REVISITING RACE/ETHNICITY AS A VARIABLE IN HEALTH RESEARCH

The recent debate promoted by the Journal stimulates fresh thinking on the evolving issues of race and ethnicity.1 Thomas' letter offers me an opportunity to clarify and develop some of my views recently published in the Journal.2

Race and ethnicity are complex, overlapping concepts that serve political, social, policy, planning, epidemiologic, public health, and other purposes.3 I analyzed the strengths and limitations of race and ethnicity in the public health arena, in the context of politics, health policy, health planning, clinical care, surveillance, health services research, and causal research. I concluded that both race and ethnicity have a role to play, and the choice of one or the other is dependent on context and purpose.3

In my view, the priority is to achieve widespread understanding of the strengths and limitations of the concepts of race and ethnicity in the fields of public health, health care, and research. The need for shared concepts and vocabulary, which will require regular review and revision, will then be readily appreciated,2,4 as will the vital role of editors.5

Thomas's letter exemplifies my own current view that neither race nor ethnicity ought to be abandoned lightly.3 Thomas wonders whether ethnicity or race would help us understand the health of Jamaicans. The key question, surely, is, In what respect? If the question is one of genetic causes of diseases, or the impact of racism based on skin color or facial features, the concept of race may serve.3 If the question is one of the causes or consequences of most complex diseases or health care delivery problems, the broader concept of ethnicity—encapsulating culture, environment, and biology—is more likely to be of value.3

A study of Jamaicans that does not differentiate between Jamaicans of Indian descent and Jamaicans of African descent is likely to mislead. Even within broadly defined ethnic or racial groups—for example, groups labeled as Indian, Asian, or South Asian6—the heterogeneity in health patterns and risk factors is immense. Those who analyze data categorized by inappropriate racial/ethnic labels, most often data captured for administrative purposes, have too readily ignored such heterogeneity. An analysis of data by catch-all categories such as “Black” and “White” is likely to be inconsequential or harmful because of the tendency of such designations to disguise important variations.

Demonstrating and understanding inequalities, whether by race or ethnicity, is, sadly, no more than a first step. At the heart of the processes initiating and perpetuating such inequalities lurk social and economic factors. To analyze ethnic or racial inequalities without paying attention to socioeconomic factors, which are inextricably entwined with ethnicity and race, is naive,7 and Thomas is correct to draw our attention to this fact.