Homelessness and Housing Insecurity: Longitudinal Investigations to Better Understand Biological Embedding During Childhood

Abstract

To date, research studying the impact of housing insecurity and homelessness on child health and development is limited. Most of the extant literature on childhood housing insecurity is cross-sectional in design and thus restricts our understanding of its causal relationships with health outcomes and of the mechanisms by which it “gets under the skin.” With three complementary chapters, my dissertation is guided by an ecobiodevelopmental framework to elucidate the complexities of childhood homelessness. In the first chapter, I use longitudinal data from the Great Smoky Mountains Study (GSMS) to investigate the impact of childhood housing insecurity on subsequent anxiety and depression symptom scores during both childhood and adulthood. To optimize the repeated measures data available in this cohort study, this chapter triangulates fixed- and random-effect models to provide strong evidence of causal relationships after rigorous adjustment of family income poverty. Specifically, I estimate that exposure to childhood housing insecurity is prospectively associated with higher childhood anxiety symptom scores (Fixed effects: standardized mean difference (SMD)=0.21; 95% CI, 0.12, 0.30; Random effects: SMD=0.25, 95% CI, 0.15, 0.35) and higher childhood depression symptom scores (Fixed effects: SMD=0.18, 95% CI, 0.09, 0.28; Random effects: SMD=0.26, 95% CI 0.14, 0.37). Additionally, I report that childhood housing insecurity is associated with higher depression symptom scores (SMD=0.11, 95% CI, 0.00, 0.21) in adulthood. In my second and third chapters, I investigate the relationships between childhood housing insecurity, homelessness, and inflammatory conditions and the related C-reactive protein (CRP) biomarker. Specifically, in my second chapter, I again use GSMS data to analyze the relationship between childhood housing insecurity and subsequent CRP levels. Importantly, I use g-estimation of structural nest mean models to rigorously explore time-varying causal relationships to critically evaluate potentially mechanisms of biological embedding. Here, I report that any exposure to childhood housing insecurity between ages 11 and 16 leads to a 6.38% increase in CRP levels. Importantly, this estimate can be interpreted as a conditional average causal effect and can thus be extrapolated to infer that multiple exposures to childhood housing insecurity during these ages would lead to a multiplicative increase in CRP levels. In my third chapter, I use longitudinal data from the Avon Longitudinal Study of Parents and Children (ALSPAC) to examine the impact of early childhood homelessness (experienced between the prenatal period and age 5 years) on respiratory functioning during primary school ages (6 to 9 years). This chapter utilizes both multilevel longitudinal modeling and generalized estimating equations (GEE) methods to strengthen the support for the observed associations. Using a random intercepts model, I show that participants who experienced early childhood homelessness had higher odds of poor respiratory functioning than participants who did not experience early childhood homelessness (conditional odds ratio [OR]: 1.63, [95% CI: 1.06- 2.53]). These results align with the marginal OR found using logistic GEE models. Taken together, the second and third papers quantify the association between childhood housing insecurity and an inflammatory biomarker (i.e., CRP) and then examine potential associations between a form of childhood housing insecurity and a chronic inflammatory condition. Together, these three chapters tell a cohesive story of the adverse impacts of housing insecurity and homelessness on short- and long-term mental and physical health outcomes and identify a potential pathway for their biological embedding during childhood. Importantly, these chapter signify that housing insecurity is more than a consequence of family income poverty and provide the arguably the strongest evidence of causal relationships with poor health outcomes that can be obtained with observational data. Ultimately, these chapters clearly demonstrate the critical need to intervene upon and prevent childhood housing insecurity and homelessness.