VOCs and "Sick Building Syndrome": Application of a New Statistical Approach for SBS Research to U.S. EPA BASE Study Data

Causal associations between "Sick Building Syndrome" (SBS) symptoms and specific environmental stressors were explored using a recently developed approach employing Principal Components Analysis (PCA) and Logistic Regression [1]. This approach was applied to data collected by the U.S. EPA from 1995 to 1996 in 28 large U.S. office buildings in the Building Assessment Survey and Evaluation study. PCA was used to apportion measurements of 13 VOC species and carbon monoxide into source-related vectors. A reduced set of four source-based exposure vectors, tentatively identified as photocopiers, automotive emissions, environmental tobacco smoke, and latex paints were derived from the data. Regression analyses indicate statistically significant associations between mucous membrane related symptoms and the PCA photocopier vector (odds ratio=1.2, p=0.04), after adjustment for age, gender, smoking status, presence of carpeting, and thermal exposure. Similar relationships (odds ratio=1.2, p=0.03) were found between sore throat symptoms and the PCA paint vector. Odds ratios are given per unit increase in PCA vector which range over 5-6 units.