This study of a large cohort of men and women diagnosed with colorectal cancer yielded several important findings. There were persistent racial/ethnic survival differences after controlling for numerous variables. Furthermore, some of the factors that appeared to substantially reduce the mortality difference between Whites and Blacks, did not impact the mortality difference between Asians and Blacks. However, adjusting for comorbidities and SES resulted in a reduction in the mortality difference regardless of reference group. Therefore, comorbidities and SES appeared to be more important explanations for the survival differences observed among Blacks relative to Asians and Whites.
Several studies have examined racial/ethnic differences in CRC survival.3, 22
Our finding that racial disparities are largely explained by socioeconomic status is consistent with most of these findings. However, to our knowledge, all prior studies have compared survival among racial/ethnic groups relative to Whites. No studies of CRC survival have used Asians (the group with the best survival in this case) as a referent group, nor examined the underlying mechanisms as they relate to specific racial groups by comparing the variation in factors contributing to survival differences using different referent groups.
In this study, we found that factors contributing to survival disparities varied by racial/ethnic group. There were no statistically significant differences between Hispanics and Whites and Hispanics and Asians; however, the survival differences between Whites and Asians widened after adjusting for a number of factors. On the other hand, SES, which is associated with survival in CRC patients,2, 3
was a key determinant of survival for Blacks. These findings are similar to a meta-analysis that demonstrated that the racial disparity in survival for colon cancer between African Americans and Caucasians was attenuated after adjusting for socioeconomic factors and treatment.3
There were large ethnic differences in SES and rural residence, and SES accounted for large reductions in CRC-specific mortality between Blacks and Whites and Blacks and Asians. Furthermore, comorbidities played a key role in survival disparities for Blacks. As in the case of SES, a larger proportion of Blacks had higher comorbidity scores compared to other racial/ethnic groups and adjusting for comorbidities reduced mortality differences between Blacks and Whites and Blacks and Asians. Although a few studies have shown that comorbidities may independently affect CRC survival,31, 32
no studies prior to this one have found that they impact racial/ethnic survival disparities.
The persistent racial/ethnic survival differences, despite controlling for numerous variables, may be explained by differences in biology,33-37
refusal of43, 44
and compliance with treatment,43, 44
post-treatment surveillance,12, 13
and access to high quality cancer care,14
which were not examined in this study.
Differences in tumor site distribution and genetics may explain the high survival rates observed among Asians. A previous study demonstrated that relative to Whites, Asians have higher rates of distal colon cancer, which is associated with a decreased risk of mortality.33
For Blacks, poor survival may be due to biologic features that may contribute to aggressive tumor behavior,7
or inherited or acquired genetic abnormalities35-37
which may impact response to therapy.37
Patients with low SES are more likely to die from CRC than patients with high SES.3
In this study, a large proportion of Blacks and Hispanics resided in low SES neighborhoods, whereas a larger proportion of Asians and Whites resided in high SES neighborhoods. Percentage of persons within a census tract living under the poverty line was used as a measure of SES; therefore, there might have been residual confounding of SES since we were not able to control for differences in SES at the individual level. In addition, other components of SES such as education were not included in our analysis but may influence diagnosis, treatment, and, ultimately, survival. Despite these limitations, there was no multi-collinearity between SES (percentage of persons in a census tract living below the poverty level) and race/ethnicity present.
Lifestyle differences may explain some of these differences in survival. Obese patients have a 50% increased risk of developing colon cancer and 30% higher risk of dying from colon cancer.45
Moreover, obese patients treated for colon cancer have poorer overall survival than normal weight patients.46
Studies have also found that higher levels of physical activity may reduce the risk of colon cancer by as much as 50%,47
and patients who engage in vigorous physical activity have lower rates of colon cancer recurrence.40
National data has shown that Blacks and Hispanics have higher rates of obesity48
and lower rates of physical activity than Whites.49
Acculturation may also explain some of the survival differences observed among these racial/ethnic groups. Relative to US-born Whites of equivalent socio-demographic backgrounds, foreign born Blacks, Hispanics and Asians, have lower mortality risks.39
However, immigrants’ risk of disability and chronic disease morbidity increases with increasing length of residence.38
Cultural beliefs and norms may be linked to racial/ethnic mortality differences. Cancer fatalism, which is the belief that death is inevitable when cancer is present,41
can be a significant barrier to early detection and treatment all of which are important for achieving optimal survival. Studies have shown that Blacks, Hispanics and Chinese are more likely to possess fatalistic views regarding cancer.41, 42
Adjusting for standard therapy yielded a small reduction (2%) in the survival disparity between Blacks and Whites in this study. However, a more complete depiction of the role of treatment in the racial/ethnic survival disparities may include accounting for differences in treatment compliance and benefit, high-quality surgical care and post-treatment surveillance. Compared to Whites, Blacks are more likely to refuse treatment10, 44
and even when Blacks receive treatment, their survival benefit from adjuvant chemotherapy is not as great.11, 34
Also, there is evidence to suggest that patients treated by a surgical specialist with high caseloads have improved CRC survival.50
Yet, Black patients are less likely to be treated by these surgeons51
or have access to high-quality subspecialists.14
Finally, post-treatment surveillance can detect CRC recurrence and lead to improved survival; however, racial/ethnic minorities are less likely to receive this care.12, 13
Additional research is needed to determine the role of each of these factors in racial/ethnic disparities in CRC survival.
In addition to the variables that were not measured in this study, another limitation of this study is that the sequence of the variables in the model may have affected the percentage reduction in hazard ratio attributable to each variable; yet, when changes in the order of the variables were made, there was little difference.
There are a number of strengths that support this study’s validity. The study included nationwide and population-based cases from 16 SEER areas, which accounts for approximately 25% of the U.S. population. The cases were ethnically diverse and included traditionally understudied racial/ethnic groups: Hispanics and Asians. Therefore, these findings may be generalizable to diverse populations 66 years of age or older residing in other areas of the U.S. Furthermore, the linked database allowed us to incorporate a number of treatment, hospital and comorbidity variables across the cancer care continuum and is an accurate and complete source of data.52, 53
In conclusion, although comorbidities and SES appear to be important factors contributing to the poorer CRC-specific survival for Blacks relative to Whites and Asians, substantial racial disparities in survival still persisted and were not fully explained by variations in a number of factors across the cancer continuum. Future research should examine the role of other factors not included in this study such as the quality of care, particularly the benefit of treatment and post-treatment surveillance.