Social Determinants Associated with COVID-19 Mortality in the United States

Shayom Debopadhaya, A. Sprague, Hongxi Mou, Tiburon L. Benavides, Sarah M. Ahn, Cole A. Reschke, John Erickson, Kristin P. BennettmedRxiv 2020 · DOI

Abstract

This study examines social determinants associated with disparities in COVID-19 mortality rates in the United States. Using county-level data, 42 negative binomial mixed models were used to evaluate the impact of social determinants on COVID-19 outcome. First, to identify proper controls, the effect of 24 high-risk factors on COVID-19 mortality rate was quantified. Then, the high-risk terms found to be significant were controlled for in an association study between 41 social determinants and COVID-19 mortality rates. The results describe that ethnic minorities, immigrants, socioeconomic inequalities, and early exposure to COVID-19 are associated with increased COVID-19 mortality, while the prevalence of asthma, suicide, and excessive drinking is associated with decreased mortality. Overall, we recognize that social inequality places disadvantaged groups at risk, which must be addressed through future policies and programs. Additionally, we reveal possible relationships between lung disease, mental health, and COVID-19 that need to be explored on a clinical level.

Introduction

By mid-2020, it was evident that COVID-19 mortality was not uniformly distributed across the United States. While clinical risk factors like age and pre-existing heart disease were well-documented, large-scale research into the social drivers of these disparities remained limited. This study addresses this gap by examining the "Social Determinants of Health" (SDOH) to uncover how systemic inequalities—ranging from language barriers to housing density—impacted mortality rates during the first wave of the pandemic. By controlling for known medical risks and outbreak timing, the researchers aimed to identify which social and economic factors independently drove significant changes in death rates.

Methods

The researchers employed a two-stage statistical approach using negative binomial mixed models, which are specifically designed to handle "overdispersed" count data where the variance is much higher than the mean. The study aggregated data for 3,093 U.S. counties from the CDC, Johns Hopkins University, and County Health Rankings.

  1. High-Risk Control Identification: The team first tested 24 established risk factors, including race, age, comorbidities, and temporal variables (e.g., "days since first infection"), to identify statistically significant controls.
  2. Social Determinant Screening: Using the significant factors from the first stage as controls, they conducted 41 separate model iterations, each testing a different social determinant, such as air pollution, graduation rates, and housing conditions.
  3. Statistical Rigor: To avoid false discoveries, the Benjamini-Hochberg Procedure was applied to adjust p-values across the multiple tests.

Results

High-Risk Factors and the "Asthma Paradox"

The initial screening of high-risk factors revealed that the timing of the outbreak was the strongest predictor of mortality, with "Days Since First Infection" associated with a 124.94% increase in the mortality rate ratio (MRR). Other significant factors included the percentage of African American residents and the prevalence of lung cancer in a county. Interestingly, higher median household income was also associated with a 22.00% increase in mortality, likely reflecting the high density of early-hit urban hubs.

Surprisingly, the model identified an inverse relationship between adult asthma prevalence and COVID-19 mortality. A standard deviation increase in county-level asthma was associated with a 20.06% decrease in mortality. This "asthma paradox" suggests a potential protective effect, possibly linked to the use of inhaled corticosteroids, which warrants further clinical investigation.

Table 1: Coefficients for high-risk factors and their impact on COVID-19 mortality rates per standard deviation increase.

Impact of Social Determinants

After controlling for high-risk factors, several social determinants emerged as powerful predictors of mortality. The percentage of the population that is uninsured and the percentage not proficient in English were among the most significant drivers, with MRR increases of 40.23% and 39.65%, respectively. Environmental factors also played a major role; higher levels of fine particulate matter (PM 2.5) were linked to a 36.78% increase in mortality.

Table 2: Social determinants associated with COVID-19 mortality rates after controlling for significant high-risk factors.

Socioeconomic factors such as overcrowded housing (24.01% increase) and lack of a high school diploma (14.13% increase) further highlighted the link between poverty and pandemic vulnerability. Conversely, higher rates of suicide (-12.13%) and excessive drinking (-17.25%) were associated with lower mortality, which may reflect lower viral spread in rural areas where these "deaths of despair" are more prevalent.

Discussion and Conclusion

These findings underscore that social inequality is a primary driver of health disparities. The strong association between limited English proficiency and mortality highlights the critical need for accessible public health communication. Similarly, the impact of overcrowding demonstrates that social distancing is often a luxury unavailable to low-income populations.

While the study is limited by the ecological fallacy—meaning county-level trends may not apply to individuals—and represents a specific snapshot in time, it provides a vital roadmap for policy. Addressing systemic racial and economic inequalities is not just a matter of social justice but a necessary component of pandemic resilience. Future work should focus on clinical-level research to explore the biological mechanisms behind the observed comorbidities and long-term policy changes to protect vulnerable communities.

Summary updated

Abstract

Abstract This study examines social determinants associated with disparities in COVID-19 mortality rates in the United States. Using county-level data, 42 negative binomial mixed models were used to evaluate the impact of social determinants on COVID-19 outcome. First, to identify proper controls, the effect of 24 high-risk factors on COVID-19 mortality rate was quantified. Then, the high-risk terms found to be significant were controlled for in an association study between 41 social determinants and COVID-19 mortality rates. The results describe that ethnic minorities, immigrants, socioeconomic inequalities, and early exposure to COVID-19 are associated with increased COVID-19 mortality, while the prevalence of asthma, suicide, and excessive drinking is associated with decreased mortality. Overall, we recognize that social inequality places disadvantaged groups at risk, which must be addressed through future policies and programs. Additionally, we reveal possible relationships between lung disease, mental health, and COVID-19 that need to be explored on a clinical level.
BibTeX
@misc{debopadhaya2020social,
  author = {Shayom Debopadhaya and A. Sprague and Hongxi Mou and Tiburon L. Benavides and Sarah M. Ahn and Cole A. Reschke and John Erickson and Kristin P. Bennett},
  title = {{Social Determinants Associated with COVID-19 Mortality in the United States}},
  year = {2020},
  doi = {10.1101/2020.08.28.20183848},
  howpublished = {medRxiv},
}
Figure