Evaluation of targets for normalization of viral pathogen concentrations in building dust

Building dust is potentially an important public health tool for monitoring human pathogens because it contains genetic material from pathogens. However, building dust is heterogeneous and presents technical challenges for environmental pathogen surveillance. To effectively use pathogen measurements in building dust for environmental surveillance, normalization procedures that account for human contribution to dust may be necessary and could help account for variability in occupancy and temporal resolution. Our analysis investigates potential indicators of human emission that could be used to normalize data on concentration of pathogens in dust. Ninety-nine vacuum dust samples were collected from 33 buildings and concentrations were measured for three emission indicators (i.e., human cytochrome B gene, Carjivirus communis (previously crAssphage), and pepper mild mottle virus) and two respiratory pathogens (i.e. SARS-CoV-2 and influenza A virus). All of these were detected in ≥91% of the samples, with the exception of influenza A virus (73%). Most pairings of indicators (nine of 10) showed a weak to moderate positive correlation with one another (Kendall's τ ≤ 0.28). Correlations between occupancy and emission indicators and respiratory pathogens were weak (Kendall's τ < |0.15|) and non-significant (p > 0.1). Among all potential markers, MT-CYB was most promising, with detection in 96% of the samples, higher concentration than others, and stronger correlation with other potential measures of occupancy.