Researchers should reconsider how they use population descriptors such as race, ethnicity, and geographic location in genomic and genetic research, according to a new report from the US National Academies of Sciences, Engineering, and Medicine.

The report, commissioned by the National Institutes of Health, says population descriptors do not fully capture the range of human genetic diversity and their use should be carefully considered.

“The problem had been raised, well before our report,” said committee member Ann Morning, a sociology professor at New York University. “There is a concern that researchers are using population descriptors like race and ethnicity, without thinking very carefully about what they might actually represent,” she added.

According to the report, race and ethnicity have a long history of being identified incorrectly in genomic research as reasons for average differences between groups — and using such socially constructed descriptors in genetic and genomic research can reinforce the misconception that social inequities are caused by biological differences.

The system of fitting all human beings into categories based on race and ethnicity is simply not a very good measure of human genetic diversity, Morning added.

“Classifying people by race is a practice entangled with and rooted in racism, and the pernicious effects of applying this classification to genetics and genomics research have undeniably caused harm over the last century,” committee member Charmaine Royal, a professor of African and African-American studies, biology, global health, and family medicine and community health at Duke University, said in a statement. The report recommends against using race as a proxy for human genetic variation and for researchers to be mindful of terminologies used to label groups. The report said the term “Caucasian,” for instance, should not be used under any circumstance, as it was originally coined to convey the notion of white supremacy.

“Because it comes with those connotations of white supremacy and of the idea that there’s somehow scientific grounding for racial categories, we recommend against using that term,” Morning said.

To rectify the problematic usage of population descriptors, the committee put forth recommendations to help researchers to avoid typological thinking — the belief that humans can be grouped into discrete, innate categories based on race, ethnicity, or ancestry.

The report recommends that genomics and genetics researchers tailor their use of population descriptors based on the type and purpose of their study and explain why and how those descriptors were selected in their work. It also offers a decision tree to help researchers choose whether race, ethnicity or indigeneity, geography, genetic ancestry, or genetic similarity would be the most appropriate descriptor to include in their work.

The committee members also noted that, for many studies, no population descriptor may be needed at all. “We basically ask the researcher to consider a series of questions about the research, and hopefully that will help lead them to making some decision about which, if any, population descriptor would be appropriate for their study,” Morning said.

The committee further encourages researchers to consider the effect of environmental factors in their study design, as such nongenetic factors can influence the development of genetic disease, and to engage with the communities and participants.