Long-term indoor VOC concentrations assessment a trend analysis of distribution, disposition, and personal exposure in cohort study samples

Inhalation is one of the entry ports for different chemicals into the human body. In order to investigate this application route and its negative health effect to humans, the presence of volatile organic compounds (VOCs) in indoor air is monitored since many years. To assess global trends and changes of the distribution and disposition of VOCs and the corresponding personal exposure, this study analyzed annual indoor air concentrations collected over a period of 9 (2006-2014) years in the context of a birth control cohort study of 72 VOCs. Additionally, Short Time-series Expression Model (STEM) was used to identify certain correlation for VOCs from different compound classes. For similar to 42 % of the compounds, a tendency to lower annual median indoor air concentrations was found, and for similar to 10 % of the VOCs, a trend to higher annual median indoor air concentrations. No such tendencies were observed for similar to 22 % of the investigated compounds. For similar to 26 % of the VOCs, the applied linear regression model was not suitable to predict global trends as annual median values were not linearly distributed. Mann-Kendall test was used to (i) confirm the results from the linear regression model and to (ii) calculate trends for those compounds, where linear regression was found to be unsuitable. Thus, for only approximately four of the investigated VOCs, no prediction was possible using both statistical approaches. STEM analysis revealed the connection of benzene, ethylbenzene, m+p xylene, alpha-pinene, 3-carene, pentadecane, and decamethlycyclopentasiloxane, in addition to the correlation of 1-butanol, chlorobenzene, heptanal, and 2-ethyl-1-hexanol concentrations