Nationally Representative Levels of Selected ... - ACS Publications

Oct 28, 2013 - organic carbons (VOCs) from 3218 houses, 546 apartments, and 93 other dwelling types were measured using passive sampling followed by ...
1 downloads 0 Views 1005KB Size
Article pubs.acs.org/est

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 dx.doi.org/10.1021/es403055e | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology

Article

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

dx.doi.org/10.1021/es403055e | Environ. Sci. Technol. 2013, 47, 13276−13283

Environmental Science & Technology

Article

Table 1. Descriptive Statistics (%) of Smoking and Nonsmoking Homes and Age of Dwellings in the Surveya house (n = 3218) smoking home nonsmoking home > > > > > > > > > >

A A A A A A A A A A A A A A A A A A A

p

chemical

GMDa

GM

p