This proposal addresses a central challenge in cancer prevention and control: prioritizing efforts to reduce cancer disparities. Some racial and socioeconomic groups have higher rates of cancer incidence and mortality than others. However, the processes causing these disparities remain unclear. Standard epidemiologic methods describe cancer disparities but do not identify the causes of the disparities. To effectively intervene to reduce disparities, we must identify the mechanisms by which disparities arise. This grant will train the recipient to identify modifiable causes of disparities in cancer incidence and mortality using epidemiologic data. To develop the necessary statistical and substantive expertise, the recipient will receive training in three areas: (1) novel statistical techniques that are specifically suited to dissect health disparities (i.e., econometric decomposition and causal mediation analysis using marginal structural models); (2) mentored application of the statistical techniques in a racially and socioeconomically diverse cohort of U.S. men and women; and (3) biological characteristics of tumors that predict prognosis and treatment efficacy. This training will be achieved via hands-on mentored research, coursework, and limited conference attendance. Using the skills developed through the training, the grant recipient will specifically investigate causes of higher breast cancer burden in Black versus White women in the U.S. Socioeconomic status is believed to account for a large portion of racial disparities in breast cancer risk and mortality, but it is unclear by what specific pathways SES influences these racial disparities. This research will estimate the contribution of one pathway by which low SES contributes to racial cancer disparities in breast cancer risk, biology, and mortality: adult weight gain and obesity. The research uses data from the Carolina Breast Cancer Study, a population-based, case-control study conducted in 24 counties of North Carolina. Women with invasive breast cancer and population controls were enrolled between 1993 and 2011. The data consist of 1,803 cases of invasive breast cancer (787 African- American, 1,016 white) and 1,564 population-based controls (718 African- American, 846 white). A unique strength of the data is novel molecular markers of breast cancer subtypes. These markers will allow the study to quantify a potential causal pathway underlying the higher risk of the basal-like subtype of breast cancer in Black women. The results of this work will help policy-makers and the public health and clinical communities evaluate what interventions will be most effective in reducing cancer disparities.
This grant addresses a central challenge in cancer prevention and control: developing effective and efficient interventions to reduce racial and socioeconomic inequalities in rates of cancer diagnosis and death. Standard epidemiologic analysis describes population inequalities but does not identify the processes that cause the disparities. With this award, I will attain the statistical and subject-matter expertise to identify the causal mechanisms by which cancer inequalities arise and quantify the relative impacts of the different mechanisms, thereby improving efforts to prioritize the most effective interventions to reduce inequalities in cancer outcomes.
Showing the most recent 10 out of 23 publications
