Recorded presentation

Sam Harper, MS¹, Seungmi Yang, MSc¹, Sonia Angell, MD, DTM&H², Marianne Hillemeier, PhD³, Jeff Morenoff, PhD⁴, George Davey Smith, DSc, MD⁵, and John W. Lynch, PhD⁶. (1) Epidemiology, University of Michigan, Center for Social Epidemiology & Population Health, 1214 South University, 2nd Floor, Ann Arbor, MI 48104-2548, 734-615-9227, sbharper@umich.edu, (2) Robert Wood Johnson Clinical Scholars Program, University of Michigan, 6312 Medical Science Building I, 1150 West Medical Center Drive, Ann Arbor, MI 48109-0604, (3) Health Policy and Administration, The Pennsylvania State University, 116 Henderson Building, University Park, PA 16802-6500, (4) Sociology, University of Michigan, 500 South State Street, Room 3012 LSA Bldg, Ann Arbor, MI 48109-1382, (5) Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Clifton, Bristol, BS8 2PR, United Kingdom, (6) School of Public Health, University of Michigan, 109 Observatory, Room M3116, Ann Arbor, MI 48109-2029

Social capital has been proposed as a key determinant of population health. We investigated associations between three different measures of social capital – Census 2000 response rates, organizational membership, and newspaper readership in 311 US metropolitan areas. We examined race and sex-specific deaths for 13 causes occurring from 1989-98. We used multilevel Poisson regression models for each measure of social capital separately, while adjusting for age, poverty, education, income inequality, and region. The results showed heterogeneity according to the social capital indicator used and the sex-race-specific patterns of different causes of death. However, there was a consistent negative association across measures of social capital for homicide and cirrhosis among whites. In general, we found as many negative as positive associations between social capital and mortality, and different measures of social capital often gave conflicting results for the same race-sex-cause of death. For example, among white males a 1 SD increase in membership organizations decreased the risk of death from stomach cancer (RR = 0.93, 95% CI: 0.91, 0.96) and increased unintentional injury risk (RR = 1.06, 95% CI: 1.04, 1.09), while higher Census response rates increased stomach cancer risk (RR=1.05, 95% CI: 1.02, 1.08) and decreased injury risk (RR=0.95, 95% CI: 0.94, 0.98). Our results show that ecological associations between social capital and mortality depend heavily on the choice of social capital indicator and the particular sex-race-cause of death. There is no evidence for a generally beneficial effect of social capital on sex-race-cause specific mortality patterns among US metropolitan areas.

Learning Objectives: