Forty-nine nonsmoking married women participated in a home personal ex
posure study for 28 volatile organic compounds (VOCs) and total volati
le organic compounds (TVOCs). The women were selected and classified a
ccording to 18 socioeconomic categories based on age (18-34 y, 35-49 y
, 50-64 y), family income (<$25K, $25K-$40K, >$40K), and husband's smo
king status. Of the 29 analytes, 21 demonstrated no statistically sign
ificant difference in concentration between nonsmoking and smoking hom
es. One VOC, trichloroethylene, was elevated in the nonsmoking homes a
nd seven VOCs, benzene, styrene, pyridine, 2-picoline, 3-picoline, 3-e
thylpyridine, and 3-ethenylpyridine were elevated in the smoking homes
. A correlation matrix and a factor analysis indicate that benzene and
styrene were not significantly correlated or associated with 3-etheny
lpyridine, a proposed vapor phase environmental tobacco smoke (ETS) ma
rker. All of the nitrogenous bases were significantly correlated with
3-ethenylpyridine, Benzene, styrene, and TVOC were not significantly c
orrelated with the number of cigarettes smoked; however, 3-ethenylpyri
dine was significantly correlated with the number of cigarettes smoked
. A Pearson correlation analysis indicated that gas heat and smoking h
usband were significantly correlated with elevated benzene concentrati
ons, but a multiple regression model for benzene accounted for less th
an 30% of the total variance. ETS variables accounted for only 8% of t
he total variance. In the smoking homes, an apportionment technique wa
s evaluated for selected VOCs in order to determine the median percent
age of each analyte attributable to BTS. The results, with percentages
attributable to ETS were TVOC (5.5%), benzene (13.2%), styrene (12.68
), pyridine (40.7%), 2-picoline (67.1%), 3-picoline (90.1%), 4-picolin
e (37.2%), and 3-ethylpyridine (62.0%), indoor air sources other than
ETS were also identified for limonene, tetrachloroethylene, 1,4-dichlo
robenzene, and alkylbenzenes.