Community social capital and diabetes outcomes among the elderly in the State of Hawai'i

This research investigates community social capital and diabetes outcomes among the elderly in the State of Hawai`i. The study employs an ecological research design to achieve a comprehensive understanding of community social capital and diabetes among the elderly. Because ecological design aims to capture multi-layered realities, the study used a triangulation methodology. This project consisted of (1) archival data analyses, (2) small-area analysis using GIS and statistical analyses, and (3) follow-up interviews. Statistical analyses provided both individual- and community-level diabetes outcomes of the elderly. First, binary logistic regression revealed that BMI status and ethnicity (Filipino, Japanese or Native Hawaiian) were particularly strong predictors of the likelihood of diabetes. Second, multiple discriminant function analysis revealed that the predictors that best distinguished among three community groups with low, medium or high rates of diabetes-related hospitalization were bonding social capital, bridging social capital and college graduation rate. On the other hand, the study's findings show that community social capital does not contribute to distinguishing the three community groups with low, medium or high rates of diabetes-related emergency department visits. Interviews collected community-based experiences that archival data review, GIS and statistical analyses might overlook. The limitations of both data sources and research remain important areas for future research. Thus, this study can be regarded as but the first step of a much larger exploratory study. Nevertheless, its combination of qualitative and quantitative analyses drew the profound implication that community social capital is an important set of social properties and contributes to a better understanding of the community-level diabetes outcomes of the elderly. (PsycINFO Database Record (c) 2012 APA, all rights reserved).