Thread by @irisdelistmo

I just finished reading Racism: Science & Tools, and thought I'd share a summary of Chapter 12: Why Epidemiologists Must Reckon with Racism.

Our field cannot continue to disengage with racism; we need to dive in wherever we are. #epitwitter #racismisapublichealthcrisis (1/n)

Dr. Nancy Krieger gives two overarching reasons why epidemiologists cannot afford to ignore racism:

1) Racism impacts people’s health.

2) Racism impacts the science of epidemiology

(2/n)

1) Racism impacts health. Each and every generation confronts racism as it manifests in their times, so the work of epidemiologists is to make widely known the profound toll of racial injustice on disease, disability, and death, at both the individual and population level. (3/n)

2) Racism impacts the science of epi. To strive for rigor and avoid error, we must know the history of scientific racism & it's adverse effects on health. Causal analysis of the impact of injustice on health requires empirical testing of hypotheses using VALID methods. (4/n)

What happens when epi ignores the realities of racism and its impact on both people and science?

Krieger gives two case studies:
1) Cancer mortality among African Americans
2) Diabetes among American Indians

(5/n)

Case 1 -- Cancer mortality among African Americans: how categorization of race resulted in errors in monitoring trends in population health and inequities

How has race been measured since the first U.S. Census in 1790?

Interactive timeline: census.gov/data-tools/dem…

(6/n)

This timeline shows that the three most ‘stable’ racial categories since 1790 :

i. White (free white males, free white females --> White)
ii. Black (slaves --> black, negro)
iii. Indian

Our race/ethn data for health statistics are derived from census categories. (7/n)

After 1950, U.S. health stats were categorized as white/non-white – which made it impossible to explicitly deal with racism since all POC were lumped together, defined by who they were not. (8/n)

In 1954, NCI report on cancer epidemiology – reported incidence rates (not just the usual mortality rates), which was novel in the field.

Race was presented as white/non-white. This report would then set the precedent and mold for reporting race data in health. (9/n)

In 1972, researchers from Howard U, recognizing that addressing racism requires documenting specific health burdens imposed by different histories and types of racial discrimination, challenged the NCI reporting of white/non-white. (10/n)

HU researchers found that Black American incidence rates were severely under-reported, which rendered NCI findings misleading and incoherent.

Racism, when not addressed in our research, can distort monitoring and analysis of pop health and health inequities.(11/n)

Case 2 - Diabetes in American Indians.
How has the etiologic understanding of disease been distorted by the neglect of mainstream epi to acknowledge racism and histories of colonization?

(12/n)

A narrow biomedical and lifestyle focus on genetics and diet resulted in the “thrifty gene hypothesis” in 1960’s : genotype is frugal at metabolizing energy, which is beneficial during periods of starvation and potentially harmful when food becomes abundant. (13/n)

Racial ideology, rather than scholarship, underpins this hypothesis’ core assumptions, which are that famine is a feature of primitive culture and that diabetes is the disease of a civilization. (14/n)

A more credible scientific hypothesis focuses on the impacts of colonization and structural racism. In the case of Pima Indians, the historical and legal records are clear on who and what caused their starvation and access to poor diet – (15/n)

-- they were forcibly removed from their land by the US army. They were forced to abandon their irrigation based agriculture and access to the river. The land and farms were given to White settlers, in what is currently known as Arizona. (16/n)

Once forced to live on reservations – poor diet, reliance on US gov't for rations and commodity foods, altered diets of children forced to attend boarding school, like starch, meat, few veggies, no fresh fruit. (17/n)

This alternative hypothesis engages explicitly with biology by shifting attention from gene frequency to gene regulation, positing that any people subjected to the same history and conditions would likely have had a rapid increase in diabetes rates. (18/n)

TWO STEPS TO IMPROVE EPI RESEARCH & PRACTICE

1) Explicitly use theories of disease distribution when framing hypotheses, designing studies, developing instruments, measuring exposures, analyzing data, interpreting results, and explaining their scientific and social signif.(19/n)

Example (20/n)

2) Learn the social history of epidemiology and scientific racism. Learn the history of the times, places, and populations relevant to epi research. Use what we learn to improve study hypotheses, methods, and teaching. Keep learning. (21/n)

a) improve study hypotheses: Our grounding must be relevant to theories of disease distribution AND history pertaining to specific populations, exposure, and biology. Including the various levels of exposure/outcome, and understanding the relative temporal dimensions. (22/n)

b) improve methods: measuring exposures in relation to relevant domains, levels (systems and policies to interpersonal & internalized), temporal period, etc. READ: Counterfactual Causal Inference and Engaging with Racism as a Causal Determinant tiny.cc/0iq1pz (23/n)

c) improve teaching – teaching about racism and health cannot be confined to specialized courses. It is essential as an introductory course at all levels of education.

*before we can teach, we need to undergo the process of unlearning and learning* (24/n)

CONCLUSION
Epidemiologists must reckon with racism to 1) avoid harm and do better science 2) situate our research in historical, societal, and ecological context, 3) become aware of the deep rooted ways in which racism harms populations and leads to disparities in health. (25/n)