Social Drivers of Mental Health: A U.S. Study Using Machine Learning

Area of Science:

• Public Health
• Data Science
• Social Epidemiology

Background:

• Social determinants significantly influence population mental health.
• Aggregated-level comparisons of social drivers are feasible.
• Machine learning offers a novel approach to analyze these complex relationships.

Purpose of the Study:

• To identify and rank social drivers of mental health at the census tract level in the U.S.
• To analyze the influence of behavioral, environmental, and social domains on mental health outcomes.
• To explore how poverty and racial segregation moderate these effects.

Main Methods:

• Utilized data from 38,379 U.S. census tracts (2021).
• Employed Extreme Gradient Boosting machine learning (2022) to analyze two mental health indicators (depression, poor mental health) against three social driver domains.
• Examined leading drivers within main and subsample analyses stratified by poverty and racial segregation.

Main Results:

• Social drivers explained over 90% of the variance in mental health indicators.
• Smoking (behavioral) was a common driver for both depression and poor mental health.
• Climate zone (environmental) and racial composition (social) emerged as significant correlates, varying by census tract characteristics like poverty and segregation.

Conclusions:

• Population mental health is deeply contextualized by local social drivers.
• Census tract-level analysis is crucial for understanding upstream causes of mental health issues.
• Tailored interventions based on these granular analyses can improve population mental health outcomes.