Exhaled aerosol increases with COVID-19 infection, age, and obesity

Abstract

COVID-19 transmits by droplets generated from surfaces of airway mucus during processes of respiration within hosts infected by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus. We studied respiratory droplet generation and exhalation in human and nonhuman primate subjects with and without COVID-19 infection to explore whether SARS-CoV-2 infection, and other changes in physiological state, translate into observable evolution of numbers and sizes of exhaled respiratory droplets in healthy and diseased subjects. In our observational cohort study of the exhaled breath particles of 194 healthy human subjects, and in our experimental infection study of eight nonhuman primates infected, by aerosol, with SARS-CoV-2, we found that exhaled aerosol particles vary between subjects by three orders of magnitude, with exhaled respiratory droplet number increasing with degree of COVID-19 infection and elevated BMI-years. We observed that 18% of human subjects (35) accounted for 80% of the exhaled bioaerosol of the group (194), reflecting a superspreader distribution of bioaerosol analogous to a classical 20:80 superspreader of infection distribution. These findings suggest that quantitative assessment and control of exhaled aerosol may be critical to slowing the airborne spread of COVID-19 in the absence of an effective and widely disseminated vaccine.

Keywords: COVID-19; aerosols; respiratory medicine; superspreaders.

Fig. 1.

Exhaled breath particles of 74 essential workers at No Evil Foods and of 120 volunteers at Grand Rapids Community College. (A) All participants; (B) “superspreader” (of aerosol particles) participants (first decile); (C) “superspreader” (of aerosol particles) participants (second decile); and (D) “low spreader” participants. Data represent particle counts per liter of exhaled air (particle diameter larger than 300 nm) for each of the 194 individuals. Error bars represent SD sample calculations based on 3 to 12 exhaled aerosol count measurements, with each measurement an average of counts over a 5-s time interval.

Fig. 2.

Exhaled breath particles as a function of BMI-years for volunteers reporting age and BMI (n = 146). Results of linear regression analysis are shown for the exhaled aerosol numbers from the superspreader and low spreader (of aerosol particles) subjects showing significant correlation, particularly for the superspreader subjects (r² = 0.98).

Fig. 3.

Exhaled breath particles and corresponding genomic SARS-CoV-2 viral RNA in experimentally infected (A) rhesus macaques (RM) and (B) African green monkeys (AGM). Both groups are segregated by species (n = 4; n = 8). The corresponding color-matched box-and-whisker plots of total exhaled breath particles represent iterative five 1-min sampling events to genomic viral RNA (color-matched circles) for each animal at each respective time point. Mean calculated correlation between time point-matched exhaled breath particle production and genomic viral RNA showed statistically significant correlations in 75% of the RM (RM01, r² = 0.93, P < 0.03; RM02, r² = 0.99, P < 0.004; RM04, r² = 0.98, P < 0.0008) and 50% of the AGM (AGM02, r² = 0.91, P < 0.04; AGM03, r² = 0.97, P < 0.01).

Fig. 4.

Exhaled breath particles and corresponding particle size distributions in experimentally infected (A) rhesus macaques (RM) and (B) African green monkeys (AGM); dpi, days postinfection.

Fig. 5.

Exhaled breath particles and corresponding particle size distributions in rhesus macaques (n = 4) infected with Mtb. Total particle counts per liter of air sampled as a measure of production during 10 min of continuous mask sampling for (A) all exhaled aerosol particles of >0.5 μm and (B) all exhaled aerosol particles of >1.0 μm. The total number of particles increased with time postinfection (PI) (in A), with the fraction of particles larger than 1 μm increasing less significantly, reflecting a high submicron fraction (> 90%) from 3 wk PI.