The concept of "race" emerged in the 1600s with the trans-Atlantic slave trade and persists as a social construct despite lacking a scientific basis as a biological category.
What this shows
Braveman and Parker Dominguez argue that the term "race" should be abandoned while retaining "racism" as a tool for dismantling oppression. They show how official U.S. classifications treat some groups as "races" and others as "ethnicities" without scientific justification. Geographic ancestry remains socially meaningful for monitoring health inequities produced by racism — but the label "race" reinforces false ideas of inherent biological difference.
"Race" reinforces notions of inherent biological differences based on physical appearance.
— Braveman & Parker Dominguez (2021)
genetics
Lewontin
(1972)
At a typical genetic locus, roughly 85% of human genetic diversity falls within populations rather than between socially defined racial groups.
What this shows
Lewontin's analysis found that most genetic variation exists among individuals within any population, with only a small fraction aligning with conventional racial classifications. The pattern undercuts treating "race" as a biological taxonomic unit. Later studies have replicated the general finding, with exact percentages varying depending on how populations are grouped.
Less than 15% of all human genetic diversity is accounted for by differences between human groups!
— Lewontin (1972)
consensus
Fuentes et al.
(2019)
The American Association of Physical Anthropologists states that race does not accurately represent human biological variation.
What this shows
The AAPA's 2019 institutional statement holds that humans are not divided into distinct continental racial genetic clusters. The Western concept of race emerged from colonialism and discrimination rather than biological discovery. The statement affirms that racism has real biological consequences for health even though racial categories do not map onto biological groups.
Race does not provide an accurate representation of human biological variation.
— Fuentes et al. (2019)
current research
Rotimi et al.
(2025)
In the All of Us cohort, participants within self-identified race and ethnicity groups show gradients of genetic variation rather than discrete genetic clusters.
What this shows
Using whole-genome data from more than 230,000 participants, the authors report that self-identified race and ethnicity categories contain wide internal genetic diversity. Continental and subcontinental ancestries vary considerably within those categories nationally and by state. The results suggest race and ethnicity labels are imperfect proxies for genetic ancestry in research and clinical contexts.
participants within self-identified race and ethnicity groups exhibit gradients of genetic variation rather than discrete clusters.
— Rotimi et al. (2025)
medicine
Vyas et al.
(2020)
Race correction in clinical algorithms may direct more medical attention or resources toward patients racialized as white than toward patients racialized as Black, Hispanic, or Asian American.
What this shows
Vyas et al. document widely used decision tools that treat race as a biological input — adjusting estimates for kidney function, obstetric risk, and other outcomes. The authors describe how race-based clinical algorithms can route patients racialized as Black, Hispanic, or Asian American toward different care paths than white patients. The paper helped catalyze specialty-wide review of race correction in medicine.
may direct more attention or resources to white patients than to members of racial and ethnic minorities.
— Vyas et al. (2020)
medicine
Inker et al.
(2021)
The authors report that race in estimated glomerular filtration rate equations is a social construct, not a biological one, and present new equations that omit race.
What this shows
Inker et al. developed creatinine- and cystatin C–based kidney function equations without a race variable. They found that prior equations with race correction overestimated measured filtration for participants the study grouped as Black and that removing the adjustment without a replacement formula underestimated it. New combined creatinine–cystatin C equations without race showed smaller gaps between those groupings in validation cohorts.
race in eGFR equations is a social and not a biologic construct.