The data were analysed with extended linear mixed-effects models in order to identify influent variables and elaborate a multi-sector picture of formaldehyde exposures.