Changes in the Neighborhood Built Environment and Chronic Health Conditions in Washington, DC, in 2014-2019:...

Area of Science:

• Urban planning and public health research.
• Utilizing geospatial technology and deep learning for environmental health studies.

Background:

• Traditional neighborhood studies often use cross-sectional designs.
• Google Street View (GSV) images offer scalable, longitudinal data for built environment assessment.

Purpose of the Study:

• To analyze longitudinal changes in Washington, DC's built environment and demographics from 2014-2019.
• To examine associations between these changes and health outcomes using GSV and computer vision.

Main Methods:

• Systematic sampling of 434,115 GSV images at 100m intervals.
• Convolutional neural networks (CNNs) for built environment feature extraction.
• Multilevel mixed-effects models to assess built environment-health outcome associations.

Main Results:

• Shift towards higher-density housing (66%-72%) and single-lane roads (37%-42%).
• Gentrification indicated by increased college-educated residents (16%-41%), income, and property values.
• Increased construction linked to lower obesity, diabetes, high cholesterol, and cancer rates; higher African American neighborhoods saw less construction.

Conclusions:

• Washington, DC underwent significant urban transformation with demographic shifts.
• Urban development is linked to reduced chronic disease prevalence.
• Highlights the complex interplay of urban development, demographics, and health, necessitating further research on community impacts.