Race/Ethnicity, Socioeconomic Status and Breast Cancer

Institution: Public Health Institute
Investigator(s): William Wright, Ph.D. -
Award Cycle: 2000 (Cycle VI) Grant #: 6IB-0071 Award: $38,258
Award Type: IDEA
Research Priorities
Community Impact of Breast Cancer>Health Policy and Health Services: better serving women's needs



Initial Award Abstract (2000)
Women in California from different race/ethnic groups or socioeconomic levels are not equally affected by breast cancer. The California Cancer Registry (CCR) has collected statewide breast cancer data since 1988 and continues to be a valuable resource for research on cancer incidence and mortality. The Women’s Health Survey (WHS) is an annual telephone survey that collects health practices data on a random sample of California Women. The proposed research will utilize CCR data and WHS data to investigate timely and important questions about the relationship between race/ethnicity and socioeconomic status in the risk of developing breast cancer. We will also investigate how these variables are related to the stage of breast cancer at diagnosis, which is an important prognostic factor for survival.

The three specific aims that will be addressed are:
  1. To determine whether the relationship between socioeconomic status (SES) and stage at diagnosis for breast cancer in California varies for different race/ethnic groups.
  2. To investigate whether the relationship between SES and reproductive risk factors for breast cancer differ for four race/ethnic groups in California.
  3. To evaluate the importance of considering race/ethnicity when calculating breast cancer incidence rates.
Several fairly innovative statistical methods will be employed. We will use principal component analysis to measure SES. We feel this is a more scientifically valid method of measuring SES than methods used by other researchers. Another innovative method is log-multinomial regression, which will be used to evaluate stage at diagnosis under Objective 1. To our knowledge, this method has not been previously used to analyze stage at diagnosis for breast or any other cancer. Other methods to be used include Poisson regression and multiple linear regression.

The results of this study may be useful for designing more effective intervention programs and for designing larger scale research aimed at investigating these disparities in the breast cancer experience in California. The results of this research will provide insight on the unequal burden of cancer in California, which addresses a priority of the BCRP to support research that may lead to reducing inequities in health care delivery and health care policy. There is a trend in epidemiologic research to evaluate the relationship between behavioral or social (e.g., SES) factors and disease. Thus, the proposed research is timely and consistent with state and national research priorities.


Final Report (2002)
Women in California from different race/ethnic groups or socioeconomic levels are not equally affected by breast cancer. The California Cancer Registry (CCR) has collected statewide breast cancer data since 1988 and continues to be a valuable resource for research on cancer incidence and mortality. The California Women's Health Survey (CWHS) is an annual telephone survey that collects health practices data on a random sample of California women. Our research utilized CCR and CWHS data to investigate timely and important questions about the relationship between race/ethnicity and socioeconomic status (SES) in the risk of developing breast cancer and in the risk of late stage at diagnosis. We investigated how these variables are related to stage of breast cancer at diagnosis, an important prognostic factor for survival.

We addressed three specific aims:

1. Determine whether the relationship between SES and stage at diagnosis for breast cancer in California varies for different race/ethnic groups and determine the degree to which this relationship depends on how stage at diagnosis is measured.

2. Investigate the relationship between SES and risk factors for breast cancer for four race/ethnic groups in California.

3. Evaluate the importance of considering race for monitoring breast cancer incidence rates.

We found that SES was positively related to the probability of being diagnosed with early stage breast cancer for all race/ethnic groups regardless of how stage was defined. The relationship between SES and stage at diagnosis was strongest for Hispanic women, followed by black, white, then Asian/other women.

We also showed that the strength of the relationship between SES and stage at diagnosis depends on how stage is defined. Specifically, we evaluated stage at diagnosis defined as early and late, where in situ breast cancer cases were included in the early category. We also defined stage at diagnosis in three categories: in situ, early, and late. We employed an innovative statistical method to evaluate stage defined in three categories. We showed that the strength of the relationship between SES and stage at diagnosis depends on how stage at diagnosis is defined; specifically, the relationship is enhanced when in situ cases are combined with early stage cases in one category.

We evaluated the relationship between known and suspected breast cancer risk factors and SES using data collected in the CWHS. The found that number of live births and, to a lesser extent, age at first birth and body mass index are risk factors that were significantly related to the SES index and roughly exhibited the anticipated race/ethnic patterns in the strength of the association. Thus, SES may act as a surrogate for these variables in analyses involving SES and breast cancer incidence.

Finally, we evaluated the consequences of the new race classifications in the 2000 US Census. Our findings here may have important implications across a broad range of future breast cancer studies ranging from epidemiologic studies through health services and health policy studies which use incidence calculations. Beginning in 2000, individuals could select more than one race group on the US Census forms, which compromises our ability to obtain the race-specific population estimates needed for breast cancer incidence rate calculations. We demonstrated that because of this new method of reporting race in the Census, the CCR may be unable to detect significant changes in race-specific breast cancer incidence rates over time.