tetano
Editor, Senior Moderator
Front Public Health
. 2024 Jun 11:12:1359192.
doi: 10.3389/fpubh.2024.1359192. eCollection 2024. Socioeconomic and sociodemographic correlations to COVID-19 variability in the United States in 2020
Nikolay Golosov[SUP] 1 [/SUP], Shujie Wang[SUP] 1 2 [/SUP], Manzhu Yu[SUP] 1 [/SUP], Nakul N Karle[SUP] 3 [/SUP], Oye Ideki[SUP] 3 [/SUP], Bishara Abdul-Hamid[SUP] 4 [/SUP], Christopher Blaszczak-Boxe[SUP] 3 [/SUP]
Affiliations
The COVID-19 pandemic provided an additional spotlight on the longstanding socioeconomic/health impacts of redlining and has added to the myriad of environmental justice issues, which has caused significant loss of life, health, and productive work. The Centers for Disease Control and Prevention (CDC) reports that a person with any selected underlying health conditions is more likely to experience severe COVID-19 symptoms, with more than 81% of COVID-19-related deaths among people aged 65 years and older. The effects of COVID-19 are not homogeneous across populations, varying by socioeconomic status, PM[SUB]2.5[/SUB] exposure, and geographic location. This variability is supported by analysis of existing data as a function of the number of cases and deaths per capita/1,00,000 persons. We investigate the degree of correlation between these parameters, excluding health conditions and age. We found that socioeconomic variables alone contribute to ~40% of COVID-19 variability, while socioeconomic parameters, combined with political affiliation, geographic location, and PM[SUB]2.5[/SUB] exposure levels, can explain ~60% of COVID-19 variability per capita when using an OLS regression model; socioeconomic factors contribute ~28% to COVID-19-related deaths. Using spatial coordinates in a Random Forest (RF) regressor model significantly improves prediction accuracy by ~120%. Data visualization products reinforce the fact that the number of COVID-19 deaths represents 1% of COVID-19 cases in the US and globally. A larger number of democratic voters, larger per-capita income, and age >65 years is negatively correlated (associated with a decrease) with the number of COVID cases per capita. Several distinct regions of negative and positive correlations are apparent, which are dominated by two major regions of anticorrelation: (1) the West Coast, which exhibits high PM[SUB]2.5[/SUB] concentrations and fewer COVID-19 cases; and (2) the middle portion of the US, showing mostly high number of COVID-19 cases and low PM[SUB]2.5[/SUB] concentrations. This paper underscores the importance of exercising caution and prudence when making definitive causal statements about the contribution of air quality constituents (such as PM[SUB]2.5[/SUB]) and socioeconomic factors to COVID-19 mortality rates. It also highlights the importance of implementing better health/lifestyle practices and examines the impact of COVID-19 on vulnerable populations, particularly regarding preexisting health conditions and age. Although PM[SUB]2.5[/SUB] contributes comparable deaths (~7M) per year, globally as smoking cigarettes (~8.5M), quantifying any causal contribution toward COVID-19 is non-trivial, given the primary causes of COVID-19 death and confounding factors. This becomes more complicated as air pollution was reduced significantly during the lockdowns, especially during 2020. This statistical analysis provides a modular framework, that can be further expanded with the context of multilevel analysis (MLA). This study highlights the need to address socioeconomic and environmental disparities to better prepare for future pandemics. By understanding how factors such as socioeconomic status, political affiliation, geographic location, and PM[SUB]2.5[/SUB] exposure contribute to the variability in COVID-19 outcomes, policymakers and public health officials can develop targeted strategies to protect vulnerable populations. Implementing improved health and lifestyle practices and mitigating environmental hazards will be essential in reducing the impact of future public health crises on marginalized communities. These insights can guide the development of more resilient and equitable health systems capable of responding effectively to similar future scenarios.
Keywords: COVID-19; PM2.5; ordinary least squares (OLS); random forest (RF); socioeconomic variables; statistical correlation.
. 2024 Jun 11:12:1359192.
doi: 10.3389/fpubh.2024.1359192. eCollection 2024. Socioeconomic and sociodemographic correlations to COVID-19 variability in the United States in 2020
Nikolay Golosov[SUP] 1 [/SUP], Shujie Wang[SUP] 1 2 [/SUP], Manzhu Yu[SUP] 1 [/SUP], Nakul N Karle[SUP] 3 [/SUP], Oye Ideki[SUP] 3 [/SUP], Bishara Abdul-Hamid[SUP] 4 [/SUP], Christopher Blaszczak-Boxe[SUP] 3 [/SUP]
Affiliations
- PMID: 38919927
- PMCID: PMC11196783
- DOI: 10.3389/fpubh.2024.1359192
The COVID-19 pandemic provided an additional spotlight on the longstanding socioeconomic/health impacts of redlining and has added to the myriad of environmental justice issues, which has caused significant loss of life, health, and productive work. The Centers for Disease Control and Prevention (CDC) reports that a person with any selected underlying health conditions is more likely to experience severe COVID-19 symptoms, with more than 81% of COVID-19-related deaths among people aged 65 years and older. The effects of COVID-19 are not homogeneous across populations, varying by socioeconomic status, PM[SUB]2.5[/SUB] exposure, and geographic location. This variability is supported by analysis of existing data as a function of the number of cases and deaths per capita/1,00,000 persons. We investigate the degree of correlation between these parameters, excluding health conditions and age. We found that socioeconomic variables alone contribute to ~40% of COVID-19 variability, while socioeconomic parameters, combined with political affiliation, geographic location, and PM[SUB]2.5[/SUB] exposure levels, can explain ~60% of COVID-19 variability per capita when using an OLS regression model; socioeconomic factors contribute ~28% to COVID-19-related deaths. Using spatial coordinates in a Random Forest (RF) regressor model significantly improves prediction accuracy by ~120%. Data visualization products reinforce the fact that the number of COVID-19 deaths represents 1% of COVID-19 cases in the US and globally. A larger number of democratic voters, larger per-capita income, and age >65 years is negatively correlated (associated with a decrease) with the number of COVID cases per capita. Several distinct regions of negative and positive correlations are apparent, which are dominated by two major regions of anticorrelation: (1) the West Coast, which exhibits high PM[SUB]2.5[/SUB] concentrations and fewer COVID-19 cases; and (2) the middle portion of the US, showing mostly high number of COVID-19 cases and low PM[SUB]2.5[/SUB] concentrations. This paper underscores the importance of exercising caution and prudence when making definitive causal statements about the contribution of air quality constituents (such as PM[SUB]2.5[/SUB]) and socioeconomic factors to COVID-19 mortality rates. It also highlights the importance of implementing better health/lifestyle practices and examines the impact of COVID-19 on vulnerable populations, particularly regarding preexisting health conditions and age. Although PM[SUB]2.5[/SUB] contributes comparable deaths (~7M) per year, globally as smoking cigarettes (~8.5M), quantifying any causal contribution toward COVID-19 is non-trivial, given the primary causes of COVID-19 death and confounding factors. This becomes more complicated as air pollution was reduced significantly during the lockdowns, especially during 2020. This statistical analysis provides a modular framework, that can be further expanded with the context of multilevel analysis (MLA). This study highlights the need to address socioeconomic and environmental disparities to better prepare for future pandemics. By understanding how factors such as socioeconomic status, political affiliation, geographic location, and PM[SUB]2.5[/SUB] exposure contribute to the variability in COVID-19 outcomes, policymakers and public health officials can develop targeted strategies to protect vulnerable populations. Implementing improved health and lifestyle practices and mitigating environmental hazards will be essential in reducing the impact of future public health crises on marginalized communities. These insights can guide the development of more resilient and equitable health systems capable of responding effectively to similar future scenarios.
Keywords: COVID-19; PM2.5; ordinary least squares (OLS); random forest (RF); socioeconomic variables; statistical correlation.