key: cord-0754420-5ava6xbk
authors: KC, M.; Oral, E.; Straif-Bourgeois, S.; Rung, A. L.; Peters, E. S.
title: The Effect of Area Deprivation on COVID-19 Risk in Louisiana
date: 2020-08-26
journal: nan
DOI: 10.1101/2020.08.24.20180893
sha: dfc447ebef9c01e2e963fc0ade13b21c87c19036
doc_id: 754420
cord_uid: 5ava6xbk

Purpose: Louisiana currently has the highest per capita case count for COVID-19 in the United States and disproportionately affects the Black or African American population. Neighborhood deprivation has been observed to be associated with poorer health outcomes. The purpose of this study was to examine the relationship between neighborhood deprivation and COVID-19 in Louisiana. Methods: The Area Deprivation Index (ADI) was calculated and used to classify neighborhood deprivation at the census tract level. A total of 17 US census variables were used to calculate the ADI for each of the 1148 census tracts in Louisiana. The data were extracted from the American Community Survey (ACS) 2018. The neighborhoods were categorized into quintiles as well as low and high deprivation. The publicly available COVID-19 cumulative case counts by census tract was obtained from the Louisiana Department of Health website on July 31, 2020. Descriptive and Poisson regression analyses were performed. Results: Neighborhoods in Louisiana were substantially different with respect to deprivation. The ADI ranged from 136.00 for the most deprived neighborhood and -33.87 in the least deprived neighborhood. We observed that individuals residing in the most deprived neighborhoods had a 45% higher risk of COVID-19 disease compared to those residing in the least deprived neighborhoods. Conclusion: While the majority of previous studies were focused on very limited socio-environmental factors such as crowding and income, this study used a composite area-based deprivation index to examine the role of neighborhood environment on COVID-19. We observed a positive relationship between neighborhood deprivation and COVID-19 risk in Louisiana. The study findings can be utilized to promote public health preventions measures besides social distancing, wearing a mask while in public and frequent handwashing in vulnerable neighborhoods with greater deprivation.

On March 09, 2020, Louisiana reported its first case of COVID-19 and soon 60 thereafter appeared to be a hot spot of the coronavirus pandemic in the US [1] . Within 61 two weeks of the initial confirmed case, the state had one of the world's highest average 62 daily growth rate [2] [3] [4] . As of July 31, 2020, the state of Louisiana has the highest per 63 capita case count in the United States with a total of 116,280 confirmed cases and 64 3,835 deaths [2] . The incidence and mortality rates of COVID-19 has been 65 disproportionate across racial and ethnic groups [5, 6] . Specifically, non-Hispanic 66

African Americans have higher rates of incidence, hospitalization, and death from 67 COVID-19 compared to non-Hispanic Whites. In early July, the US Centers for Disease 68

Control and Prevention (CDC) estimated that non-Hispanic African Americans have 4.7 69 times the rate of age-adjusted COVID-19 related hospitalization rates than non-Hispanic 70

Whites [7] . The sources of disparities in COVID-19 outcomes might be explained from a 71 social determinants of health perspective. Non-Hispanic African Americans are more 72 likely to have vulnerable and low-paying jobs that don't allow remote work, which 73 increases risk of contracting 9] . Furthermore, non-Hispanic African 74

Americans are more likely to rely on public transportation and to live in crowded housing 75 or work in crowded worksite that places them an increased risk for COVID-19 disease. 76

African Americans exhibit a greater burden of chronic medical conditions, such as 77 hypertension, diabetes, heart disease, chronic disease, and obesity that increase the 78 severity of COVID-19 illness [10] [11] [12] . In Louisiana, 2.9 million people have at least one 79 chronic condition, and a total of 68 percentage of Louisiana adults are overweight or 80 obese [13] . Furthermore, the poverty rate is much higher among African Americans as 81 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020 . . https://doi.org/10.1101 those of non-Hispanic Whites, and are concentrated in neighborhoods with high poverty 82 [14, 15] . The neighborhood socioeconomic status (SES) is linked to access to health 83 care services, people residing in low SES neighborhoods are less likely to have access 84 to health care services, which further increases the risk of adverse health outcomes 85 related to COVID-19, such as higher hospitalizations and mortality [16, 17] is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 26, 2020 . . https://doi.org/10.1101 Socioeconomic characteristics of residential neighborhoods influence health-related 104 behaviors, conditions, and health outcomes [32, 33] . Deprived neighborhoods are 105 correlated with health risk behaviors, overcrowding, less social cohesion, and higher 106 levels of environmental pollutants, and has been identified as a critical social 107 determinant of health [34] [35] [36] [37] . Low socioeconomic status (SES), often regarded as a 108 fundamental cause of disease, has been shown to increase the risk of COVID-19 109 because it impacts access to fundamental resources that an individual or a 110 neighborhood may require to avoid 38] . 111

Neighborhoods with a higher number of people per household or room tend to 112 have a higher rate of confirmed COVID-19 cases than neighborhoods with fewer 113 residents [25, 39, 40] . Individuals who share a room or live in overcrowded housing and 114 the use of public transportation often spread the disease rapidly as distancing 115 preventive measures are impossible to adopt. 116

In this study, we used the Area Deprivation Index (ADI) to measure 117 neighborhood deprivation. The ADI is a composite measure of neighborhood 118 socioeconomic disadvantage, created by Gopal K Singh in 2003 [41] . The ADI, 119 composed of 17 education, employment, housing-quality, and poverty census derived 120 measures, is a robust metric measuring many relevant social determinants of health 121 that may help explain the socio-biologic mechanisms of disease [41, 42] . We disease. The studies that exist are limited, examining only a couple of specific risk 126 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint factors, such as overcrowding and income. The use of the Area Deprivation Index (ADI) 127 in the present study includes 17 neighborhood-level factors and provides a robust 128 measure of neighborhood deprivation. The primary purpose of this paper is to 129 investigate the relationship between neighborhood deprivation and COVID-19 risk in 130

Louisiana. 131

Publicly available data on cumulative COVID-19 cases by census tract was 134 

The main outcome in this study was COVID-19 cases per 1,000 persons in 142

Louisiana census tracts as of July 31, 2020. 143

Neighborhood deprivation was measured by the ADI, as described by Singh in 145 2003 [41] . ADI is a validated, factor-based deprivation index that uses 17 census 146 derived measures of poverty, education, housing, and employment indicators at the 147 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint census tract level to classify the neighborhoods [41, 45] . More-disadvantaged 148 neighborhoods are those with a higher ADI score. 149

The census indicators used in the calculation of ADI include educational 150 distribution (percentage of the population with less than 9 years and with 12 or more 151 years of education), median family income, median home value, median gross rent, 152 median monthly mortgage, income disparity, unemployment, percent employed person 153 in white-collar occupation, percent families below poverty, percent population below 154 150% poverty threshold, single-parent household rate, homeownership rate, percent 155 household without a telephone, percent household without a motor vehicle, percent 156 occupied housing units without complete plumbing, and household crowding [41, 45] . 157

Data from the Census Bureau's American Community Survey (ACS) 2018 was 159 used to calculate the ADI score. The 17 US census indicators were multiplied by the 160

Singh's coefficients (factor weights) for all census tracts in Louisiana [41, 46] . 161

The base score of each indicator was summed to get the total base score for a census 162 tract. Each census tract's base score was standardized by dividing the difference 163 

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10. 1101 where j represents the j th census tract, and k is the total number of census tracts in Louisiana.

Finally, the standardized values were adjusted to a base mean of 100 and a standard deviation 168 of 20 as suggested by Knighton et al [46] . 169 = ( + 100) * 20.

The details of ADI calculation and a list of variables included in the calculations can be 171 found in Knighton et al [46] . 172

Based on the ADI scores, the census tracts were categorized into quintiles of 173 deprivation; they were also dichotomized as either low deprivation or high deprivation. 174

The median Louisiana ADI was used to dichotomize the census tracts. Census tracts 175

with missing values for the indicators were excluding while calculating the ADI. to estimate risk ratio. An offset variable was used, and the model was corrected for over 182 dispersion. 183

There was a substantial difference between the ADI of the least deprived and most 185 deprived neighborhoods. The overall median (IQR) ADI for Louisiana was 104.32 186 (76.00), with the most deprived neighborhood having an ADI of 136.00, and the least 187 deprived neighborhood having an ADI of -33.87. While the median ADI of the least 188 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint deprived neighborhood was 76.00, the median ADI of the most deprived neighborhood 189 was 118.45 (Table 1) . 190 CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020 . . https://doi.org/10.1101 In figure 1 , the census tracts in red represent the most deprived neighborhoods, 219 while the census tracts in green are the least deprived neighborhoods in Louisiana. In 220 figure 2, the census tracts in yellow represent census tracts with fewer COVID-19 cases 221 per 1,000 persons as of July 31, 2020, while the census tracts in brown and dark brown 222 represent higher COVID-19 cases per 1,000 persons. Figure 3 shows the distribution of 223 ADI and COVID-19 cases per 1,000 persons simultaneously in Louisiana by census 224

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020 . . https://doi.org/10.1101 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint status and overcrowded housing may explain why non-Hispanic African American and 254

Hispanic populations are at higher risk of getting In addition to overcrowding and neighborhood-level SES, the disparities in 256 COVID-19 cases between neighborhoods might be directly related to the nature of 257 residents' occupations, a lack of telecommunication infrastructure, use of public 258 transportation, and utility disruptions. Areas with concentrated poverty and extreme 259 racial segregation had a higher incidence of 50, 51] . Since low-income isolating themselves, however, asymptomatic cases and symptomatic individuals who 273 don't get paid sick leave or are essential workers are likely to spread the disease more 274 rapidly. As of July 29, 2020, there were more than 100 worksite outbreaks in Louisiana 275

[2]. The majority of studies have emphasized how adversely affected by COVID-19 276 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint certain racial and ethnic communities are; however, these groups of people may have 277 differential exposure to the virus due to long-standing systemic health and social 278

This study has several limitations. Due to a lack of data, we were unable to 280 account for COVID-19 testing per census tract in our statistical analysis or perform a 281 time series analysis of COVID-19 case counts. Similarly, we couldn't include the data on 282 COVID-19 testing in a deprived neighborhood and the availability of free testing clinics 283 in our analysis. This study is limited only to COVID-19 cases per 1,000 persons in 284

Louisiana census tracts, the severe outcomes such as hospitalizations including 285

Intensive Care Units (ICUs) admissions, and mortality were not assessed. Another 286 limitation is that the impact of race couldn't be examined due to the lack of data at the 287 census tract by race. 288

A key strength of this study is the use of the ADI to characterize neighborhood 289 disadvantage. The ADI is a validated and becoming more widely used composite index 290 of neighborhood disadvantage. The ADI provides a robust method to identify and 291 classify deprived neighborhoods. The use of the most relevant social determinants of 292 health in the calculation of ADI allows for better contextualization of the neighborhood. 293

Despite these limitations, we believe that this study contributes to the literature on social 294 determinants of health and COVID-19 disease in the neighborhood by establishing the 295 relationship between the neighborhood deprivation and COVID-19 cases in Louisiana. 296

Findings may help authorities to prioritize the public health response especially by 297 increasing free testing sites and contact tracing in the targeted areas. In addition, it is 298 important to promote public health preventions measures for case isolation and 299 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 26, 2020. . https://doi.org/10.1101/2020.08.24.20180893 doi: medRxiv preprint quarantine of close contacts besides social distancing, wearing a mask while in public 300 and frequent handwashing to ultimately reduce the spread of COVID-19 in the most 301 vulnerable populations. 302

We observed a great disparity in deprivation among Louisiana neighborhoods. We also 304 found an association between neighborhood deprivation and the COVID-19 cases per 305 1,000 persons in Louisiana. There have been many studies on how COVID-19 is 306 clustered in neighborhoods, however, future studies should explore specific 307 mechanisms behind this association. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 26, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 26, 2020 . . https://doi.org/10.1101 

Louisiana Department of Health. Gov. Edwards Confirms Louisiana's First 321

Presumptive Positive Case of COVID-19

COVID-19) Information

The Next Coronavirus Hot Spot: Louisiana Races to 325 Prepare for the Worst: As number of cases jump

Racial Health Disparities and Covid-19 -Caution 330 and Context

COVID-19 And Racial/Ethnic Disparities In Health Risk

And Household Composition. Health Aff (Millwood)

Centers for Disease Control and Prevetion. A weekly Surveillance Summary of 335 U.S. COVID-19 Activity

338 Determinants of disparities in COVID-19 job losses

343 13. Partnership to fight chronic disease. WHAT IS THE IMPACT OF CHRONIC 354 DISEASE ON LOUISIANA? : FightChronicDisease.org

Poverty among minorities in the United States: explaining the racial 356 poverty gap for Blacks and Latinos

For blacks in America, the gap in neighborhood poverty 358 has declined faster than segregation

The Disproportionate Impact of COVID-19 on Racial and Ethnic 360 Minorities in the United States

The Disproportionate Impact 362 of COVID-19 on Racial and Ethnic Minorities in the United States

Low-Income and 364 Communities of Color at Higher Risk of Serious Illness if Infected with Coronavirus

Assessing 367 Differential Impacts of COVID-19 on Black Communities

COVID-19 Cases Heaviest in Poor Neighborhoods. SF Weekly. 369 21. Roux AVD, Mair C. Neighborhoods and health

The 373 development of a standardized neighborhood deprivation index

Assessing capacity to social distance and neighborhood-level health disparities during 377 the COVID-19 pandemic. medRxiv

Spatial Inequities in COVID-19 outcomes in 379 Three US Cities. medRxiv

The COVID-19 Pandemic Predominantly Hits Poor 381 Neighborhoods, or does it? Evidence from Germany

Towards Health Equity in Connecticut: The 387 Role of Social Inequality and the Impact of COVID-19

Social Determinants of 390 Health: Future Directions for Health Disparities Research

The Detrimental Impact of the 393 COVID-19 Crisis on Health Equity and Social Determinants of Health

COVID-19 and the need to prioritize health 396 equity and social determinants of health

Neighborhood Socioeconomic Status 398 and Health: A Longitudinal Analysis

Sorting out the connections between the 402 built environment and health: a conceptual framework for navigating pathways and 403 planning healthy cities

The Role of 405 the Built Environment on Health Across the Life Course: A Call for CollaborACTION. Am 406

The impact of the built environment on health: an emerging field

Health-Risk Behaviour in 410 Deprived Neighbourhoods Compared with Non-Deprived Neighbourhoods: A 411 Systematic Literature Review of Quantitative Observational Studies

Disparities in the Population at Risk of Severe Illness 414

From COVID-19 by Race/Ethnicity and Income

Revealing the unequal burden of COVID-19 by income

race/ethnicity, and household crowding: US county vs. ZIP code analyses

Crowding and the 419 epidemic intensity of COVID-19 transmission

Area deprivation and widening inequalities in US mortality

Accessible -The Neighborhood Atlas

Geocoding and monitoring of US socioeconomic inequalities in mortality and cancer 426 incidence: does the choice of area-based measure and geographic level matter

Public Health Disparities Geocoding Project

Neighborhood 432 socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study

Area Deprivation Index Measuring Patient Socioeconomic Status in an Integrated 436 Health System: Implications for Population Health

Rapid Emergence of SARS-CoV-2 in the

Positivity Rates, and Hospitalization for 46,793 Persons 440 Tested by Northwell Health

Neighborhood Socioeconomic Status, and 443 SARS-CoV-2 Infection Among Pregnant Women in

The neighborhoods where COVID collides with overcrowded homes. 445 California CalMatters

COVID-19 in Context: Racism, 447 Segregation and Racial Inequities in Philadelphia2020

The impact on African

Spatial transmission of COVID-19 via 452 public and private transportation in China

Workers who could work at home, did work at 454 home, and were paid for work at home, by selected characteristics