Exploring the Association Between Structural Racism and Mental Health: Geospatial and Machine Learning Analysis

Area of Science:

• Public Health
• Geospatial Analysis
• Deep Learning

Background:

• Structural racism is a known driver of mental health disparities.
• Previous research has focused on individual factors, with less exploration of their collective impact as manifestations of structural racism.
• Milwaukee County's diversity provides a unique setting for investigating these multifactorial influences.

Purpose of the Study:

• To delineate the association between structural racism and mental health disparities in Milwaukee County.
• To employ geospatial and deep learning techniques for a comprehensive analysis.
• To quantify the impact of structural racism on poor mental health prevalence.

Main Methods:

• Compiled 217 georeferenced variables, initially excluding race to identify nonracial determinants.
• Utilized tree-based methods (random forest) and conventional techniques for variable selection, addressing multicollinearity.
• Applied geographically weighted random forest and self-organizing maps with K-means clustering to analyze spatial heterogeneity and quantify racism's impact.

Main Results:

• Twelve influential factors explained 95.11% of mental health variability.
• Top factors included smoking, poverty, insufficient sleep, lack of health insurance, employment, and age.
• African American neighborhoods showed a 2.23 times higher likelihood of high-risk mental health clusters.

Conclusions:

• Structural racism demonstrably shapes mental health disparities, with Black communities bearing a disproportionate burden.
• The integrated geospatial and deep learning approach effectively elucidates complex social determinants of mental health.
• Findings underscore the necessity for targeted interventions addressing both individual and systemic factors to reduce mental health inequities.