Associations of socioeconomic status (SES) with breast cancer risk and survival are well described. The risk of breast cancer is lower in women of lower SES, while survival after breast cancer in these groups is poorer for certain racial/ethnic groups, in particular African-Americans. However, the reasons behind these associations are still not well understood, likely because prior research lacked the necessary multilevel approach for joint examination of individual and neighborhood (contextual) characteristics. This study, based on previously obtained interview data, proposes to enhance individual-level socioeconomic, immigration/acculturation, and behavioral data on breast cancer cases and controls from two large, multiethnic, population-based studies by linking them to contextual geospatial data on SES, immigration/acculturation, and the social and built environment. These individual and contextual variables will be used to examine their independent and joint effects on cancer risk and survival, and the extent to which these effects are due to behavioral, social, and established breast cancer risk or prognostic factors.
The specific aims are to: 1a) quantify the independent and joint effects of individual- and contextual-level SES on breast cancer risk within racial/ethnic groups;1b) quantify the independent and joint effects of individual- and contextual-level immigration/acculturation factors on breast cancer risk among Hispanic women;2a) quantify the independent and joint effects of individual- and contextual-level SES on stage at diagnosis, breast-cancer specific survival and overall survival within racial/ethnic groups;2b) examine the extent to which individual-level and contextual-level SES explain racial/ethnic variation in survival;and 2c) quantify the independent and joint effects of individual- and contextual-level immigration/acculturation factors on survival among Hispanic and Asian-American women. Breast cancer cases (n=4205) and controls (n=3236) from the San Francisco Bay Area Breast Cancer Study and the Northern California Family Registry for Breast Cancer will be geocoded based on residence at diagnosis (cases), or selection into the study (controls) using a common protocol, and then linked to clinical data from the regional population-based cancer registry, to contextual geospatial data, and to data from hospital and radiation facilities, in order to produce a detailed, multilevel dataset. Incorporating novel, state-of-the-art analytic methods, including multilevel regression analyses, geographic information systems (GIS), and tree-based discriminant analyses (recursive partitioning), these analyses will identify joint effects among individual and contextual measures and specific populations at highest risk of developing breast cancer and/or having the worst survival. Such information has strong translational potential for reducing disparities in the burden of breast cancer by informing prevention and intervention efforts targeted toward specific populations for reducing incidence, ensuring early detection, and improving survival.
Incorporating novel, state-of-the art analytic methods, applied to a multilevel dataset comprising individual-level data on socioeconomic status, immigration/acculturation, behavioral factors, and linked to contextual-level data on socioeconomic status, immigration/acculturation features of neighborhoods, and social and built environment, this study will identify joint effects among these individual and contextual measures and specific populations at highest risk of developing breast cancer and/or having the worst survival. Such information has strong translational potential for reducing disparities in the burden of breast cancer by informing prevention and intervention efforts targeted toward specific populations for reducing incidence, ensuring early detection, and improving survival.
