Multilevel modeling to estimate mediation of perceived stress on the relationship between allostatic load and neighborhood environment

Purpose and Study Question:

Studies of neighborhoods have found that physical and social environments are associated with a variety of health outcomes, from mortality to cardiovascular health to depression. Recent studies have used allostatic load as an indicator of accumulated physiologic stress and possible preclinical disease, yet there remain many questions about the pathways through which neighborhood environment might influence allostatic load (and health). This study examines the relationship between neighborhood environment, perceived stress, and allostatic load using a 13-item measure of allostatic load, three separate neighborhood quality measures, and a unique neighborhood definition.    

Methods:

Multilevel models on a sample of 1,307 residents of Texas City, Texas, were used to estimate the association of both individual and neighborhood characteristics with allostatic load. Mediation analysis was conducted to examine the role of perceived stress as a possible pathway between neighborhood environment and allostatic load.

Results:

Preliminary results indicate that there is considerable variation in SES and resident neighborhood perception at the neighborhood level in Texas City, which is significantly associated with allostatic load. Significant individual-level predictors of increased allostatic load included gender, age, chronic conditions, income, smoking behaviors, and drinking behaviors, but did not include perceived stress.

Conclusions:

The strengths of this study are the use of several types of neighborhood measures and a unique, grounded definition of “neighborhood” that may better represent actual, rather than administratively-derived, neighborhood boundaries.  Further analyses of these data will provide additional insight into whether neighborhood effects on health work through perceived stress.