dcyphr | American Indian Reservations and COVID-19: Correlates of Early Infection Rates in the Pandemic


In this study, the researchers wanted to determine characteristics associated with the high COVID-19 incidence on Native American reservations. They examined 287 Native American Reservations and tribal homelands, which had 861 cases as of April 10th, 2020. The relationship between COVID-19 infection rate and certain community and household characteristics was examined. The researchers found that COVID-19 cases were more likely to be found on Native American Reservations with a higher proportion of households without indoor plumbing and those that were English-only. They found no relationship between COVID-19 cases and household overcrowding measures. In conclusion, the lack of indoor plumbing and access to drinkable water may be an important determinant of COVID-19 infection on Native American reservations. Access to information about COVID-19 transmission and prevention in Native American languages may also be important in reducing COVID-19 cases in these communities. Previous studies have pointed to household plumbing overcrowding, and language as being risk factors for other infectious diseases and these risk factors are still relevant for COVID-19. Investing in public health and household infrastructure in these communities are necessary to protect them from high COVID-19 incidence.


The researchers wanted to understand household- and community-level risk factors for COVID-19 infection on Native American reservations.


The United States is experiencing one of the world’s largest COVID-19 outbreaks. COVID-19 in the U.S is intensifying long-standing racial and class health and economic disparities. The U.S Native American population experiences significant and unique economic and health hardships, placing them at higher risk for COVID-19.

There are 574 federally recognized tribal nations across 326 American Indian reservations, often located in rural areas.. Tribal nations are independent nations that communicate with the U.S government and maintain their own healthcare and public health systems. However, these systems are almost always underfunded and have been focused on chronic, not infectious, disease.

 Inadequate public health infrastructure, limited medical resources, and high rates of poverty make American Indian reservations poorly prepared to face the COVID-19 pandemic. Especially concerning in these communities are lack of household plumbing, lack of information about COVID-19 in Native American languages, and household overcrowding. Native American households are 3.7 times more likely to lack indoor plumbing compared to the general U.S public and the prevalence of English use ranges from 7-100%. In addition, Native American reservations have more crowded households than the general U.S public, at 6.9% compared to 3.4%. Thus, it is important to examine how these factors may play a role in COVID-19 incidence in these communities.


Sample Overview

287 tribal reservations were included in the analysis. On average, 60% of the population in each reservation was Native American. The median age was 34 years old and the population was equally male and female. 12% of the sample held a Bachelor’s degree or higher. The average household size was 3, and more households were overcrowded than the rest of the U.S. Median income for households was $20,000 lower than the national average, and marital status was also lower. The percentage of English-speaking households is greater than average, at 82% compared to 78%; however, this does not account for large rates of non-English households in Navajo and Pueblo nations. Finally, more reservation households had incomplete plumbing, at 1% compared to the 0.4% national average. The rate of COVID-19 cases per 1000 people, as of April 10th, is 4 times higher than the national average.

Analysis for all tribal nations in the lower 48 states

COVID-19 cases per 1000 people was higher on reservations with more households that lacked indoor plumbing. It was lower on reservations with a high percentage of English-language only households. Household overcrowding was not significantly associated with COVID-19 cases per 1000 people. The Navajo Nation and Oklahoma tribes had a high level of COVID-19 cases, and were outliers in population size and the community- and household-level characteristics studied. The researchers found that the highest rate of COVID-19 was in the Pueblo tribes.


COVID-19 cases on Native American reservations are strongly associated with lack of indoor plumbing and English-only households. There was no relationship with household overcrowding. These findings show that structural and social determinants play an important role in COVID-19 transmission in these communities.

Lack of indoor plumbing had the most significant relationship with COVID-19 cases. Solutions for the pandemic include providing drinking water and hand sanitizer to fulfill critical sanitation needs.

The significant association between English-only households and decreased COVID-19 cases show the importance of catering to Indigenous language speakers with public health campaigns. There are 150 Indigenous languages spoken by more than 350,000 people in the U.S.

Household overcrowding was not significantly associated with COVID-19 cases. More research is necessary to confirm this result, and overcrowding may play a more significant role as the pandemic progresses.

The results of this study have important public health implications. Before physical distancing and stay-at-home orders, hand-washing was considered the most effective prevention method against COVID-19; however, this ignores households that do not have reliable access to water. In addition, not disseminating public health announcements, warnings, and directives in local languages may prevent communities to prepare effectively for pandemic situations.

The federal government has not prioritized the needs of tribal and reservation communities. Reservations are poorly prepared to respond to large numbers of COVID-19 cases and future pandemics.


Study population

The number of COVID-19 cases and deaths was collected online from American Indian tribal nations and communities, as well as from data compiled by Indian Country Today. COVID-19 cases at the reservation and homeland level were linked to reservation/homeland characteristics, which were derived from the 5-Year American Community Survey (ACS). This linked dataset was analyzed for 48 states excluding Alaska and Hawaii, as data about COVID-19 cases was not available at the time. 287 tribal nations were included, and analysis was weighted for population size to account for large reservations such as the Navajo Nation, and smaller Oklahoma tribes.

Outcome of interest

The outcome variable was COVID-19 cases per 1000 individuals. All instances of verified COVID-19 cases were compiled and counted for each tribal nation separately.

Focal independent variables

Population characteristics were derived from the 2018 5-Year ACS data. The researchers hypothesized three independent variables that would have a relationship with COVID-19 cases on reservations. The first is the proportion of households lacking complete plumbing, as sanitation and access to running water are important determinants of COVID-19 transmission. The second is the proportion of households with more than 1 person per room, as it is difficult to self-quarantine in crowded conditions. The last variable is the proportion of households that speak English-only, as it may indicate how many households receive information about COVID-19 that is actually understood.

Control variables

Control variables included several economic and social factors known to influence health. These variables included percentage of reservation households that are Native American, the U.S state that the majority of the tribal reservation is located in, average household size, median household age, gender, household income, percentage of households married, and percentage of the population with a Bachelor’s degree or higher.

Statistical analyses

The researchers conducted a least-squares model and Poisson regression for analysis. All analyses were conducted in STATA.