Race/Ethnicity As A Variable In Epidemiological Research

​
​Wei Perng, PhD, MPH

Following a series of racist events and civil unrest, increased attention is being given to the interpretation of “race/ethnicity” – a construct that correlates with a person’s self-identified geographical origins and/or cultural affiliations – in health science research. In particular, researchers are making a conscious effort to recognize and articulate that racial/ethnic disparities in health and disease risk are due to social experiences and structural inequities that operate at institutional and interpersonal levels, and influence physiology.

A recent example was a symposium review published in the Journal of Internal Medicine in April 2020 that discussed physiologic reasons for racial differences in glucose-insulin homeostasis.  The authors proposed that there may be innate genetic differences in Black vs. white women that make Black women more prone to development of type 2 diabetes, when they develop excess adiposity and consume a high-glycemic diet. While such a proposition is not necessarily problematic, these assertions did not include references to studies involving genetic data. This article was followed by a letter to the editor in August 2020, calling attention to the assumptions about racial essentialism: the view that specific racial/ethnic groups have a set of homogenous biological attributes. The letter emphasized several points: the inappropriateness of making conclusions about genomic racial differences without genetic data; the relevance of social determinants of health as a strong upstream risk factor for disease; and problems with use of racially inflammatory language with roots in archaic colonial-era rhetoric. The authors of the original review paper responded respectfully to the letter, acknowledged the relevance of a sociocultural basis for disease risk, conveyed a shared desire to eliminate health disparities, and expressed their commitment to facilitating research that overcomes the barriers of genetic determinism.
               After reading this discussion, I made it a goal to incorporate these concepts into my research and teaching. How can one objectively quantify and appropriately interpret the health effects of a self-identified social construct? It turns out there are a few ways. First, while there may be consistently detectable differences in physiology across racial/ethnic categories, researchers and educators should avoid making conclusions about biological causes without the appropriate data. Racial/ethnic health differences arise through varied mechanism (e.g., psychosocial stress, structural and political barriers to resources), with genetics as just one possible pathway. Second, given that race/ethnicity is a self-identified construct that “gets under the skin” through perceptions of bias/discrimination, whenever possible, studies should gather information from qualitative assessments and questionnaires about lived experiences.  Third, I was excited to learn that there are ways to quantify racial/ethnic disparities using objective metrics. For example, Dr. Loni Tabb uses spatial analyses techniques to evaluate whether spatial heterogeneity (i.e., census tract measures of neighborhood socioeconomic status, density of fresh food stores, physical activity resources) accounts for some of the variation in racial/ethnic differences for cardiovascular health outcomes. Such findings have important implications to target interventions that seek to reduce health inequities – a tall task that will require interdisciplinary collaboration and rigorous exploration of holistic, multifactorial pathways underlying disease risk.