Abstract

Introduction: As climate change accelerates severe weather events, electrical grid failures increasingly threaten public health and community stability. Socially vulnerable populations often bear a disproportionate burden of power outages and their associated health consequences. This dissertation examines energy resilience and health disparities in the St. Louis Metropolitan Statistical Area (MSA) through three interconnected aims: (1) mapping the spatial distribution of power outage frequency and duration across levels of social vulnerability and urbanicity; (2) evaluating the impact of long-duration power outages on emergency room (ER) visits and testing moderation by social vulnerability; and (3) applying community-engaged group model building (GMB) to map the systems influencing inequities in power outages in the region. Methods: For Aim 1, power outage data from 2022 to 2024 were collected across zip codes in the St. Louis MSA. Spatial econometric models were used to analyze outage frequency and duration relative to the CDC Social Vulnerability Index (SVI), urban-rural designations, total population, Medicare participation, and durable medical equipment (DME) reliance. For Aim 2, a narrative literature review informed the selection of health conditions sensitive to power loss, followed by spatial analysis to assess the relationship between long-duration outages (6+ hours) and ER visit rates. For Aim 3, a multi-stakeholder GMB workshop was conducted with regional experts to map the dynamic, interrelated factors underlying regional energy disparities. Results: Spatial modeling demonstrated that zip codes in the highest social vulnerability quartile (SVI Q4) experienced 22% more power outages and an average of 677.51 additional outage minutes compared to the lowest vulnerability quartile (Q1), holding covariates constant. Urban-rural designation showed no significant main effect in adjusted spatial models, while higher proportions of DME reliance were associated with lower outage frequencies. The narrative review confirmed links between outages and increased hospitalizations and ER visits for respiratory, cardiovascular, carbon monoxide, temperature-related, and medical device-dependent complications. Long-duration outages (6+ hours) were associated with increased ER visit rates but were not statistically significant. There was a high baseline ER rate in high vulnerability zip codes and no evidence that social vulnerability moderates the relationship between outages and ER visits except in the quartile with the lowest vulnerability. The community-engaged systems model identified reinforcing drivers that could be perpetuating disparities, including cycles of neighborhood disinvestment, utility service gaps, a disaster attention window affected by community perceptions and conflicting narratives, and systemic tensions between private market utility structures and the public good. Discussion: The findings demonstrate that power outage exposure and its downstream health impacts in St. Louis are spatially concentrated in historically marginalized, high-SVI neighborhoods, reflecting enduring legacies of urban segregation. Electrical grid vulnerabilities may exacerbate disease. Other studies have identified effects on the day of an outage and up to seven days following, as power disruptions increase ER visits for climate- and outage-sensitive health conditions. However, further analyses using models that adjust for high baseline health burden in socially vulnerable areas may be needed to understand the true effect of outages on the health of communities. Systems modeling reveals that technical grid upgrades alone cannot resolve inequities without addressing the underlying policy landscape and historic disinvestment patterns that stall community resilience. Conclusion: As climate change increases the risk of power disruptions, the ability of communities to power local systems will become increasingly important to advancing health in the U.S. This dissertation advances understanding of the relationships between power outages, social vulnerability and health in the St. Louis region. The findings demonstrate the value of considering energy resilience as a multidimensional public health issue shaped by interacting social, environmental and regulatory factors. They also highlight the value of engaging local organizations to better understand these issues within the context of local policy and planning. This dissertation contributes to a broader research framework for understanding how energy issues shape population health in the U.S. Incorporating health into energy resilience policy and planning could create opportunities to protect population health in a changing climate.

Committee Chair

Ross Brownson

Committee Members

Andrew Linhares; Ginger McKay; Jean-Francois Trani; Kim Johnson; Ross Hammond

Degree

Doctor of Philosophy (PhD)

Author's Department

Systems Science for Health Equity(SS4HE) Concentration

Author's School

Bursky School of Public Health

Document Type

Dissertation

Date of Award

8-12-2026

Language

English (en)

Download

Available for download on Monday, February 28, 2028

Included in

Public Health Commons

Share

COinS