A Consideration of Racism in Pediatric Epidemiologic Studies
2021; Elsevier BV; Volume: 239; Linguagem: Inglês
10.1016/j.jpeds.2021.08.004
ISSN1097-6833
AutoresKarl Kuban, Renée Boynton‐Jarrett, Timothy Heeren, T. Michael O’Shea,
Tópico(s)Migration, Health and Trauma
ResumoSee related article in 236:4. In epidemiologic research, race is considered often as a covariate of an outcome because such is customary practice, is seemingly easy to measure, is stable over time, and often is associated with variation in the outcome.1Jones C.P. Invited commentary: "race," racism, and the practice of epidemiology.Am J Epidemiol. 2001; 154 (discussion 305-6): 299-304Crossref PubMed Scopus (320) Google Scholar The complexities and limitations in using race, a socially constructed way of grouping people for these purposes, however, perceptively elaborated 20 years ago,1Jones C.P. Invited commentary: "race," racism, and the practice of epidemiology.Am J Epidemiol. 2001; 154 (discussion 305-6): 299-304Crossref PubMed Scopus (320) Google Scholar unmasks a dogma and challenges the neutrality of the traditional rationale for using race as a covariate. Jones underscores the points that definition and delineation of race are highly heterogeneous, contextually based, and subject to change over time, and race's association with outcome is difficult to distinguish from a broad2Zysk G. Bruck W. Gerber J. Bruck Y. Prange H.W. Nau R. Anti-inflammatory treatment influences neuronal apoptotic cell death in the dentate gyrus in experimental pneumococcal meningitis.J Neuropathol Exp Neurol. 1996; 55: 722-728Crossref PubMed Scopus (149) Google Scholar range of underlying racist policies and practices. Yet, because analyzing the underlying social and structural causes for health outcome inequities, a practice suggested by Jones, may both be difficult and influenced by preconceived biases, epidemiologic studies often simply adjust or control for race without exploring its foundation. Such rote adjustment both impedes efforts to understand better the basis for racial inequities, and this practice may reinforce the false notion of biological determinism.3Chandran A. Knapp E. Liu T. Dean L.T. A new era: improving use of sociodemographic constructs in the analysis of pediatric cohort study data.Pediatr Res. 2021; ([Online ahead of print])https://doi.org/10.1038/s41390-021-01386-wCrossref PubMed Scopus (4) Google Scholar Jones's seminal commentary includes a set of recommendations for research practice to address these concerns and make strides to document, measure, and understand the impact of racism. Many pediatric epidemiologic researchers, however, have been slow to recognize and accept the impact of inherent biases, or to institute recommendations proffered by Jones, and rather persist in "customary" analytic practices. We have been guilty of this in a number of our own studies.4Leviton A. Joseph R.M. Allred E.N. O'Shea T.M. Kuban K.K.C. Antenatal and neonatal antecedents of learning limitations in 10-year old children born extremely preterm.Early Hum Dev. 2018; 118: 8-14Crossref PubMed Scopus (3) Google Scholar,5Leviton A. Joseph R.M. Allred E.N. O'Shea T.M. Taylor H.G. Kuban K.K.C. Antenatal and neonatal antecedents of executive dysfunctions in extremely preterm children.J Child Neurol. 2018; 33: 198-208Crossref PubMed Scopus (5) Google Scholar In the context of growing efforts to confront and dismantle racism in US society, including racism in healthcare research,6Boyd R.W. Lindo E.G. Weeks L.D. McLemore M.R. On racism: a new standard for publishing on racial health inequities. Health Affairs Blog.https://www.healthaffairs.org/do/10.1377/hblog20200630.939347/full/Date: 2020Date accessed: August 26, 2021Google Scholar,7Krieger N. Boyd R.W. De Maio F. Maybank A. Medicine's privileged gatekeepers: producing harmful ignorance about racism and health. Health Affairs Blog.https://www.healthaffairs.org/do/10.1377/hblog20210415.305480/full/Date: 2021Date accessed: August 26, 2021Google Scholar we wish to re-sensitize clinical epidemiologic–child health researchers to the implications of using race as a covariate (ie, not the primary focus of antecedent risk) in analyses. Drawing on examples from our work within the ELGAN (Extremely Low Gestational Age Newborn) Study as a springboard, we reflect in retrospect that perhaps the most powerful impetus for inclusion of race when considering neurologic and cognitive outcome in our studies was its apparent association with these outcomes. In attempting to address potential confounding when considering antecedent risks related to being born extremely preterm with later cognitive outcomes in our analyses, we often included race, as well as a handful of other variables encompassing socioeconomic status (SES), usually maternal education, marital status, and age, as well as source of medical insurance. Conducting analyses in this way often yielded residual disparate risk of adverse outcome associated with Black race compared with White subjects, the reference group. We wish to highlight a few of the fallacies in our decision to approach analyses in this manner. These include overly simplistic assumptions that a few measures of SES are sufficient to account truly for confounding of the association between race and outcome; the mistaken and biased consideration of evaluating odds or risks by assigning White participants as the "norm" reference group, comparison group; and failure to recognize that simply reporting residual risk associated with race does a disservice to understanding fundamental risk–outcome associations and that ignoring outcome disparities by race masks and diminishes an opportunity, although imperfect, to document the social underpinnings of risk. The first fallacy relates to the assumption that including a handful of SES variables in analyses, consisting of maternal education, health insurance source, marital/partner status, and maternal age, would suffice to represent the breadth of socioeconomic-related factors that might confound the association between race and outcome. To the contrary, research indicates that the potential for social-related characteristics to influence outcome is much broader and includes factors such as unequal socioeducational opportunities,8Teasley M.L. Education policy and outcomes within the African American population.Soc Work Public Health. 2019; 34: 61-74Crossref PubMed Scopus (1) Google Scholar sociomedical inequities,9Kirby J.B. Kaneda T. 'Double jeopardy' measure suggests blacks and hispanics face more severe disparities than previously indicated.Health Aff (Millwood). 2013; 32: 1766-1772Crossref PubMed Scopus (17) Google Scholar, 10Odonkor C.A. Esparza R. Flores L.E. Verduzco-Gutierrez M. Escalon M.X. Solinsky R. et al.Disparities in health care for black patients in physical medicine and rehabilitation in the united states: a narrative review.PM R. 2021; 13: 180-203Crossref PubMed Scopus (22) Google Scholar, 11Rivera L.M. Ethnic-racial stigma and health disparities: from psychological theory and evidence to public policy solutions.J Soc Issues. 2014; 70: 198-205Crossref PubMed Scopus (13) Google Scholar sociohousing inequalities,12Bailey Z.D. Krieger N. Agénor M. Graves J. Linos N. Bassett M.T. Structural racism and health inequities in the USA: evidence and interventions.Lancet. 2017; 389: 1453-1463Abstract Full Text Full Text PDF PubMed Scopus (1747) Google Scholar socioemployment exclusions,12Bailey Z.D. Krieger N. Agénor M. Graves J. Linos N. Bassett M.T. Structural racism and health inequities in the USA: evidence and interventions.Lancet. 2017; 389: 1453-1463Abstract Full Text Full Text PDF PubMed Scopus (1747) Google Scholar sociocriminal justice inequalities,13Hetey R.C.E.J. The numbers don't speak for themselves: racial disparities and the persistence of inequality in the criminal justice system.Curr Dir Psychol Sci. 2018; 27: 183-187Crossref Scopus (81) Google Scholar sociotransportation limitations,14Joyce N.R. Pfeiffer M.R. Zullo A.R. Ahluwalia J. Curry A.E. Individual and geographic variation in driver's license suspensions: evidence of disparities by race, ethnicity and income.J Transp Health. 2020; 19: 100933Crossref PubMed Scopus (4) Google Scholar sociovoting disenfranchisement,15Cottrell D. Herron M.C. Rodriguez J.M. Smith D.A. Mortality, incarceration, and African American disenfranchisement in the contemporary United States.Am Politics Res. 2018; 47: 195-237Crossref Scopus (11) Google Scholar and sociostigmatization in the market place,16Schreer G. Smith S. Thomas K. "Shopping while black": examining racial discrimination in a retail setting.J Appl Soc Psychol. 2009; 39: 1432-1444Crossref Scopus (62) Google Scholar among others. Beyond the inequities are the resulting biological consequences of race-based discrimination, to include epigenomic changes that enhance risk of schizophrenia, bipolar disorder, and asthma,17Barcelona de Mendoza V. Huang Y. Crusto C.A. Sun Y.V. Taylor J.Y. Perceived racial discrimination and DNA methylation among African American women in the InterGEN study.Biol Res Nurs. 2018; 20: 145-152Crossref PubMed Scopus (41) Google Scholar as well as brain changes associated with stress that predict a greater risk of mental health disturbance.18Berger M. Sarnyai Z. "More than skin deep": stress neurobiology and mental health consequences of racial discrimination.Stress. 2015; 18: 1-10Crossref PubMed Scopus (269) Google Scholar In a recent ELGAN study analysis example, we sought to analyze the role of perinatal intracranial hemorrhage as a risk for cognition outcomes at 10 years and considered race as a covariate.19Campbell H. Check J. Kuban K.C.K. Leviton A. Joseph R.M. et al.Neonatal cranial ultrasound findings among infants born extremely preterm: associations with neurodevelopmental outcomes at 10 years of age.J Pediatr. 2021; 237: 197-205.e4Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar We found that race was associated with adverse outcomes when we considered only maternal education, health insurance source, and marital/partner status but disappeared once we simply added food insecurity to the set of SES variables. Another recent analysis in a longitudinal study of 1055 maternal–child dyads identified risk factors associated with cognitive performance at 4-6 years. Twenty-four variables were associated with cognitive outcomes. For all but 2 of the 24 exposures, Black children were less likely to be exposed to factors that favored better outcome. Although the unadjusted mean difference in test scores by race was 13.3 points, once the model was adjusted for the 24 exposures, there were no differences in cognitive test scores as a function or race.20LeWinn K.Z. Bush N.R. Batra A. Tylavsky F. Rehkopf D. Identification of modifiable social and behavioral factors associated with childhood cognitive performance.JAMA Pediatr. 2020; 174: 1063-1072Crossref PubMed Scopus (21) Google Scholar Second, in several ELGAN study analyses, the investigators assessed the potential significance of race as a covariate of cognitive and other neurologic outcomes by considering ORs of race categories on outcome risk. In so doing, investigators selected as reference the least at-risk group, in this case being White, as the reference of 1. The mistaken implication of setting White as the normal reference, a conventional approach taken in many epidemiologic studies, tends to perpetuate the notion that White is the measuring stick of normal against which being Black, or other non-White group, is measured and obscures the more fundamental social and experiential underpinning of the observed differences. Third, although it is clear that using race as a risk category tends to obscure the powerful socially and structural underpinnings of risk, a tension necessarily remains between use of overly simplistic race–outcome categorization with minimizing the presence and impact of outcome inequities. As many as one-third of publications in major pediatrics journals that report race and/or ethnicity data do not discuss socioeconomic or structural factors leading to the outcome.21Walsh C. Ross L.F. Are minority children under- or overrepresented in pediatric research?.Pediatrics. 2003; 112: 890-895Crossref PubMed Scopus (42) Google Scholar To institute social changes required to achieve equity, structural racism and systemic factors that perpetuate inequities must be acknowledged. So, what do we do about this in relation to epidemiologic studies in which the primary risk, outcome, or analyses are not focused on race? First, do not consider race as a covariate or as a specific risk factor for the outcome without specifying the reason for its use and the social and/or the structural factors for which race may be function as a proxy. Second, in place of using race as a covariate, use broader and more inclusive constructs of social, socioeconomic, socioeducational, sociomedical, and other social-based risks, which are likely to explain race-based differential risks. Third, avoid the implicit assumption of a normative group by not setting a particular race as the default referent in all analyses. An alternative analytic approach in regression modeling is to implement population-average coding as the reference and then to look at deviation toward elevated or reduced risk relative to this reference group by other groups. This moves away from making any assumptions about normality of a group. Fourth, in relation to when residual differences occur among racial/ethnic groups, provide discussion and interpretation of findings beyond report of statistical findings.22Ross P.T. Hart-Johnson T. Santen S.A. Zaidi N.L.B. Considerations for using race and ethnicity as quantitative variables in medical education research.Perspect Med Educ. 2020; 9: 318-323Crossref PubMed Scopus (32) Google Scholar Adoption of these strategies for determining when, why, and how to include race in research will help to ethically and thoughtfully contextualize findings associated with race in epidemiologic analyses and should enhance efforts to identify solutions to the well documented racial inequities in health outcomes in the US. It will also minimize direct or indirect implications of a biological basis for this social construct. Diversity, Equity, and Inclusion in The Journal of PediatricsThe Journal of PediatricsVol. 236PreviewDiscrimination permeates everyday life and underscores the need to create a culture of universal respect for each person. The Journal of Pediatrics embraces this culture and is committed to establishing trust and transparency in the scientific process. In the pursuit of knowledge, we must avoid as much as possible discrimination, exclusivism, and implicit bias. We also need to work toward unambiguous implementation of Diversity, Equity, and Inclusion (DEI) principles throughout our editorial processes. Full-Text PDF
Referência(s)