Impact of altitude on COVID-19 infection and death in the United States: A modeling and observational study

Abstract

Background: COVID-19, the disease caused by SARS-CoV-2, has caused a pandemic, sparing few regions. However, limited reports suggest differing infection and death rates across geographic areas including populations that reside at higher elevations (HE). We aimed to determine if COVID-19 infection, death, and case mortality rates differed in higher versus low elevation (LE) U.S. counties.

Methods: Using publicly available geographic and COVID-19 data, we calculated per capita infection and death rates and case mortality in population density matched HE and LE U.S. counties. We also performed population-scale regression analysis to investigate the association between county elevation and COVID-19 infection rates.

Findings: Population density matching of LA (< 914m, n = 58) and HE (>2133m, n = 58) counties yielded significantly lower COVID-19 cases at HE versus LE (615 versus 905, p = 0.034). HE per capita deaths were significantly lower than LE (9.4 versus 19.5, p = 0.017). However, case mortality did not differ between HE and LE (1.78% versus 1.46%, p = 0.27). Regression analysis, adjusted for relevant covariates, demonstrated decreased COVID-19 infection rates by 12.82%, 12.01%, and 11.72% per 495m of county centroid elevation, for cases recorded over the previous 30, 90, and 120 days, respectively.

Conclusions: This population-adjusted, controlled analysis suggests that higher elevation attenuates infection and death. Ongoing work from our group aims to identify the environmental, biological, and social factors of residence at HE that impact infection, transmission, and pathogenesis of COVID-19 in an effort to harness these mechanisms for future public health and/or treatment interventions.

Fig 1. High and low altitude counties by elevation (m).

Red lines mark the definition of high altitude (>2133m) and low altitude (<914m).

Fig 2. COVID-19 infection and death in matched high and low elevation counties.

A) Mean COVID-19 cumulative per capita incidence per 100,000 population B) and lack of correlation with population density. C) Mean COVID-19 cumulative per capita death per 100,000 population D) positively correlated with population density at both high- and low-elevation counties. E) COVID-19 case mortality in high and low elevation counties of similar population density. N = 58 for both high elevation and low elevation counties. *p<0.05 by one-sided t-test.

Fig 3. U.S. county-level regression models, with incident COVID-19 cases as the outcome.

A) 120-day incidence decreased by 11.72% (16.07%, 7.14%) on average, B) 90-day incidence by 12.01% (16.10%, 7.72%) on average, and C) 30-day incidence decreased by 12.82% (17.18%, 8.23) on average per 495 meters of elevation on average, after adjustment for covariates.

Fig 4. Summary of proposed mechanisms of difference in COVID-19 infection, transmission, and pathogenesis at high altitude.

Factors which may impact COVID-19 infection and death at high altitude include host, environmental, viral, and healthcare factors.