About this srcd poster session

---

Panel information
Panel 28. Solicited Content: Climate Change

Abstract

---

Introduction: Neighborhoods are important development contexts which are stratified by structural disadvantage and characterized by the clustering of potential risk and protective factors. However, existing studies often examine neighborhood exposures separately, while much research on neighborhood types (NTs) and developmental health includes neighborhood disadvantage (ND) as a characteristic rather than a separate upstream stratification factor that may influence or interact with neighborhood types. Consequently, this study sought to (a) identify neighborhood types (NTs) across the US incorporating a range of characteristics, separating upstream ND factors; (b) examine the relationship between ND and NTs; and (c) determine the association between NTs and children’s physical, mental, and neurocognitive health.

Methods: We used baseline data for 9- to 10-years-olds from the national Adolescent Brain and Cognitive Development (ABCD) Study® (n=11,192) across 21 U.S. sites. Primary residential address was geocoded and linked to the American Community Survey and additional datasets capturing environmental data, largely at the census tract level (Cardenas-Iniguez et al., 2024; Fan et al., 2021). ND was measured as a factor based on the percentage of families below the federal poverty level, single-parent households, residents with high school diploma, unemployment rate, and median family income. Outcome measures included neurocognitive performance (NIH Toolbox), internalizing / externalizing behavior (Child Behavior Checklist), body-mass-index (BMI), and sleep problems (Sleep Disturbance Scale). First, we conducted latent profile analysis (LPA) to identify NTs based on 26 variables across eight domains (air pollution, heat, greenness, urbanicity, services, housing, infrastructure, insurance) selecting the optimal number of NTs (i.e., profiles) using statistical indicators (e.g., BIC, Lo-Mendell-Rubin test) and theory. Second, we performed ANOVA to test differences in ND scores across NTs. Finally, we employed linear mixed effects models to examine associations between NTs and physical, mental and neurocognitive outcomes, adjusting for age, sex, race/ethnicity, household income-to-needs ratio, and parents’ educational attainment, and a random intercept for site.

Results: The LPA identified an eight-NT solution as the best fit (BIC: 715570, LMR test: p<0.001, entropy= 0.925), with class characteristics and size outlined in Figure 1a. As illustrated in Figure. 1b, the NTs capture national and within-site variations in neighborhoods. The ND score differed significantly between all NTs (p<0.001), with two classes (6,8) capturing the most disadvantaged neighborhoods (Figure 1c). Overall, NTs significantly predicted 5 of 7 NIH toolbox tasks and all three composites, BMI, sleep, and externalizing problems (all p<0.05). While the full pattern of comparisons across NTs is varied (Table 1), Class 3 exhibited the highest and Class 6 exhibited the lowest level of neurocognitive performance across tasks, and Class 8 exhibited the highest BMI and fewest sleep disturbances.

Conclusion: These findings reveal distinct clusters of neighborhood exposures, referred to as neighborhood types, which vary by U.S. region are stratified by ND level. NTs predict neurocognitive performance, mental and physical health, though the full pattern is nuanced. Future research should further investigate the potential mechanistic or interactive relationships between ND and NTs to enhance understanding of neighborhood effects and inform policies targeting upstream and place-based determinants of health and development.

Author information

Author	Role
Jiahao Wang, University of Southern California	Presenting author
Jared N. Schachner, University of Southern California	Non-presenting author
David J.M. Kraemer, Dartmouth College	Non-presenting author
Annie Kwon, University of Southern California	Non-presenting author
Morgan S. Polikoff, University of Southern California	Non-presenting author
Megan M. Herting, University of Southern California; Children’s Hospital Los Angeles	Non-presenting author
Elizabeth R. Sowell, University of Southern California; Children’s Hospital Los Angeles	Non-presenting author
Kiros Berhane, Columbia University	Non-presenting author
Santiago Morales, University of Southern California	Non-presenting author
Daniel A. Hackman, University of Southern California	Non-presenting author

⇦ Back to session