Response to Krieger

We thank Dr Krieger for commending our study of secular trends in estrogen receptor (ER)–negative breast cancer incidence in the United States by race, age, and geographic region (1). We appreciate Dr Krieger (2,3) drawing attention to her prior work and agree that structural racism is an important social determinant of health that warrants further evaluation as a factor contributing to breast cancer inequities.

Using a specialized The Surveillance, Epidemiology, and End Results (SEER) Program 13 dataset, Dr Krieger (2) found that being born in a state with a history of Jim Crow ordinances was associated with an elevated risk of ER–negative breast cancer among Black women but not White women. In contrast, we studied state or registry of residence at diagnosis in SEER 17. Dr Krieger’s (2) work and that of others evaluating state-level measures of structural racism (4) and other breast cancer risk factors [eg, excess body weight (5)] may offer insights into some of the signals we observed. A strength of our study is the use of novel multilevel models to identify heterogeneity in ER–negative breast cancer incidence among non-Hispanic White, non-Hispanic Black, and Hispanic women across different US regions, represented by 17 SEER registries. We found that age-adjusted ER–negative breast cancer incidence decreased over time in each age group and racial or ethnic group. We also identified important regional heterogeneity for both the age-adjusted risk and its rate of decline. Whereas non-Hispanic White women experienced declines in all age groups and registries, non-Hispanic Black women aged 50-84 years had the slowest declines or even slight increases within registries that included states with (Atlanta metro, Greater Georgia, and Louisiana) and without (California and Los Angeles) a history of Jim Crow ordinances. Furthermore, Hispanic women had the lowest ER–negative incidence rates in all age groups, and their trends lie between non-Hispanic White and non-Hispanic Black women, with younger Hispanic women having trends more similar to non-Hispanic White women and older Hispanic women having trends more similar to non-Hispanic Black women. Our descriptive study provides a basis for generating hypotheses and designing analytical studies with well-characterized exposures to identify underlying determinants of observed patterns in ER–negative breast cancer incidence, for which non-Hispanic Black women experience the highest rates.

As a Black woman scientist, Dr Davis Lynn, the first author of our report, is keenly aware that systemic racism is a nationwide public health crisis. We are committed to continued surveillance of diverse populations and to the conduct of studies that use smaller area exposures estimates and person-level data to understand factors, including structural racism, that may contribute to dissimilar breast cancer incidence trends by race and ethnicity. Given the increase in availability of population-level datasets for analytical linkage studies and rapid advances in statistical methodology, a transformative period for the field of descriptive epidemiology is within reach (6), with immense opportunities for epidemiologists, sociologists, biostatisticians, and data scientists to join hands to advance integrative descriptive epidemiologic research. This collaborative interdisciplinary approach has the potential to improve our understanding of—and ultimately, reduce—inequities in breast cancer and population health (7).

Funding

This research is supported by the Intramural Research Program of the National Cancer Institute at the National Institutes of Health.

Notes

Role of the funder: The funder had no role in the design of the study, collection, analysis, and interpretation of the data; the writing of this response; or the decision to submit the response for publication.

Figure 1.

Net drifts (model‐based estimates of annual percent change in age‐adjusted rates) for estrogen receptor (ER)–negative breast cancer incidence among women aged 50‐69 years and 70‐84 years by race or ethnicity and The Surveillance, Epidemiology, and End Results (SEER) Program registry. Posterior medians with 95% credible intervals are shown. SF = San Francisco; SJ = San Jose.

Disclosures: All authors declare no competing financial interests.

Author contributions: Supervision: GLG, PSR. Conceptualization: All authors. Formal analysis: BCDL, PC, and PSR. Writing—original draft: All authors. Writing—review and editing: All authors.

Data Availability

All data used in this work are publicly available from the US Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Specifically, this work used data from the SEER-18 registries database. These data can be downloaded using the software SEER*Stat, which may be downloaded from https://seer.cancer.gov/seerstat/.