Georgia State University Center of Excellence: Syndemics of Disparities

Georgia State University, Atlanta GA

Investigators

Linked publications & trials

Abstract

Much ofour knowledge about health disparities is based on epidemiologic investigations that use race or ethnicity as an important distinguishing variable. Their results demonstrate important differences in the effect of risks and conditions on persons of different ethnicities. A few examples: African Americans have higher rates of HIV and STIs than do Whites or Hispanics;^' ^ African American women with breast cancer have shorter survival than their White counterparts;3 African American women, regardless of income or education, have consistentiy higher frequencies of low birth weight pregnancies and infant mortality than do White women.4 Perhaps the most important aspects of these disparities are their strength (large relative risks) and consistency (they occur over a broad range of risks and conditions). Counterfactuals, in their simplest form, are hypothetical suppositions¿ if...then statements that attempt to address causality by supposing an alternate reality. The epidemiologic studies that have produced revealing health disparities are based on a hidden counterfactual: What if African American people were White? Then what would be their risk? Such studies often attempt to conttol for other factors (that is, to make the African American and White groups as similar as possible), but rarely can account for the historical, economic, social, and cultural differences that are encompassed by broad-brush ethnic descriptors.s The tacit assertion is that if black were white, there would be no difference¿a result that, in addition to its veiled racism, gives little direction for correcting disparities. We are left with real disparities and phantom solutions. Research in the proposed NCMHD Center of Excellence wdll focus on the health disparities that affect inner city African American people, but wdU attempt a broader and more comprehensive approach to defining health disparity populations. Since race is non-explanatory, we will seek a broad range of information about populations at risk by assembling relevant data sets from a variety of sources.^ Though race/ethnicity designations are powerful epidemiologic variables, and may be important as problem locators, we seek to avoid the implicit trap of using race/ethnicity as a convenient explanation. We plan to use three major approaches to issues of health disparity populations: (1) Broad-based data compUation; (2) Multilevel modeling and analysis; (3) Analysis of syndemic phenomenon in populations at risk.