Date of Award


Document Type

Open Access Dissertation


Health Promotion, Education and Behavior

First Advisor

Heather M. Brandt


In the United States, racial and ethnic minorities, economically disadvantaged and medically underserved groups bear a disproportionate amount of the cancer burden. Myriad social and environmental factors attribute to these disparities including disparate exposures to environmental pollutants, which account for two percent of all cancer deaths nationally. There is empirical evidence demonstrating risk perceptions and cancer worry are shaped by race/ethnicity and social and environmental experiences. Cancer risk perceptions among Non-Whites, especially Blacks compared to Whites is lower for various reasons. Low perceived cancer risk may explain persistent cancer disparities, since protective health behaviors are higher among persons who perceive their risk of cancer is higher. In addition to findings of lower perceived cancer risk, studies have shown that Blacks compared to Whites perceive their environmental health risks such as exposures to air and water pollution and other unhealthy environmental conditions are high even when they do not reside in an area with known issues. A paucity of research has explored the interplay between these factors among Blacks in metropolitan areas with disparate environmental conditions and cancer outcomes. This study explored perceived and actual cancer risk using an environmental health survey and geospatial methods in Metropolitan Charleston, South Carolina. The survey was used to document perceptions of cancer risk, neighborhood environmental health risks, and risk- reducing health behaviors. In addition, it evaluated the association between low perceived cancer risk and health behaviors among Blacks. Geospatial methods were used to analyze and map environmental cancer risk from 1996-2005, identify cancer clusters and hotspots, and to determine if cancer risk and outcomes vary spatially by racial and socioeconomic characteristics. Descriptive statistics, bivariate and multivariate analyses were performed in SAS 9.3. Total cancer risk from the National-Scale Air Toxics Assessment for 1996 to 2005 was georeferenced and analyzed in ArcGIS 10.2. Cancer clusters and hot spots were identified using Anselin’s Local Moran’s I and Getis-Ord Gi* statistic. Correlations were performed in SPSS 22.0.

Survey respondents (N=405) were 100% Black, 81% female (n=323), 19% male (n=75), and ranged from 18 to 87 years of age. Low perceived cancer risk (absolute risk) was associated with daily alcohol consumption and having had a colon cancer screening female, and older age (24-65, p<.05). Worry about cancer was significantly associated with being a current smoker, fair diet, non-alcohol consumption, and colon cancer screening tests (p<.05). The Spearman’s rho test revealed a statistically significant relationship between cancer risk and five-year incidence (p=.043). No significant relationship was observed between cancer risk and five-year mortality. However, incidence and mortality were significantly correlated with one another (p<.001). We detected a positive association (p < .001) between cancer risk and % Black and %poverty and a negative association with %income. Our findings suggest that perceived cancer risk is an important indicator of health behaviors among Blacks. Direct or indirect experiences with cancer and/or the environment, as well as awareness of family history of cancer are viable explanations of cancer risk perceptions. We believe our findings have implications for reducing place-based environmental cancer disparities and developing policies to reduce environmental and cancer burden in underserved and economically disadvantaged groups. Geographic variability in cancer risk may partially explain cancer disparities between groups.