Interpretation
These findings reinforce information available from other contexts. Population studies from the US have consistently shown that socially and financially disadvantaged endometrial cancer patients have more limited prognosis. Two large studies using the National Cancer Database (NCDB) to evaluate endometrial cancer patients found that Black race, insurance status, lower income and education were all independently associated with increased mortality 18,20. Similarly, a Surveillance, Epidemiology and End Results (SEER) cancer database study evaluating a large endometrial cancer cohort with long term follow up26 reported higher overall and cancer-specific mortality among Black patients, and tied this to disparities in presentation and treatment. However, publications from the US may not be transferrable to the Canadian context, due to the complexity of the American healthcare payer system and financially-driven barriers to access.
American data on the association between SDH and endometrial cancer outcomes focuses heavily on racial marginalization13,18,20,24,26,27. Race differs from ethnicity6 and is not routinely collected in Canadian administrative and healthcare databases. Our findings did not support an association between ethnic marginalization and overall survival in Ontarian women with endometrial cancer. The ethnic concentration domain of the Canadian Marginalization Index reflects the concentration of self-identified ethnic minorities and new immigrants in the community35. We postulate that the healthy immigrant effect may partially counterbalance the effects of social, cultural and financial marginalization often associated with ethnic concentration36. In support of this, a targeted analysis of immigration status in our patient population found that recent immigration to Canada is, in fact, associated with improved overall survival (HR=0.58 for new residents within 5 years, p=0.002).
Data from other jurisdictions with public healthcare systems are also consistent with our findings in Ontario: Population-based studies from Sweden 17 and from the UK 16 have described higher rates of advanced-stage endometrial cancer among socially disadvantaged women; however, disease outcomes were not compared in these studies. Cooper et al report data from national statistics, including over 53,000 endometrial cancer patients, in England and South Wales 37 between 1986-1999; they note a deprivation gap in endometrial cancer survival which has not narrowed over the 15-year study period. Finally, a comprehensive cancer registry study from Australia 38 demonstrates a significant excess risk of death in disadvantaged endometrial cancer patients from communities with low levels of educational attainment and skilled employment.
Disparities in cancer incidence and outcomes in marginalized populations have been linked to adverse health behaviors 39, and many publications on the association of SDH with cancer outcomes evaluate diseases with behavioral risk factors, such as lung cancer, head and neck tumors and cervical cancer rates7,40–42. We chose to focus on a neoplasm which is not associated with smoking or with sexual behavior in an attempt to mitigate confounders. However, obesity is a well-established risk factor for endometrial cancer 43–46 and has consistently been linked to social marginalization 47. We did not have access to individual-level BMI, but were able to capture severe obesity (BMI>40) through billing codes. We found that severe obesity was inversely associated with the risk of death; this may be explained by the fact that obesity increases the risk for estrogen-dependent tumors 43–46, which are primarily well-differentiated and carry a better prognosis. This hypothesis is supported by the findings on multivariable analysis, where the association between obesity and improved survival did not persist when adjusted by disease histology and stage.