School of Public Health

Hannah Cohen-Cline

Neighborhood characteristics, social capital, and depression: a twin study


Depression is one of the leading contributors to the global burden of disease, and can have a debilitating effect on quality of life. While it is already well-accepted that individual-level factors can influence depression, there is growing recognition of the important role the social and built environment plays in exacerbating or alleviating mental health problems. Depression risk varies across neighborhoods, and differences in neighborhood context may impact depression through diverse pathways, including reduced social services and lack of infrastructure, fear of crime and victimization, and a break down of social trust and community participation. Social capital, a measure of the social environment that encompasses the emotional, economic, and informational resources available to an individual or a group through their social networks, has been hypothesized as a protective factor against depression. It has further been hypothesized as a potential explanatory factor for the association between neighborhood context and depression. There is, however, little consensus about which domains of social capital are most relevant to depression, and while much of the previous literature has shown how social capital differs across neighborhoods, relatively less research is devoted to understanding what causes this variation. Understanding how the neighborhood social and built environment influences depression risk can inform decisions about investing scarce resources in community-based mental health promotion, and may ultimately contribute to a reduction in the burden depression places on individuals and the healthcare system. The goal of this dissertation was to explore pathways linking neighborhood characteristics, social capital, and depression. Although previous literature has evaluated these potential associations, inferences are limited due to concerns about unmeasured genetic and childhood environment confounding and self-selection into neighborhoods. This dissertation contributes to the literature by partially addressing these methodological concerns through the use of a twin study.


This study uses data from the community-based University of Washington Twin Registry (UWTR). The UWTR contains over 8,000 monozygotic (identical, MZ) and dizygotic (fraternal, DZ) adult twins. It contains extensive survey data on sociodemographics, health behaviors and outcomes, and built environment measures linked to geocoded residential addresses. All twins included in the study were from same-sex pairs. Chapter One assesses the association between five neighborhood environment factors (neighborhood socioeconomic depression, crime, residential instability, gentrification, and income inequality) and depression. Chapter Two examines the association between different domains of social capital (cognitive and structural) and depression, and investigates whether social capital serves as a moderator or a mediator in the neighborhood socioeconomic deprivation-depression pathway. Chapter Three explores possible reasons for the variation of social capital across neighborhoods by assessing the associations between three built environment domains (neighborhood composition, pedestrian-oriented design, and commercial diversity) and social capital in the Puget Sound region of Washington State. All analyses employed a random intercept “within-between” twin model with the outcome hypothesized to follow a Poisson distribution. In this model, the individual-level outcome is regressed on the twin-pair mean exposure and the individual twin’s deviation from their twin pair mean. The within-pair effect is the main parameter of interest because it inherently controls for potential confounding due to genetic and childhood environment factors shared between twins within a pair.


In Chapter One, only neighborhood socioeconomic deprivation showed an association with depression in the unadjusted and adjusted models; specifically, higher deprivation was linked to greater depressive symptoms, independent of individual-level sociodemographic characteristics and population density. In Chapter Two, greater cognitive social capital, which refers to how individuals perceive their environment and was measured by sense of belonging, neighborhood social cohesion, workplace connections, and trust, was associated with fewer depressive symptoms. In contrast, structural social capital, which refers to social actions and behaviors and was measured by community participation, volunteerism, and social interactions, was not significantly associated with depression. Further, no social capital measure served as a mediator or a moderator in the neighborhood socioeconomic deprivation-depression association. Finally, in Chapter Three, only property values were linked to any of the neighborhood social capital variables, and these results were mixed. In the unadjusted models, property values were associated with greater sense of belonging, neighborhood social cohesion, and perceived safety; however, the associations were no longer significant in the models adjusted for sociodemographic characteristics. Further, no measure of pedestrian-oriented design or commercial diversity was associated with social capital.


Overall, these results indicate that both neighborhood socioeconomic deprivation and cognitive social capital are important risk factors for depression, independent of individual-level sociodemographics characteristics. These two factors do not, however, contribute to depression risk through the same pathway, nor does social capital influence the neighborhood socioeconomic deprivation-depression association. Additionally, the study does not provide evidence that differences in characteristics of the built environment can explain differences in the distribution of depression across neighborhoods. This suggests that there is not a strong association between these measures of social capital and built environment, and that specific interventions targeting the built environment may not improve social capital directly. The finding that social capital and neighborhood socioeconomic deprivation do influence depression risk is of great importance; however, other pathways to improving social capital, and therefore depression, will need to be explored.