated that a significant difference existed between the expected frequencies and the observed frequencies indicating that self-identified African Americans cannot be characterized solely by markers linked to African populations. In conclusion, a selected group of ancestry informative markers offers a better approach for assessing diversity among research participants than racial/ancestral categories and offers a safe, effective way of identifying who can benefit from personalized medicine.