Nationally Representative Levels of Selected Volatile Organic

Oct 28, 2013 - Environmental Science & Technology 2018 Article ASAP ... of a short path thermal desorption–gas chromatography/mass spectrometry meth...
0 downloads 0 Views 1005KB Size


Nationally Representative Levels of Selected Volatile Organic Compounds in Canadian Residential Indoor Air: Population-Based Survey Jiping Zhu,*,† Suzy L. Wong,§ and Sabit Cakmak‡ †

Exposure and Biomonitoring Division and ‡Air Health Effects Research Section, Health Canada, Ottawa, Ontario, Canada K1A 0K9 Health Analysis Division, Statistics Canada, Ottawa, Ontario, Canada K1A 0T6


S Supporting Information *

ABSTRACT: A comprehensive, population-based national indoor air survey was conducted in 2009−2011 in Canada. A total of 84 volatile organic carbons (VOCs) from 3218 houses, 546 apartments, and 93 other dwelling types were measured using passive sampling followed by thermal desorption GC/MS. A total of 12 VOCs were measured in both this study and the 1992 Canadian national study. Arithmetic means of VOCs in this study were 2−5 times lower than those in the 1992 study with the exception of a higher styrene level (1.13 μg·m−3). Comparing the geometric means of the 24 VOCs showed that levels for the VOCs in this study were comparable to those reported in Europe. They were generally within a factor of 2; 1,4-dichlorobenzene (0.21 μg·m−3) and 1,2,4-trimethylbenzene (0.51 μg·m−3) were noticeably lower in this study than in the European studies. There were 47 VOCs detected in more than 50% of Canadian households; 33 of them were higher in houses than in apartments for all nonsmoking homes, while only 4 were lower in houses than in apartments. A total of 11 of 47 VOCs were higher in smoking homes compared to nonsmoking homes, while the rest had similar levels. Principal component analysis identified several groups of VOCs with possible common sources.

effects.15 The domestic environment is most often studied16−20 because it is the environment where people spend the majority of their time and therefore it is important to estimate total inhalation exposure to VOCs.21,22 Other indoor environments such as offices,23 schools,24,25 and transportation vehicles26 have also been studied. Several studies investigating VOCs in indoor air have been carried out in Canada. These include the first Canadian national survey that was conducted in 1992 and monitored 26 VOCs in 750 houses across Canada27 and a 2003 regional study that conducted concurrent measurements of 37 VOCs in both indoor air and outdoor air of 75 houses in the city of Ottawa.5 In addition, indoor exposure to VOCs has been evaluated in recent years in other cities in Canada, including Regina,28 Windsor,29 and Quebec.30 Since 1994, health risk assessment of new chemicals used in Canada has been conducted according to regulations established under the Canadian Environmental Protection Act (CEPA), 1999.31 For the estimated 23 000 substances on the Domestic Substances List that were manufactured in, imported to, or used in Canada on a commercial scale between 1984 and

INTRODUCTION Volatile organic compounds (VOCs) are chemicals that have sufficiently high vapor pressure to allow them to partition primarily to the gas phase. The World Health Organization uses an upper boiling point limit of 240−260 °C to define VOCs.1 Indoor VOCs have many emission sources, including building materials and consumer products.2 Human exposure to VOCs in a nonoccupational environment occurs predominately through inhalation of air, particularly indoor air; in countries in North America, for example, people tend to spend the majority of their time indoors.3,4 In general, VOCs are present more frequently and at higher concentrations in indoor air than in outdoor air.5 Many VOCs have known toxicities.6 The negative health impact of occupational exposure to VOCs is well established, and regulations in many jurisdictions around the world have been implemented to minimize such exposure.7−9 Additionally, several studies have also suggested that exposure to indoor air VOCs can result in a variety of negative health outcomes.10 For example, it was recently reported that maternal exposure to trichloroethylene and perchloroethylene is associated with adverse birth outcomes.11 Residential exposure to VOCs has been linked to respiratory symptoms12,13 and to asthma in young children in particular.14 Globally, there is a continued interest in studying the levels of VOCs in indoor environments due to their potential health Published 2013 by the American Chemical Society

Received: Revised: Accepted: Published: 13276

July 10, 2013 October 6, 2013 October 28, 2013 October 28, 2013 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


with an internal diameter of 5.0 mm and a tube length of 90 mm and was packed with Carbopack B 60/80 (part no. N9307002, Perkin-Elmer, Inc., Shelton, CT). All tubes were cleaned in the laboratory, and both ends were sealed with brass hexagonal caps (part no. L1003015, Perkin-Elmer) prior to use. Collection of VOCs in indoor air has been described in detail elsewhere.36 Briefly, cleaned tubes were shipped first to the CHMS operation center in Ottawa, Ontario, for coding and identification and then to the CHMS MEC at the collection site using a courier ground-shipping service. At the MEC, the exposure end of the tube was changed to a Teflon analytical cap by a field technician and given to CHMS participants. The participant was asked to change the analytical cap to a meshscreen diffusion cap and deploy the tube in their living room or family room for seven consecutive days. At the end of the exposure period, the diffusion cap was changed back to the analytical cap for sealing, and the tube was placed in a sealed small aluminum container and shipped back to the testing laboratory in a prepaid Canada Post envelope. Starting and ending times of the exposure were recorded by the participants. Samples were analyzed within 72 h of arrival in the laboratory. Selected VOCs in the tube were desorbed through a thermal desorber unit (model ATD650, Perkin-Elmer) onto a gas chromatography/mass spectrometry (GC/MS) system (Agilent 7890A gas chromatograph coupled with an Agilent 5975C mass spectrometer, Agilent Inc., Santa Clara, CA). The primary thermal desorption was conducted at an initial purge time of 3 min at room temperature and then at 330 °C for 12 min, while the internal cold trap was set at −30 °C. The secondary desorption of VOCs from the cold trap to the GC column was carried out at 300 °C for 3 min. The inlet split was 20 mL·min−1 with a total desorb flow of 60 mL·min−1, while the secondary desorb flow was 6 mL·min−1. The desorbed VOCs were separated through a capillary GC column (DB-624, 60 m × 250 μm × 1.4 μm). The GC oven temperature was initially set at 45 °C for 3 min, increased by 6 °C·min−1 to 180 °C and then 30 °C·min−1 to 250 °C, and kept there for 10 min. The mass spectrometer was operated in electronic impact mode (70 eV), and detection was in the full scan mode (mass range 33−400 amu). Analytical Performance and QA/QC. Stock solution was prepared by mixing 30 μL of each neat chemical that was in liquid form. Chemicals in solid form were weighed separately and added to the liquid mixture. The working solution was diluted from the stock solution using methanol. A nine-level calibration (0.3−1000 ng) was conducted weekly (over the weekend). A regression coefficient (r2) of greater than 0.99 was achieved. Daily single-level (100 ng) calibration was conducted for each batch. The response of the single level was controlled within ±20% of the average response of the multicalibrations. Sample tubes were mailed back by the respondents. Over 90% of the IASs were returned to the testing laboratory within 7 days; only less than 1% of the tubes were returned longer than 25 days following the termination of exposure. Upon receipt of the IAS, the conditions of the tube were inspected by the laboratory staff. Tubes that were damaged were not analyzed. Results from samples that did not have information available on the exposure time or did not have an exposure time between 4 and 10 days (5760−14400 min) were not included in the survey data files and were not analyzed in this study. Analytical method performance and quality control data are available in Table S1 in the Supporting Information. The limit of detection (LOD) of the instrument was obtained by

1986 and not already assessed as new substances, the Canadian government completed in September 2006 the process of categorizing those that should be subjected to a screening-level risk assessment. Under the Chemicals Management Plan (CMP), screening assessments of about 200 prioritized chemicals have been completed,32 with additional priority substances to be assessed by 2020.33 Knowledge of the concentrations of indoor air VOCs is important to assess the health risks to humans. Although studies in Canada and globally have provided useful data for the assessment of human inhalation exposure to VOCs, the number of VOCs with adequate information for human exposure assessment is still limited. Specifically, only limited data are available on the presence and levels in indoor air for a number of VOCs that are considered priorities for assessment under the CEPA and CMP. Additionally, some reported data from previous studies were based on a small sample size, and the confidence in their representativeness is low. Therefore, in response to the need for nationally representative data for VOCs that are being assessed by the Canadian government, a national indoor air survey of 84 selected VOCs was conducted in Canada as part of the 2009−2011 Canadian Health Measures Survey (CHMS).34 The objective of the survey was to generate nationally representative data for the levels of selected VOCs in Canadian homes to support government efforts to assess the health risks of these VOCs to Canadians.

METHODS Survey Design. CHMS is an ongoing Canadian survey designed to provide comprehensive, direct health measures data at the national level.34 It uses two sampling frames for selecting its samples, an area frame of geographic units (clusters) for constructing and selecting collection sites and a list frame of the dwellings within each site. Samples were collected between September 2009 and November 2011. A general description of the sampling strategy for the survey is available elsewhere.35 Canadians aged 3−79 living in private households were included in the survey. Residents of Indian Reserves, institutions, and some remote regions and fulltime members of the Canadian Forces were excluded.35 The collection sites were allocated by five regions: Atlantic (2), Quebec (4), Ontario (6), Prairies (3), and British Columbia (3) (Figure S1, Supporting Information). Collection of data was scheduled to have each region sampled in both year 1 and year 2 of the survey and was spread among seasons, taking into consideration minimizing the movement of staff and equipment between sites. At each site, a mobile examination center (MEC), where physical measures and health examination were performed, was set up, and homes within a radius of about 50 km (or up to 100 km for rural areas) from the MEC were included.35 A total of 8520 households were selected, with respondents from 6465 households agreeing to participate. Of those who agreed to participate, respondents from 4722 households reported to MECs. At the MEC, an indoor air sampler (IAS) was given to one respondent per household. Respondents from 4686 households received an IAS. Either respondents from the remaining households refused to take an IAS or an IAS was not available to give to the respondent because of supply problems.36 Collection and Analysis of Indoor Air VOCs. VOCs in residential indoor air were collected using a passive sampling tube serving as an IAS. The tube was made of stainless steel 13277 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


Table 1. Descriptive Statistics (%) of Smoking and Nonsmoking Homes and Age of Dwellings in the Surveya house (n = 3218) smoking home nonsmoking home <10 years old 10 to <20 years old 20 to <30 years old 30 to <40 years old 40 to <50 years old 50 or more years old

apartment (n = 546)

other (n = 93)







9.1 90.9 13.7 15.0 16.5 17.4 11.9 23.2

12.2 87.8 13.3 13.2 15.5 19.2 12.1 26.7

17.9 82.1 9.0 11.7 14.8 20.1 11.5 18.9

21.0 79.0 11.4E 9.9E 15.4E 25.3 13.4E 24.6E

14.0 86.0 14.0 25.8 20.4 16.1 4.3 14.0

F 83.8 F F 21.1E F F F


House = single detached, double, row or terrace, or duplex. Apartment = low-rise apartment or high-rise apartment. Other = hotel, rooming/ lodging house, camp, mobile home, or other. Original = percentage values calculated from original sample numbers. Weighted = percentage values based on weighted data at the household level. Superscript “E” indicates use with caution (coefficient of variation 16.6−33.3%). Online “F” indicates the coefficient of variation was too high to reliably report the estimate.

that had at least one person that answered the survey. This weight corresponds to the number of households represented by that household for all of the households in Canada that are not an institution or a collective dwelling. Statistical analyses were performed using SAS version 9.2 and SUDAAN version 10 software. Standard errors, coefficients of variation, and 95% confidence intervals were calculated with the bootstrap technique.39 t tests were used to compare weighted geometric mean values between categories. Principal component analysis (PCA) with varimax rotation was carried out on the VOC measurements using R-2.15 software to find underlying components that may be used to identify common sources of the VOCs.

analyzing seven laboratory blank tubes, tubes that had been cleaned through the laboratory tube cleaning procedure, spiked with standards at several low levels (0.1, 0.3, 0.5, or 1 ng, depending on the individual VOCs). Each analytical batch of samples contained one laboratory blank tube. The amount (ng) of an analyte detected in the samples was first subtracted by the amount in the laboratory blank tube. If the blank-subtracted value was less than the LOD, the value was coded “< LOD”. For values above the LOD, the air concentration (μg·m−3) was calculated as follows: [blankcorrected amount in the sample (ng) × 1000]/[(exposure time (min) × uptake rate (mL·min−1)]. The uptake rates of VOCs were determined experimentally in a previous study.37 For VOCs whose uptake rates were not experimentally determined, the equation of 11.912 × diffusivity (cm2·s−1) − 0.3089 derived from the previous study was used.37 Field quality control samples for the survey included 173 field blanks, 74 travel blanks, and 79 pairs of duplicates deployed at the MEC.36 Nine VOCs, benzaldehyde, benzene, 2propanol, acetone, 2-methyl-1,3-butadiene, decane, ethylbenzene, limonene, and o-xylene, were detected in 50% or more of the field blanks. The amount of these nine VOCs in the samples was further subtracted by the median amount of the 173 field blanks, and the concentration was recalculated. Any resulting values below the LOD were coded “< LOD”.36 Statistical Analysis. Of the 4686 samplers given to the respondents, indoor air data from 3857 households were valid and included in the analysis in this study. Since the data were collected in 18 collection sites in 5 regions in Canada, the number of degrees of freedom was specified as 13 (18 collection sites minus 5 regional strata) for statistical analysis.38 A value of 1/2LOD was used to substitute “< LOD” for statistical analysis of the data set. The method detection limit (MDL) was calculated from log-transformed field blank data as the arithmetic mean plus the standard deviation multiplied by Student’s t value at the 99% confidence level for a one-tailed test. If the resulting MDL was lower than the LOD, the LOD value was used as the MDL. All estimates were produced using the household-level survey weights. In accordance with the weighting strategy, the weights for collection sites were multiplied by the selection weights for dwellings (households) and adjusted for nonresponse. Following the conversion of household weights into person weights, the latter were adjusted for nonresponse at the interview stage and the MEC stage.38 A survey weight was given to each household in the final sample, that is, the sample of households

RESULTS AND DISCUSSION Dwelling Types and Selection of VOCs. This population-based national indoor air survey covered many types of dwellings. It included not only 3218 houses (single detached, double, row or terrace, and duplex dwellings) but also 546 lowrise and high-rise apartments (Table 1). The latter were lacking in the 1992 Canadian national survey.27 The current survey also included 93 other private dwellings such as hotels, rooming/ lodging houses, camps, and mobile homes, which are often not included in indoor air surveys.16,20 The survey also covered many different ages of dwellings, including both newer and older ones. There were almost twice as many older dwellings (>50 years) than the newest (<10 years) in both house and apartment categories. The similar distribution of the age of dwellings between the original (nonweighted) and weighted dwelling numbers shows a random selection of the residence in the survey. In addition, both smoking homes, which are homes where someone smoked inside the home every day or almost every day, and nonsmoking homes were represented in the survey. The smoking rates from surveyed dwellings were similar to the Canadian smoking prevalence rate reported by Health Canada.40 Rates of smoking homes in this survey, however, were almost doubled in apartments as compared to houses. A total of 84 VOCs were measured in this survey (the full list of VOCs is available in Table S2, Supporting Information). The selection of VOCs was based on a combination of the need for nationally representative indoor air data to support government regulatory work on priority chemicals31−33 and the suitability of these chemicals to be measured by the chosen analytical method. The presence of VOCs in indoor air is an important source of exposure to humans. However, measured concen13278 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


Table 2. Indoor Air (μg·m−3) of 47 Volatile Organic Compounds with above 50% Detection Frequencies in Canadian Homes





1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

camphene limonene α-pinene tetrahydrofuran 2-butoxyethanol 1,4-dichlorobenzene naphthalene benzaldehyde 2-methyl-1,3-butadiene decamethylcyclopentasiloxane octamethylcyclotetrasiloxane cyclohexanone acetone 4-methyl-2-pentanone 2-pentanone 2-butanone bromodichloromethane trichloroethylene chloroform perchloroethylene tetrachlorocarbon (1-methylethyl)benzene 1,2,4-trimethylbenzene benzene 1,2,3-trimethylbenzene styrene m/p-xylene toluene ethylbenzene o-xylene dodecane undecane cyclohexane hexane decane heptane pentane decanal octanal nonanal 2-furancarboxaldehyde hexanal 1-propanol 1-pentanol 2-propanol 1-butanol 2-methyl-2-propanol

99.25 99.84 99.92 59.92 92.53 94.68 98.60 99.59 99.92 97.17 98.76 87.87 95.90 96.27 96.55 99.17 53.75 75.50 93.86 98.76 99.33 79.65 98.96 99.69 99.74 99.77 99.90 99.95 99.95 99.95 64.17 74.98 96.89 97.56 99.20 99.53 99.92 82.14 92.97 97.12 97.95 98.83 51.62 54.26 80.74 83.95 89.63

arithmetic meanb 0.77 37.89 13.10 0.34 6.16 5.52 2.62 3.98 5.83 41.32 6.70 0.98 9.00 0.54 0.57 3.23 0.07 0.21 0.62 1.94 0.32 0.25 1.37 1.93 4.33 1.13 14.44 17.80 4.22 4.33 1.71 5.27 1.95 4.33 6.81 5.64 7.96 1.45 2.10 3.45 2.99 20.61 0.33 0.71 7.31 2.26 0.66

(0.64−0.90) (32.78−42.99) (9.01−17.18 (0.25−0.43) (4.77−7.54) (4.06−6.97) (1.51−3.74) (3.20−4.77) (4.83−6.83) (35.47−47.18) (5.45−7.94) (0.67−1.30) (7.54−10.46) (0.40−0.68) (0.46−0.68) (2.49−3.98) (0.05−0.10) (0.06−0.36) (0.47−0.78) (1.08−2.79) (0.29−0.35) (0.17−0.32) (1.01−1.73) (1.56−2.30) (3.24−5.43) (0.94−1.32) (10.22−18.65) (14.10−21.51) (2.87−5.57) (2.97−5.68) (0.89−2.53) (3.76−6.79) (1.14−2.76) (2.99−5.68) (4.60−9.01) (3.54−7.74) (5.96−9.96) (1.07−1.83) (1.66−2.54) (2.65−4.25) (2.37−3.62) (13.32−27.90) (0.21−0.46) (0.48−0.94) (5.60−9.02) (1.75−2.77) (0.45−0.86)

geometric meanb 0.46 21.30 5.62 0.05 3.02 0.21 0.85 2.76 3.06 13.11 2.98 0.38 3.78 0.23 0.36 1.14 0.03 0.04 0.29 0.31 0.29 0.09 0.51 1.04 1.58 0.72 4.88 7.94 1.35 1.41 0.46 1.52 0.50 1.21 1.38 1.27 3.17 0.83 1.22 2.24 1.84 11.13 0.09 0.25 1.42 1.19 0.20

(0.38−0.54) (18.32−24.75) (4.40−7.19) (0.04−0.06) (2.30−3.82) (0.15−0.28) (0.73−1.00) (2.20−3.46) (2.60−3.59) (10.90−15.77) (2.57−3.46) (0.29−0.49) (3.13−4.57) (0.19−0.27) (0.29−0.44) (0.92−1.43) (0.02−0.04) (0.03−0.04) (0.22−0.40 (0.25−0.39) (0.27−0.32) (0.06−0.12) (0.42−0.61) (0.86−1.26) (1.31−1.90) (0.60−0.83) (4.05−5.87) (6.64−9.50) (1.12−1.63) (1.16−1.72) (0.39−0.54) (1.28−1.82) (0.36−0.69) (0.98−1.51) (1.13−1.70) (1.09−1.47) (2.56−3.93) (0.59−1.18) (0.91−1.65) (1.69−2.98) (1.48−2.98) (7.46−16.60) (0.07−0.12) (0.19−0.34) (0.98−2.04) (0.95−1.51) (0.16−0.26)








0.09 1.68 0.40 0.01 0.26 0.01 0.04 0.29 0.34 0.22 0.15 0.03 0.05 0.01 0.07 0.02 0.01 0.01 0.05 0.01 0.02 0.01 0.10 0.10 0.15 0.10 0.46 1.09 0.12 0.11 0.13 0.29 0.05 0.07 0.18 0.09 0.48 0.11 0.03 0.07 0.08 0.09 0.02 0.06 0.05 0.17 0.02

0.09 3.60 0.73 0.01 0.29 0.01 0.18 0.73 0.63 0.60 0.57 0.04 0.16 0.03 0.07 0.20 0.01 0.01 0.05 0.04 0.17 0.01 0.10 0.23 0.32 0.19 0.91 1.69 0.25 0.26 0.14 0.32 0.05 0.20 0.18 0.21 0.74 0.13 0.03 0.26 0.40 1.64 0.02 0.07 0.05 0.20 0.02

0.24 10.74 2.53 0.01 1.51 0.06 0.40 1.78 1.55 4.64 1.45 0.16 1.98 0.10 0.20 0.50 0.01 0.02 0.15 0.11 0.23 0.03 0.21 0.51 0.64 0.40 1.91 3.50 0.56 0.58 0.15 0.39 0.21 0.50 0.48 0.53 1.38 0.40 0.91 1.40 1.04 5.51 0.03 0.08 0.36 0.57 0.09

0.45 21.47 5.47 0.03 3.30 0.14 0.72 2.88 2.79 15.92 2.75 0.41 4.53 0.22 0.38 1.02 0.02 0.03 0.37 0.22 0.28 0.08 0.43 0.88 1.25 0.68 3.94 6.74 1.15 1.16 0.41 1.30 0.50 0.96 1.18 1.04 2.50 1.01 1.65 2.68 1.87 12.41 0.05 0.17 2.03 1.29 0.19

0.84 43.58 13.00 0.13 6.23 0.42 1.50 4.60 5.46 41.71 5.58 0.91 9.06 0.47 0.70 2.36 0.09 0.06 0.73 0.72 0.36 0.23 1.06 2.01 3.22 1.23 10.04 15.29 2.62 2.86 0.85 3.11 1.20 2.35 3.32 2.55 5.63 1.94 2.62 4.54 3.30 24.52 0.25 0.71 6.40 2.60 0.40

2.44 126.2 43.45 1.26 18.97 12.30 7.74 9.96 23.44 180.2 20.15 3.26 29.45 1.62 1.58 9.76 0.29 0.27 2.02 5.64 0.54 0.99 5.18 7.38 18.75 3.24 56.32 73.42 13.63 16.52 3.66 20.50 5.91 13.65 23.68 14.35 31.50 3.72 4.87 8.28 8.60 63.92 1.23 2.85 28.03 6.71 2.49

5.87 228.1 129.3 6.12 49.66 185.0 29.93 19.64 50.33 411.7 70.91 9.53 87.14 4.65 2.83 38.69 0.49 2.18 4.17 25.10 0.82 2.52 14.96 15.19 61.61 7.91 199.3 191.8 52.50 57.46 26.68 71.60 21.26 50.55 127.9 97.51 96.20 8.14 10.89 15.00 18.97 153.5 5.20 7.00 86.59 15.64 9.40

DF = percentage of detection frequency. bRanges in parentheses are the 95% confidence intervals. cPercentile of concentration distribution.

trations of VOCs in Canadian homes are not always available. For example, of the VOCs assessed by the government of Canada under the CMP,33 only about half of them actually had available indoor air data from Canada and/or elsewhere. VOCs in Canadian Residences. Of the 84 target VOCs, 37 were detected in less than 50% of Canadian homes (Table S2, Supporting Information). Among these 37 VOCs, almost half had detection frequencies of less than 1%. Two VOCs, 1,2dibromoethane and 1,2-dimethoxyl-4-(2-propenyl)benzene, were not detected in any of the homes. Only five VOCs (4methyl-3-penten-2-one, dibromochloromethane, 1,1′-biphenyl, 2-ethyl-1-hexanol, and 1-methyl-2-pyrrolidinone) had detection frequencies above 10%. Although these VOCs were detected

infrequently, the confirmation of their absence or the upper range distribution values (at the 99th percentile) are still valuable for assessing upper boundary estimates of human exposure to these substances. The rest of the 47 VOCs were detected in more than 50% of Canadian homes (Table 2). The detection frequencies ranged from 51.62% for 1-propanol to 99.95% for toluene, ethylbenzene, and o-xylene. Table 2 includes the arithmetic mean and geometric mean with their respective 95% confidence intervals (95% CI), as well as the concentrations at several key distribution percentiles. Levels at the 1st percentile and 99th percentile were reported as lower and upper limits of the data set. A comparison of arithmetic mean values for 12 VOCs that 13279 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


Table 3. Comparison of Indoor Air VOC Levels (μg·m−3) of This Study with Those of Three Other Population-Based Studies arithmetic mean

geometric mean


this study (n = 3857)

Canada41 (1992) (n = 757)

this study (n = 3857)

Helsinki16 (1997) (n = 201)

France20 (2005) (n = 490)

trichloroethylene perchloroethylene chloroform 1,4-dichlorobenzene benzene toluene ethylbenzene o-xylene m/p-xylene styrene 1,2,4-trimethylbenzene 1,2,3-trimethylbenzene decane hexane undecane cyclohexane α-pinene limonene hexanal octanal benzaldehyde 1-butanol 2-methyl-2-propanol 2-butoxyethanol naphthalene

0.21 1.94 0.62 5.52 1.93 17.8 4.22 4.33 14.4 1.13 1.37

0.50 2.70 1.50 18.9 5.40 40.8 8.20 5.60 20.7 0.30 11.5

<1.39 <1.28

1.0 1.4

1.57 14.62 2.17 1.88 6.13 0.84

4.2 2.1 12.2 2.3 2.3 5.6 1.0 4.1



0.04 0.31 0.29 0.21 1.04 7.94 1.35 1.41 4.88 0.72 0.51 1.58 1.38 1.21 1.52 0.50 5.62 21.3 11.1 1.22 2.76 1.19 0.20 3.02 0.85

2.43 2.65 <2.38 2.61 <2.0 9.09 11.57 8.66 3.59 4.18 7.02 3.37 1.62 0.55

5.3 1.2 6.2



Table 4. Statistically Significant Differences in Geometric Mean (GM) Values between Houses (H, n = 2925) and Apartments (A, n = 528), Excluding Smoking Homes




α-pinene o-xylene m/p-xylene 1,2,3-trimethylbenzene ethylbenzene 1,2,4-trimethylbenzene toluene hexane pentane 2-butanone (1-methylethyl)benzene tetrahydrofuran hexanal cyclohexane 2-furancarboxaldehyde camphene heptane 2-methyl-2-propanol 4-methyl-2-pentanone

3.87 0.76 2.56 0.83 0.66 0.25 3.48 0.59 1.39 0.49 0.04 0.03 4.67 0.14 0.45 0.14 0.43 0.07 0.06


> > > > > > > > > > > > > > > > > > >







<0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

styrene cyclohexanone 2-butoxyethanol benzene decane benzaldehyde undecane 2-pentanone 1-pentanol tetrachlorocarbon 2-propanol 1-butanol decamethylcyclopentasiloxane dodecane decanal bromodichloromethane 2-methyl-1,3-butadiene 1,4-dichlorobenzene

0.14 0.23 0.78 0.20 0.44 0.48 0.40 0.07 0.07 0.02 0.37 0.16 2.51 0.05 0.14 0.01 0.41 0.23


<0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.05 <0.05 <0.05 <0.01 <0.01 <0.01 <0.01

GMD = difference in geometric mean values between H and A homes.

for most of the compared VOCs (Table 3).16,20,42 Differences however do exist. The most noticeable examples were 1,4dichlorobenzene and 1,2,4-trimethylbenzene, where the geometric means found in our study (0.21 and 0.51 μg·m−3, respectively) were many times lower than the values reported in the French study (0.42 and 0.41 μg·m−3, respectively). The French study and Helsinki study were conducted in 2003− 2005 and 1996−1998, respectively. Our study has shown that

were measured both in this study and in the 1992 Canadian national survey27,41 showed, in general, 2−5-fold decreases in indoor air VOC levels in Canada over the past two decades, except for styrene levels, which had a higher value (1.13 μg· m−3) in this study than that (0.30 μg·m−3) in the 1992 study (Table 3). Internationally, indoor air levels from this study were generally within a factor of 2 of the reported values in Europe 13280 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


Table 5. Statistically Significant Differences in Geometric Mean (GM) Values between Smoking Homes (S) and Nonsmoking Homes (NS) house




2-methyl-1,3-butadiene 2-furancarboxaldehyde benzene styrene 2-pentanone 2-butanone 1,2,4-trimethylbenzene ethylbenzene toluene m/p-xylene naphthalene

8.81 2.03 1.12 0.44 0.14 0.53 0.13 0.29 1.55 1.01 0.12



> > > > > > > > > > >




<0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.05 <0.05 <0.05 <0.05

12.2 2.88 1.96 0.70 0.27 0.89 0.26 1.05 4.77 3.88 0.36


> > > > > > > > > > >


p <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

GMD = difference in geometric mean values between S and NS homes.

of indoor air VOCs can be made using principal component analysis (PCA). The first two principal components (PC1 and PC2) accounted for 13% and 8% of the total variation, respectively, and showed eight common clusters (letters A−H, circled) of VOCs (Figure 1). The numbers in Figure 1 correspond to the

VOC levels have declined in Canada and elsewhere, consistent with trends toward reducing indoor air sources of VOCs43 and reductions in automobile emissions through improved emissions technology.44 VOC Levels and the Home Environment. The inclusion of both houses and apartments in the study allowed us to compare VOC levels between these two types of dwellings. t tests showed that, for all nonsmoking homes, 33 of the 47 tested VOCs that had a detection frequency above 50% had lower geometric means in apartments than in houses; only 4 VOCs had higher values in apartments (Table 4). Smoking homes were excluded in the t test to avoid potential confounding from smoking indoors. Comparisons of VOC concentrations between houses and apartments are rarely reported. One of the factors that may influence the levels of certain VOCs in indoor air is the garage configuration. Wheeler et al. have compared BTEX (benzene, toluene, ethylbenzene, and xylenes) levels between all apartments and houses, including smoking homes, and found that the presence of an attached garage on the houses, especially an attached garage with a connecting door to the living space of the house, can lead to higher BTEX levels indoors.45 Effects of smoking on indoor air VOCs levels were tested between houses and apartments. Cigarette smoke contains many VOCs, including the BTEX compounds.46 Among the 11 VOCs that showed a statistically significant difference in both dwelling types, many were benzene derivatives, including BTEX and styrene (Table 5). Findings of higher indoor air levels of these VOCs in smoking homes are consistent with the observation that cigarette smoking and blood levels of BTEX and styrene were highly correlated in the United States National Health and Nutrition Examination Survey.47 Cigarette smoke was also identified as a possible indoor source for naphthalene, attributing to naphthalene levels in indoor air.48 Principal Component Analysis of VOCs. VOCs are present in many building materials and consumer products, resulting in complex profiles of VOCs in indoor air.2 The U.S. Department of Health & Human Services has established a comprehensive Household Products Database containing information on many chemicals found in different products.49 A VOC can be present in many household products, and a product may contain numerous VOCs, adding to the difficulties in identifying the contributing factors affecting VOC levels in indoor air. Despite the complex nature of VOC mixtures in indoor environments, source identification and apportionment

Figure 1. Principal component analysis of 47 VOCs that had detection frequencies above 50% in Canadian residential indoor air. The numbers in the figure correspond to the numbers in Table 3 for the identification of VOCs. Inset: A separate principal component analysis was run on the remaining 22 VOCs that had not been assigned to a cluster (indicated by the arrow). The first and third principal components are shown in the inset.

numbers of VOCs listed in Table 3. Compounds in clusters A and B are probably linked to the use of water-based cleaning agents in homes, as these compounds are often present in these products.50,51 Four clusters (C, D, E, and F) contained alkylbenzenes and aliphatic hydrocarbons that are often found in gasoline and in oil-based solvents.49 The fact that they were separated in several clusters indicates that although these VOCs are primarily used as solvents, different products may contain different VOCs. Notably, the simple aliphatic hydrocarbons appeared in cluster C and separated from aromatic hydro13281 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology

carbons in clusters D, E, and F (Figure 1), indicating the sources of aliphatic hydrocarbons and aromatic hydrocarbons might be different. On the left side is cluster G that comprises three halogenated hydrocarbons, namely, bromodichloromethane, trichloroethylene, and tetrachlorocarbon, despite the fact that these three chemicals may have quite different sources. Bromodichloromethane is a disinfection byproduct of chlorination of drinking water, while trichloroethylene and tetrachlorocarbon are solvents in cleaning products.49 Another VOC (cluster H) that was isolated from the rest was decamethylcyclopentasiloxane (DPS), which is predominantly found in personal care products.52 A subsequent analysis conducted using PC1 and PC3 on the remaining VOCs that were not included in the clusters separated a further four VOCs, namely, perchloroethylene (PCE), octamethylcyclotetrasiloxane (OTS), naphthalene (NAP), and 1,4-dichlorobenzene (D4B), from the rest of the VOCs (inset in Figure 1). These four VOCs have unique sources in indoor air. For example, perchloroethylene is primarily used in dry cleaning of clothing, and 1,4dichlorobenzene is a common component in mothballs.49 Strength and Limitations of the Study. One strength of this study is the data source. This is so far the largest population-based survey of VOCs in residential indoor air. The comprehensive study design and chosen analytical method allowed us to produce national estimates for a large number of VOCs; the study not only monitored VOCs that are commonly present and measured in other surveys, but also included several VOCs that were monitored for the first time. The unique data generated under this study can be used in the risk assessment and risk management of these VOCs. The large sample size also enabled us to examine statistically significant differences in indoor air VOC levels between houses and apartments and between smoking and nonsmoking homes. The survey design does not allow the production of reliable, representative estimates at collection site level or regional level. Therefore, comparisons of geographic differences and seasonal differences of indoor air VOCs were not made. Due to constraints in CHMS field operation, outdoor air was not measured in this study. As a consequence, we did not know the concentration ratios of VOCs in indoor air and outdoor air; such information is essential to understand the sources of VOCs in indoor air. One of the unique aspects of the survey was relying on home occupants to collect indoor air VOCs and to mail the samples to the testing laboratory for analysis. This approach avoided the need for the presence of laboratory technicians in homes for sampling; this is particularly useful for conducting a large-scale survey involving homes in rural areas. The sample collection and analysis methodology used in this study has been applied to a subsequent national indoor air survey in CHMS cycle 3, which started in January 2012, with the objective of generating national estimates of a new set of VOCs in Canadian residential indoor air by collecting indoor air samples from approximately 4000 homes over a two-year period.34



Corresponding Author

*Phone: 613-946-0305. E-mail: [email protected] Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This study was financially supported by the Canadian federal government. Indoor air samples were analyzed by CASSEN Testing Laboratories, Toronto, Ontario, Canada, under a contract with Health Canada. We sincerely thank many colleagues at Health Canada and Statistics Canada, particularly Leona Mackinnon, Xinghua Fan, Deborah Schoen, and Chris Hebbern, for reviewing, commenting on, and editing the manuscript during the institutional internal review process.


(1) Indoor Air Quality: Organic Pollutants. Report on a WHO Meeting, Berlin, Aug 23−27, 1987; EURO Reports and Studies 111; World Health Organization Regional Office for Europe: Copenhagen, 1989. (2) Otson, R.; Fellin, P. Volatile organics in the indoor environment: sources and occurrence. In Gaseous Pollutants: Characterization and Cycling; Nriagu, J. O., Ed.; John Wiley & Sons, Inc.: New York, 1992; pp 335−421. (3) Leech, J. A.; Wilby, K.; McMullen, E.; Laporte, K. The Canadian Human Activity Pattern Survey: report of methods and population surveyed. Chronic Dis. Can. 1996, 17, 118−123. (4) Guo, H.; Lee, S. C.; Chan, L. Y.; Li, W. M. Risk assessment of exposure to volatile organic compounds in different indoor environments. Environ. Res. 2004, 94, 57−66. (5) Zhu, J.; Newhook, R.; Marro, L.; Chan, C. C. Selected volatile organic compounds in residential air in the city of Ottawa, Canada. Environ. Sci. Technol. 2005, 39, 3964−3971. (6) List of Air Toxics in the 2005 NATA Assessment. http://www. (accessed July 2, 2013). (7) Hertsenberg, S.; Brouwer, D.; Lurvink, M.; Rubingh, C.; Rijnders, E.; Tielemans, E. Quantitative self-assessment of exposure to solvents among shoe repair men. Ann. Occup. Hyg. 2007, 51, 45−51. (8) Wilson, M. P.; Hammond, S. K; Nicas, M.; Hubbard, A. E. Worker exposure to volatile organic compounds in the vehicle repair industry. J. Occup Environ. Hyg. 2007, 4, 301−310. (9) Wang, S. W.; Qian, H.; Weisel, C.; Nwankwo, C.; Fiedler, N. Development of solvent exposure index for construction painters. J. Occup. Environ. Hyg. 2011, 8, 375−386. (10) Zhang, J.; Smith, K. Indoor air pollution: a global health concern. Br. Med. Bull. 2003, 68, 209−225. (11) Forand, S. P.; Lewis-Michl, E. L.; Gomez, M. I. Adverse birth outcomes and maternal exposure to trichloroethylene and tetrachloroethylene through soil vapor intrusion in New York State. Environ. Health Perspect. 2012, 120, 616−621. (12) Casset, A.; de Blay, F. Health effects of domestic volatile organic compounds. Rev. Mal. Respir. 2008, 25, 475−485. (13) Choi, H.; Schmidbauer, N.; Sundell, J.; Hasselgren, M.; Spengler, J.; Bornehag, C.-G. Common household chemicals and the allergy risks in pre-school age children. PLoS One 2010, 5 (10), e13423. (14) Rumchev, K.; Spickett, J.; Bulsara, M.; Phillips, M.; Stick, S. Association of domestic exposure to volatile organic compounds with asthma in young children. Thorax 2004, 59, 746−751. (15) Rumchev, K.; Brown, H.; Spickett, J. Volatile organic compounds: do they present a risk to our health? Rev. Environ. Health 2007, 22 (1), 39−55. (16) Edwards, R. D.; Jurvelin, J. K.; Jantunen, M. VOC concentrations measured in samples and residential indoor, outdoor and microenvironments in EXPOLIS-Helsinki, Finland. Atmos. Environ. 2001, 35, 4531−4543.


* Supporting Information S

Collection sites of the 2009−2011 CHMS (Figure S1), analytical method performance (Table S1), and indoor air levels of each measured VOC (Table S2). This material is available free of charge via the Internet at 13282 | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology


(17) Son, B.; Breysse, P.; Yang, W. Volatile organic compounds concentrations in residential indoor and outdoor and its personal exposure in Korea. Environ. Int. 2003, 29, 79−85. (18) Hippelein, M. Background concentrations of individual and total volatile organic compounds in residential indoor air of SchleswigHolstein, Germany. J. Environ. Monit. 2004, 6, 745−752. (19) Weisel, C. P.; Alimokhtari, S.; Sanders, P. F. Indoor air VOC concentrations in suburban and rural New Jersey. Environ. Sci. Technol. 2008, 42, 8231−8238. (20) Billionnet, C.; Gay, E.; Kirchner, S.; Leynaert, B.; AnnesiMaesano, I. Quantitative assessments of indoor air pollution and respiratory health in a population-based sample of French dwellings. Environ. Res. 2011, 111, 425−434. (21) Saarela, K.; Tirkkonen, T.; Laine-Ylijoki, J.; Jurvelin, J.; Nieuwenhuijsen, M. J.; Jantunen, M. Exposure of population and microenvironmental distributions of volatile organic compound concentrations in the EXPOLIS study. Atmos. Environ. 2003, 37, 5563−5575. (22) Ilacqua, V.; Hänninen, O.; Kuenzli, N.; Jantunen, M. F. Intake fraction distributions for indoor VOC sources in five European cities. Indoor Air 2007, 17, 372−383. (23) Salonen, H. J.; Pasanen, A. L.; Lappalainen, S. K.; Riuttala, H. M.; Tuomi, T. M.; Pasanen, P. O.; Bäck, B. C.; Reijula, K. E. Airborne concentrations of volatile organic compounds, formaldehyde and ammonia in Finnish office buildings with suspected indoor air problems. J. Occup. Environ. Hyg. 2009, 6 (3), 200−209. (24) Godwin, C.; Batterman, S. Indoor air quality in Michigan schools. Indoor Air 2007, 17 (2), 109−121. (25) Scheepers, P. T.; Konings, J.; Demirel, G.; Gaga, E. O.; Anzion, R.; Peer, P. G.; Dogeroglu, T.; Ornektekin, S.; van Doorn, W. Determination of exposure to benzene, toluene and xylenes in Turkish primary school children by analysis of breath and by environmental passive sampling. Sci. Total Environ. 2010, 408, 4863−4870. (26) Geiss, O.; Tirendi, S.; Barrero-Moreno, J.; Kotzias, D. Investigation of volatile organic compounds and phthalates present in the cabin air of used private cars. Environ. Int. 2009, 35 (8), 1188− 1195. (27) Fellin, P.; Otson, R. Assessment of the influence of climatic factors on concentration levels of volatile organic compounds (VOCs) in Canadian homes. Atmos. Environ. 1994, 28 (22), 3581−3586. (28) Héroux, M.-E.; Clark, N.; van Ryswyk, K.; Mallick, R.; Gilbert, N. L.; Harrison, I.; Rispler, K.; Wang, D.; Anastassopoulos, A.; Guay, M.; MacNeill, M.; Wheeler, A. J. Predictors of indoor air concentrations in smoking and non-smoking residences. Int. J. Environ. Res. Public Health 2010, 7, 3080−3099. (29) Stocco, C.; MacNeill, M.; Wang, D.; Xu, X.; Guay, M.; Brook, J.; Wheeler, A. J. Predicting personal exposure of Windsor, Ontario residents to volatile organic compounds using indoor measurements and survey data. Atmos. Environ. 2008, 42, 5905−5912. (30) Héroux, M.-E.; Gauvin, D.; Gilberta, N. L.; Guay, M.; Dupuis, G.; Legris, M.; Levesque, B. Housing characteristics and indoor concentrations of selected volatile organic compounds (VOCs) in Quebec City, Canada. Indoor Built Environ. 2008, 17, 128−137. (31) The Canadian Environmental Protection Act, 1999. http:// (accessed July 2, 2013). (32) The Government of Canada “Challenge” for chemical substances that are a high priority for action. http://www. (accessed July 2, 2013). (33) Chemicals Management Plan. http://www. (accessed July 2, 2013). (34) Canadian Health Measures Survey Web site. http://www. (accessed July 2, 2013). (35) Canadian Health Measures Survey (CHMS): definitions, data sources and methods.

pl?Function=getSurvey&SDDS=5071&lang=en&db=imdb&adm= 8&dis=2 (accessed July 2, 2013). (36) Patry-Parisien, P.; Zhu, J.; Wong, S. L. Implementation of the indoor air component of cycle 2 of the Canadian Health Measures Survey. Health Rep. 2013, 24 (5), 3−10. (37) Xian, Q.; Feng, Y.; Chan, C. C.; Zhu, J. Use of reference chemicals to determine passive uptake rates of common indoor air VOCs by collocation deployment of active and passive samplers. J. Environ. Monit. 2011, 13, 2527−2534. (38) Statistics Canada. Canadian Health Measures Survey (CHMS) Data User Guide: Cycle 2; Statistics Canada: Ottawa, 2013. (39) Rust, K. F.; Rao, J. N. K. Variance estimation for complex surveys using replication techniques. Stat. Methods Med. Res. 1996, 5, 281−310. (40) Canadian Tobacco Use Monitoring Survey: Summary of Annual Results for 2011. (accessed July 2, 2013). (41) Davis, C. S.; Otson, R. Estimation of emissions of volatile organic compounds (VOCs) from Canadian residences. In Volatile Organic Compounds in the Environment, ASTM STP 1261; Wang, W., Schnoor, J., Doi, J.; Eds.; American Society for Testing and Materials: West Conshohocken, PA, 1996; pp 55−65. (42) Weisel, C. P.; Zhang, J.; Turpin, B. J.; Morandi, M. T.; Colome, S.; Stock, T. H.; Spektor, D. M.; Korn, L.; Winer, A. M.; Kwon, J.; Meng, Q. Y.; Zhang, L.; Harrington, R.; Liu, W.; Reff, A.; Lee, J. H.; Alimokhtari, S.; Mohan, K.; Shendell, D.; Jones, J.; Farrar, L.; Maberti, S.; Fan, T. Relationships of indoor, outdoor, and personal air (RIOPA). Part I. Collection methods and descriptive analyses. Res. Rep.Health Eff. Inst. 2005, 130 (1), 1−127. (43) Wolkoff, P. Trends in Europe to reduce the indoor air pollution of VOCs. Indoor Air 2003, 13 (Suppl. 6), 5−11. (44) Pokharel, S. S.; Bishop, G. A.; Stedman, D. H.; Slott, R. Emissions reductions as a result of automobile improvement. Environ. Sci. Technol. 2003, 37 (22), 5097−5101. (45) Wheeler, A. J.; Wong, S. L.; Khoury, C.; Zhu, J. Predictors of indoor BTEX concentrations in Canadian residences. Health Rep. 2013, 24 (5), 11−17. (46) Polzin, G. M.; Kosa-Maines, R. E.; Ashley, D. L.; Watson, C. H. Analysis of volatile organic compounds in mainstream cigarette smoke. Environ. Sci. Technol. 2007, 41, 1297−1302. (47) Chambers, D. M.; Ocariz, J. M.; McGuirk, M. F.; Blount, B. C. Impact of cigarette smoking on volatile organic compound (VOC) blood levels in the U.S. population: NHANES 2003−2004. Environ. Int. 2011, 37 (8), 1321−1328. (48) Batterman, S.; Chin, J. Y.; Jia, C.; Godwin, C.; Parker, E.; Robins, T.; Max, P.; Lewis, T. Sources, concentrations, and risks of naphthalene in indoor and outdoor air. Indoor Air 2012, 22 (4), 266− 278. (49) Household Products Database. http://householdproducts.nlm. (accessed July 2, 2013). (50) Zhu, J.; Cao, X.-L.; Beauchamp, R. Determination of 2butoxyethanol emissions from selected consumer products and its application in assessment of inhalation exposure associated with cleaning tasks. Environ. Int. 2001, 26, 589−597. (51) Wolkoff, P.; Schneider, T.; Kildesø, J.; Degerth, R.; Jaroszewski, M.; Schunk, H. Risk in cleaning: chemical and physical exposure. Sci. Total Environ. 1998, 215, 135−156. (52) Wang, R.; Moody, P.; Koniecki, D.; Zhu, J. Low molecular weight cyclic volatile methylsiloxanes in cosmetic products sold in Canada: implication for dermal exposure. Environ. Int. 2009, 35, 900− 904.

13283 | Environ. Sci. Technol. 2013, 47, 13276−13283