2

Foundations and Background

This chapter provides the foundations for the report, beginning with a set of scientific and ethical principles that grounded the committee’s work. Race and ethnicity are often used and defined differently across fields and by different individuals. To develop a common understanding from which to build analysis and recommendations, this chapter presents the committee’s definitions of race and ethnicity and includes background on how these concepts have been commonly used in the United States. The chapter defines biomedical research and concludes with a brief discussion of the complexities that arise at the intersection of social context and biology.

GUIDING PRINCIPLES

Chapter 1 described injustices that have contributed to health disparities and mistrust of the research establishment among some racial and ethnic groups. Researchers and other parties in the biomedical research ecosystem have a responsibility to uphold values and principles that merit trust in scientific findings. The trustworthiness of biomedical research depends on both its integrity and ethical foundations. In the context of race and ethnicity, trust influences how these data about identity are ascertained, disclosed, and documented, which in turns influences the quality of data that will be used in research. In keeping with the National Academies of Sciences, Engineering, and Medicine’s (the National Academies’) commitment to responsible and ethical science, this report and its recommendations are built on a set of guiding principles, comprehensive propositions that reflect the highest ideals of science (NASEM, 2017). For biomedical research to be ethical, it must both be scientifically valid and have social value (Freedman, 1987). The guiding principles that frame this report are the scientific standards of validity, objectivity, rigor, reproducibility, replicability, openness,

and transparency, together with the ethical principles of justice, respect for persons, beneficence, equity, and inclusion.

Principles are comprehensive, fundamental truths, laws, or assumptions that serve as the rationale for a system of reasoning, belief, or behavior. Principles reflect values and experience but are also based in evidence, logic, and reason. The principles of responsible and ethical science often overlap, and related terms may be used in multiple ways. Moreover, in some situations, the demands of two or more principles may conflict (e.g., beneficence and autonomy in ethics). Thus, the committee recognizes that its principles should be understood in context—not as absolutes, but as guides to further analysis and reasoned action.

Scientific Principles

• Validity refers to the extent to which a measure, test, or study correctly or truthfully represents the concept, characteristic, or phenomenon intended to be assessed. Validity is a benchmark for the integrity of research (George et al., 2003).
• Objectivity, or freedom from bias, is a fidelity to facts and evidence arising from research that is largely free of biases, emotions, and untestable beliefs. It is key to the validity and reproducibility of research, particularly awareness and minimization of bias in the framing of a research question, the design and conduct of the study, and the interpretation and reporting of results.
• Rigor, the strict and controlled application of careful research design and scientific method across all stages of a study, is a foundation of responsible biomedical science (NIH, 2023).
• Reproducibility is the consistency of the results obtained using the same data, methods, and conditions of analysis, which may confirm the validity of prior work. Reproducibility is particularly important for computational data, as detailed in reports from the National Academies (e.g., NASEM, 2019).
• Replicability is the consistency of research results from work that asks the same question with different methods. The term replicability is often conflated with reproducibility in common use, but reproducible research may not be replicable (NASEM, 2019).
• Openness is the willingness to share methods, processes, and data, which is fundamental to the reproducibility and replicability of research within and across disciplines and contexts.
• Transparency, which is related to but distinct from openness, is the disclosure of information necessary for the reproduction, verification, and evaluation of research and the evaluation of its potential biases (Resnik, 2023).

Ethical Principles

The well-known Belmont principles for ethical research with human beings are at the core of this report’s framework (HEW, 1979), complemented by specific considerations for larger ethical questions related to race and ethnicity in research.

• Respect for persons in the context of general research refers both to the conviction that individuals should be treated as autonomous agents whose considered opinions and choices should be honored and to the commitment that persons with diminished autonomy should be protected, even to the point of excluding them from activities that may do them harm (Emanuel et al., 2000). In the context of this report, as in the 2023 National Academies report on population descriptors, respect calls for understanding the preferences, norms, and values of individuals and communities to inform the ways in which people are described in research (NASEM, 2023a).
• Beneficence in research encompasses both the commitment to prevent and minimize possible harms and the need to maximize possible benefits. Beneficence is linked to the social value of research, in that ethical studies should produce meaningful benefit for the public at large and, to the greatest extent possible, benefit for individuals and communities that participate in it. In the context of this report, beneficence requires consideration of how the use of race and ethnicity as research variables may harm or benefit a study’s participants, their communities, and wider society.
• Justice is a multifaceted principle that, in the context of research with human participants, addresses the fair distribution of the benefits and burdens of research across communities and populations (NASEM, 2023b). Unlike respect for persons and beneficence, the principle of justice requires inclusion of disadvantaged groups for potentially beneficial research, but this inclusion must be balanced to protect groups from being overburdened (Yearby, 2017, 2021). Here, justice particularly involves determinations of whom to study and how to characterize them related to a particular biomedical question. Justice requires that researchers do not use certain populations for research that may perpetuate the belief that there are biological differences between racial and ethnic populations, but it also requires that these populations can participate in beneficial biomedical research.
• Equity is a dimension of justice that reflects not only fairness in the distribution of the benefits and burdens of biomedical research, but also the fairness and unfairness of the social systems that affect health and in which biomedical research is conducted (Yearby, 2017, 2021; see also NASEM, 2023b). A commitment to equity calls for researchers to recognize and redress their biases regarding the populations they study and the power differentials between them. It also requires researchers to consider whether their study will further perpetuate or exacerbate unfairness in social systems.
• Inclusion is similarly a dimension of both justice and respect for persons. Inclusion requires recognition of the identities, strengths, needs, and lived experiences of members of racial and ethnic minority communities and populations, particularly as these groups see themselves, and works to create productive and respectful partnerships (NIMHD, 2024).

A COMMON UNDERSTANDING OF RACE AND ETHNICITY

The committee is comprised of members from a variety of disciplines—medicine, biomedical research, social sciences, and more. Race and ethnicity are defined and used differently across fields such as clinical practice and sociology. To assess the current use

of race and ethnicity in biomedical research and to provide guidance for future use, it was necessary to grapple with the terms “race” and “ethnicity” and develop a common understanding of these concepts as well as their relationship to racism. Although there are many existing definitions of race and ethnicity, they are not necessarily consistent with one another; for example, some definitions treat the concepts as largely interchangeable while others highlight key differences between the two concepts. Thus, the committee consulted recent National Academies reports¹ on related topics as well as the broader scientific literature to develop a shared understanding of these terms.

Defining race and ethnicity can be elusive because the concepts are dynamic, highly contextual, and multidimensional, incorporating social, political, and geographic factors. Race is, to borrow a term from computer science, an overloaded word, indicating that the word has multiple meanings that depend on the context. The social context and related factors give meaning and vibrancy to the definition of race, affecting how race is conceptualized and operates in real life (Duany, 1998; Leeman, 2018). Even the words “race” and “ethnicity” are entangled. For instance, surveys in the United States commonly provide two categories for ethnicity, Hispanic/Latino and Not Hispanic/Latino. Yet, many people with Latin American ancestry consider their race to be Hispanic/Latino and find it difficult to answer a separate question about their racial identity in contexts like the U.S. census (Leeman, 2018; OMB, 2024). In addition, many people identify with a race or ethnic category that was not among the options included in the census. It should be noted that the 2024 revisions to the Office of Management and Budget (OMB) standards on race and ethnicity have combined race and ethnicity under one question, further enmeshing these concepts (see the following section, “U.S. Office of Management and Budget Race and Ethnicity Categories”). Failing to recognize the political context of race can also have far-reaching implications. American Indian or Alaska Native, for instance, has been a single racial category on the last three U.S. censuses, but that does not recognize an inherent complexity—that each of the 574 federally recognized Tribes is an independent nation and political body (Cherokee Nation, 2024; Library of Congress, n.d.).

Race and ethnicity are also difficult to define in a research context. Different academic disciplines do not share a common history or usage of the terms, which have evolved over time (Hammonds and Herzig, 2009; Morning, 2011; Roberts, 2012). Sociology has long defined race as a social construct; however, other fields have been slower to arrive at this conclusion (Morning, 2007). Anthropology, for example, historically had a concept of race rooted in shared physical characteristics or features that has since developed to incorporate other social and cultural aspects, in turns critiquing and reinforcing race science over time (Baker, 1998; Bashkow, 2020; Gravlee, 2009; Jobson, 2019; Morning, 2007). For these reasons, the committee discussed the meaning of race and ethnicity as part of their deliberations.

In brief, the committee defined race as a sociopolitical construct conceived to describe and categorize people hierarchically (Box 2-1). Race is not valid as a biological concept; race is a dynamic social division that has been used to include or exclude

___________________

¹ Advancing Antiracism, Diversity, Equity, and Inclusion in STEMM Organizations: Beyond Broadening Participation (2023); Using Population Descriptors in Genetics and Genomics Research: A New Framework for an Evolving Field (2023); Federal Policy to Advance Racial, Ethnic, and Tribal Health Equity (2023).

individuals and groups, and it varies across historical, political, and geographic contexts. Historically, race has been used to create and justify advantage or disadvantage for some groups over others. Ethnicity can be defined as a socially and politically constructed term used to describe people from a similar national or regional background who share common cultural, historical, and social experiences. An ethnic group is often defined based on a belief in shared values, behaviors, heritage, or language. Ethnic categories also vary across historical, political, and geographic contexts (see “Ethnicity” in Chapter 5 for more). Lastly, both race and ethnicity are intertwined with the concept of ancestry—that is, a person’s origin or descent, lineage, “roots,” or heritage (NASEM, 2023a). All three concepts are part of a family of descent-associated descriptors that attempt to represent aspects of common origin (NASEM, 2023a).

The definition of race provides a connection to the systems, institutions, beliefs, and processes that underpin racism. Racism, one form of discrimination, is rooted in a belief in innate differences between groups of people. Understandings of race based on physical features, including perceived inherent and biological differences, date to colonialism and the Transatlantic slave trade and are inextricable from the history of the United States. Yet, the history of race can be traced farther into the past. Historians note that the term race (raza) was first used with this connotation during the Spanish Inquisition, when a belief in “limpieza de sangre” (blood purity) led Spanish rulers to question the loyalty of Jewish and Moorish converts to Catholicism (Fredrickson, 2002). While a comprehensive review of the origins of race and racism are beyond the scope of this report, it is important to recognize how this history has shaped scientific knowledge and medical practice to this day. Racism can manifest in various forms, including structural, institutional, and interpersonal racism (see Box 2-2). Beyond interpersonal

acts of discrimination or prejudice, racism can persist through systems and processes that reinforce inequity among racial and ethnic groups (Gee and Ford, 2011; Jones, 2000).

Among various forms of racism, scientific racism is particularly salient to this report. A pernicious ideology that sought to legitimize racism and White hegemony via the guise of pseudoscientific methods and evidence, scientific racism emerged in Western science as the ideas of evolutionary theory and the scientific impulse to categorize came together and resulted in the false notion that humans could be divided into distinct biological groups that could be ranked hierarchically. For instance, in the 19th century, notable physicians tried to identify the physical characteristics of Black individuals that could “serve to distinguish him from the white man” (Tidyman, 1826). Such spurious differences included thicker bones (Cartwright, 1851) and skulls, less sensitive nervous systems, and diseases intrinsic to darker skin (Tidyman, 1826). (See Chapter 3 for additional examples and discussion.) This type of biased research has long been embedded in the biomedical evidence base, shaping medical knowledge and practice, and it continues to affect science and medicine today. An assessment of publications from 1950 to 2000 found that biological theories of race and biological essentialism (e.g., that there are “African” and “White” genes) have evolved but persisted in biomedical and life science journals, despite the prevailing belief that the scientific community has moved away from these notions (Obasogie et al., 2015; see also Jones et al., 2024).² These purportedly “scientific” attitudes normalized the use of race to stratify groups of people and compare their risk of disease development or prognosis.

While the focus of this report is not on the origins of scientific racism or other forms of racism, the committee recognizes the importance of this history and its enduring implications. Indeed, other scholars have written extensively on these topics, and readers interested in a more comprehensive assessment of these topics, including the various forms and evolution of racism, are encouraged to consult the following reference list:

• Bonilla-Silva, E. 2021. Racism without Racists: Color-blind Racism and the Persistence of Racial Inequality in America. Lanham, MD: Rowman & Littlefield.
• Feagin, J. 2006. Systemic Racism: A Theory of Oppression. New York: Routledge.
• Fredrickson, G. M. 2002. Racism: A Short History. Princeton, NJ: Princeton University Press.
• Hammonds, E. M., and R. M. Herzig. 2009. The Nature of Difference: Sciences of Race in the United States from Jefferson to Genomics. Cambridge, MA: MIT Press.
• Hogarth, R. A. 2017. Medicalizing Blackness: Making Racial Difference in the Atlantic World, 1780-1840. Chapel Hill, NC: The University of North Carolina Press.
• Jablonski, N. G. 2012. Living Color: The Biological and Social Meaning of Skin Color. Berkeley, CA: University of California Press.

___________________

² See a recent series of articles published by The New England Journal of Medicine exploring the history of mistreatment of groups of people on the basis of their race, ethnicity, religion, gender, and physical or mental conditions: https://www.nejm.org/recognizing-historical-injustices (accessed October 16, 2024).

• Keevak, M. 2011. Becoming Yellow: A Short History of Racial Thinking. Princeton, NJ: Princeton University Press.
• Morning, A. 2011. The Nature of Race: How Scientists Think and Teach About Human Difference. Berkeley, CA: University of California Press.
• National Academies of Sciences, Engineering, and Medicine. 2023. Chapter 2 The Historical and Contemporary Context for Structural, Systemic, and Institutional Racism in the United States. Advancing Antiracism, Diversity, Equity, and Inclusion in STEMM Organizations: Beyond Broadening Participation. Washington, DC: The National Academies Press.
• Roberts, D. 2011. Fatal Invention: How Science, Politics, and Big Business Re-create Race in the Twenty-first Century. New York: The New Press.
• Saini, A. 2019. Superior: The Return of Race Science. Boston, MA: Beacon Press.
• Smedley, A., and B. Smedley. 2012. Race in North America: Origin and Evolution of a Worldview. Boulder, CO: Westview Press.

Moving from Concepts of Race and Ethnicity to Measurements

Adding to the complexity of working with race and ethnicity, definitions of these constructs can be framed in a conceptual way or, alternatively, in the context of operationalizing (measuring) them. Distinct from conceptual definitions that describe the meaning of each term (such as those in Box 2-1), operational definitions explain how the concept is measured or how a variable might be defined in a research context. An operational definition delineates how a concept, such as race, was measured—for example, that race was measured by allowing participants to choose among a set of labels to answer a question about their racial identity. Operational definitions will, of necessity, vary across studies and contexts. In the United States, one of the most well-known operational definitions of ethnicity was the former federal definition, consisting of only two categories—Hispanic/Latino or Not Hispanic/Latino, until OMB released its revisions in March 2024.

The process of moving from the identification and definition of a relevant concept to its measurement is a key step in research methodology. Researchers often do this implicitly or without thinking much about it, but it is worth making the process explicit to highlight how a single concept can be measured in different ways. Take, for example, the concept of weight. Weight can be defined as the force acting on an object due to gravity; this is an abstract definition. The concept of weight is made concrete (operationalized) in how it is measured. For example, weight is often measured by putting an object on a scale. When determining the weight of a person, weight can also be assessed through self-report, which might be an estimate of how much that person thinks they weigh. Weight can also be described in different units, such as pounds or kilograms. Each of these are measures of weight, but they will not necessarily be equally well suited to different situations.

The same kind of methodological thinking is relevant when measuring concepts like race, ethnicity, nationality, or ancestry. People’s nationality can be measured by the passport they carry, the place of their current residence, or their self-reported identification.

Similarly, ancestry can be measured in multiple ways, such as using genetic assessments of similarity or social perceptions of ancestry through genealogical records or family lore. Depending on the research question or context, some measures may be more appropriate than others. For example, genetic measures of relatedness may help identify allelic similarities (or differences) that have associations with particular illnesses, but social measures of ancestry may be more appropriate for understanding how particular health beliefs or practices are passed down in families.

Focusing on the decision to use a particular measure or set of measures helps identify many of the key debates that surround the use of race in science and medicine. Should a measure of race be self-reported or recorded as perceived by others? Is it better proxied by measuring visible, physical characteristics such as skin tone, or is it more closely related to ancestry? Should it be measured as an individual characteristic, or is it (also) a part of the social context in which people live? Today, most large-scale datasets used for health research rely on self-reported measures of race that draw on a set of categorical distinctions that have been defined by the government, but which are sometimes contested through political processes. This type of measure has been criticized for not reflecting how people actually identify themselves in their everyday lives (Atkin and Minniear, 2023) and for implying that race is something one is rather than a social position that is negotiated through interactions as part of a broader system of hierarchy and inequality. Self-reported race is also often uncritically applied, as when a measure intended to ensure inclusion during study recruitment is later treated by default as a relevant measure of “difference” during analysis (Bentz et al., 2024) or inappropriately interpreted as causal (Holland, 2001; Kaufman, 2008; VanderWeele and Robinson, 2014). Taking these critiques seriously, more recent work has focused on understanding race and racialization not as a static individual characteristic but as a dynamic process that hinges on racial appearance or treatment by others (Rose, 2023; Saperstein and Penner, 2012; Vargas et al., 2019), has been built up historically (Hudson, 2021; Nagata et al., 2024; Wrigley-Field, 2024), and is reflected in highly unequal contexts (Brown and Homan, 2024; O’Brien et al., 2020; Tan et al., 2022; Torche and Sirois, 2019). Although this does not render measures of racial self-identification irrelevant, it suggests their use should be carefully considered and explicitly matched to a particular purpose.

U.S. OFFICE OF MANAGEMENT AND BUDGET RACE AND ETHNICITY CATEGORIES

Background of the OMB Categories and the Census

In 1977, the OMB Statistical Policy Directive No. 15 created federal standards for reporting race and ethnicity data to provide information needed for enforcing civil rights laws (OMB, 1977, 1997). The four race categories and two ethnicity categories soon became known popularly as “the OMB categories” (see Box 2-3). Directive 15 was updated in 1997 to include five race categories, two ethnicity categories, and the option to report more than one race. In 2024, the OMB categories were updated to include seven combined race and ethnicity categories. Directive 15 provides a minimum

set of categories for federal agencies to use in collecting and reporting data on race and ethnicity. The categories have been widely used across government agencies to the extent that the categories have become ubiquitous and synonymous with the conception of race in the United States. Appearing in multiple contexts, the OMB categories have purposes across federal agencies and sectors, including in the census and for inclusion purposes in federally funded research.

Every 10 years, the U.S. Census Bureau collects information about the country’s population. The OMB categories and census categories are often believed to be one and the same, but they are, in fact, distinct. The Census Bureau must use the OMB categories at a minimum but is ultimately accountable to Congress. As such, the census

may also include additional categories or questions, but only as needed by government agencies or as legislated by Congress. For instance, the inclusion of “Some Other Race” was mandated by Congress (P.L. 109–108; U.S. Census Bureau, 2021). To provide recommendations for changes to the content of census questions, the Census Bureau runs content tests to evaluate different wording and formatting for the questions (e.g., Mathews et al., 2017).

The census offers a prime example of how the categories have evolved over U.S. history. “The list of racial groups on the U.S. Census, for example, has changed nearly every decade since the first enumeration in 1790, with categories like ‘mulatto,’ ‘Mexican,’ and ‘Hindu’ appearing and disappearing (Lee, 1993; Prewitt, 2005)” (NASEM, 2023a, p. 73;

see also Figure 2-2 in NASEM, 2023a). Further demonstrating how racial and ethnic categories evolve, Italian, Jewish, and Irish population groups were commonly considered to be racial groups in the early 20th century (NASEM, 2023a; see also Jacobson, 1999). The names of categories also evolve, reflecting changing attitudes and politics of the day; for instance, “colored” became “Black,” then “Negro” became common for a time, and the term later evolved to “Black or African American” in the 2020 Census.³ Thus, race and ethnicity categories are not static, and, indeed, OMB convened the Federal Interagency Technical Working Group on Race and Ethnicity Standards to revise Directive 15. In 2023, OMB held a public comment period on initial proposals for revisions. In 2024, OMB published revisions to Directive 15 that included adding the geographically defined category of Middle Eastern or North African, combining the race and ethnicity questions into a single question stem, requiring data collection for more detailed subcategories, and updating specific terminology to provide greater consistency and clarity of the minimum category definitions (OMB, 2024).

Requiring the collection of more detailed subcategories unless an agency applies for an exemption marks a significant departure from the previous standards. For each of the minimum race and ethnicity categories, the OMB standards include six detailed subcategories, based on the largest subpopulations in the United States, and an option to select “Another group” or complete a write-in field. The policy suggests offering a write-in box whenever possible to enable greater self-identification. It should be noted that there are not standard subgroups for the category American Indian or Alaska Native, so offering a write-in field will be necessary. As in previous iterations of the OMB standards, any additional detailed categories used must be able to “roll up” into the set of seven minimum categories. Although effects of this change remain to be seen, it is an attempt to respond to a core criticism of the minimum categories—that they aggregate many different groups and fail to capture information distinct to various populations. For example, the Native Hawaiian and Pacific Islander category represents over 20 ethnicities, each with its own language, history, and culture.

Nomenclature for American Indians and Alaska Natives

The minimum OMB categories include American Indian or Alaska Native, but Indigenous people of the United States are unique among racial or ethnic minority groups in the United States because Tribes are sovereign nations with distinct legal status. Related considerations that affect collaborations with Tribal nations for research purposes are covered in Chapter 4. It is also important for health practitioners, researchers, and the public to understand the terms used to designate and distinguish between the First Peoples found in northern North America. Indigenous is used broadly and will be used in this report when speaking about people who had been on this continent for millennia before European colonization and during early contact with European colonizers. Because relationship to place is foundational to the concept of indigeneity, the discussion of indigeneity in this report focuses primarily on the U.S. context. At times

___________________

³ https://www.pewresearch.org/social-trends/feature/what-census-calls-us/ (accessed August 27, 2024).

Indigenous will be used inclusive of First Peoples across Canada and the United States. Indian is the actual legal term in both Canada⁴ and the United States today. In Canada, there are three groups denoting populations indigenous to what is now Canada: First Nations, Metis, and Inuit. In the United States, in modern vernacular the term Native encompasses any persons with close heritage and lineage to current or historical Tribes, even though—through the vast purposeful interference by the U.S. federal government—they may have lost any connections to their heritage groups and are not enrolled members of their Tribes. In fact, they may not know their Tribes. And in fact, their Tribes may no longer exist in the eyes of Congress. Native will also encompass self-identified Indigenous peoples whether enrolled or not. There are finely nuanced distinctions among these terms. The term “American Indian”—and, at times, “American Indian or Alaska Native” (AIAN)—indicates enrolled members of one of the 574 federally recognized Tribes in the United States. Today many AIAN choose to directly be identified by their Tribe’s name rather than a blanket “global” indication of Indigeneity. Even this can be confusing as anglicized Tribal names are increasingly being rejected for the names in the Tribes’ own language. Some examples are Dine’ as opposed to Navajo, A:Shiwi instead of Zuni, and Anishinaabe not Chippewa. Thus, it is important to take into account a Tribe’s and an individual’s descriptor preferences when referring to a population.

DEFINING BIOMEDICAL RESEARCH

Biomedical research is by nature broad and multidisciplinary, drawing on expertise across fields of biology, medicine, epidemiology, social sciences, behavioral sciences, and many other disciplines. In the context of this report, biomedical research is scientific research across biological, social, and behavioral disciplines that pertains to human health, ranging from preclinical methods to population health. The committee’s definition is intentionally broad, encompassing many related subfields—human physiology, clinical epidemiology, biomedical informatics, comparative effectiveness research, and numerous others (see Box 2-4). This expansive definition is intended to be inclusive and to avoid reinforcing scientific silos among disciplines while emphasizing areas of research that are most germane to human health and so may involve race and ethnicity. Importantly, the behavioral, social, and biomedical sciences often influence one another, and much of biomedical research operates at their intersection. Thus, this definition acknowledges the interaction of biological and social factors. Of note, although genetics and genomics research may fall under the umbrella of biomedical research, the committee did not focus specifically on these fields because they were addressed by a 2023 National Academies report (NASEM, 2023a). See Chapter 4 for a brief summary.

Biomedical research operates along a translational spectrum, ranging from basic or discovery science to translational research to clinical trials and implementation science. This research gives rise to a range of medical applications including pharmaceuticals, biotechnology, diagnostics, surgical interventions, clinical tools, and medical devices.

___________________

⁴ Indian means a person who pursuant to this Act is registered as an Indian or is entitled to be registered as an Indian; (Indien), Indian Act R.S.C., 1985, c. I-5, Definitions.

These would all be included under this umbrella of biomedical research. The extent to which this encompasses early-stage bench science, including work with mammalian animal models, is more ambiguous because some preclinical methods fall within scope of this definition while others do not. For example, human-derived cell lines (e.g., HeLa cells), human organoids, and preclinical computational modeling using databases of human samples would all have bearing on human health and might be considered within scope. In contrast, non-human experiments (e.g., with C. elegans, a common model organism) would not likely have direct relevance to issues of race, ethnicity, and human health. Therefore, this report is primarily concentrated on the bulk of research downstream of preclinical models and further along on the translational spectrum.

Artificial intelligence (AI) is increasingly being used in biomedical research applications and in medicine as AI underpins a growing set of data science methods and clinical decision tools that are intended to aid medical professionals in their care of patients. The tools are employed in medical imaging, surgery, health monitoring, personalized treatment, and disease diagnostics (Raz et al., 2022; Varghese et al., 2024). AI-based methods are also used preclinically in, for example, drug development and genomics analysis. Although not the primary focus of this report, AI will likely play an increasingly large role in biomedical research, health care, and other related sectors, and is, thus, considered in the relevant research contexts throughout the report. In addition, the committee notes that the roles and impacts of AI in the clinic and biomedical research are a rapidly moving target. Though an in-depth examination of AI in health care is beyond the scope of this report, investigation of the impact of AI in this space is ongoing (e.g., Lee et al., 2024; Li et al., 2022; Ratwani et al., 2024).

The consideration of medical devices and medical instrumentation, in many cases, is distinct from the evaluation of how clinical diagnostic and decision-making tools incorporate racial and ethnic biases. Many of these devices, especially ones that employ optical sensors, use light readings to make assessments but historically have often not accounted for how optical physics interacts with attributes that are typically associated with race and ethnicity, such as skin pigmentation. Notably, these differences can affect how accurately the devices work and the outputs that the devices deliver. Medical

devices are discussed in several places in the report and offer a specific use case for the committee’s recommendations.

INTERACTION OF SOCIAL CONTEXT AND BIOLOGY WHEN USING RACE AND ETHNICITY IN BIOMEDICAL RESEARCH

As discussed in earlier sections of this chapter, race and ethnicity are dynamic, contextual, and difficult to define. In addition, biomedical research is intrinsically complex and interdisciplinary. The intersection of these domains presents unique challenges and potential pitfalls, such as oversimplification and misinterpretation of research findings.

Studies of human health involve social and environmental context in addition to biology. Race and ethnicity have long been assumed to be useful approximations of social context and are often used as proxies for other variables, ranging from socioeconomic status and environmental exposures to experiences of discrimination. It is widely understood that these social factors affect health, but they can be easily overlooked or obscured by an excess focus on race and ethnicity. It is important to recognize that race itself does not cause health differences; rather, factors such as social determinants of health, racism, and discrimination affect biological systems and health (see Chapter 3, section “Health Disparities and the Study of Racism” and Chapter 5). This complex interaction of biological and social factors can be difficult to tease apart.

Moreover, deep-seated misconceptions about race and ethnicity continue to affect science today and make these issues all the more challenging. The historical tendency in Western science to use categorization to understand the world essentialized race and reinforced the erroneous idea that people could be grouped into distinct categories (Hammonds and Herzig, 2009; Morning, 2011). Exacerbating the confusion, differences in physical appearance—such as skin color, which is commonly viewed as synonymous with race—are partially explained through genetic inheritance. But equating this biological phenotype with race is a misconception. Skin color is, in fact, a complex trait resulting from the contributions of many genes and the environment. Variation in skin color does not follow a clear distribution based on racial and ethnic categories (Jablonski, 2021), and the seeming connection between race and biology falls apart completely when examining complex traits and genetic variation (see Chapter 5, sections “Skin Color and Pigmentation” and “Genetic Markers and Ancestry”). Genetics research has made clear that human genetic variation is continuous, refuting the existence of distinct human races (Duello et al., 2021; Jorde and Wooding, 2004; NASEM, 2023a; see also Fullwiley, 2024 and Nelson, 2016). In addition, though there may be epidemiological differences in disease prevalence, fundamental molecular and cellular mechanisms are the same across racial and ethnic groups and are, moreover, often shared across species. Despite the accumulation of these lines of evidence over decades, it has been difficult to root out old beliefs, including the misattribution of biological differences to race, and their lasting impact on science and medicine.

This chapter began by laying a foundation of ethical and scientific principles for the committee’s work. It defined race and ethnicity and introduced the OMB system for collection of race and ethnicity data in the United States. The chapter concluded with a discussion of the nuances that arise when bringing together biomedical research with race

and ethnicity. Subsequent chapters will explore this complexity in more detail. The next chapter examines current uses of race and ethnicity in biomedical research, beginning with a general overview and then assessing examples throughout the areas of race correction in clinical practice, medical devices, secondary data use, and emerging AI applications.

REFERENCES

Freedman, B. 1987. Scientific value and validity as ethical requirements for research: A proposed explication. IRB: Ethics & Human Research 9(6):7–10.

Gee, G. C., and C. L. Ford. 2011. Structural racism and health inequities: Old issues, new directions. Du Bois Review: Social Science Research on Race 8(1):115–132.

George, K., A. Batterham, and I. Sullivan. 2003. Validity in clinical research: A review of basic concepts and definitions. Physical Therapy in Sport 4(3):115–121.

Gravlee, C. C. 2009. How race becomes biology: Embodiment of social inequality. American Journal of Physical Anthropology 139(1):47–57.

Hammonds, E. M., and R. M. Herzig. 2009. The nature of difference: Sciences of race in the United States from Jefferson to genomics. Cambridge, MA: MIT Press.

HEW (U.S. Department of Health, Education, and Welfare). 1979. The Belmont report. Washington, DC: Government Printing Office. https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/read-the-belmont-report/index.html (accessed August 2, 2024).

Holland, P. 2001. The false linking of race and causality: Lessons from standardized testing. Race and Society 4:219–233.

Hudson, D. 2021. Achieving health equity by addressing legacies of racial violence in public health practice. Annals of the American Academy of Political and Social Science 694(1):59–66.

Jablonski, N. G. 2021. Skin color and race. American Journal of Physical Anthropology 175(2):437–447.

Jacobson, M. F. 1999. Whiteness of a different color: European immigrants and the alchemy of race. Cambridge, MA: Harvard University Press.

Jobson, R. C. 2019. Race: Oxford Bibliographies. https://www.oxfordbibliographies.com/display/document/obo-9780199766567/obo-9780199766567-0206.xml (accessed August 29, 2024).

Jones, C. P. 2000. Levels of racism: A theoretic framework and a gardener’s tale. American Journal of Public Health 90(8):1212–1215.

Jones, D. S., E. Hammonds, J. P. Gone, and D. Williams. 2024. Explaining health inequities—The enduring legacy of historical biases. New England Journal of Medicine 390(5):389–395.

Jorde, L. B., and S. P. Wooding. 2004. Genetic variation, classification and “race.” Nature Genetics 36(S11):S28–S33.

Kaufman, J. S. 2008. Epidemiologic analysis of racial/ethnic disparities: Some fundamental issues and a cautionary example. Social Science & Medicine 66(8):1659–1669.

Krieger, N. 2014. Discrimination and health inequities. International Journal of Health Services 44(4):643–710.

Lee, S. M. 1993. Racial classifications in the U.S. census: 1890–1990. Ethnic and Racial Studies 16:75–94.

Lee, S. W., M. Morcos, D. W. Lee, and J. Young. 2024. Demographic representation of generative artificial intelligence images of physicians. JAMA Network Open 7(8):e2425993–e2425993.

Leeman, J. 2018. Becoming Hispanic: The negotiation of ethnoracial identity in US census interviews. Latino Studies 16(4):432–460.

Li, F., N. Ruijs, and Y. Lu. 2022. Ethics & AI: A systematic review on ethical concerns and related strategies for designing with AI in healthcare. AI 4(1):28–53.

Library of Congress. n.d. U.S. Reports: Worcester v. The state of Georgia, 31 U.S. (6 pet.) 515 (1832). https://www.loc.gov/item/usrep031515/ (accessed August 26, 2024).

Mathews, K., J. Phelan, N. A. Jones, S. Konya, R. Marks, B. M. Pratt, J. Coombs, and M. Bentley, 2017. 2015 National Content Test: Race and eth nicity analysis report. Washington, DC: U.S. Census Bureau. https://www.census.gov/programs-surveys/decennial-census/decade/2020/planning-management/plan/final-analysis/2015nct-race-ethnicity-analysis.html (accessed August 2, 2024).

Morning, A. 2007. “Everyone knows it’s a social construct”: Contemporary science and the nature of race. Sociological Focus 40:436–454.

Morning, A. 2011. The nature of race: How scientists think and teach about human difference. Oakland, CA: University of California Press.

Nagata, D. K., J. H. J. Kim, and J. P. Gone. 2024. Intergenerational transmission of ethnoracial historical trauma in the United States. Annual Review of Clinical Psychology 20(1):175–200.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2017. Fostering integrity in research. Washington, DC: The National Academies Press.

NASEM. 2019. Reproducibility and replicability in science. Washington, DC: The National Academies Press.

NASEM. 2023a. Using population descriptors in genetics and genomics research: A new framework for an evolving field. Washington, DC: The National Academies Press.

NASEM. 2023b. Toward equitable innovation in health and medicine: A framework. Washington, DC: The National Academies Press.

NASEM. 2023c. Federal policy to advance racial, ethnic, and tribal health equity. Washington, DC: The National Academies Press.

NASEM. 2023d. Advancing Antiracism, Diversity, Equity, and Inclusion in STEMM Organizations: Beyond Broadening Participation. Washington, DC: The National Academies Press.

Nelson, A. 2016. The social life of DNA: Race, reparations, and reconciliation after the genome. Boston, MA: Beacon Press.

NHGRI (National Human Genome Research Institute). n.d. The human genome project. https://www.genome.gov/human-genome-project (accessed September 9, 2024).

NIH (National Institutes of Health). 2023. Enhancing reproducibility through rigor and transparency. https://grants.nih.gov/policy/reproducibility/index.htm (accessed January 26, 2023).

NIMHD (National Institute on Minority Health and Health Disparities). 2024. Diversity and inclusion in clinical trials. https://www.nimhd.nih.gov/resources/understanding-health-disparities/diversity-and-inclusion-in-clinical-trials.html (accessed August 28, 2024).

Obasogie, O. K., J. N. Harris-Wai, K. Darling, and C. Keagy. 2015. Race in the life sciences: An empirical assessment, 1950-200. Fordham Law Review 83(6):3089–3114.

O’Brien, R., T. Neman, N. Seltzer, L. Evans, and A. Venkataramani. 2020. Structural racism, economic opportunity and racial health disparities: Evidence from U.S. Counties. SSM Population Health 11:100564.

OMB (U.S. Office of Management and Budget). 1977. Directive No. 15: Race and ethnic standards for federal statistics and administrative reporting. https://spd15revision.gov/content/spd15revision/en/history/1977-standards.html (accessed January 8, 2025).

OMB. 1997. Revisions to the standards for the classification of federal data on race and ethnicity. https://www.whitehouse.gov/wp-content/uploads/2017/11/Revisions-to-the-Standards-for-the-Classification-of-Federal-Data-on-Race-and-Ethnicity-October30-1997.pdf (accessed August 2, 2024).

OMB. 2024. Revisions to OMB’s Statistical Policy Directive No. 15: Standards for maintaining, collecting, and presenting federal data on race and ethnicity. Federal Register. https://www.federalregister.gov/documents/2024/03/29/2024-06469/revisions-to-ombs-statistical-policy-directive-no-15-standards-for-maintaining-collecting-and (accessed August 2, 2024).

Prewitt, K. 2005. Racial classification in America: Where do we go from here? Daedelus 134(1):5–17.

Priest, N., and D. R. Williams. 2018. Racial discrimination and racial disparities in health. In B Major, J. F. Dovidio, and B. G. Link (eds.), The Oxford handbook of stigma, discrimination, and health. New York: Oxford University Press. Pp. 163–182.

Ratwani, R. M., D. Classen, and C. Longhurst. 2024. The compelling need for shared responsibility of AI oversight: Lessons from health IT certification. JAMA 332(10):787–788.

Raz, M., T. C. Nguyen, and E. Loh. 2022. Artificial intelligence in medicine: Applications, limitations and future directions. New York: Springer Nature.

Reskin, B. 2012. The race discrimination system. Annual Review of Sociology 38(1):17–35.

Resnik, D. 2023. Openness in scientific research: A historical and philosophical perspective. Journal of Open Access Law 11(1):132.

Roberts, D. 2012. Fatal invention: How science politics and big business re-create race in the twenty-first century. New York: New Press.

Rose, E. K. 2023. A constructivist perspective on empirical discrimination research. Journal of Economic Literature 61(3):906–923.

Roundtable on Community Change. 2017. Glossary for understanding the dismantling structural racism/promoting racial equity analysis. https://www.aspeninstitute.org/wp-content/uploads/files/content/docs/rcc/RCC-Structural-Racism-Glossary.pdf (accessed November 25, 2024).

Saperstein, A., and A. M. Penner. 2012. Racial fluidity and inequality in the United States. American Journal of Sociology 118(3):676–727.

Tan, S. B., P. deSouza, and M. Raifman. 2022. Structural racism and COVID-19 in the USA: A county-level empirical analysis. Journal of Racial and Ethnic Health Disparities 9(1): 236–246.

Tidyman, P. 1826. Art. VI. A sketch of the most remarkable diseases of the Negroes of the Southern States, with an account of the method of treating them, accompanied by physiological observations. Philadelphia Journal of the Medical and Physical Sciences (1820–1827) 3(6):306.

Torche, F., and C. Sirois. 2019. Restrictive immigration law and birth outcomes of immigrant women. American Journal of Epidemiology 188(1):24–33.

U.S. Census Bureau. 2021. 2020 Census frequently asked questions about race and ethnicity. https://www.census.gov/programs-surveys/decennial-census/decade/2020/planning-management/release/faqs-race-ethnicity.html (accessed November 26, 2024).

VanderWeele, T. J., and W. R. Robinson. 2014. On the causal interpretation of race in regressions adjusting for confounding and mediating variables. Epidemiology 25(4):473–484.

Vargas, E., M. Juarez, L. Cacari-Stone, and N. Lopez. 2019. Critical “street race” praxis: Advancing the measurement of racial discrimination among diverse Latinx communities in the U.S. Critical Public Health 31:1–11.

Varghese, C., E. M. Harrison, G. O’Grady, and E. J. Topol. 2024. Artificial intelligence in surgery. Nature Medicine 30(5):1257–1268.

Williams, D. R., J. A. Lawrence, and B. A. Davis. 2019. Racism and health: Evidence and needed research. Annual Review of Public Health 40:105–125.

Wrigley-Field, E. 2024. Stolen lives: Redress for slavery’s and Jim Crow’s ongoing theft of lifespan. Russell Sage Foundation Journal of the Social Sciences 10(2):88–112.

Yearby, R. 2017. Exploitation in medical research: The enduring legacy of the Tuskegee syphilis study. Case Western Reserve Law Review 67(4):1171–1276.

Yearby, R. 2020. Structural racism and health disparities: Reconfiguring the social determinants of health framework to include the root cause. Journal of Law, Medicine & Ethics 48(3):518–526.

Yearby, R. 2021. Race based medicine, colorblind disease: How racism in medicine harms us all. American Journal of Bioethics 21(2):19–27.

This page intentionally left blank.