A Favorable Effect of Higher Ambient Temperature on COVID-19 Deaths in the USA

Background: A scientific question of considerable importance is whether the SARs-CoV-2 has a seasonal effect, given the current COVID-19 pandemic. Methods: We test for a temperature effect of the deleterious effect of SARS-CoV-2 in the USA by grouping states by their ambient winter temperature. The death rate for each of the 50 states was used as a measure of the deleterious effect of SARS-CoV-2 in the USA. The multiple linear regression approach models the state death rate as a function of their winter temperature zone and population density. Findings: Using a progressive set of four regression models, both winter temperature, as measured by hardiness zone, and population density were strongly and independently associated with state deaths attributed to SARS-CoV-2 per million residents through April 13, 2020 (p-values <.0001). The model explains 64% of the variation in the death rates, and 74% of the variation when Louisiana, an outlier state, is removed from the model. The 5 warmest states had ≤22% the death rate of the 10-state temperature zone group that includes NY, after adjusting for population density. Overall, the states from other zones, both colder and warmer than zone 6, had approximately equivalent 40% lower death rate than zone 6. Interpretation: A strong temperature association has been identified for SARS-CoV-2. This new understanding could assist in determining a risk-mitigation strategy for re-opening the country based on population density, temperature, and known hotspots. Funding: None