County-level social environment determinants of health-related quality of life among US adults: a multilevel analysis.

To show that an individual's health-related quality of life (HRQOL) is not determined only by their personal-level characteristics, but also is socially determined by both physical and social environmental characteristics of their communities. This analysis examined the association of selected county-level indicators on respondents' unhealthy days and assessed the utility of mean unhealthy days for US counties as community health indicators. Data came from the 1999-2001 Behavioral Risk Factor Surveillance System. We used multilevel models to calculate the proportion of between-county variation in HRQOL that was explained by county-level contextual variables and examine the causal heterogeneity of some personal-level factors modified by these contextual variables. Counties with worse socioeconomic indicators, high mortality rate, and low life expectancy were associated with higher numbers of unhealthy days. These indicators explained 13-22% variance of county-level physically unhealthy days and 4.5-9.5% variance of county-level mentally unhealthy days. The GINI index, suicide rate, percent uninsured, primary care facilities-to-population ratio, and most county-level demographic and housing indicators also had significant but smaller impact on respondents' unhealthy days. Also, the counties with poorer socioeconomic scores had additional negative HRQOL impact on older persons. This study provides important new empirical information on whether various commonly-measured characteristics of the social environment, which are believed to be social determinants of health, are in fact associated with the perceived physical and mental health of its residents. Our findings provide additional support for the construct validity of county-level HRQOL as a community health indicator.