Introduction

The SARS-CoV-2 coronavirus disease 2019 (COVID-19) pandemic continues to be a significant public health concern globally. Severe COVID-19 patients primarily suffer from respiratory system failures that can lead to death [1, 2]. In Louisiana, African Americans (AA) account for 32.2% (1.5 million) of the population, and 21 of the 64 counties have more than 40% AA with poor health conditions [3]. The Louisiana population of 4.67 million people includes White Americans (WA) (59.3%), AA (32.2%), and others (Asian, Hispanic, Alaskan, Hawaiian) (8.5%) (Fig. S1a) [3, 4]. According to the Healthy People 2020, Louisiana residents, especially AA, are more vulnerable to adverse health outcomes due to higher percentages of underlying health conditions, poverty, and other health disparities. Health disparity is defined as a race or ethnic group who experience systematically greater challenge to health conditions that are linked with social, economic, and environment factors [5]. About 25% (n = 16) of the 64 counties were ranked as having the poorest health outcomes in Louisiana. [6,7,8].

The clinical and epidemiological studies of COVID-19 have shown comorbidities associated with increased risk of infection and mortality outcomes of those with hypertension (30%), diabetes (19%), and coronary heart disease (8%) [3, 9,10,11]. The most common underlying health conditions that are associated with Louisiana’s COVID-19 deaths are hypertension (59.4%), diabetes (36.7%), cardiac disease (21%), kidney disease (20%), obesity (20%), heart failure (13%), chronic obstructive pulmonary disease (COPD) (11.7%), neurological (8.43%), cancer (7.51%), and asthma (4.15%) [12] (Fig. S1). Social factors that contribute to health outcome such as poverty, lack of vehicle access, and crowded housing were established as social vulnerability index (SVI) by CDC/ATSDR, which refers to the negative effects on communities caused by external stresses such as disease outbreaks on human health scores 0 (low) to 1 (high), using 15 US census data. The average county level of SVI index in Louisiana is 0.9. [13, 14].

In China, diabetic individuals infected with COVID-19 showed higher hospital admission rates and were more likely to develop severe pneumonia and higher mortality rates compared to those without underlying health conditions [15,16,17]. A meta-analysis of 33 case–control studies in adults (~16,000 patients, excluding pregnant women) has shown that diabetes was significantly associated with a two-fold increase in COVID-19 mortality [17]. World Health Organization (WHO) has acknowledged that hypertension and cardiovascular disease also increase the risk of severe COVID-19 and mortality [18]. The mortality, morbidity, and complications of COVID-19 are also associated with the increased risk of developing acute kidney injury (AKI) among chronic kidney disease (CKD) patients during the progression of COVID-19. [18,19,20,21].

In the USA, 89.3% of adult patients had one or more underlying health conditions; the most common were hypertension (49.7%), obesity (48.3%), chronic lung disease (34.6%), diabetes mellitus (28.3%), and cardiovascular disease (27.8%) [22, 23]. About 4,103 patients with COVID-19 in New York City found that the most frequent factors leading to hospital admission were age > 65 years and obesity [24]. Patients with obesity are more prone to developing serious illness that requires hospitalization and invasive ventilation [24]. COVID-19 is among the viruses that trigger COPD; however, patients with COPD do not appear to have increased risk of COVID-19 infection, but slightly increased risk of hospitalization. [25].

Several COVID-19 and health disparity studies in the UK have shown disproportionate impacts in minority groups with poorer cardiometabolic profiles and likely to live in crowded household conditions [25]. High numbers of those critically ill with COVID-19 were Black, Asian, or minority ethnic backgrounds [26]. Frydman et al. [27] studied COVID-19 morbidity and mortality and found that AA are more likely to have severe inflammation and blood clots associated with COVID-19 than other racial and ethnic groups. A recent study in the USA found that 95% of patients had at least one underlying medical condition, commonly hypertension (50.5%) or obesity (33.0%) [28]. The 70.6% of the 326 patients seen within an integrated-delivery health system (Ochsner Health) in Louisiana died from COVID-19 were Black. However, there was no difference between AA and WA in-hospital mortality after adjusting for sociodemographic and clinical characteristics on admission. [11, 28].

Prior health disparity and COVID-19 studies in Louisiana are limited to comprehensively evaluate the risk of underlying health conditions on COVID-19 incidence and death in minority communities, especially in AA. This study targets the pandemic period between March 1, 2020 and March 1, 2021 (1-year period), and aims to (1) determine the associations between health and social disparity factors and COVID-19 infection and mortality in Louisiana and (2) build a GIS database for geospatial analysis of environmental factors and underlying chronic health conditions to identify communities at risk of COVID-19 infection and death. The results of this study will help to inform the public, scientific communities, and policy makers to make data-driven decisions for planning effective response to prevention, allocating resources, and reducing the risk of COVID-19 or other potential infectious diseases.

Methods

Data Source

COVID-19 Data Source

On March 8, 2020, the Louisiana Department of Health (LDH) reported the first COVID-19 case in Jefferson County. During the stay-at-home order (phase 1), the highest COVID-19 cases and deaths were found in older (> 50 years) individuals. After reopening (phase 2), more positive cases were found in the 18–29 years old with lower death rates, and most deaths occurred in individuals > 50 years. [12] County-level COVID-19 daily reports of the 64 counties in Louisiana have been released weekly, including the number of daily testing results, number of deaths, and cases by race, ethnicity, age, and sex. COVID-19 epidemiological descriptive analysis was calculated for COVID-19 case and death counts and rates per 10,000 population in all races, AA, WA, and others (e.g., Asian, Hispanic, Hawaiian).

Health and Sociodemographic Data

Potential COVID-19 health and social disparity factors included are (1) underlying health conditions: hypertension, diabetes, cardiovascular, kidney, obesity, COPD, cancer, asthma, and body mass index (BMI) from CDC-PLACES: Local Data for Better Health and Behavioral Risk Factor Surveillance System [29, 30]; (2) demographic: age, race, and ethnicity from US Census Bureau 2020 [3]; (3) socioeconomic: income, household size, insurance availability, poverty, and population density from Statistical Atlas database [4]; and (4) behavioral: social vulnerability index (SVI). [13, 14].

Dependent and Explanatory Variables

The dependent variables are African American mortality rate (model A) and cumulative incidence rate (model B). Explanatory variables include underlying health conditions, sociodemographic, and behavior factors that are significantly associated with AA mortality and incidence rates. A total of nine underlying health conditions: heart disease, diabetes, obesity, cancer, hypertension, asthma, chronic obstructive pulmonary disease (COPD), kidney disease, and BMI. Adjusted variables include median income, age, percentage of poverty, percentage of uninsured persons, household size, population density, and social vulnerability index (SVI).

Data Analysis

Descriptive analysis (minimum, 25th, 50th, 75th, maximum, and mean ± standard deviation) was used to report COVID-19 cases, deaths, incidence, and mortality rates. The incidence and mortality rates were reported as the number of cases and deaths per 10,000 people per county. Spatial distribution and high burden of COVID-19 incidence and mortality rates were mapped in quantile range using ArcGIS Pro software. Statistical Analysis System (SAS) software was used to conduct correlation, multivariable generalized linear models with negative binomial distribution and log link function, unadjusted and adjusted odds ratios (AORs) of interest among COVID-19 incidence and mortality rates for each underlying health condition and other variables. We defined each model using known variables to be significant from previous analysis and disease comorbidity. Model A consists of crude (model A1), role of diabetes and obesity adjusting for statistically significant social and behavior variables (model A2), and role of diabetes and obesity in counties with prevalence 75th percentile and higher, adjusting for statistically significant social and behavior variables (model A3). Similarly in statistical analysis method, AA incidence rate models comprise nine models (B1–B9). All significantly associated underlying health conditions variables were categorized into four groups based on their pathological functions: diabetes and obesity (models B2–B3); asthma and COPD (models B4–B5); heart disease and hypertension (models B6–B7), and kidney disease (models B8–B9).

We tested for multicollinearity setting a variance inflation factor (VIF) cut-off at 2.5. We presented odds ratio (OR) for each exposure with the corresponding 95% confidence interval (CI) and p value. We defined counties with AA percentage in 75th percentile and higher (40% of county population) as higher COVID-19 risk and others as lower COVID-19-risk counties, respectively (Fig. 1). The data distribution and differences between variance and mean were tested; over-dispersions were detected, suggesting that a negative binomial model is appropriate for this analysis. Additionally, the multicollinearity analysis was conducted to measure correlations between each independent variable and outcome; frequency analysis was also used to identify the numbers of counties that were in higher and lower risk groups. The counties with zero deaths were excluded in sensitivity analyses.

Fig. 1
figure 1

Counties with Black population 40% and higher (left) and Counties in 49th – 64th ranking (lowest) of health factors and outcome

Full size image

Results

COVID-19 Cases and Deaths in Louisiana

From March 1, 2020 to March 1, 2021, a total of 9,621 deaths were reported in Louisiana, which accounts for AA (3,702), WA (5,778), and other (141), representing 2.6%, 2.3%, and 0.03% of corresponding infected patients, respectively (Fig. S1c). A total of 408,047 COVID-19 infection cases were reported: AA (136,460), WA (218,332), other (53,255), and unknown (22,150). County-level case and death counts were the highest in Cancer Alley counties such as Jefferson, Orleans, and East Baton (Fig. 2), with a 40% and higher AA population.

Fig. 2
figure 2

a Total COVID-19 case counts and b death counts in Louisiana between March 1, 2020-March 1, 2021

Full size image

The state’s incidence and mortality rates per 10,000 population were 876 and 21, respectively. County-level mortality rate was in the range of 0–56, whereas incidence rate was 530–1,511 per 10,000 population. African American population accounts for 32.2% of the total Louisiana population; however, AA had the highest mortality rate (9 deaths per 10,000 persons; 0.26% of total AA population) compared to WA (7 deaths per 10,000 persons; 0.23% of total WA population) and other (1 death per 10,000 persons; 0.03% of total other races population). The incidence rate was higher in WA (462 cases per 10,000 persons), whereas AA (324 cases per 10,000 population) and others (142 cases per 10,000 population) were 1.4- and threefold lower, respectively (Table 1). However, the percentage of COVID-19 cases was highest in AA (9%), followed by WA (7.4%) and others (7.8%).

Table 1 Characteristics of state and county level of COVID-19 deaths, cases, death rate, and incidence rate per 10,000 population by race in Louisiana between March 1, 2020 and 2021
Full size table

Associations Between COVID-19 Cases, Deaths, and Underlying Health Conditions in AA

Total COVID-19 case and death counts were highly associated with all eight underlying health conditions (p < 0.0001) (Table 2). The odds of incidence and mortality rates in counties with higher AA density are at least 10 times higher than counties with lower AA density, [OR = 14.4 (95% CI: 3.9,54.5) and OR = 10.6 (95% CI: 3.2,35.6), respectively (Fig. 3). The mortality rate of the AA population was positively associated with diabetes, hypertension, and overweight BMI range (p < 0.05) (Fig. 4 and Table 2) while the AA COVID-19 incidence rate was significantly associated with heart disease, diabetes, obesity, hypertension, asthma, COPD, and kidney disease (p < 0.05) (Fig. 5 and Table 2). The AA had 1.3 and 9 times greater risk of COVID-19 death than WA and other races, respectively (Fig. S1b). Similarly, AA COVID-19 incidence was 1.2 and 1.7 times higher than WA and other races, respectively (Fig. S1c). We included significantly correlated variables from previous analysis (Table 2) in the model to test the role of underlying health conditions on AA mortality and incidence rates (models A1–B9) (Fig. S3-4). The results were demonstrated robustness across sensitivity analyses.

Table 2 Pearson correlation coefficients between number of people with underlying health conditions and COVID-19 total deaths and cases and county underlying health condition prevalence (%) with COVID-19 mortality and incidence rates per 10,000 population (N = 64)
Full size table
Fig. 3
figure 3

The odds ratio of COVID-19 incidence rate between counties with African American density higher than 40% and lower (left) and the odds ratio of COVID-19 mortality rate between counties with African American density higher than 40% and lower (right)

Full size image
Fig. 4
figure 4

Correlations of African American COVID-19 mortality rate with diabetes prevalence (r = 0.3, p = 0.01) (left) and hypertension prevalence (r = 0.4, p = 0.004) (right)

Full size image
Fig. 5
figure 5

Correlations of African American COVID-19 incidence rate with diabetes prevalence (r = 0.8, p < 0.0001) (left) and obesity prevalence (r = 0.4, p < 0.0001) (right)

Full size image

Association Between Diabetes and Hypertension and COVID-19 Mortality Rate

The risk of COVID-19 mortality was higher in counties with higher AA population density than lower AA population density [OR = 1.7 (95% CI: 1.2, 2.4, p = 0.002)]. Diabetes and hypertension in the higher risk group was significantly associated with greater odds of COVID-19 mortality rate, after adjusting for BMI and household size [OR = 1.5 (95% CI: 1.05, 2.25, p = 0.016)]. Higher odds of COVID-19 mortality rate was also found in the high Black density counties with high burdens of diabetes and hypertension (diabetes 14% and hypertension 45%) [OR = 1.8 (95% CI: 1.1, 3.0, p = 0.02)] (Fig. S3; Table 3: models A1–A3).

Table 3 Comparison of crude and adjusted odds ratio of African American COVID-19 death and incidence between counties with higher than 40% AA population (high risk group) and others (lower risk group) in Louisiana between March 1, 2020 and March 1, 2021
Full size table

Associations Between Underlying Health Conditions and COVID-19 Incidence Rate

In the nine models (Table S1: B1–B9) generated, higher risk groups had significantly greater odds of COVID-19 incidence rate than lower risk counties [crude OR = 1.9 (95% CI: 1.6, 2.4, p < 0.0001)] (Fig. S4a; Table 3: model B1). However, the effect size varied across the eight models when adjusting for underlying health conditions, BMI, income, poverty, and household size.

Diabetes and Obesity

In multivariate logistic regression, including diabetes, obesity, BMI, income, poverty, and household size, the higher Black density counties were associated with higher odds of COVID-19 incidence rate [OR = 1.6 (95% CI: 1.3, 2.0, p < 0.0001)]. Additional adjustment for higher burden of diabetes and obesity (14% and 41%, respectively) resulted in a significantly higher OR than the crude model [OR = 2.3 (95% CI: 1.6, 3.2, p < 0.0001)] (Fig. S4a; Table 3: models B2–B3).

Asthma and COPD

The effects of asthma and COPD were tested with adjustment for BMI, income, poverty, and household size. There was a statistically significant association between higher risk counties (i.e., > 40% AA population) and higher odds of COVID-19 incidence rate [OR = 1.6 (95% CI: 1.3, 2.1, p < 0.0001)]. Counties with higher Black populations that had higher burden of asthma and COPD (10.5% and 10.4%, respectively) were significantly associated with higher odds of COVID-19 incidence [OR = 1.8 (95% CI: 1.2, 2.7, p = 0.003)] (Fig. S4b; Table 3: models B4–B5).

Heart Disease and Hypertension

There was a statistically significant association between counties with higher AA density and higher odds of COVID-19 incidence rate [OR = 1.6 (95% CI: 1.3, 2.1, p < 0.0001)]. Counties with higher AA population with heart disease and hypertension prevalence higher than 75th percentile (8.4% and 45%, respectively) were significantly associated with higher odds of COVID-19 incidence [OR = 1.8 (95% CI: 1.2, 2.7, p = 0.008)] (Fig. S4c; Table 3: models B6–B7).

Kidney Disease

There was a statistically significant association between counties with higher Black density and greater odds of COVID-19 incidence rate [OR = 1.7 (95% CI: 1.3, 2.1, p < 0.0001)]. Counties with higher Black population with kidney disease prevalence higher than 75th percentile (3.6%) were significantly associated with higher odds of COVID-19 incidence [OR = 2.1 (95% CI: 1.5, 2.9, p < 0.0001)] (Fig. S4d; Table 3: models B8–B9).

Spatial Distribution of High Burden Areas

The high-risk areas for AA mortality rates (Fig. 6b) are associated with counties that had diabetes and hypertension higher than 75th percentile (Fig. 6a). Therefore, the high burden of COVID-19 mortality rates in AA was found in four counties: Claiborne, Bienville, Morehouse, and Madison (Fig. 6c). High burden areas for AA COVID-19 incidence rates (Fig. 7b) are associated with the highest risk of COVID-19 infection rate, diabetes, and obesity (Fig. 7a). Therefore, high burdens of COVID-19 cumulative incidence rates were found in Claiborne, East Carroll, Madison, Tensas, and St. Helena (Fig. 7c).

Fig. 6
figure 6

a High African American density counties with higher prevalence of diabetes and hypertension, b counties with higher than 75th percentile of African American death rate (deaths per 10,000 population), and c African American COVID-19 death rate high burden counties

Full size image
Fig. 7
figure 7

a High African American density counties with higher prevalence of diabetes and obesity, b counties with higher than 75th percentile of African American incidence rate (cases per 10,000 population), and c African American COVID-19 incidence rate high burden counties

Full size image

Discussion

Underlying health conditions (diabetes, hypertension, obesity, cancer, cardiovascular disease, asthma, COPD, and kidney disease) and social disparities (age, income, poverty, and population density) were significantly correlated with increased numbers of COVID-19 mortality and incidence in Louisiana. Other studies show similar findings [22, 31,32,33]. A disproportionate number of COVID-19 cases and deaths in AA communities were also reported in other US states such as New York and Illinois. [23, 34].

Louisiana’s most prevalent chronic health conditions among adults age 18–44 are hypertension, obesity, and high cholesterol [29, 30]. Hypertension and diabetes were the top two conditions linked to COVID-19 deaths [12]. Our study observed strong associations for mortality and infection among those with diabetes, hypertension, and obesity (p < 0.0001). About 60% of the COVID-19 deaths had pre-existing hypertension, followed by 36.7% with diabetes and 20% with obesity. These are observed mostly in AA population. A study in England found Blacks had more than 3.5 times greater risk of COVID-19 death than other races [35]. We found that AA COVID-19 deaths are 1.3 and 9 times greater than WA and other races, respectively (Fig. S1b). We found that high mortality and incidence rates of COVID-19 among AA in Louisiana is consistent with other studies in the USA, Ireland, and UK. [12, 25,26,27, 30, 34, 36].

Among the 50 states in the USA, Louisiana’s overall poor health status is ranked 41st [7]. Forty percent of AA adults have hypertension, which is higher than the national average level (30%). Higher prevalence of hypertension was found in individuals with low education status and annual income below $25,000 [29]. Diabetes in adults (14%) is higher than the national average level (8.9%) in which Louisiana was ranked 47th among 50 states [8]. Our study found that diabetes is the only underlying health condition that is highly associated with both COVID-19 infection and death. Diabetes is highly associated with obesity, [37,38,39,40,41,42,43] poor diet such as high in fat and cholesterol that lead to obesity and increase risk of diabetes [44]. Previous study in the USA found that Black adults have higher risk of developing diabetes than White adults, especially between women [45]. Health disparity factors including neighborhood, psychosocial, socioeconomic, and behavioral factors were linked to diabetes’s biological risk factors [45]. Diabetes may enhance complications in individuals through an imbalance in angiotensin-converting enzyme 2 (ACE2) activation pathways leading to a well-described hormonal pathway and inflammatory response [46]. A two-fold increase in COVID-19 severity and mortality was observed compared to non-diabetics [17]. However, patients with diabetes and well-controlled blood glucose had lower mortality than those with diabetes and poorly controlled blood glucose. [10, 11, 17, 47].

Age groups (18–64 years and 65 years and older) were significantly correlated with COVID-19 total cases and deaths; however, underlying health conditions were stronger predictors of COVID-19 incident and mortality rates than age groups in AA populations (Table 2). Based on CDC definitions, [31] COVID-19 patients were categorized into eight age groups. Therefore, the age groups data in this study was widely categorized (18–64 years and 65 years and older) and caused a redundancy issue in our analysis. The actual age group with highest rates of infection in Louisiana is 40–60 years. [12].

In Louisiana, the percentage of adults with heart disease (10.4%), COPD (8.6%), and kidney disease (4%) is higher than national prevalence of 8.4%, 6.5%, and 2.9%, respectively, while asthma is lower than the national prevalence (7.9%) [7]. Our study found that among COVID-19 deaths, heart disease, COPD, asthma, and kidney disease were present in 21%, 20%, 11.7%, and 4.2% of the patients, respectively. Asthma and kidney disease prevalence are higher in AA compared to WA. Similar results have been found in US studies [1, 2, 48]. Severity of COVID-19 in AA includes kidney inflammations and coagulopathies with elevated levels of high-sensitivity C-reactive protein, developed proteinuria (sign of kidney damage), and serum lactose dehydrogenase [15, 20, 37]. Additionally, the study in Ireland also found higher risk of mortality, hospitalization, and ICU admission were significantly associated with cardiovascular disease, BMI, neurological condition, chronic kidney disease, and cancer [49]. Previous studies in the USA showed that severe illness and ICU admission were highly associated with moderate to severe asthma, COPD such as emphysema and chronic bronchitis, and heart disease such as myocarditis, which causes inflammation in the heart muscle [29, 37]. Twenty-one of Louisiana’s 64 counties have high AA density. Counties with high burden of COVID-19 incidence and death are mostly located in the northern part of Louisiana and had higher AA density, even though each county has lower total populations. These counties were ranked in the lowest 49th–64th for health factors (health behaviors, clinical care, social and economic factors, and physical environment) and outcome (length of life, quality of life) in Louisiana (Fig. 1). [8]

Conclusion

The study identified at-risk groups for COVID-19 infection and death at the early stage of the outbreak when individual data was limited. Reducing incidence of underlying health conditions, improving access to health care, quality of treatment, and environmental status are important approaches to improve health outcomes that increase Louisiana’s population resilience against COVID-19 and other infectious diseases.

Limitations

Our data were observational, and we used county-level rather than individual (COVID-19, health, and sociodemographic) data. Individual data could provide a higher level of sensitivity and validity. Age adjustment is important for COVID-19 infection, hospitalization, and death. Because individual age data cannot be confirmed in the county data, the results may be under or overestimated [30]. Finally, more underlying medical conditions and behavior such as neurological disorder and smoking should be included in the estimations of risk, which could have caused us to omit less prevalent risk factors of severity of infection or death.