A retrospective analysis assessing 15 months of COVID-19 cases and deaths reveals small but significant reductions in cases among counties with increasing college enrollment. However, little to no change has been noted with respect to death rates. Other analyzes focusing on differences between age groups also reveal little or no variation in each group’s case and death rates by college enrollment size. However, notable age-related trends were discovered. First, the highest case rates were in young to middle-aged adults (20-59), with 20-29 year olds experiencing the highest case rates of any other age group – a finding also found by Monod et al.30 Second, although the 0-9 and 80+ age groups had the lowest case rates, the 80+ age group’s risk of death was significantly higher than all. other age groups. These trends support widely observed findings regarding COVID-19 deaths.31.32.
Time series plots of daily aggregate means of county case and death rates (Fig. 4) show observable differences in the magnitude of the spikes with increasing college enrollment across counties. However, when assessing COVID transmission by wave, the critical differences become more apparent. Three waves of COVID-19 occurred before May 2021, with each wave being more severe than the previous wave. However, the third wave saw an approximately 3-6 fold increase in cases and deaths compared to waves 1 and 2. Although all counties experienced this drastic increase in COVID-19, those with more college enrollments had significantly lower case rates (5.3, 10.6, and 27.2% for counties with small, medium, and large college enrollments, respectively) compared to counties with no college enrollment. Counties with medium or large college enrollment experienced significantly lower COVID-19-related death rates (average of 12.8 and 29.8% lower, respectively) than counties with low or no college enrollment.
The second epidemiological period of interest was for counties before and after the fall 2020 semester. As wave analysis shows, all counties experienced a rapid increase in COVID-19 cases and deaths after the onset of the fall 2020 semester. However, counties with increasing college enrollment experienced significantly lower case rates (3.7, 10.8, and 27.1 percent lower for counties with small, medium, and large, respectively) compared to counties without college enrollment. Counties with medium and large college enrollments also resulted in significantly lower mortality rates (average of 13.2 and 30.2 percent lower, respectively) compared to counties with little or no college enrollment.
Our subgroup analysis of COVID-19 mitigation strategies for the fall 2020 semester provides additional evidence that colleges and universities were not associated with an increase in county-level cases, even before the outbreak. establishment of coherent containment plans. Cluster analysis identified that the schools identified in cluster 2 tended to be larger land-grant universities that maintained fully online or hybrid course instruction during the fall 2020 semester and were likely to find in counties with lower population-adjusted COVID cases. These schools were more likely to have mandatory testing for students which was significantly associated with lower overall case counts in the counties. In the fall 2020 semester, county- and state-level factors (e.g., mask use, mask mandates, and median household income) were much more predictive of overall cases at the county level, which held true over the broader analysis period.
Overall, the COVID-19 pandemic through March 30, 2021 has spread rapidly through all U.S. counties with similar patterns in the timing and intensity of cases and deaths, regardless of size. university registrations. However, the magnitude of county incidence of cases and deaths is strongly associated with college enrollment. Although there were minimal differences in death rates by university enrollment, large enrollment universities were hardest hit by COVID-19 in the early stages of the pandemic – a strength presumed driving force behind the lack of significant differences in overall death rates. However, as the pandemic progressed in intensity (cases and deaths per day), counties with increasing college enrollment experienced decreased risks of acquiring and dying from COVID-19.
Together, these two analyzes strongly suggest that community-level variables — not universities — are driving COVID-19 cases during this period. Despite larger populations, counties with large college enrollments have fared better than counties with little or no college enrollment, especially as COVID-19 cases spiked in the winter of 2020-2021. (wave 3). Compared to counties with little or no college enrollment, larger counties with college enrollment contained higher household incomes, less unemployment, and had higher immunization rates (% with at least 1 dose). These counties also tended to apply statewide mandates more frequently and longer throughout the pandemic compared to counties with little or no college enrollment (Tables S1 and S2).
In addition to the differences noted above, public health decisions depended on several political, economic and social factors. It is evident that this pandemic has fueled political divisions, which have influenced both public health decisions and compliance. From respect for social distancing and mask-wearing to vaccination rates, political associations seem to be a strong force of influence33. Using the 2020 US presidential election as a proxy for determining a county’s political affiliation, the associations of COVID-19 cases and deaths are moderately correlated (Figs. S1 and S2). Counties with higher college enrollments are also correlated with increased overall education rates, which have been shown to be associated with pandemic response utilization.34,35,36.
To date, several studies have analyzed the transmission risk associated with universities and/or college-age populations, but all have been limited to no more than 4-5 institutions.2,7,37,38. Moreover, very few studies have estimated the attributable risk between universities and students within the communities where they reside.39. Studies that looked at community transmission and associations with students were systematic reviews, mathematical simulations, and/or focused on the age of primary or elementary school students10,40,41. Our retrospective analysis is novel in that all data collected is specific to our central question, narrowing the scope of our investigation to produce tangible estimates of transmission risk at high spatial and temporal scales.
This study is limited in its analysis by aggregating non-lineage-specific individual-level COVID-19 case data to county of residence. In doing so, generalizations are made across different population sizes and strains of COVID-19 and cannot capture subtle differences. Many reporting agencies submit varying levels of completeness to the CDC, which leads to a high risk of verification bias. Additionally, due to the rapid onset of cases, particularly during peak periods, significant delays in reporting have also occurred. This analysis attempted to establish etiology-specific dates of onset given the data made available. In some cases, differences of a week or more between the reported onset date and the true onset date probably exist and are unavoidable. However, the data represents a significant portion of the US population, reducing generalization errors and representing the best source for such a study. Limitations of the subgroup analysis included an incomplete picture of mitigation strategies for some universities (e.g. Group 1 was defined by schools that did not fully report their COVID-19 plans). Counties with no tertiary enrollments were also very limited in this dataset, which could have impacted the analysis. We also assumed that the mitigation plans would be static during the fall semester 2020 due to the lack of data available on the changes. Given that this was the first semester with some in-person schools, we felt it was likely that a university would maintain the proposed plans, and if any changes did occur, they would be for further mitigation and not to be reduced.
This study incorporates key differences between US counties stratifying on a range of enrolled college students. In general, counties with growing school populations tend to be more populated and urban. Thus, several potentially important factors associated with COVID transmission, reporting, knowledge, attitudes, and practices have been generalized or overlooked. Future studies would do a great service to public health by expanding on the methodologies and results of this study on several of these key factors. For example, counties with large college enrollments appear to have many key protective factors in place to mitigate COVID-19. However, these counties also contain much larger populations of Black, Indigenous, and People of Color (BIPOC) who suffer from disproportionate racial and health inequalities. Therefore, social vulnerabilities are significantly higher in counties with large college enrolments. Preliminary analyzes of this dataset, with respect to COVID-19 outcomes by race and ethnicity, suggest corroborating evidence for these inequalities (data not shown). Future studies should also assess co-evolution and/or competition between COVID-19 and other respiratory viruses, particularly seasonal influenza and respiratory syncytial virus, conducted in longitudinal analyzes similar to those of this study.