The relationship between five different measures of structural social capital, medical examination outcomes, and mortality

• Published in: Social science & medicine (1982) 2013-05, Vol.85, p.18-26
• Main Authors: Muennig, Peter, Cohen, Alison K., Palmer, Aileen, Zhu, Wenyi
• Format: Article
• Language: English
• Subjects: Adult; Biological and medical sciences; Biomarkers; Blood Pressure; C-Reactive Protein - analysis; Cardiovascular Diseases - mortality; Cause of Death; Cholesterol - blood; Clinical outcomes; Cross-Sectional Studies; Diabetes. Impaired glucose tolerance; Endocrine pancreas. Apud cells (diseases); Endocrinopathies; Epidemiology; Etiopathogenesis. Screening. Investigations. Target tissue resistance; Female; Fibrinogen - analysis; Follow-Up Studies; General aspects; Health; Health Status Disparities; Humans; Male; Medical sciences; Miscellaneous; Mortality; Mortality - trends; National Death Index; NHANES III; Nutrition Surveys; Physical examinations; Public health. Hygiene; Public health. Hygiene-occupational medicine; Regression analysis; Religion & psychology; Risk Factors; Social capital; Social Support; Socioeconomic Factors; United States - epidemiology
• ISSN: 0277-9536; 1873-5347
• DOI: 10.1016/j.socscimed.2013.02.007

Higher social capital is associated with improved mental and physical health and reduced risk of premature mortality. We explored the relationship between five measures of structural social capital and 1) intermediate health outcomes (elevated C-reactive protein, cholesterol, blood pressure, and serum fibrinogen) and 2) distal outcomes (cardiovascular and all cause mortality). We did so using the National Health and Nutrition Examination Survey III 1988–1994 linked to the National Death Index with mortality follow-up through 2006. We employed ordinary least squares regression for the intermediate outcomes, seemingly unrelated regression (SUR) to consider combined effects, and Cox proportionate hazards models for mortality outcomes. We then performed extensive sensitivity analyses, exploring the contribution of various variables and reverse causality. We find that our measures of social capital did not predict statistically significant changes in the laboratory biomarkers we study. Nevertheless, belonging to organizations or attending church >12 times per year were associated with reduced all cause mortality (hazard ratio [HR] = 0.81, 95% confidence interval [CI] = 0.70–0.93 and HR = 0.72, 95% CI = 0.60–0.86, respectively). In SUR analyses, however, combined laboratory values were significant for all measures of social capital we study with the exception of visits to neighbors. This suggests that some forms of structural social capital improve survival through small changes in multiple measures of biological risk factors rather than moderate or large changes in any one measure. ► We employ a powerful, nationally-representative medical examination dataset linked to 18-years of mortality data in the US. ► We explore differences in 5 separate measures of bonding social capital, and make comparisons across these measures. ► We dispatch a powerful statistical tool that estimates the joint effects of small changes in multiple measures of health. ► We estimate reverse causality and cohort effects. ► We investigate the impact of health insurance as well as first and second hand smoking on outcomes.