Impact of race and socioeconomic status on breast cancer mortality in USA: A cross-sectional time-series analysis.

Abstract

Purpose: Racial, ethnic, and socioeconomic (SES) disparities in incidence and mortality exist among all breast cancer types and stages, particularly between African American (AA), Caucasian American (CA), and Hispanic American (HA) women. The goal of our study is to investigate all three dimensions of SES to identify occupation, education, and income as predictors for breast cancer mortality among racial groups. Methods: A pooled cross-sectional time-series analysis is used to determine the statistical significance of the variables predicting age-adjusted breast cancer mortality in the 50 U.S. states from 1999-2015. Results: This study used beta coefficients, b, as indicators to determine the strength of the association between the dependent variables (race/ethnicity, income, education, and occupation) and breast cancer mortality. African American women showed the greatest statistical significance and strongest beta coefficient (p=0.000, b=0.061). Caucasian women exhibited a p-value of 0.007 and a beta of 0.037. The positive coefficients exhibited in both AA and CA women suggests that the probability of dying from breast cancer increases as the number of women in each racial group increases. The high beta coefficient shown in AA women illustrates a stronger relationship among this racial group, signifying the AA race alone may independently predict breast cancer death. Hispanic women demonstrated intermediate effects with statistical significance (p=0.004) and produced a negative beta coefficient (b=-0.057). The HA statistics illustrate an increase in this ethnic group would decrease breast cancer mortality, implying the presence of a protective factor. When analyzed independently, race or ethnic group alone showed significant relations to total age-adjusted breast cancer mortality, holding constant influences of education, income and employment. Also, under control were health care expenditures and lifestyle risks. Conclusion: This is the first breast cancer study to include all 50 U.S. states longitudinally and cross-sectionally to provide a large scale, population-based analysis. Additionally, this study is one of the first to consider all three dimensions of SES as predictors for breast cancer mortality. The results found that race and SES independently serve as strong predictors of breast cancer mortality, and when both are exhibited, the strength of the association is greater than having one characteristic alone.