Table 1 Human panel studies with good subject exposure methodologies associating measures of oxidative stress with different pm 2.5 species or specific sources
Study | Oxidative stress measures | Subjects | PM 2.5 species, other measures of exposure | Associations found | Accuracy of subject exposure to BC/EC 1 |
|---|---|---|---|---|---|
Delfino et al., 2008 [74] | Circulating GPx-1, Cu, Zn-SOD | 29 non-smoking elderly w/CAD in Los Angeles area, many on statins or anti-hypertension meds; 12 weekly blood draws | Ambient PM0.25, PM2.5, PN, EC, OC, BC, OCpri, SOA | For Cu, Zn-SOD: Associations found with PM0.25, PM2.5, EC, BC, OCpri, but not with OC, PN, SOA | Very good: monitors inside and outside retirement communities |
Delfino et al., 2009 [75] | Circulating GPx-1, Cu, Zn-SOD | 60 non-smoking elderly w/CAD in Los Angeles area, many on statins or anti-hypertension meds; 5-12 weekly blood draws | Ambient PM0.25, PM2.5, PN, EC, OC, OCpri, SOA | For Cu, Zn-SOD: Associations (stronger in cool season) with PM0.25, PM2.5, EC, PN, OCpri, but not with OC, SOA | Very good: monitors inside and outside retirement communities |
Kim et al., 2004 [76] | 8-OHdG | 20 boilermakers, average 45.5 year of age, repairing oil-fired boilers in Boston area | Industrial levels of PM2.5, PM2.5 metals V, Cr, Mn, Ni, Cu, Pb (PM2.5 = 440 μg/m3, V = 1.23 μg/m3, other metals < 1.0 μg/m3) | Post work shift 8-OHdG levels significantly higher than pre shift; increased PM2.5, V, Mn, Ni, Pb concentrations significantly associated with increased 8-OHdG | Excellent: personal monitors used |
Loft et al., 1999 [77] | 8-oxodG | 57 non-smoking diesel bus drivers in greater Copenhagen | Ambient air in urban (30 drivers) vs. rural/suburban (27 drivers) areas | Significantly higher 8-OHdG excretion in urban drivers vs. rural/suburban drivers | Very good: exposure is based upon comparison of urban (heavily trafficked) vs. rural/suburban areas |
Sauvain et al., 2011 [78] | 8-OHdG | 32 Swiss bus maintenance workers | PM4, OC, EC, PM metal (Fe, Mn, Cu) content, PAHs (PM4 between 25 and 71 μg/m3) | 8-OHdG excretions significantly increased within each shift and between two consecutive work days; increases in non-smokers associated with increases in OC and particulate Cu | Very good: associations based upon exposure measured indoors, as nearby as possible to work stations |
Wei et al., 2009 [79] | 8-OHdG | 2 non-smoking young security guards working by a major road in Beijing, noon to 8 PM; worksite ambient PM2.5 = 243 μg/m3, background PM2.5 = 104 μg/m3 | Pre and post-shift 8-OHdG samples collected for 29 days; associations with four “clusters” of PM2.5species (PM2.5 mass, PAHs, metals, polar organic species) | Post work shift increases in 8-OHdG significantly associated with PM2.5 mass, PAHs, and metals, but not with polar organic species, as measured at worksite | Very good exposure : taken at worksite |
Lee et al., 2011 [80] | 8-OHdG | 28 diesel exhaust inspectors in Taiwan, 38 age and gender matched controls office workers, monitored during 3 consecutive day work periods | Diesel PM2.5 emissions (DEP2.5), PAH content of DEP2.5 (personal daily PM2.5 = 86 to 94 μg/m3, PAHs = 3.04 to 4.11 ng/m3) | 8-OHdG levels significantly higher for inspectors vs. controls on days 2 and 3; increased PAH concentrations in DEP2.5 significantly associated with increased 8-OHdG in exposed group, after adjusting for smoking status and BMI | Very good: personal monitors and ambient samplers installed within meters of work sites |
Lai et al., 2004 [81] | 8-OHdG | 47 female highway toll workers in Taiwan, 27 female office workers as controls | Exposed (toll workers, 8 hour shifts) vs. office workers | Toll workers had significantly higher 8-OHdG levels than controls (86% higher 8-OHdG comparing non-smokers to non-smokers); increases in 8-OHdG almost 5 times higher per 1000 trucks/buses than per 1,000 cars, result not significant (separate lanes for different vehicle types) | Very good: exposure defined by whether working in traffic or not; for those working in traffic, further defined by car lane or truck/bus lane workers and by traffic density |
Sorensen et al., 2005 [82] | 8-oxodG | 49 non-smoking students in Copenhagen, median age 24, in each of two seasons (summer, autumn) | PM2.5, six metals in PM2.5 fraction (V, Cr, Pt, Ni, Cu, Fe); PM2.5 = 20.7 μg/m3 (fall), 12.6 μg/m3 (summer) | No associations found with urinary 8-oxodG; significant positive associations found for V, Cr with 8-oxodG in lymphocytes. 98 samples available for urinary 8-oxodG, 52 for lymphocytes | Excellent; weekly personal monitors concentrations for each season |
Allen et al., 2009 [83] | 8-OHdG, F2-isoprostanes | 10 adults, ages 18-49, with metabolic syndrome but no history of ongoing medical care for heart disease, hypertension, asthma, diabetes, or other chronic condition | 2 hour exposure to either diesel exhaust (200 PM2.5) from engine operating at 75% of rated output, or filtered air | No significant increase in 8-OHdG or F2-isoprostanes levels after exposure, in double blind crossover experiment | Excellent: exposure to diesel emissions or filtered air occurred in chamber designed for purpose |
Jacobs et al., 2011 [84] | Oxidized LDL | 79 non-smoking diabetics, average age 56.5 years, in urban Belgium | Carbon area in airway macrophages, distance of residences from major roads | Increase in IQR carbon loading, decrease in distance from major roads both associated with increase in oxidized LDL | Excellent: carbon loading in personal macrophages is a precise measure of particulate carbon, which in urban area is from traffic, mainly diesels |
Kipen et al., 2011 [85] | White blood cell (WBC), red blood cell (RBC) proteasome activity, | 38 healthy young subjects in New Jersey | Diesel exhaust (DE), laboratory-generated secondary organic aerosol (SOA, based upon gas phase reactions of ozone and d-limonene), both ~ 190 μg/m3 | Compared to filtered air exposure, significant decreases of WBC proteasome activity after exposure to either DE or SOA, decrease of RBC proteasome activity after exposure to DE, “presumably via induction of oxidative stress” | Excellent: exposure chamber used |
Adar et al., 2007 [86] | HRV measures: SDNN, r-MSSD, PNN50 + 1, LF, HF, LF/HF, HR | 44 non-smoking subjects over age 60 living in 4 seniors’ residences in St. Louis | BC, PM2.5 levels as recorded by mobile monitor which followed subjects, including in traffic on bus | For 24 average BC concentrations, for 5 minute concentrations on diesel bus or not on bus, and for 5 and 24 hour means, the great majority of many BC associations are significant; PM2.5 highly correlated with BC and associations similar to those for BC | Very good (monitor follows subjects during day, stays in residence of subjects at night) |
Schwartz et al., 2005 [87] | 4 measures of HRV | 27 subjects living adjacent to major urban road in Boston, age 61-89 | BC, PM2.5, “secondary PM” derived by subtracting BC mass from PM2.5 mass | BC: 7 of 8 associations significant | Very good (monitor adjacent to same major urban road to which subjects’ residences also adjacent) |
PM2.5: 3 of 8 associations significant | |||||
“Secondary PM”: no significant associations of 8 | |||||
Ebelt et al., 2005 [88] | SDNN, R-MSSD measures of HRV | 16 non-smoking patients with COPD in Vancouver, average age = 74 | PM2.5, sulfate, estimated non-sulfate PM2.5 | Significant associations only with non-sulfate PM2.5 | Very good (personal monitoring information combined with ambient monitoring information, to separate effects) |
For SDNN, 2 of 2 tests | |||||
For R-MSSD, 1 of 2 tests | |||||
Creason et al., 2001 [89] | HF, LF measures of HRV | 56 elderly, non-smoking men, residents of a retirement center near Baltimore, near commuter roads | PM2.5 (sulfate monitored but not used in models; expressed as percentage of PM2.5 in discussion) | For complete dataset of 24 days, insignificant associations with either HF or LF, with a “U” shaped function where HRV measures are at normal levels only at lowest and highest PM2.5 concentrations; when a two day period with no effects on HRV is removed, remaining 22 days have a significant, linear reduction in both HF and LF, for both indoor and outdoor PM2.5; two days removed had highest and 3rd highest PM2.5 levels, were high in sulfate but came from rural areas with no apparent urban or industrial source of harmful PM2.5 | Reasonably good in that study was able to determine by wind back-trajectory analysis that for 22 days where PM2.5 reduced HRV measures, air parcels passed over either urban or industrial areas, but that for the 2 days without effect on HRV, parcels passed over more rural areas |
Suh and Zanobetti, 2010 [72] | 4 measures of HRV: SDNN, RMSSD, PNN50, HF, LF/HF | Same as in Wheeler et al. (2006) | EC, sulfate, PM2.5 | IQR increase in personal monitored EC significantly associated with decreases in SDNN, RMSSD, PNN50, and HF, and with increase in LF/HF; IQR increase in ambient monitored EC not associated with changes in any HRV measures; Neither sulfate nor PM2.5 (only personal monitored available for sulfate) significantly associated with any HRV measures | Excellent for personal monitored EC, PM2.5, and sulfate; Poor (horizontal exposure misclassification; central monitor reading for people across metro Atlanta area) for ambient EC |
Park et al. (2007) [90] | SDNN, HF, LF, LF/HF | 497 subjects of Normative Aging Study living across greater Boston area | BC, sulfate, PM2.5, combined with wind trajectory information showing source directions | Four of six trajectories had high and nearly equal concentrations of BC, sulfate, and PM2.5 - associations are for these four trajectories. | Poor for BC (horizontal exposure misclassification; central monitor reading for people across metro Boston area); however, use of wind trajectory analysis mitigates poor exposure by allowing interpretation of rural vs. urban sources |
For local (stagnant) wind trajectory: BC (2 significant tests of 4); Sulfate (1 of 4); PM2.5 (1 of 4) | |||||
For long distance urban wind trajectory from southwest (over major urban areas): BC (3 of 4 tests significant); | |||||
For two non-urban trajectories: BC, sulfate, PM2.5 (none significant) | |||||
Langrish et al., 2009 [91] | SDNN, LF | 15 healthy, non-smoking young volunteers in Beijing | Personal monitor PM2.5, randomized cross over study design: volunteers walked a predetermined city center route with or without a face mask | 24 hour SDNN significantly lower when face mask not used; LF significantly lower (one test) without mask, but interpretation not straightforward | Excellent; most important finding is corroboration that urban particulate matter, not gases, drives reduction in SDNN, suggesting associations in other studies with BC/EC indicate direct PM effects, BC/EC not a proxy for vehicular gaseous emissions |
Langrish et al., 2012 [92] | HF, RMSSD | 98 patients with coronary artery disease, average age 62 | Personal monitor PM2.5, randomized cross over study design: volunteers walked a predetermined city center route with or without a face mask | HF, RMSSD significantly lower when face mask not used | Excellent; most important finding is corroboration that urban particulate matter, not gases, drives reduction in RMSSD, HF, suggesting associations in other studies with BC/EC indicate direct PM effects, BC/EC not a proxy for vehicular gaseous emissions |