Basu and Siddiqi Respond to “Context, Behaviors, and Hypertension Inequalities”

We thank Dr. Hicken for her commentary (1) on our article regarding disparities in the distributions of blood pressure across racial/ethnic groups in the United States (2). She correctly places our results, which largely show that proximate risk factors are inadequate to explain disparities, in the context of a larger body of literature that suggests that broader social determinants of health must be addressed for such disparities to be reduced. Indeed, we have written extensively about social welfare policies, political inequalities, and broader socioeconomic power disparities as the ultimate drivers of poor health and of health inequalities (3, 4). Although we agree with Dr. Hicken that multilevel analyses, including analyses of social and political determinants of health, are therefore more useful for modifying disparities, our aim in our article was primarily methodological. Our goal was to illustrate the implementation of decomposition methods that can be used to study disparities at any level of analysis. Unfortunately, our nation's largest surveys of hypertension primarily record data at the level of proximate risk factors for hypertension; hence, we focused on the most commonly utilized metrics of risk for a demonstration analysis. However, decomposition is not limited to proximate factors, and our illustration is intended as a starting point for the more comprehensive analysis of disparities of outcomes that are measured as continuous variables, which we hope will add a new tool for epidemiologic analysis at any level.