Multivariate evaluation of VOCs in buildings where people with non-specific building-related symptoms perceive health problems and in buildings where they do not

Abstract

Volatile organic compounds (VOCs) were sampled in buildings where people with non-specific building-related symptoms perceive health problems and in buildings where they do not. In total, nine persons and 34 buildings were included in the study. The obtained VOC data was evaluated using multivariate methods, to investigate possible systematic differences in air quality of 'problem' and 'non-problem' buildings. All individual compounds were included as variables in a multivariate partial least squares (PLS) data analysis. 'Problem' and 'non-problem' buildings separated into two distinct groups, showing that air samples of the two groups of building were chemically different. The study showed that there was an identifiable systematic difference in the collected VOC data between buildings causing and not causing problems for people with non-specific building-related symptoms (also called sick building syndrome, SBS). This is an important indication that even such volatile organic compounds that can be sampled by commonly used adsorbents are of importance for the presence of such symptoms. By coordination of procedures for sampling and analysis of VOCs in buildings between laboratories, to get large datasets and more general models, the method can become a useful diagnostic measure in evaluating indoor air and to identify chemical compounds and sources that contribute to SBS problems.