Use of Race in Kidney Research and Medicine

Abstract

Black Americans and other racially and ethnically minoritized individuals are disproportionately burdened by higher morbidity and mortality from kidney disease when compared with their White peers. Yet, kidney researchers and clinicians have struggled to fully explain or rectify causes of these inequalities. Many studies have sought to identify hypothesized genetic and/or ancestral origins of biologic or behavioral deficits as singular explanations for racial and ethnic inequalities in kidney health. However, these approaches reinforce essentialist beliefs that racial groups are inherently biologically and behaviorally different. These approaches also often conflate the complex interactions of individual-level biologic differences with aggregated population-level disparities that are due to structural racism (i.e., sociopolitical policies and practices that created and perpetuate harmful health outcomes through inequities of opportunities and resources). We review foundational misconceptions about race, racism, genetics, and ancestry that shape research and clinical practice with a focus on kidney disease and related health outcomes. We also provide recommendations on how to embed key equity-enhancing concepts, terms, and principles into research, clinical practice, and medical publishing standards.

Introduction

Scientists have reached general agreement in [recognizing] that mankind is one: that all men [people] belong to the same species, Homo sapiens. The myth of “race” has created an enormous amount of human and social damage. In recent years it has taken a heavy toll in human lives and caused untold suffering. It still prevents the normal development of millions of human beings and deprives [civilization] of the effective co-operation of productive minds. —“The Race Question” (UNESCO 1950)

Race disparities in kidney disease are profound and have long been a focus in research and clinical practice. However, there is a lack of clarity regarding how several related concepts, including race, racism, genetics, and ancestry, should be defined and operationalized in research, clinical practice, and medical publishing. Consideration of key concepts and their appropriate use can help to align an evolving discourse in these arenas and to advance kidney health equity (Table 1).

Table 1.

Recommendations for use of race in kidney research, medicine, and medical publishing

Race, race essentialism, and racism
Recognize race as an omnipresent sociopolitical, nonbiologic construct
Recognize that racial differences in health outcomes are due to historical and persistent racialized policies and actions, not innate group differences
Recognize that race is a risk factor for racism, which in turn, is a risk factor for poor health
Recognize that race is only indirectly related to population-level ancestry genes
Acknowledge race essentialism as a threat to patient-centered care and rigorous science
Acknowledge the multiple pathways through which racism may influence biology (e.g., epigenetic modifications, allostatic load)
Race, genetics, and ancestry
Recognize that race is not directly related to population ancestry
Knowledge of population ancestry should not be used to infer an individual’s genetic polymorphism profile
Precisely define ancestry in studies, including the exact methods of its determination
Race in clinical and epidemiologic studies
Precisely define, contextualize, and provide the reasons for including race and ethnicity (and other social variables) in studies
Avoid aggregation bias and the ecologic fallacy (e.g., applying group-level inferences to individuals)
Carefully consider race adjustment and consider how other social (e.g., education, employment, insurance, and residential location) variables may be influenced by racialized policies
Avoid use of race or ethnicity in clinical algorithms, risk prediction models, or formulas in medical decision making
Identify the socioecologic levels (interpersonal, institutional, or internalized) through which racism may be acting in causal models
Peer review and publication practices
Ensure that reviewers have appropriate expertise to review studies on race and have diverse lived experiences
Consider engaging reviewers from diverse academic disciplines (e.g., social science, law, or genetics) when insights on race or racism are needed
Ensure that race differences are appropriately contextualized in published reports
Prioritize studies that “center at the margins” to include diverse patient and community voices
Recognize that words matter, including thoughtful and humble use of terminology

Defining Race, Race Essentialism, and Racism

Concepts of race, race essentialism, and racism are embedded throughout research, clinical care, and medical publishing.

Race

In the United States, conceptualizations of race are nearly ubiquitous in medicine and society (1). The concept of race was originally conceived through a taxonomy that categorized humans into phenotypically defined (e.g., by skin color and hair texture) hierarchical groupings in which White/European individuals were deemed superior (2–5). Linnaeus, Blumenbach, and others conceived and embraced the concept of race and advanced it as a foundation for a racialized social stratification, and they added legitimacy to its use by weaving it into scientific discourse and thinking (6–11). For centuries, the promotion of race as a biologic construct has been the basis for race-based scientific inquiry, abuse (e.g., eugenics), and differential medical treatment of minoritized individuals (12,13). Despite this, evidence has clearly demonstrated that race is not a genetic or biologic characteristic that inherently confers health risk (14–16). Rather, racialized societal policies and practices that subject individuals to unequal resources and treatment in society on the basis of assigned racial designations (which are neither scientific nor internally consistent) provide a powerful substrate for health disparities (17).

Race Essentialism

Race essentialism is defined as the view that “racial groups possess an underlying essence or fixed uniform and defining characteristics that represent informative unalterable traits [beyond phenotype] including genetics, behaviors, and ability” (18–20). Essentialist thinking has validated and reinforced the use of race as a basis for categorizing individuals in the design, conduct, analysis, and interpretation of biologic research (21,22). This ideology has legitimized social hierarchies and fueled misguided race-based medical practice, including pharmacogenomics (23,24) and estimations of biologic processes including kidney function (25). It has also exacerbated racial health inequities (6,26). Race essentialist reasoning inappropriately points to perceived innate biologic inferiority of socially oppressed individuals (12,27,28) while failing to recognize rectifiable root causes of inequitable health outcomes, including structural racism (29) and associated racial inequities (e.g., access to health care, housing, food availability, monetary resources, etc.) (30,31) that affect health.

Racism

Racism is defined as beliefs, actions, and policies through which differential value, resources, opportunities, and power are assigned to individuals or groups on the basis of their socially assigned race. It is a primary mechanism through which racial and other health inequities are generated (17,32–34). Racism operates at individual (internalized), interpersonal (personally mediated), cultural, and structural or institutional levels (17). Internalized racism refers to how minoritized groups accept negative and stigma-laden messages about their otherness, ability, and value (e.g., minoritized individuals feeling inherently less self-efficacy for disease management) (17,35). Interpersonal racism may be intentional or unintentional and overt or covert, and it reflects a system of power aligned with prejudice, bias, and/or discrimination on the basis of negative assumptions about the value, abilities, motives, intentions, and other attributes of a group (e.g., viewing behaviors of a minoritized group as lazy or irresponsible) (36–39). Cultural racism refers to how media representations and other social messaging can convey the idealization of Whiteness as superior and normal (e.g., representations of professionalism as White) (40). Structural or institutionalized racism refers to the ways in which racialized laws, policies, practices, and norms result in differential access to resources, opportunities, and services and afford greater privilege and opportunity to White individuals independent of personal work ethic (17,29,41–46). Structural racism is also systemic (i.e., reflecting linked and interdependent social systems, including finance, transportation, education, health, etc.) (29). It is difficult to quantify because it operates across multiple dimensions, and it persists despite legal or policy shifts that may leave behind a façade of equality or meritocracy. For example, although the US federally sanctioned practices of residential neighborhood redlining and racially restrictive housing covenants of the 1930s and 1940s were eliminated, these policies have continued to generate racial inequity manifest as lesser wealth (47,48), fewer resources for public education (49), poorer access to health-promoting food and other resources (50,51), political disenfranchisement (52), criminalization/overexposure to incarceration (53), exposure to harmful toxins (54), and segregated/suboptimal health care access (55,56).

Although race is associated with health, it often serves as an imprecise proxy for the cumulative effect of systemic racism, reflected through the “repeated experience” among racial minorities with “social, economic, and political adversity and marginalization” (57–60). For instance, “weathering” or high-effort coping with acute and chronic stressors may have a profound effect on minoritized individuals at a physiologic level (weathering as measured in part by allostatic load, overactivation of neurohormonal factors, increased expression of inflammatory mediators, and reduced immune/antiviral gene expression) (57,61). Similarly, minoritized individuals’ limited access to recommended treatments or health services, mediated through racialized policies (e.g., limiting access to employer-based health insurance and fewer health facilities in low-income and mostly minoritized communities), has a profound effect on health (62–65).

Differentiating Concepts of Race, Genetics, and Ancestry

The American Society of Human Genetics, the National Institutes of Health, and others have denounced the use of genetics to reinforce concepts of racial purity or superiority or to advance false notions that different “races” are biologically or genetically distinct (16,66).

Race and Ancestry

“Ancestry” lacks a standard definition and may refer to genealogic ancestry (descriptions of ancestors by geography or continent of origin on the basis of a family pedigree) or genetic ancestry (a reference to the paths through which genetic material has been inherited, often using haploid markers such as mitochondrial DNA or autosomal diploid markers thought to be “ancestry informative”) (67,68). Ancestry has been touted as a solution to pitfalls associated with race-based medicine and research (69), yet it has a very complex relationship to race. For example, in the United States, a recent Nigerian immigrant, an Aboriginal Australian, and a Black descendent of the United States are each likely to be classified as being of Black or African American race. Yet, these individuals may each have very distinct ancestral characteristics (70). Several key shortcomings must be acknowledged with ancestry-focused research. First, ancestry, like race, lacks standardized definitions (71). As a result, the ways in which information is gathered in commercial studies of ancestry are highly variable among laboratories (72). Second, participants in existing genetic databases used to determine ancestry often lack substantial demographic diversity (73), potentially limiting the discerning insights drawn from these data. Third, self-reported ancestry does not perfectly correlate with genomic data, and individuals with multiple ancestries (e.g., many populations in the United States that have “admixed” ancestry) cannot be grouped into discrete categories (72,74). Large ancestral categories (which may capture genetic variation in local populations), like racial categories, do not explain variation within or between local populations (75).

Generalizing Ancestry and Genetics to Racialized Populations

The terms “race,” “genetics,” and “ancestry” are frequently conflated (68). Yet, >2000 different ethnolinguistic groups, >14 identified “ancestral population clusters,” and substantial genetic variability are recognized within the African continent, with an evolution of these groups over time (76). Hence, it is problematic to generalize about Black or African American individuals with regard to continental ancestry–mediated disease-associated genetic variants (76,77). Although there may be noncausal correlations between genomic variations and socially designated racial categories (e.g., hemoglobin S and APOL1 risk alleles), genetic variations are not demarcated by distinct population-level boundaries or clusters (e.g., defined by race, ethnicity, or geography) (78–80). The mere correlation between population-level allelic variations and disease risk should not be interpreted as validly generalizable to all individuals within a population without rigorous evaluation of social and environmental mediators of observed correlations or their ubiquity among individuals within populations of interest. In sum, racial association with disease via genomic studies “runs the risk of re-inscribing race and genetics” and race-essentialist thinking (81,82).

Attempts to characterize race by gene profiles perpetuate racial essentialism (83), ignore genetic and social heterogeneity within “racial” groups, obscure the environmental mechanisms linking structural racism to health disparities, and slow our progress to understanding the root causes of health disparities (84). If the terms race, genetics, and/or ancestry are expressed jointly in studies, they should be clearly defined, and the intended implications and rationale for inclusion of data elements should be delineated clearly (70). Finally, statistical correlations between population-level allelic variations and disease risk should not be interpreted as consistently generalizable to all individuals within a population (67,68) but only to those persons with that identified allelic variation.

Race in Clinical and Epidemiologic Studies

Researchers should carefully examine when, how, and why race is incorporated into epidemiologic and population-based clinical research studies and how race is defined (85). In general, race and ethnicity should be treated much differently (86) (e.g., to broadly understand socially driven group-level differences and not to infer biologic information about individuals) than traditional biomedical and health assessment variables.

Accounting for Race

Race should be accounted for in studies when considerations regarding the health effects of socially imposed racialized policies/practices on individuals are relevant to the scientific question of interest. This may include studies that seek to describe differences in an outcome (e.g., kidney failure incidence) between racialized groups or to describe the interaction of race with other variables of interest (e.g., referrals for kidney transplants). Investigators should carefully describe how race has been defined (e.g., via self-report or externally derived/socially assigned through government, health care providers, or others) (87). Investigators should also identify the socioecologic level (interpersonal, institutional, or internalized) and mechanisms through which they hypothesize racism may operate with relation to the primary scientific question (88). It is also important to consider whether other intersecting forms of social oppression and discrimination (e.g., on the basis of sex, sexual orientation, age, nationality, religion, and income) may compound the effects of racism (89). Health risks due to gene variants (e.g., APOL1) should be attributed to genetic polymorphisms as they cannot be attributed to socially assigned race (80).

Race may be commonly “adjusted for” in clinical and epidemiologic studies seeking to identify potential causal associations between various exposures and health outcomes. Investigators who adjust for race without careful consideration of the mechanisms by which racism influences may fail to recognize how potential confounders, such as education, employment, insurance status, income, wealth, and more, have been racialized in and of themselves (49,90–92). Investigators should carefully consider how including these other racialized variables could affect both the health of minoritized individuals and racial health equity in intended and unintended ways. Investigators should also consider the merits of stratifying analyses by race instead of adjusting for race in an effort to examine the effect of the differential distribution of other risk factors (e.g., social determinants of health) by race that is most often the result of structural racism (17,70). Additional attention should be paid to disaggregating data when possible to account for heterogeneity, which may exist within the social experiences of racial or ethnic groups (e.g., differences between Black and non-Black Hispanics or Asian subgroups) (93).

Moreover, although race and socioeconomic status (SES) are correlated due to the omnipresence of structural racism, they should not be conflated. Group-level SES differences are driven by racialized policies and not innate racial and ethnic attributes (91). For instance, numerous studies have demonstrated racial differences in kidney health outcomes, even after accounting for individuals’ SES (94,95). These racial differences in outcomes should be examined carefully to consider both the multicollinearity and additional unmeasured effects (e.g., residual confounding) of other racialized social policies (e.g., insurance access and eviction policy) and resulting environments, which could influence health outcomes (70,91). To avoid aggregation bias and ecologic fallacy (i.e., assumptions that observations at the group population–level are uniformly applicable to individuals), population inferences should not be assigned to each individual in the group regarding their risk for health conditions (e.g., diabetes or CKD risk) (96), and careful attention should paid toward potential within-group confounding or effect modification (91,97).

Statistical Models, Risk Prediction Equations, Clinical Algorithms, and Formulas

The use of race in studies to identify population-level health associations may be appropriate for identifying the need for group-level public health messaging, education, interventions, policy planning, and needs for resource allocation as well as outcome monitoring. However, it is not appropriate to include race or ethnicity in risk prediction equations, clinical algorithms, or formulas used to uniformly guide clinical decision making for individual patients. This is because race ignores individuals’ genetic and social heterogeneity and obscures mechanisms linking racism to health disparities, and its inclusion can introduce aggregation bias and ecologic fallacy into medical decision making (98). Use of race in algorithms to guide clinical decision making may also lead to unintended consequences as it did when hypertension guidelines were adjusted on the basis of race (e.g., Black individuals were considered to benefit less from angiotensin-converting enzyme inhibitor/angiotensin receptor blocker) (99). For instance, numerous studies investigating race-based eGFR equations demonstrated unintended harmful consequences of these equations on the health care of Black individuals, including under-recognition of kidney disease and delayed receipt of guideline-based care (100–103).

Race in Clinical Care

Many Americans harbor negative subconscious attitudes toward minoritized groups as exemplified by moderate to significant anti-Black implicit bias (104). Despite publicly expressed beliefs that physicians are not racist (105), physicians score higher on anti-Black implicit bias tests than the general population, lawyers, and doctoral prepared researchers do (104). These biases contribute to racial and ethnic discrimination in the health care workforce and in patient care (106). Additionally, the indoctrination of racial essentialism in medical education and research may enable unempathetic interactions and clinicians’ biased, paternalistic, or disparate care (106–108).

As clinicians acknowledge how racism and health disparities are mediated through layers of inequities produced by a race-stratified society, they may gain enhanced empathy for racially minoritized patients. For instance, excess stressors disproportionately imposed upon minoritized individuals have been associated with substance abuse, poor dietary patterns, and cognitive impairment (109–111). This understanding may help clinicians better understand how attitudes toward health care, such as mistrust (e.g., recognizing that “medical mistrust” is often the by-product of current and historically racist abuses in science and medicine), or behaviors, such as adherence (e.g., recognizing financial barriers to obtaining medications), and more are driven by patients’ experiences with racialized systems.

Race in Scientific Peer Review and Publication Practices

Academic medicine and biomedical research are not immune to biases and racialized structures. Hence, peer review and publication processes should ensure key concepts are considered, defined, and operationalized in a responsible manner (88,112,113).

Peer Review

Editors, reviewers, and investigators can take several actions to promote equity and embed antiracism into their processes. First, editors should ensure that reviewers have appropriate expertise and represent diverse lived experiences and social identities, which inevitably affect the interpretation and adjudication of work (88,114). In the case of studies that directly interrogate race or racism in their relation to health, editors should seek reviewers from diverse academic disciplines (e.g., social science, law, genetics, etc.) to provide insight regarding whether structural racism or gene-environment interactions have been adequately considered. Second, peer reviewers can examine and acknowledge their own positionality (e.g., personal biases and prior personal experiences) in relation to the work being reviewed. Peer reviewers can also help ensure that race differences are appropriately contextualized by study investigators. For instance, when race is associated with unequal outcomes or other findings in a study, reviewers can ask study investigators to discuss how and/or when racism may be operating. In responding to these reviews, investigators should avoid deficit narratives that blame minoritized groups for having innate deficiencies (88). The review process can also enhance value by inviting commentary on how race relates to a study, including acknowledging the long and violent history of exclusion or abuse of marginalized populations in the research enterprise when appropriate. Editors and reviewers can elevate and prioritize studies that “center at the margins” so that minoritized patient and community voices are valued for their critical insights and to emphasize the importance of these lived experiences in influencing health outcomes (26). Finally, editors and reviewers can encourage authors to expand their authorship groups to include those in intersecting disciplines as well as nonacademic partners (e.g., community health workers, individuals working at federal qualified health centers, etc.) who have contributed substantially to the work but are often overlooked during the drafting and publication of findings (88).

Words Matter

Language shifts with history, political climates, and other contexts. However, words always convey meaning and can sometimes exacerbate stigmatization of marginalization of groups (115). When referring to socially assigned racial categories, several terms that reinforce the ideologic legacy of White supremacy (and other forms of oppression) actively or passively should be avoided. For instance, terms such as “vulnerability” or “adversity” often represent a negative description of a condition that society has created for a group of people and that often gets ascribed as an innate group trait. This could be replaced with terms such as “under-resourced,” “disinvested,” or “historically marginalized,” which provide context about the power dynamics and root causes of a condition (113). It is also important to carefully deploy current terms describing racial categorizations, particularly in locales where terminology changes frequently, such as in the United States. For example, in the 1950s, individuals who are currently labeled as “Black or African American” were previously labeled as “Negro” by the US Census Bureau (116). Terms used to describe racial groups (e.g., Black or White) should be capitalized because they are proper nouns and not colors, and they should not be made plural. For instance, instead of referring to “Blacks or Whites,” investigators should refer to “Black or White individuals, study participants, or patients.” The terms “racialized” or “minoritized” are often used to emphasize the oppression and marginalization of racial or minority groups.

Because words and their meanings will continue to evolve over time, achieving complete competence or omniscience about all words may be impossible. However, investigators, editors, reviewers, and journals should strive to continually ensure that published words embody cultural humility and equity and do not harm and stigmatize individuals or groups.

Toward Equity-Focused Race Consciousness

Equity-focused race consciousness promotes the three espoused core principles for achieving health equity, including providing resources according to need, valuing all individuals and populations equally, and recognizing and rectifying historical injustices (41). Race consciousness is a key tenet required to promote equity, justice, and fairness in the aspiration to achieve health (117,118). Equity-focused race consciousness requires that fundamental premises regarding the origins and causes of race-related individual-level health differences be interrogated with an eye toward upending race-essentialist thinking (6,27). Equity-focused race consciousness also requires recognizing racism as a root cause of kidney and other health inequities (e.g., in contrast to exclusively focusing on genetic factors, like high-risk APOL1 alleles, as a primary cause of racial disparities in kidney health). Similarly, implementing a race-conscious equity approach into research requires avoiding race-evasive “color blindness,” which purports that race should not be considered when determining opportunity or outcomes (119). “Color-blind” approaches disguised as a desire to combat prejudice/discrimination/bias (e.g., proclaiming that race does not matter) (120) can mask race-driven inequities and oppression in a racialized society (119,121,122). For example, inequities could widen or be masked around kidney transplantation or, more visibly, coronavirus disease 2019 vaccine distribution if the race of the recipient was obscured or not collected (123,124) (contrarily, studies that examine racial differences in purported biologic factors [e.g., receptor variations across a population that increased the likelihood of dying outside of the hospital]). Early research focused on coronavirus disease 2019 mortality in which data were not disaggregated by race/ethnicity obscured the severity of disparities, with downstream effects on resource distribution (125).

The use of race in research and medicine is complex yet necessary. Race should be recognized in medicine—not because it is a fixed trait approximating a biologic characteristic but because it is a latent group concept that helps to capture group-level health outcomes that are linked to racism, which selectively and differentially exerts its effects on individuals’ health on the basis of their racially stratified social status (84). Responsible, equity-focused, and race-conscious approaches are required to disrupt long-standing and harmful conventions. Clinicians, researchers, and publishers all have a shared responsibility to ensure that race is operationalized responsibly (84). Through intentional and thoughtful action, there is hope to forge a meaningful path toward racial equity in kidney health.

Disclosures

L.E. Boulware reports receiving honoraria from the Robert Wood Johnson Clinical Scholars Program and various universities for visiting professorships; serving as a scientific advisor or member of the Association for Clinical and Translational Science and the Robert Wood Johnson Clinical Scholars National Advisory Committee; and serving on the editorial boards of Journal of the American Medical Association and Journal of the American Medical Association Network Online. C. Jones reports receiving honoraria from Merck Center for Observational and Real-World Evidence and speaking on “Naming Racism and Moving to Action” on February 24, 2021. D. Mohottige reports serving as an National Kidney Foundation (NKF) Health Equity Advisory Committee member. D. Mohottige's partner owns minimal stocks in Gilead and Pfizer totaling <$5000. K.C. Norris reports consultancy agreements with Atlantis Healthcare for compliance, research, and quality care for dialysis and CKD care in Puerto Rico; serving as a scientific advisor or member of American Association of Kidney Patients (AAKP), Association of American Medical Colleges (AAMC), Atlantis Healthcare, CJASN, ESRD Network Forum, Ethnicity & Disease, International Society of Nephrology (ISN), JASN, and National Kidney Foundation Kidney Early Evaluation Program (NKF-KEEP); and other interests/relationships with AAKP, the American Society of Nephrology, ESRD Network Forum, NKF, and Society of General Internal Medicine (SGIM). The remaining author has nothing to disclose.

Funding

D. Mohottige is supported by the Duke Center for Research to Advance Healthcare Equity, which is supported by National Center on Minority Health and Health Disparities award U54MD012530. K.C. Norris is supported in part by National Center for Advancing Translational Sciences grant UL1TR001881 and National Institute on Aging grant P30AG021684.