Race and Ethnicity, Measurement Issues with

Race and ethnicity are controversial variables in epidemiological studies. Most of the controversy comes from the misuse of these variables as risk factors and from issues concerning validity and consistency of data over time and territory. Substantial inconsistencies in the categorization of race and ethnicity can be found in the literature. For these reasons, some journals have written policies and published glossaries to better define these variables. However, revisions of criteria are often required due to the dynamics of social and demographic change, such as migrations, globalization, and other cultural movements that may change the perception of group identity.

Epidemiological studies may use race and ethnicity variables in several situations. In the sampling process, these variables may be used to determine whether the true diversity of the total population is being represented by the sample and to audit the randomization process. For example, National Institutes of Health (NIH) requires the assessment of these variables to ensure that the traditionally understudied minorities are sufficiently represented. However, the validity of using race and ethnicity as causal explanatory variables or risk factors is very questionable. The detection of statistical differences between racial or ethnic groups should be considered as a starting point to better understand the true underlying genetic, environmental, or socioeconomic risk factors.

Once the relevance of the use of race and ethnicity is established, the measurement of these variables needs to be planned, validated, and analyzed with caution. The main issues are the difficulty in separating the concept of race from the concept of ethnicity, the nonequivalence of data collection methods, and the mutability of use and meaning of terminology.