The intersection of adverse childhood experiences and neighborhood determinants of health: An exploratory spatial analysis

The Association Between a Neighborhood Adverse Childhood Experiences Index and Body Mass Index Among New York City Youth

Background: The role of neighborhood factors in the association between adverse childhood experiences (ACEs) and body mass index (BMI) has not been widely studied. A neighborhood ACEs index (NAI) captures neighborhood environment factors associated with ACE exposure. This study examined associations between BMI and an NAI among New York City (NYC) youth. An exploratory objective examined the NAI geographic distribution across NYC neighborhoods. Methods: Data for students attending NYC public general education schools in kindergarten-12th grade from 2006-2017 (n = 1,753,867) were linked to 25 geospatial datasets capturing neighborhood characteristics for every census tract in NYC. Multivariable hierarchical linear regression tested associations between BMI and the NAI; analyses also were conducted by young (<8 years), school age (8-12 years), and adolescent (>12 years) subgroups. In addition, NAI was mapped by census tract, and local Moran's I identified clusters of high and low NAI neighborhoods. Results: Higher BMI was associated with higher NAI across all sex and age groups, with largest magnitude of associations for girls (medium NAI vs. low NAI: unstandardized β = 0.112 (SE 0.008), standardized β [effect size]=0.097, p < 0.001; high NAI vs. low NAI: unstandardized β = 0.195 (SE 0.008), standardized β = 0.178, p < 0.001) and adolescents (medium NAI vs. low NAI: unstandardized β = 0.189 (SE 0.014), standardized β = 0.161, p < 0.001, high NAI vs. low NAI: unstandardized β = 0.364 (SE 0.015), standardized β = 0.334, p < 0.001 for adolescent girls; medium NAI vs. low NAI: unstandardized β = 0.122 (SE 0.014), standardized β = 0.095, p < 0.001, high NAI vs. low NAI: unstandardized β = 0.217 (SE 0.015), standardized β = 0.187, p < 0.001 for adolescent boys). Each borough of NYC included clusters of neighborhoods with higher and lower NAI exposure, although clusters varied in size and patterns of geographic dispersion across boroughs. Conclusions: A spatial index capturing neighborhood environment factors associated with ACE exposure is associated with higher BMI among NYC youth. Findings complement prior literature about relationships between neighborhood environment and obesity risk, existing research documenting ACE-obesity associations, and the potential for neighborhood factors to be a source of adversity. Collectively, evidence suggests that trauma-informed place-based obesity reduction efforts merit further exploration as potential means to interrupt ACE-obesity associations.

Can adverse childhood experiences predict chronic health conditions? Development of trauma-informed, explainable machine learning models

Introduction

Decades of research have established the association between adverse childhood experiences (ACEs) and adult onset of chronic diseases, influenced by health behaviors and social determinants of health (SDoH). Machine Learning (ML) is a powerful tool for computing these complex associations and accurately predicting chronic health conditions.

Methods

Using the 2021 Behavioral Risk Factor Surveillance Survey, we developed several ML models-random forest, logistic regression, support vector machine, Naïve Bayes, and K-Nearest Neighbor-over data from a sample of 52,268 respondents. We predicted 13 chronic health conditions based on ACE history, health behaviors, SDoH, and demographics. We further assessed each variable's importance in outcome prediction for model interpretability. We evaluated model performance via the Area Under the Curve (AUC) score.

Results

With the inclusion of data on ACEs, our models outperformed or demonstrated similar accuracies to existing models in the literature that used SDoH to predict health outcomes. The most accurate models predicted diabetes, pulmonary diseases, and heart attacks. The random forest model was the most effective for diabetes (AUC = 0.784) and heart attacks (AUC = 0.732), and the logistic regression model most accurately predicted pulmonary diseases (AUC = 0.753). The strongest predictors across models were age, ever monitored blood sugar or blood pressure, count of the monitoring behaviors for blood sugar or blood pressure, BMI, time of last cholesterol check, employment status, income, count of vaccines received, health insurance status, and total ACEs. A cumulative measure of ACEs was a stronger predictor than individual ACEs.

Discussion

Our models can provide an interpretable, trauma-informed framework to identify and intervene with at-risk individuals early to prevent chronic health conditions and address their inequalities in the U.S.