Shimotsu, Scott Toshiro2011-06-302011-06-302011-05https://hdl.handle.net/11299/108284University of Minnesota Ph.D. dissertation. May 2011. Major: Epidemiology. Advisors Dr. Rhonda Jones-Webb, Dr. Jean Forster 1 computer file (PDF); : ix, 189 pages, appendices A-G.Objective. Excessive drinking contributes to 79,000 excess deaths annually and is associated with cardiovascular disease, several cancers, liver cirrhosis and social problems including drinking and driving, homicide, and other types of crime. The overarching goal of this dissertation was to test a model that examines how individual and neighborhood level characteristics contribute to excessive drinking. Methods. In Manuscript 1, we investigated the relationship between fruit and vegetable intake and alcohol consumption, and whether the inverse relationship between low fruit and vegetable intake and higher alcohol consumption was more pronounced among the poor in a large multi-ethnic sample of 9,959 adults from Hennepin County, MN using the Survey of the Health of Adults, the Population, and the Environment (SHAPE). In Manuscript 2, we explored whether food and alcohol access are related, and whether the relationship between food and alcohol access differs in poorer neighborhoods in Hennepin County, MN using Census Decennial and InfoUSA business data. The 3rd manuscript investigated whether living in a low SES neighborhood was associated with excessive drinking and if the retail environment (e.g., mix of food and liquor stores) mediated this relationship using SHAPE, Census Decennial, and InfoUSA data. A variety of statistical methods were used to answer our research questions including hierarchical Poisson and linear regression models. Results. In Manuscript 1, we found higher fruit and vegetable intake was associated with lower alcohol consumption and this relationship was more pronounced among individuals with lower household incomes. In Manuscript 2, we found the relationship between food and alcohol access differed by neighborhood SES, with higher income neighborhoods having more supermarkets and grocery stores, and liquor stores (RR=1.47; 95% CI: 1.21, 1.80). In Manuscript 3, we found that individuals living in census tracts with only liquor stores had a 46% higher risk of binge drinking than individuals living in neighborhoods with only food stores (RR=1.46; 95% CI: 1.03,2.07) after adjusting for demographic and lifestyle factors. Conclusion. Neighborhood characteristics such as the mix of food and liquor in neighborhoods are important in understanding excessive drinking above and beyond demographic and lifestyle factors. Future research on social conditions impacting alcohol consumption should explore the mix of stores, not just the over-concentration of liquor stores in neighborhoods.en-USAlcohol accessAlcohol consumptionDietExcessive drinkingFood accessNeighborhoodEpidemiologyThesis or Dissertation