Abstract B25: Characterizing the neighborhood environment for studies of cancer risk: The Multiethnic Cohort

Abstract

Abstract Introduction: While considerable progress has been made in identifying individual-level risk factors for cancer etiology, the contextual features of the neighborhood environments, where these factors operate have yet to be thoroughly investigated. This is particularly relevant for minority and low socioeconomic populations that have limited neighborhood resources and experience higher levels of obesity, smoking, and other high risk behaviors. Recognizing the need for neighborhood studies of cancer in minority populations, we have linked the well-characterized resources of the Multiethnic Cohort Study (MEC) and the California Neighborhoods Data System to enable prospective studies of the neighborhood environment and cancer risk. The objective of this analysis was to characterize the contextual features of the neighborhood environments of MEC participants, residing in California at baseline, and assess the relationship between these neighborhood characteristics, demographic factors, and health behaviors. Methods: Study subjects included 111,697 MEC participants, ages 45-75, residing in California (93% from Los Angeles County) that completed a baseline questionnaire from 1993-1996. Residential baseline addresses were geocoded to latitude and longitude coordinates and assigned 1990 Census block groups to characterize their neighborhood social (e.g., socioeconomic status (SES) distributed into quintiles based on statewide distribution, racial/ethnic composition, percent foreign-born) and built environment (e.g., population density, urbanicity, street connectivity, housing) characteristics using the California Neighborhoods Data System. Linear regression and Cochran-Mantel-Haenszel tests were conducted adjusting for age, sex, race/ethnicity to examine the relationship between neighborhood characteristics, demographic factors, and health behaviors at baseline. Results: Ninety-eight percent of California residences of MEC participants at baseline (n=109,682, 30.3% African Americans, 41.3% Latinos, 13.1% Whites, 11.3% Japanese Americans; and 42.8% men and 57.2% women) were successfully geocoded and linked to 7,947 Census block groups, our neighborhood unit. African Americans and Latinos were more likely to reside in low SES neighborhoods (67% and 61%, respectively, in quintiles 1 and 2). In contrast, Whites and Japanese Americans were less likely to reside in low SES neighborhoods (27% and 19%, respectively, in quintiles 1 and 2). Higher levels of neighborhood SES were significantly associated with higher levels of education, higher mean intake of fruits (g/kcal/day), lower mean intake of red meat (g/kcal/day), higher alcohol intake (g/day), lower proportions of smoking, higher mean levels of moderate/vigorous activity, and lower mean levels of body mass index (BMI) in comparison to low SES neighborhoods (Ps<0.001). Higher levels of population density were significantly associated with lower education, lower mean intake of fruits (g/kcal/day), higher mean intake of red meat (g/kcal/day), and lower mean levels of moderate physical activity in comparison to low population density neighborhoods (Ps<0.001). Residing in neighborhoods with higher levels of commuting by car or motorcycle were significantly associated with higher mean levels of moderate and moderate/vigorous physical activity and lower mean intake of vegetables compared to residing in neighborhoods with less commuting by car or motorcycle (Ps<0.001). Residing in neighborhoods of higher levels of Hispanic acculturation were significantly associated with lower mean levels of moderate physical activity, lower mean intake of fruits (g/kcal/day), higher mean intake of red meat (g/kcal/day), and higher mean levels of BMI in comparison to residing in neighborhoods of low levels of Hispanic acculturation (Ps<0.001). Conclusion: These findings demonstrate the complex relationships between specific social and built environment features and several health behaviors among MEC participants, residing primarily in Los Angeles County at baseline. Future work will evaluate additional neighborhood characteristics such as retail and restaurant food environments, parks, and traffic density. The diversity and range of neighborhood exposures of MEC participants coupled with our understanding of the relationship between these exposures and individual-level factors establish the framework for future studies of their independent and joint effects on cancer risk. Citation Format: Iona Cheng, Salma Shariff-Marco, Juan Yang, Andrew Hertz, Jana Koerte, Hang Le Galanis, Myles Cockburn, Christina Clarke, Clayton Schupp, Cheryl Albright, Loic Le Marchand, Laurence Kolonel, Brian Henderson, Scarlett Lin-Gomez, Kristine Monroe, Lynne Wilkens. Characterizing the neighborhood environment for studies of cancer risk: The Multiethnic Cohort. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B25. doi:10.1158/1538-7755.DISP13-B25