Environ. Sci. Technol. 2006, 40, 2163-2170
Evaluation of Urinary Methoxyphenols as Biomarkers of Woodsmoke Exposure R U S S E L L L . D I L L S , * ,† MICHAEL PAULSEN,† JACQUI AHMAD,† DAVID A. KALMAN,† FADY N. ELIAS,‡ AND CHRISTOPHER D. SIMPSON† Department of Environmental and Occupational Health Sciences, School of Public Health and Community Medicine, University of Washington, Box 357234, Seattle, Washington 98195-7234, and Department of Orthopedic Surgery, Henry Ford Hospital, 2799 West Grand Boulevard, Detroit, Michigan 48202
Urinary methoxyphenols have been proposed as biomarkers for woodsmoke exposure, but the relationship between exposure and urinary methoxyphenol concentrations has not been characterized. We collected personal particulate matter2.5 and urine samples from 9 adults experimentally exposed to smoke from an open wood fire to characterize this relationship. Personal exposures (PM2.5 mean 1500 µg/ m3) varied 3.5-fold. Twenty-two methoxyphenols, levoglucosan, and 17 polynuclear hydrocarbons were quantified by gas chromatography/mass spectrometry assays for personal filter samples and urine samples. Most methoxyphenols had measurable preexposure levels. Propylguaiacol, syringol, methylsyringol, ethylsyringol, and propylsyringol had peak urinary concentrations after the woodsmoke exposure. Eight subjects had peak urinary elimination of methoxyphenols within 6 h (t1/2 3-5 h), whereas one had delayed elimination. Several metrics for urinary excretion were evaluated. Analyte concentration was greatly affected by diuresis. Excretion rate and analyte concentrations normalized by creatinine gave a clearer signal and were equivalent in predictive ability. Twelve-hour average creatininenormalized concentrations of each of the 5 methoxyphenols gave a Pearson correlation g 0.8 with their particlephase concentration. The sum of urinary concentrations for the 5 methoxyphenols versus levoglucosan on personal filters gave a regression coefficient of 0.75. This sum versus PM2.5 gave a regression coefficient of 0.79. The intercept of this regression suggests that the threshold for detection of an acute exposure event would be approximately 760 µg/m3 particulate matter from woodsmoke. The signalto-noise (12-h postexposure average/preexposure average) ranged from 1.1 to 8 for the 5 methoxyphenols. Analysis of multiple compounds provided assurance that elevations were not artifactual due to food or other products.
Introduction Woodsmoke exposure impacts large numbers of the world’s population. Use of wood as an energy source varies greatly * Corresponding author phone: (206)543-3263; e-mail: russ1@ u.washington.edu. † University of Washington. ‡ Henry Ford Hospital. 10.1021/es051886f CCC: $33.50 Published on Web 03/08/2006
2006 American Chemical Society
by region with developing countries using wood for a greater proportion of their total energy needs (1). About 2 billion people relied on biomass fuels for their daily energy needs in 1996 (2). Smoke from vegetation and forest fires has affected large regions and numbers of people. The 1997-1998 wild fires in Indonesia exposed an estimated 70 million Southeast Asians to woodsmoke (3); smoke was implicated in 527 deaths and 15 800 hospitalizations in eight provinces (4). Smoke from large wildfires can impair air quality at distant urban centers (5). Occupational exposure to woodsmoke is high among charcoal production workers (6) and firefighters (7). Ezzati and Kammen (8) have reviewed epidemiological studies of indoor air pollution in developing countries. Exposures to particulate matter (PM) were high (200-5000 µg/m3) when biomass was burned without venting for cooking and heat. These exposures were linked to acute lower respiratory infections, chronic obstructive pulmonary disease (COPD), asthma, low birth-weight, cataracts, lung cancer, and otitis media. In industrialized nations, short-term exposure to air pollution from residential wood combustion was associated with an increase of asthma symptoms (9). Exposures were lower in industrialized nations [20-70 µg/ m3 total suspended particles (10)] because of vented, airtight appliances and less reliance on this source of energy for cooking and heating. Wild fire smoke exposures to communities precipitated many adverse health effects: exacerbation of asthma (11), COPD (12), and bronchitis (13). Two classes of exposure biomarkers for woodsmoke have been employed in previous studies. Polycyclic aromatic hydrocarbons (PAH) are present in woodsmoke (14). PAH metabolites 2-naphthol and 1-hydroxypyrene were used as biomarkers to woodsmoke exposure in Brazilian charcoal workers (6). Methoxyphenols are derived from the pyrolysis of lignin (15) and were thought to be a specific tracer for woodsmoke (15-17). However, lignite and lesser rank coals contain methoxyphenols (18), and methoxyphenols were elevated in urine of coke workers (19). Methoxyphenols were investigated as biomarkers of woodsmoke exposure in a preliminary study (20). We extend this preliminary work to demonstrate dose-response, variability, and applicability of this family of biomarkers in a realistic exposure situation mimicking indoor open fire cooking.
Experimental Section Reagents. Sources of many chemicals were described previously (21, 22); synthesis of additional chemicals is described in the Supporting Information. For purity, retention time, and quantification mass for analytes and standards, see Table 1 in the Supporting Information. Exposures. Procedures involving human subjects were approved by the University of Washington’s Human Subjects Division. Subjects were healthy, nonsmoking adults between ages 20-65. For 2 days prior and for the duration of this study, subjects were asked not to consume foods or beverages containing woodsmoke flavorings, expose themselves to woodsmoke, or consume food cooked on open fires or charcoal barbeques. Nine adults (4 men and 5 women) were exposed to woodsmoke from a continuous open fire for 2 h. Two hours was a sufficient time to see a rise in urinary methoxyphenols based on an earlier study (20). Subjects were exposed in a hexagonal yurtlike structure (3 m sides, 2.4 m height) with a smoke hole centered in the ceiling above the fire. Subjects sat approximately 0.75 m from the fire and were allowed to moderate their exposure at will. Smoke VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2163
TABLE 1. Concentrations of Chemicals in Personal Air Particulate Samples chemical
concn (µg/m3)
chemical
concn (µg/m3)
guaiacol methylguaiacol ethylguaiacol propylguaiacol eugenol vanillin cis-isoeugenol trans-isoeugenol acetovanillone guaiacylacetone syringol 2,3-dimethoxyphenol methylsyringol ethylsyringol allylsyringol propylsyringol syringaldehyde acetosyringone coniferaldehyde propylsyringone
0.001 ( 0.002 0.006 ( 0.002 0.002 ( 0.002 0.006 ( 0.005c 0.002 ( 0.002c 0.262 ( 0.227 0.014 ( 0.015c 0.023 ( 0.014 0.200 ( 0.230 0.673 ( 0.738 0.009 ( 0.009 0.001b,c 0.046 ( 0.049 0.050 ( 0.060d 0.087 ( 0.108d 0.033 ( 0.040d 2.859 ( 1.894 1.880 ( 1.332 6.538 ( 3.771 0.620 ( 0.425
butylsyringone sinapylaldehyde levoglucosan fluorene phenanthrene anthracene fluoranthene pyrene retene benz[a]anthracene chrysene benzo[b]fluoranthene benzo[k]fluoranthene benzo[e]pyrene benzo[a]pyrene perylene indeno(1,2,3-cd)pyrene 1,2:5,6-dibenzanthracene benzo(ghi)perylene coronene
0.044 ( 0.042d 4.432 ( 2.211 123 ( 59a 0.011 ( 0.003 0.007 ( 0.001 0.006 ( 0.002 0.015 ( 0.012 0.020 ( 0.014 0.064 ( 0.108 0.081 ( 0.042 0.036 ( 0.019 0.033 ( 0.015 0.037 ( 0.024 0.021 ( 0.009 0.047 ( 0.021 0.009 ( 0.004 0.008 ( 0.003 0.012 ( 0.004 0.019 ( 0.009 0.007 ( 0.003
a Mean ( SD. N ) 9. Recoveries provided in Table 3 of the Supporting Information. b Single value g quantification limit. c Presence of peak in full scan mode could not be confirmed in three samples due to low concentration of analyte. d Presence could not be confirmed in lowest mass filter.
density increased with height and subjects less tolerant of the smoke sat lower in their chairs but all remained in the yurt. A mixture of barkless softwood (Douglas fir, Pseudotsuga menziesi) and hardwood (oak, species unknown) was burned. We examined the impact of smoked-flavored food and an over-the-counter drug on the urinary methoxyphenols in a separate experiment in a extension of earlier work (20). Analysis of urinary methoxyphenols was performed by the unmodified earlier method (20). Following a 4-day control period, 1 day passed between the ingestion of realistic quantities of each item. Items consumed were an expectorant (active ingredient: guaifenesin [glyceryl guaiacolate]; 200 mg), potato chips (alder-smoke flavor, 2 servings, 42.5 g/serving), beef jerky (hickory-smoke flavor, 5 servings, 53 g/serving), and barbeque sauce (1 serving, 30 mL, 35 g/serving). Personal Monitoring. One personal PM2.5 sample was collected for each subject using the Harvard Personal Environmental Monitor for PM2.5 (HPEM2.5; Harvard School of Public Health, Boston, MA) (23). Air was sampled at breathing zone level. Details are provided in the Supporting Information. Area Monitoring. One subject was also fitted with an integrating nephelometer (model pDR-1200, 10 s intervals, maximum response with 0.1-10 µm particles; ThermoElectron, Waltham, MA) for a continuous estimate of particle exposure, and a continuous monitor for CO, CO2, and temperature (Model 8551 Q-Track; TSI, Shoreview, MN). Subjects were within 2 m of each other. Urine Sampling. For 24 h prior to the exposure, subjects collected all urine voids at will in separate containers for a baseline of methoxyphenol excretion. Subjects then collected all urine voids at will for 48 h postexposure for measuring woodsmoke biomarker elimination. Samples were stored frozen (-20 to -5 °C). Sample Analyses. Filters were extracted by a previously described protocol (21). Extracts were analyzed for levoglucosan (22), methoxyphenols (21), and PAH (24) by gas chromatography/mass spectrometry. Urine samples were analyzed by an earlier procedure (20) with extensive modifications (see the Supporting Information).
Results Personal PM2.5 spanned about 3.5-fold (840-3000 µg/m3; mean ( SD, 1515 ( 682 µg/m3). Area PM2.5 spanned 4 orders 2164
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
of magnitude (0.02-33 µg/m3; arithmetic mean ( SD, 1.58 ( 0.72 µg/m3; geometric mean ( SD, 2.79 ( 3.45 µg/m3). Exposure intensity was variable (see Figure 1 in the Supporting Information). Temperature, which could affect phase distribution of methoxyphenols, rose through the exposure period (21-36 °C; mean 32 °C). The particulate concentration of levoglucosan (123 ( 59 µg/m3) was 102-104 higher than other chemicals (Table 1). Benzo[a]anthracene (0.08 µg/m3) and retene (0.06 µg/m3) were 2-10-fold higher in concentration than other PAH. Of the PAH, retene had the greatest variability (170% CV); two subjects had personal particulate concentrations of g0.2 µg/ m3, and the rest had concentrations 10-100-fold less. Otherwise, PAH variation was generally < 60% CV. See Table 2 in the Supporting Information for recoveries. Less volatile methoxyphenols had higher particulate concentrations (Table 1). Another trend was apparents carbonyls (e.g. vanillin) had higher concentrations than those with alkyl or alkenyl substituents. The confluence of the two trends was seen by low volatility carbonyls having highest concentrations, while the corresponding alkylsyringols had concentrations 50-1000-fold less. Three samples were reanalyzed in full scan mode to confirm peak identity; exceptions to confirmations are noted in Table 1. We computed three metrics for methoxyphenol urinary elimination: urine concentration, urine concentration normalized by urinary creatinine concentration, and urinary excretion rate. Both creatinine normalization and urinary excretion rate were effective in reducing the variability due to changes in diuresis and made the peak of elimination more apparent (see Figure 2 in the Supporting Information). Generally, subjects showed similar urinary elimination profiles. For several syringols (syringol, methylsyringol, ethylsyringol, and propylsyringol) and guaiacols (propylguaiacol, cis-isoeugenol, and trans-isoeugenol), the peak of elimination was at approximately 5-6 h postexposure and was more apparent for syringols. The peak elimination was shifted toward 10 h in one subject (see Figure 3 of the Supporting Information). Eugenol, vanillin, acetovanillone, allylsyringol, syringaldehyde, acetosyringone, coniferaldehyde, propylsyringone, and butylsyringone had no clear peak of elimination. Since taking a timed sample to capture a peak apex after an uncontrolled exposure event would be difficult in a
FIGURE 1. Comparison of the amount of methoxyphenols inhaled (based on methoxyphenols concentrations in the filter samples) to the amount of methoxyphenols eliminated in urine. Also displayed are the vapor pressures of the chemicals and the mass eliminated corrected for the nominal background elimination of the methoxyphenols. population, average urinary concentration rather than a peak concentration would be more useful. With a threshold set to the mean + 3 SD of the preexposure levels for each subject, this subject-specific threshold was exceeded by 12- and 24-h average concentrations in approximately 35% of the cases (9 subjects × 22 compounds). Excretion rate and creatinine concentration gave a higher incidence than unadjusted concentrations. Also, 12-h averages had slightly higher incidence than 24-h averages (see Table 4 in the Supporting Information). Subject data for three metrics (concentration, creatinine adjusted concentration, and excretion rate) are presented in Table 2. Background levels of methoxyphenols varied greatly between subjects (40-200% CV). Guaiacol, eugenol, vanillin, and acetovanillone had the highest background levels. Other compounds were 1-2 orders of magnitude lower with the higher molecular weight carbonyls being the lowest. The maximum values of elimination were only about twice that of the 12- or 24-h averages. Pearson correlations were performed for various urinary metrics with particle phase concentration of individual methoxyphenols (see Table 5 in the Supporting Information). No urinary metric was significantly correlated with the particle-phase air concentration for guaiacol, allylsyringol, and most carbonyl methoxyphenols (vanillin, acetovanillone, coniferaldehyde, syringaldehyde, and acetosyringone). Among the urinary metrics, those based on creatinine-adjusted concentration were most frequently correlated (p e 0.05) with concentrations in the air particulate. Particulate concentrations of propylguaiacol, the isoeugenols, guaiacylacetone, syringol, methylsyringol, and propylsyringol were significantly correlated with their urinary 12- and 24-h average creatinine-adjusted concentrations. Among the methoxyphenols with the highest response for 12-h average creatinine-adjusted concentrations (propylguaiacol, cis-isoeugenol, trans-isoeugenol, syringol, methylsyringol, ethylsyringol, and propylsyringol), propylsyringol was the only compound that did not have a significant correlation (R ) 0.7-0.9) with PM2.5 or levoglucosan concentrations in air (see Table 6 in the Supporting Information).
Pearson correlations among the particle-bound methoxyphenols were calculated (see Table 7 in the Supporting Information). With three exceptions (the most volatile methoxyphenols), all of the methoxyphenols were highly correlated (R ) 0.91-1.00; p e 0.05) in the particulate samples (see the Supporting Information). With the same exceptions individual methoxyphenol concentrations in particulate were significantly correlated with levoglucosan (R ) 0.81-0.96) and PM2.5 (R ) 0.83-0.97). Urinary methoxyphenols (average 12-h postexposure creatinine-adjusted concentration) had some similarities in correlations among marker compounds to those found with the particulate measurements. The majority of urinary methoxyphenols were significantly correlated (R ) 0.670.91; p e 0.05) but less strongly than within the particulate samples (see Table 8 in the Supporting Information). We compared the total mass eliminated in 24 h after exposure for each methoxyphenol with the mass of that methoxyphenol inhaled assuming complete absorption, standard ventilation rate for adults at rest [0.5 m3/h (25)], and no contribution to exposure from vapor phase methoxyphenols. For chemicals with vapor pressures (see Table 9 in the Supporting Information) less than 10-4 Torr the mass eliminated was within an order of magnitude of the mass assumed absorbed (Figure 1). The mass eliminated corrected by the average preexposure elimination of methoxyphenols was also compared. In cases where the preexposure value was greater than the postexposure value, the data were not included in the calculation of the mean. Eugenol and acetosyringone had the highest number (6) of data points excluded for this reason. As the vapor pressure of the methoxyphenols increased, much more methoxyphenol was eliminated than could be accounted for in the particulate. For allylsyringol and methoxyphenols of greater volatility (chemicals to the right of allylsyringol in Figure 1), 99% or more of the mass eliminated in urine could not be attributed to particulate matter. Random fluctuations or bias in any single methoxyphenol concentration would be made less influential if the sum of several chemicals’ concentrations were calculated as the VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2165
2166 9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
TABLE 2. Measures of Methoxyphenol Elimination in Urine guaiacols
H-
methyl-
ethyl-
propyl-
eugenol
vanillin
cis-isoeugenol
trans-isoeugenol
acetovanillone
guaiacylacetone
coniferaldehyde
concentration preexposurea 12-h maxb 12-h averageb 24-h averageb creatinine adjusted preexposure 12-h max 12-h average 24-h average excretion rate preexposure 12-h max 12-h average 24-h average
(µg/mL) 0.691 ( 0.322c 0.924 ( 0.680 0.486 ( 0.279 0.524 ( 0.280 (µg/g) 102.73 ( 76.10 156.11 ( 174.96 95.06 ( 66.30 85.66 ( 53.41 (µg/min) 0.846 ( 0.517 1.550 ( 1.630 0.808 ( 0.578 0.786 ( 0.488
(µg/mL) 0.075 ( 0.073 0.135 ( 0.093 0.085 ( 0.059 0.083 ( 0.060 (µg/g) 8.35 ( 7.30 21.67 ( 11.54 14.97 ( 8.92 12.03 ( 6.30 (µg/min) 0.062 ( 0.051 0.201 ( 0.096 0.120 ( 0.057 0.103 ( 0.051
(µg/mL) 0.025 ( 0.021 0.038 ( 0.033 0.027 ( 0.024 0.029 ( 0.027 (µg/g) 3.31 ( 2.58 7.88 ( 4.99 4.93 ( 3.27 4.93 ( 3.03 (µg/min) 0.035 ( 0.034 0.071 ( 0.059 0.041 ( 0.028 0.044 ( 0.032
(µg/mL) 0.001 ( 0.001 0.006 ( 0.004 0.004 ( 0.003 0.004 ( 0.002 (µg/g) 0.20 ( 0.12 1.33 ( 0.69 0.88 ( 0.53 0.63 ( 0.31 (µg/min) 0.002 ( 0.001 0.012 ( 0.007 0.007 ( 0.004 0.006 ( 0.003
(µg/mL) 0.237 ( 0.324 0.185 ( 0.278 0.068 ( 0.082 0.110 ( 0.121 (µg/g) 79.11 ( 194.77 19.83 ( 23.91 8.91 ( 7.01 11.24 ( 7.82 (µg/min) 0.501 ( 1.058 0.239 ( 0.361 0.093 ( 0.102 0.118 ( 0.107
(µg/mL) 0.057 ( 0.029 0.126 ( 0.173 0.062 ( 0.059 0.060 ( 0.047 (µg/g) 9.44 ( 6.46 14.29 ( 8.51 9.55 ( 3.51 8.50 ( 2.93 (µg/min) 0.077 ( 0.054 0.156 ( 0.104 0.084 ( 0.043 0.078 ( 0.034
(µg/mL) 0.002 ( 0.003 0.016 ( 0.009 0.009 ( 0.006 0.007 ( 0.003 (µg/g) 0.44 ( 0.71 3.01 ( 1.44 1.83 ( 1.08 1.29 ( 0.49 (µg/min) 0.003 ( 0.004 0.027 ( 0.014 0.015 ( 0.007 0.011 ( 0.004
(µg/mL) 0.014 ( 0.010 (8) 0.039 ( 0.026 0.025 ( 0.018 0.020 ( 0.017 (µg/g) 2.59 ( 3.48 (8) 7.37 ( 3.80 4.56 ( 2.70 3.27 ( 1.73 (µg/min) 0.017 ( 0.020 0.068 ( 0.044 0.037 ( 0.021 0.029 ( 0.018
(µg/mL) 0.210 ( 0.170 0.433 ( 0.198 0.213 ( 0.097 0.199 ( 0.084 (µg/g) 25.24 ( 6.05 49.97 ( 13.97 31.85 ( 9.47 28.22 ( 8.55 (µg/min) 0.217 ( 0.087 0.543 ( 0.221 0.289 ( 0.094 0.266 ( 0.084
(µg/mL) 0.020 ( 0.013 0.026 ( 0.018 0.019 ( 0.012 0.019 ( 0.011 (µg/g) 2.94 ( 2.73 5.20 ( 2.52 3.79 ( 1.87 3.39 ( 1.79 (µg/min) 0.024 ( 0.017 0.046 ( 0.027 0.030 ( 0.015 0.028 ( 0.014
(µg/mL) 0.003 ( 0.002 0.005 ( 0.007 0.002 ( 0.002 0.002 ( 0.002 (µg/g) 0.72 ( 1.25 1.77 ( 3.01 0.63 ( 0.90 0.67 ( 0.91 (µg/min) 0.005 ( 0.008 0.010 ( 0.014 0.004 ( 0.006 0.006 ( 0.007
syringols
H-
methyl-
ethyl-
propyl-
allyl-
syringaldehyde
propylsyringone
concentration preexposure 12-h max 12-h average 24-h average creatinine adjusted preexposure 12-h max 12-h average 24-h average excretion rate preexposure 12-h max 12-h average 24-h average
(µg/mL) 0.007 ( 0.005 0.053 ( 0.030 0.032 ( 0.022 0.022 ( 0.014 (µg/g) 1.11 ( 0.85 11.28 ( 6.92 6.53 ( 5.04 4.21 ( 2.57 (µg/min) 0.009 ( 0.005 0.098 ( 0.061 0.052 ( 0.032 0.035 ( 0.019
(µg/mL) 0.006 ( 0.008 0.030 ( 0.019 0.018 ( 0.013 0.014 ( 0.010 (µg/g) 0.71 ( 0.86 5.98 ( 3.94 3.74 ( 2.92 2.58 ( 1.83 (µg/min) 0.007 ( 0.009 0.057 ( 0.036 0.031 ( 0.018 0.023 ( 0.014
(µg/mL) 0.007 ( 0.005 0.024 ( 0.019 (8) 0.014 ( 0.010 (8) 0.011 ( 0.007 (8) (µg/g) 1.06 ( 1.13 3.10 ( 2.30 (8) 2.24 ( 1.93 (8) 1.57 ( 0.93 (8) (µg/min) 0.008 ( 0.006 0.032 ( 0.026 (8) 0.019 ( 0.014 (8) 0.015 ( 0.008 (8)
(µg/mL) 0.001 ( 0.002 0.003 ( 0.002 0.002 ( 0.002 0.002 ( 0.001 (µg/g) 0.36 ( 0.70 0.67 ( 0.50 0.45 ( 0.32 0.31 ( 0.16 (µg/min) 0.004 ( 0.007 0.006 ( 0.005 0.004 ( 0.003 0.003 ( 0.002
(µg/mL) 0.043 ( 0.062 0.103 ( 0.126 0.053 ( 0.063 0.048 ( 0.037 (µg/g) 7.22 ( 8.64 20.73 ( 36.43 10.98 ( 18.60 9.09 ( 9.47 (µg/min) 0.066 ( 0.093 0.179 ( 0.265 0.103 ( 0.176 0.078 ( 0.088
(µg/mL) 0.003 ( 0.003 0.011 ( 0.006 0.005 ( 0.004 0.005 ( 0.002 (µg/g) 0.87 ( 0.78 2.24 ( 1.53 1.06 ( 0.81 0.97 ( 0.51 (µg/min) 0.007 ( 0.007 0.018 ( 0.011 0.008 ( 0.007 0.009 ( 0.005
(µg/mL) 0.003 ( 0.003 0.003 ( 0.001 (7) 0.002 ( 0.001 (7) 0.002 ( 0.001 (8) (µg/g) 0.60 ( 0.82 0.60 ( 0.44 (7) 0.45 ( 0.30 (7) 0.41 ( 0.31 (8) (µg/min) 0.004 ( 0.006 0.007 ( 0.006 (7) 0.004 ( 0.003 (7) 0.004 ( 0.003 (8)
a
The three background preexposure samples were averaged.
b
2,4-dimethoxyphenol acetosyringone (µg/mL) 0.028 ( 0.038 0.038 ( 0.057 0.028 ( 0.052 0.029 ( 0.039 (µg/g) 4.08 ( 4.19 6.65 ( 10.04 4.79 ( 8.35 4.31 ( 5.59 (µg/min) 0.025 ( 0.021 0.048 ( 0.049 0.030 ( 0.038 0.032 ( 0.025
(µg/mL) 0.012 ( 0.004 0.024 ( 0.015 0.014 ( 0.008 0.012 ( 0.007 (µg/g) 1.80 ( 0.82 3.39 ( 1.33 2.43 ( 0.95 1.96 ( 0.78 (µg/min) 0.015 ( 0.007 0.035 ( 0.020 0.020 ( 0.009 0.018 ( 0.009
The maximum or and averages were postexposure. c Mean ( SD (N); N was 9 if not given.
butylsyrigone
sinapylaldehyde
(µg/mL) 0.001 ( 0.001 (2) 0.0006 (1) 0.0002 (1) 0.0001 (1) (µg/g) 0.13 ( 0.02 (2) 0.379 (1) 0.133 (1) 0.069 (1) (µg/min) 0.0009 ( 0.0002 (2) 0.001 (1) 0.0004 (1) 0.0002 (1)
(µg/mL) 0.002 ( 0.002 0.002 ( 0.003 (8) 0.001 ( 0.001 (8) 0.002 ( 0.004 (µg/g) 0.31 ( 0.45 0.93 ( 1.94 (8) 0.42 ( 0.72 (8) 0.49 ( 0.70 (µg/min) 0.002 ( 0.002 0.004 ( 0.006 (8) 0.002 ( 0.003 (8) 0.003 ( 0.004
distance of the rightmost point in each plot was 3-5-fold higher than the next highest Cook’s distance. Elimination half-lives (see Table 10 in the Supporting Information) were generally shorter for syringols (3 h) compared to guaiacols (5 h). We confirmed that guaiacol was a urinary metabolite of guaifenesin (an expectorant in over-the-counter cold remedies). Consumption of smoke flavorings increased many of the methoxyphenols in urine (see Figure 4 in the Supporting Information). Guaiacol and eugenol were elevated after consumption of any of the flavored products. Other methoxyphenols were only elevated for certain foods, for example barbeque sauce only elevated guaiacols. FIGURE 2. The sum of urinary creatinine-adjusted concentrations for selected methoxyphenols. Propylguaiacol, syringol, methylsyringol, ethylsyringol, and propylsyringol were summed. Exposure started at 0 h. A void was not forced at the beginning of exposure, thus the point immediately before 0 h may be elevated by the exposure.
FIGURE 3. Regression of the sum of indicator urinary methoxyphenols (12-h average postexposure) against PM2.5 and levoglucosan on personal filters. metric for exposure. We chose the five methoxyphenols with the highest response (ratio of maximum response to average background) after exposure: propylguaiacol, syringol, methylsyringol, ethylsyringol, and propylsyringol. A plot (Figure 2) of this sum by subject shows that the maximum occurred within 12 h postexposure. The linear regression of the sum of urinary methoxyphenol concentrations (12-h average, creatinine adjusted) versus PM2.5 gave a R2 of 0.738 (p ) 0.003) with an intercept of 760 ( 210 µg/m3 (Figure 3). Regression against levoglucosan concentration on the personal filters gave a R2 of 0.775 (p ) 0.002) with an intercept of 56 ( 17 µg/m3. No point in these regressions gave standardized residuals greater than 1.4. If the rightmost point in the plot is deleted, then both regressions become nonsignificant (p ≈ 0.10). The Cook’s
Discussion In our original study, elimination rate and concentration adjusted for flow rate were considered the most sensitive metrics for detection of an exposure to particulate matter from woodsmoke (20). In our current study with a greater number of subjects and samples, creatinine-adjusted concentration and elimination rate appear equivalent in detecting the exposure event. Creatinine-adjusted concentration clearly has practical advantages of not needing to know the urine formation period. Because of the rapid clearance of the methoxyphenols and apparent differences between subjects in time of maximum elimination, timing of a spot sample collection could be problematic. We found that average concentrations of individual methoxyphenols were correlated with exposure levels. A 12-h average was found more responsive than a 24-h average most likely because of the rapid clearance (700 µg/m3) to woodsmoke.
Acknowledgments This work was funded by the National Institute of Occupational Safety and Health (#R03-OH007656) and the Northwest Center for Particulate Air Pollution and Health (U.S. EPA Grant #CR827355).
(20) (21)
Supporting Information Available Synthesis of additional chemicals, sampling details, Tables 1-11, and Figures 1-4. This material is available free of charge via the Internet at http://pubs.acs.org.
(22) (23)
Literature Cited (1) FAO Economics of wood energy. In State of the World’s Forests, 2005; ORGANIZATION, F. A. A., Ed.; United Nations Publications: New York, 2005. (2) UNDP Energy after Rio: Prospects and Challenges; United Nations Development Programme, United Nations Publications: New York, 1996. (3) Aiken, S. R. Runaway fires, smoke-haze pollution, and unnatural disasters in Indonesia. Geographical Rev. 2004, 94, 55-79. (4) Barber, C. V.; Schweithelm, J. Trial by Fire: Forest Fires and Foresty Policy in Indonesia’s Era of Crisis and Reform; World Resources Institute: 2000. (5) Sapkota, A.; Symons, J. M.; Kleissl, J.; Wang, L.; Parlange, M. B.; Ondov, J.; Breysse, P. N.; Diette, G. B.; Eggleston, P. A.; Buckley, T. J. Impact of the 2002 Canadian forest fires on particulate matter air quality in Baltimore city. Environ. Sci. Technol. 2005, 39, 24-32. (6) Kato, M.; Loomis, D.; Brooks, L. M.; Gattas, G. F.; Gomes, L.; Carvalho, A. B.; Rego, M. A.; DeMarini, D. M. Urinary biomarkers in charcoal workers exposed to wood smoke in Bahia State, Brazil. Cancer Epidemiol., Biomarkers Prev. 2004, 13, 10051012. (7) Reinhardt, T. E.; Hanneman, A.; Ottmar, R. Smoke Exposure in Prescribed Burns; USDA, Forest Service, Pacific Northwest Research Station: 1994. (8) Ezzati, M.; Kammen, D. M. The health impacts of exposure to indoor air pollution from solid fuels in developing countries: knowledge, gaps, and data needs. Environ. Health Perspect. 2002, 110, 1057-1068. (9) Boman, B. C.; Forsberg, A. B.; Ja¨rvholm, B. G. Adverse health effects from ambient air pollution in relation to residential wood combustion in modern society. Scand. J. Work Environ. Health 2003, 29, 251-260. (10) Larson, T. V.; Koenig, J. Q. Wood smoke: Emissions and noncancer respiratory effects. Annu. Rev. Public Health 1994, 15, 133-156. (11) Emmanuel, S. C. Impact to lung health of haze from forest fires: the Singapore experience. Respirology 2000, 5, 175-182. (12) Mott, J. A.; Mannino, D. M.; Alverson, C. J.; Kiyu, A.; Hashim, J.; Lee, T.; Falter, K.; Redd, S. C. Cardiorespiratory hospitalizations associated with smoke exposure during the 1997, Southeast Asian forest fires. Int. J. Hyg. Environ. Health 2005, 208, 75-85. (13) Sorensen, B.; Fuss, M.; Mulla, Z.; Bigler, W.; Wiersma, S.; Hopkins, R. Surveillance of morbidity during wildfires - Central Florida, 1998. MMWR Weekly 1999, 48, 78-79. (14) Freeman, D. J.; Cattell, F. C. R. Woodburning as a source of atmospheric polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 1990, 24, 1581-1585. (15) Simoneit, B. R. T.; Rogge, W. F.; Mazurek, M. A.; Standley, L. J.; Hildemann, L. M.; Cass, G. R. Lignin pyrolysis products, lignans, and resin acids as specific tracers of plant classes in emissions from biomass combustion. Environ. Sci. Technol. 1993, 27, 2533-2541. (16) Hawthorne, S. B.; Krieger, M. S.; Miller, D. J.; Mathiason, M. B. Collection and quantitation of methoxylated phenol tracers for atmospheric pollution from residential wood stoves. Environ. Sci. Technol. 1989, 23, 470-475. (17) Hawthorne, S. B.; Miller, D. J.; Barkley, R. M.; Krieger, M. S. Identification of methoxylated phenols as candidate tracers for atmospheric wood smoke pollution. Environ. Sci. Technol. 1988, 22, 1191-1196. (18) Sykes, R.; Fowler, M. G.; Pratt, K. C. A plant tissue orign for Ulminites A and B in Saskatchewan Lignites and Implications for R0. Energy Fuels 1994, 8, 1402-1416. (19) Bieniek, G. Simultaneous determination of 2-methoxyphenol, 2-methoxy-4-methylphenol, 2,6-dimethoxyphenol and 4′-hy-
(24)
(25) (26)
(27) (28)
(29) (30) (31) (32) (33) (34) (35) (36)
(37)
(38)
(39)
(40) (41)
droxy-3′-methoxyacetophenone in urine by capillary gas chromatography. J. Chromatogr., B: Biomed. Sci. Appl. 2003, 795, 389-394. Dills, R. L.; Zhu, X.; Kalman, D. A. Measurement of urinary methoxyphenols and their use for biological monitoring of wood smoke exposure. Environ. Res. 2001, 85, 145-158. Simpson, C. D.; Paulsen, M.; Dills, R. L.; Liu, L. J.; Kalman, D. A. Determination of methoxyphenols in ambient atmospheric particulate matter: tracers for wood combustion. Environ. Sci. Technol. 2005, 39, 631-637. Simpson, C. D.; Dills, R. L.; Katz, B. S.; Kalman, D. A. Determination of levoglucosan in atmospheric fine particulate matter. J. Air Waste Manage. Assoc. 2004, 54, 689-694. Sioutas, C.; Chang, M.; Kim, S.; Koutrakis, P.; Ferguson, S. T. Design and experimental characterization of a PM1 and a PM2.5 personal sampler. J. Aerosol Sci. 1999, 30, 693-707. USEPA Method TO-13: The Determination of Benzo(a)pyrene and Other Polynuclear Aromatic Hydrocarbons in Ambient Air using Gas Chromatogrphic and High Performance Liquid Chromatographic Analysis. In Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air (EPA/ 625/R-96/010b), 2nd ed.; Cincinnati, OH, 1999; pp TO13-11TO13-107. USEPA. Exposure Factors Handbook; EPA/600/8-89/043 ed.; NTIS: Washington, DC, 1989. Adams, T. B.; Cohen, S. M.; Doull, J.; Feron, V. J.; Goodman, J. I.; Marnett, L. J.; Munro, I. C.; Portoghese, P. S.; Smith, R. L.; Waddell, W. J.; Wagner, B. M. The FEMA GRAS assessment of cinnamyl derivatives used as flavor ingredients. Food Chem. Toxicol. 2004, 42, 157-185. Badger, D. A.; Smith, R. L.; Bao, J.; Kuester, R. K.; Sipes, I. G. Disposition and metabolism of isoeugenol in the male Fischer 344 rat. Food Chem. Toxicol. 2002, 40, 1757-1765. Smith, R. L.; Adams, T. B.; Doull, J.; Feron, V. J.; Goodman, J. I.; Marnett, L. J.; Portoghese, P. S.; Waddell, W. J.; Wagner, B. M.; Rogers, A. E.; Caldwell, J.; Sipes, I. G. Safety assessment of allylalkoxybenzene derivatives used as flavouring substances methyl eugenol and estragole. Food Chem. Toxicol. 2002, 40, 851-870. Fischer, I. U.; von Unruh, G. E.; Dengler, H. J. The metabolism of eugenol in man. Xenobiotica 1990, 20, 209-222. Ogata, N.; Matsushima, N.; Shibata, T. Pharmacokinetics of wood creosote: glucuronic acid and sulfate conjugation of phenolic compounds. Pharmacology 1995, 51, 195-204. Miller, J. J.; Powell, G. M.; Olavesen, A. H.; Curtis, C. G. Fate of 2,6-dimethoxyphenol-U-14C in the rat. Xenobiotica 1974, 4, 285-289. Roberts, M. S.; Magnusson, B. M.; Burczynski, F. J.; Weiss, M. Enterohepatic circulation: physiological, pharmacokinetic and clinical implications. Clin. Pharmacokinet. 2002, 41, 751-790. Bullingham, R. E.; Nicholls, A. J.; Kamm, B. R. Clinical pharmacokinetics of mycophenolate mofetil. Clin. Pharmacokinet. 1998, 34, 429-455. Phillips, D. H. Polycyclic aromatic hydrocarbons in the diet. Mutat. Res. 1999, 443, 139-147. Jongeneelen, F. J. Biological monitoring of environmental exposure to polycyclic aromatic hydrocarbons; 1-hydroxypyrene in urine of people. Toxicol. Lett. 1994, 72, 205-211. Arnarp, J.; Bielawski, J.; Dahlin, B. M.; Dahlman, O.; Enzell, C. R.; Pettersson, T. Tobacco smoke chemistry. 2. Alkyl and alkenyl substituted guaiacols found in cigaret smoke condensate. Acta Chem. Scand. 1989, 43, 44-50. Bieniek, G.; Kurkiewicz, S.; Wilczok, T.; Klimek, K.; Swiatkowska, L.; Lusiak, A. occupational exposure to aromatic hydrocarbons at a coke plant: Part II. Exposure assessment of volatile organic compounds. J. Occup. Health 2004, 46, 181-186. Bieniek, G.; Kurkiewicz, S.; Wilczok, T. Occupational exposure to aromatic hydrocarbons at a coke plant: Part I. Identification of hydrocarbons in air and their metabolites in urine by a gas chromatography-mass spectrometry method. J. Occup. Health 2004, 46, 175-180. Hatcher, P. G.; Lerch, H. E., III; Kotra, R. K.; Verheyen, T. V. Pyrolysis g.c.-m.s. of a series of degraded woods and coalified logs that increase in rank from peat to subbituminous coal. Fuel 1988, 67, 1069-1075. Maykut, N. N.; Lewtas, J.; Kim, E.; Larson, T. V. Source apportionment of PM2.5 at an urban IMPROVE site in Seattle, Washington. Environ. Sci. Technol. 2003, 37, 5135-5142. Larson, T.; Gould, T.; Simpson, C.; Liu, L. J.; Claiborn, C.; Lewtas, J. Source apportionment of indoor, outdoor, and personal PM2.5 in Seattle, Washington, using positive matrix factorization. J. Air Waste Manage. Assoc. 2004, 54, 1175-1187.
VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2169
(42) Sedivec, V.; Flek, J. Determination of toxic substances and their metabolites in biological fluids by gas chromatography. III. Occurrence and determination of guaiacol in urine. Pracovni Lekarstvi (AN 1970:496901) 1970, 22, 176-181. (43) Gibaldi, M.; Perrier, D. Pharmacokinetics, 2nd ed.; Marcel Dekker: New York, 1982. (44) Davidson, C. I.; Lin, S.-F.; Osborn, J. F.; Pandey, M. R.; Rasmussen, R. A.; Khalll, M. A. K. Indoor and outdoor air pollution in the Himalayas. Environ. Sci. Technol. 1986, 20, 561-567. (45) Robin, L. F.; Lees, P. S.; Winget, M.; Steinhoff, M.; Moulton, L. H.; Santosham, M.; Correa, A. Wood-burning stoves and lower respiratory illnesses in Navajo children. Pediatr. Infect. Dis. J. 1996, 15, 859-865. (46) Albalak, R.; Bruce, N.; McCracken, J. P.; Smith, K. R.; De Gallardo, T. Indoor respirable particulate matter concentrations from an open fire, improved cookstove, and LPG/open fire combination in a rural Guatemalan community. Environ. Sci. Technol. 2001, 35, 2650-2655.
2170
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
(47) Naeher, L. P.; Leaderer, B. P.; Smith, K. R. Particulate matter and carbon monoxide in highland Guatemala: indoor and outdoor levels from traditional and improved wood stoves and gas stoves. Indoor Air 2000, 10, 200-205. (48) Materna, B. L.; Jones, J. R.; Sutton, P. M.; Rothman, N.; Harrison, R. J. Occupational exposures in California wildland fire fighting. Am. Ind. Hyg. Assoc. J. 1992, 53, 69-76. (49) Bolstad-Johnson, D. M.; Burgess, J. L.; Crutchfield, C. D.; Storment, S.; Gerkin, R.; Wilson, J. R. Characterization of firefighter exposures during fire overhaul. Am. Ind. Hyg. Assoc. J. 2000, 61, 636-641.
Received for review September 23, 2005. Revised manuscript received January 17, 2006. Accepted February 14, 2006. ES051886F
