Methods: Using data from the 2005 California Health Interview Survey, we analyzed responses from 17,472 men to examine the extent to which a usual source of health care mediates the relationship between health insurance and diabetes.
Results: Logistic regression analyses were calculated to determine whether health insurance enrollment is associated with diabetes, controlling for demographic characteristics. The findings show that health insurance status did relate to the likelihood of diabetes even after controlling for demographic differences for the overall sample and for each racial group except African American men. Specifically, controlling for other demographic factors, men had odds of probable diabetes when enrolled in public health insurance that were 2.14 times greater than those men without diabetes (OR = 2.14; p < .05). Sobel-Goodman tests for mediation indicated the largest effects between Latino and White men. For African American and Asian men, usual source of care did not serve as a significant mediation factor between health insurance and diabetes.
Conclusions and Implications: Our results suggest that major differences in diabetes prevalence are likely tied to health insurance. The finding has been repeated in several other studies. What makes this study unique is the fact that our analyses provide evidence about the link between health insurance and poor men's ability to manage their diabetes in the medical care system which other studies allude to but do not show through analyses. More research is needed to identify not only determinants of racial disparities in diabetes but also to deepen our understanding of how health policy and the social environment can eliminate elevated risks for diabetes among individual and groups of men.