Special Issue of Aerosol Science and Technology for Air Pollution and Health: Bridging the Gap from Sources-to-Health Outcomes

THE CONFERENCE

The U.S. Environmental Protection Agency (EPA) has established National Ambient Air Quality Standards (NAAQS) for six principal air pollutants (“criteria” pollutants): carbon monoxide (CO), lead (Pb), nitrogen dioxide (NO₂), particulate matter (PM) in two size ranges (<2.5 μm [PM2.5] and <10 μm [PM10]), ozone (O₃), and sulfur dioxide (SO₂) (U.S. EPA 2010a). Although associations have been identified between these pollutants and adverse health effects, considerable uncertainty remains regarding (a) methods and approaches to understanding relationships between air pollution and health effects; (b) which components (gas and/or particles) and sources are most toxic; (c) the mechanisms of actions of the pollutants and causal relationships; (d) effect of confounding factors, and (e) which populations are susceptible (U.S. EPA 2006a [Pb], 2006b [O₃], 2008a [NOx Integrated Science Assessment [ISA]], 2008b [SOx ISA], 2009 [PM ISA], 2010b [CO]). This holds true especially for PM, because it is composed of many components with significant spatial and temporal variations (Solomon et al. 2008; U.S. EPA 2009). Air pollution and health research continues to reduce these uncertainties across the source-to-health effects paradigm as described by the National Research Council (NRC) Research Priorities for Airborne Particulate Matter, volumes I–IV, (NRC 1998, 1999, 2001, 2004) and the U.S. EPA (2006a, 2006b, 2008a, 2008b, 2009, 2010b).

Linking air pollution and adverse health effects is complicated and requires expertise across a range of scientific disciplines—from atmospheric to exposure to health sciences, as well as inclusion of air quality managers and policy makers who implement and develop policy to reduce risk from air pollution. Interaction among these groups at different points in time helps to identify gaps in knowledge and suggest future research directions. One such opportunity was the international specialty conference “Air Pollution and Health: Bridging the Gap from Sources to Health Outcomes,” sponsored by the American Association for Aerosol Research (AAAR 2010). The conference, chaired by myself and Maria Costantini (Health Effects Institute), was designed to help disseminate and integrate results from scientific studies that cut across the range of air pollution and health-related disciplines of the source-to-health effects continuum. Conference objectives are listed in Table 1. The conference addressed the science of air pollution and health within a multipollutant framework, focusing on five key science areas—sources, atmospheric sciences, exposure, dose, and health effects—as identified by the NRC (1998). Eight key policy-relevant science questions that integrated across various parts of these science areas formed the basis of the meeting, and a ninth question addressed the policy implications of the findings (Table 2).

TABLE 1 Conference objectives

1.	Bring together researchers across the source-to-health outcomes paradigm (NRC 1998) to engage in discussion and rigorous debate regarding the latest information on linking adverse health effects of air pollution to emissions sources and atmospheric pollutants.
2.	Communicate effectively the latest information to scientists, air-quality mangers in the public and private sectors, and policy-makers in support of developing efficient and cost-effective approaches to decreasing acute and chronic adverse health effects from air pollution on community and regional scales. The conference theme is multi-pollutant focusing on the sources, fate, and health effects of ambient air pollutants. The meeting is designed to appeal to a large interdisciplinary international audience.
3.	Highlight findings from major measurement, data analysis, and modeling studies and programs conducted since the last AAAR PM and health conference in 2003. Papers and presentations that show linkages from sources to air quality and/or air quality to health outcomes are strongly encouraged.
4.	Add significantly to the peer-reviewed literature by publishing research results presented at the conference in a variety of special issues of selected journals and/or synthesis papers.

TABLE 2 Science questions

1.	Pollutants and Sources Associated with Health Effects. How does our understanding of the health effects of air pollutants (singly or in mixtures) help identify pollutants that can be linked to sources the control of which would provide maximal health benefits? (overarching theme)
2.	Reliability of Methods, Models, and Approaches. How reliable are methods (measurements and models) and approaches (epidemiological and toxicological) for studying and quantifying the links between air pollutants (species and/or sources) and adverse health effects?
3.	Pollutant Characterization and Population Exposure. How do relevant pollutant properties vary in space and time from sources and in ambient air; what are the implications of these variations for population exposure?
4.	Relation between Exposure and Dose. What advances have been made in understanding the relationships between exposures, both spatially and temporally, and estimates of dose that tie to health outcomes?
5.	Mechanisms of Action and Biomarkers of Exposure and Effects. Are patterns emerging that relate component(s) of air pollution and/or source types to mechanisms? What is the status of identifying and measuring biomarkers of exposure and/or adverse health effects from air pollution?
6.	Susceptible Populations. Who are the susceptible populations; what drives different susceptibilities to the same or different air pollutants; and are there susceptibility traits associated with specific health outcomes that are common among the subpopulations?
7.	Confounding or Other Factors. What roles do confounding or other factors have in increasing, decreasing, or obscuring attribution of the true health effects from ambient air pollutants?
8.	Accountability. Do actions taken to improve air-quality result in reduced ambient concentrations of relevant pollutants, exposure, and health effects, and have we encountered unintended consequences?
9.	Regulatory and Policy Implications. What are the policy implications of our improved understanding of the source-to-health effects paradigm?

This was the AAAR's third international specialty conference and extended the findings presented at the AAAR's first specialty conference “Particulate Matter: Atmospheric Sciences, Exposure, and the Fourth Colloquium on PM and Human Health,” held in Pittsburgh, Pennsylvania, in 2003 (Davidson et al. 2005).

Results from the 2010 AAAR Air Pollution and Health conference are being published in Aerosol Science and Technology (AS&T); Environmental Health Perspectives (Solomon 2011 ); Air Quality, Atmosphere and Health; Atmospheric Environment; and Inhalation Toxicology (Solomon 2010).

THIS ISSUE

This special issue of Aerosol Science and Technology includes selected papers from the Air Pollution and Health Conference that align with the goals and objectives of Aerosol Science and Technology. Study objectives described in each paper align with the conference science questions as indicated in Table 3. Key findings also are presented in Table 3 with brief project descriptions given below.

The first three papers in this issue, characterize different emissions sources, one located outdoors and two located indoors. Cahill et al. (2011a) examined emissions from the Roseville Rail Yard located in Sacramento, CA. Roseville is the largest rail yard on the west coast of the United States and a previous study indicated an increase in cancer risk for the neighborhood downwind of the rail yard. Aerosol samples were collected in up to 9 size ranges for five weeks in the summer and fall of 2005 and analyzed for mass, optical absorption, and trace elements at a time resolution of 3 h. Organic species were obtained separately but only during the night. Data were collected simultaneously upwind and downwind of the rail yard and the difference was estimated as the emissions contribution of the rail yard to ambient downwind concentrations. Hourly measurements of NO, NO₂, and black carbon (BC) were obtained from a collaborating study. Diesel exhaust from the trains, indicated by the observed difference in chemistry from up- versus downwind measurements primarily impacted ultrafine (UF, <100 nm) and very fine (90–260 nm) particles. Heavy PAH species, at a higher proportion than observed from diesel truck emissions also were observed in the UF size range. Anthropogenic-derived metals and petroleum-derived n-alkanes showed enrichment in the coarse particles.

Indoor sources examined included cooking and emissions of nanoparticles from a laser printer. Wang et al. (2011) measured nanoparticle emissions directly from the page output area of a laser printer and characterized these emissions for volatile organic carbon (VOC), particle number concentration, size distribution, morphology, elements, and positive and negative ion concentrations, the latter determined with an ion density meter. Similar physical and chemical properties were determined for the toner material as well. Toner usage per page was based on iron content of the toner material compared to that determined on a given page. Maximum nanoparticle emissions were observed when the internal surfaces of the printer were cold at the beginning of printing and also depended on the number of printed pages and age of the generated aerosol. Strategies to reduce emissions were described. Buonanno et al. (2011) examined the volatile fraction of cooking-generated particles by measuring the size distribution and number concentration of the emitted particles as a function of thermal conditioning. Emissions testing including several fat rich foods (cheese, pork meat, bacon) and vegetables (eggplants, onions, chips) cooked by both grilling and frying. Both categories of food contained fractions of volatile material that was measured by a shift in the size distribution to a smaller size mode. Results indicated that particles emitted when cooking vegetables had a higher fraction of volatile material as compared to cooking fatty foods, the latter explained by the presence of a solid core likely due to the partial synthesis and degradation of fatty acids to aldehydes and ketones.

TABLE 3 Relationship of special issue paper objectives to conference science questions (SQ) as given in Table 2

Category / Authors	Study objective	Key finding	SQ
Emissions Characterization
Cahill et al. (2011a)	To characterize inorganic and organic emissions and develop an emissions profile from rail yard activities. Emissions sources included diesel engines associated with the trains, reentrained soil due to rail yard activities, and other smaller sources.	Rail yard emissions consisted primarily of very fine (90–260 nm) and ultrafine (<100 nm) particles associated with diesel exhaust, including mass, organic matter, transition metals, and S with downwind upwind ratios exceeding 2. Optical absorption and NO ratios were even higher. Downwind coarse soil particles had high concentrations of anthropogenic metals associated with diesel exhaust as well as petroleum-derived n-alkanes and a higher fraction of heavier PAH species including benzo[a]pyrene, compared to diesel exhaust from trucks.	SQ3
Wang et al. (2011)	To characterize materials used in the printing process and nanoparticle and VOC emissions to obtain a better understanding of potential generation mechanisms as a function of laser printer operations as well as suggest control strategies to reduce the potentially hazardous nanoparticle emissions.	The highest nanoparticle emissions were observed at the beginning of printing, when the printer components were cold. Peak nanoparticle concentrations ranged from 10^4 to 10^6 #/cc with a single mode peak between 20–150 nm, which varied depending on the number of pages printed. Results indicate that nanoparticle formation was due to ion-induced homogeneous nucleation and condensation of vapor-phase VOC evaporated from the copolymer toner particles. Strategies suggested to reduce nanoparticle emissions included allowing a longer time for the components to heat up to reduce ion concentrations, adding a filter to the cooling air outlet, and staying at least 1 m from the printer while in operation.	SQ3
Buonanno et al. (2011)	To evaluate the amount of volatile material associated with particles generated by different cooking activities (frying and grilling) and different kinds of food (fatty and vegetable).	A significant reduction in total number concentration was observed for vegetable foods compared to fatty foods, indicating a greater amount of volatile material associated with particles emitted when vegetables are cooked and a resulting solid core remaining for fatty foods by the processes studied. Both showed lower size modes as the aerosol conditioning temperature was increased.	SQ3
Aerosol Characterization
Wheeler et al. (2011)	To employ continuous monitoring methods to examine the relationships among indoor and outdoor, and personal exposure for a variety of pollutants (PM2.5, UF PM, BC) during two winter and two summer periods in 2005 and 2006 in Windsor, Ontario, Canada.	Outdoor concentrations of the three pollutants decreased at night and increased in the morning in response to rush hour traffic. UF particles also increased in the afternoon while BC decreased; PM remained fairly constant. Higher concentrations of UF particles and fine PM were highest indoors at dinner time, indicating the impact of household cooking on indoor concentrations. Highest indoor/outdoor ratios (0.7 on average) were observed for UF particles, whereas for fine PM and BC, the average ratio was 0.5. Results indicate the importance of indoor cooking as a source of UF particles.	SQ3
Kumar et al. (2011)	To demonstrate the use of fine (2 km) versus 5 km and 10 km resolution satellite-based aerosol optical depth (AOD) and to develop an empirical model using the fine resolution AOD data to estimate spatially and temporally resolved ambient PM concentrations.	Compared to in-situ (surface sunphotometer, AERONET) measurements, the fine resolution (2 km) satellite-derived AOD data provided better AOD location precision estimates than 5 km and 10 km AOD retrievals. On average, good agreement (slope = 1) was obtained in comparing surfaced-based PM mass data to 2 km AOD resolution, although a range of slopes was observed for specific site comparisons. Error estimates (RMSE) for PM10 were smaller than for PM2.5.	SQ2
Health Effects
Faiola et al. (2011)	To investigate, via an in-vitro study, the effect of UF particles and its chemical components on the electron transport chain (ETC) in isolated mitochondria to better understand mechanisms of action of UF particles on human health and to evaluate this acellular measure of particle toxicity relative to a commonly used in-vitro approach using chemical reducing agents, such as the dithiothreitol (DTT) method.	Specific chemical components associated with UF particles inhibit ETC functionality resulting in the generation of reactive oxygen species that have been linked to inflammation and subsequently to adverse health effects. Initially Fe(II) and anthracene, which correlated with oxidized surface carbon (C‒O bonds), had the most significant correlations with ETC inhibition. After 20 min, of exposure, other PAHs became more important, but Fe(II) still contributed significantly to ETC inhibition.	SQ5
Cahill et al. (2011b)	To examine the relationship between very fine (90–260 nm) and ultrafine (<90 nm) metals in ambient particles and ischemic heart disease (IHD) across a north-south transect of California's San Joaquin Valley (2003–2007).	A high correlation was found between motor vehicular associated metals in particles less than 260 nm and IHD, in particular, non-soil Fe, Ni, Cu, and Zn as well as P. Fe, Ni, Zn, and Cu were associated with break lining, where as Zn and P were associated with zinc from Zn thiophosphate found in diesel lubricating oil. Associations of metals with break lining and diesel lubricating oil were reaffirmed by supplemental studies, such as a near roadway study in Sacramento, CA using the same measurement methods.	SQ5
Cahill et al. (2011c)	To examine the relationship between Ni and V in ambient particles and observed reductions in IHD in the southern portion of California's San Joaquin Valley as the result of switching from burning crude oil to natural gas (∼1990) used to generate steam to enhance heavy petroleum recovery.	Significant (∼30%) reductions in IHD from before approximately 1990 compared to the period 2003–2007 were associated with an order of magnitude reduction in Ni and V. A 1.8% mortality decrease per μg/m^3 of PM2.5 was observed, similar to the 6-Cities study, even though there were major differences in meteorology and pollutants. No associations were observed for CO and only a weak association was observed for IHD and ozone.	SQ5
Bioaerosols
Xu and Yao (2011)	To evaluate the sampling effects of six widely used biosamplers for biological collection efficiency and on cultural bacterial and fungal diversity. Sampler types included filtration, impaction, liquid impingement, and electrostatic and were evaluated in- and outdoors.	The different samplers resulted in a range of cultural bacterial and fungal aerosol diversity that also varied by environment tested. Various culturing methods also resulted in different results. Sampler related factors impacting culturable bioaerosols diversity included sampling mechanism, sampling time, desiccation, impaction stress, degree of particle embedding, and particle bounce as well as the interplay among these factors. Differences also were observed indoors vs. outdoors, thus, the environment (e.g., temperature, relative humidity) impacts the number of culturable bacteria and fungi counted.	SQ2
Tan et al. (2011)	To develop, optimize, and evaluate the physical and biological collection efficiencies of an automated electrostatic sampler (AES) for bioaerosol that will allow for the collection and transport of the biologically active aerosol to a variety of biosensors for real-time data, while maintaining the culturable bioaerosol activity.	The optimized sampler obtained physical collection efficiencies in the range of 70–90% at the lower flow rate and the higher voltage of the AES. A larger plate electrode also provided higher collection efficiencies. The biological collection efficiency was compared to the direct collection on MCE filters in the reference sampler with subsequent culture growth under the same conditions. The optimized AES approach consistently measured a lower biological collection efficiency, by about a factor of 3, as compared to the reference sampler for both indoor and outdoor aerosols. In this study, independent of the method, outdoor bacterial aerosol concentrations were higher than indoor.	SQ2

The next two papers in this issue characterize outdoor and/or indoor aerosols. Wheeler et al. (2011) obtained continuous measurements of fine particle (<2.5 μm) mass, BC, and UF particles over 5 days in the winter and in summer and at both indoor and outdoor residences of 48 adults in 2005 and 47 asthmatic children in 2006. Air exchange rates also were measured. Diurnal and seasonal variations were examined along with indoor/outdoor ratios. Outdoor concentrations of the pollutants measured were higher than indoor with nighttime minimums and peaks around morning rush hour, indicating motor vehicles as an important morning source of pollution. Indoor/outdoor ratios during dinner time for UF particles were higher than for BC and fine mass, indicating the importance of UF particle indoor sources, likely cooking, and the low indoor-outdoor exchange rates in this location. Kumar et al. (2011) examined improvements in spatial and temporal estimates of ambient PM2.5 and PM10 based on fine-resolution satellite-based aerosol optical depth (AOD). They computed AOD estimates at 2, 5, and 10 km using data from the MODerate Resolution Imaging Spectroradiometer (MODIS) satellite and compared the results to in-situ measurements obtained from NASA's network of sunphotometers (Aerosol RObotic NETwork [AERONET]). MODIS AOD estimates were corrected for meteorological conditions and an empirical regression model was developed to estimate ambient surface PM from the AOD data. For comparison, ambient PM2.5 and PM10 data were obtained from EPA's national Federal Reference Method network. Satellite AOD at 2 km resolution provided better spatial and temporal estimates of PM than the coarse resolution MODIS data. AOD-PM site specific regression slopes varied from 0.52 to 1.72, although on average the slope was close to 1.

Three papers addressed air pollutants and health effects. Faiola et al. (2011) describe an in vitro toxicological study designed to better understand the biological mechanisms of UF particles on adverse health effects. UF particles were collected over two week periods using a high volume size fractionating sample (ChemVol). A dense polypropylene filter, last in the series was used to collect the UF particles. Samples were collected at an urban and rural site in and near Seattle, WA. The authors examined the effect of UF bulk mass, trace metals, labile Fe(II) and easily reducible Fe(III), polycyclic aromatic hydrocarbons (PAH) particles, and surface (1–10 nm) chemistry, the latter specifically C, N, Si, O, and S on the electron transport chain (ETC) in mitochondria isolated from bovine heart tissue. A succinate oxidase activity assay was used to assess ETC function in mitochondria. Strongest correlations between ETC inhibition and measured components were obtained at 5 min with Fe(II)5 min (an aliquot taken after 5 min of reaction time) and anthracene. After 20 min the strongest correlations were observed with sum of PAH species and Fe(II)5 min. ETC inhibition did not correlate with other trace metals or with pooled redox active metals or catalytic metals. These results suggest a mechanism that allows reactive oxygen species to build up, since the O₂ normally reduced to H₂O in fully functioning mitochondria converts to a superoxide radial (O₂ ^•−), from which other ROS species are formed, such as, OH^• via the Fenton reaction.

Cahill et al. (2011b; 2011c) investigate the correlation of trace metals in PM and ischemic heart disease (IHD) in the California's Central Valley. In the first of these two papers, very fine and UF aerosol samples were collected along an N-S transect within the Central Valley from Redding to Bakersfield during 17 consecutive days in January 2009. A typical extended stagnation period occurred during this time period in the valley. Sites were located at existing California Air Resources Board (ARB) and district monitoring sites where supplemental data were collected. This study employed an 8-stage drum impactor covering the particle size range from 10 μm to 90 nm and an additional backup filter collecting particles less than 90 nm. Samples were collected every three hours and analyzed for mass and elements from Al to Mo plus Pb. Mass by the drum sampler was compared to the FRM with agreement to better than few percent during the winter, 10% on average over an independent year long evaluation. A second drum impactor was used to collect samples for organic species analysis, although the analysis was averaged over the full 17 day period. Organic species included PAHs, levoglucosan, fatty acids, and n-alkanes. IHD data were obtained from the California Department of Health Services (CA DHS). The strongest correlations of pollutants with IHD were observed between very fine and UF metals, most tied to motor vehicular sources (diesel and gasoline). Results were supported by an independent near roadway study in Sacramento, CA indicating the presence of the same motor vehicular components (non-soil Fe, Ni, Cu, Zn that are associated with brake wear and Zn thiophosphate an additive in lubricating oil) as correlated with IHD in the transect study. Removal of Zn thiophosphate from lubricating oil in motor vehicles and changes in brake drums and pads could reduce these components, and thus, potentially reduce the occurrence of IHD in the valley. Cahill et al. (2011c) compare historical Ni and V levels observed in the California Central Valley and the relationship of lower concentrations of these components in the 2003–2007 time frame versus prior to 1990 when a switch was made from burning crude oil to natural gas to generate steam to enhance petroleum recovery. In 1989–1991, IHD mortality rates were 60% higher in the southern San Joaquin Valley, where most oil production occurred, relative to the rest of the Central Valley. In 2003–2007, a 30% reduction in IHD was observed in the southern part of the valley as compared to those observed in the Sacramento valley, suggesting that the reduction in Ni and V due to the switch from oil to natural gas may have resulted in the reduction of IHD in the southern valley.

The final two papers in this special issue describe the development and/or evaluation of bioaerosol samplers (Xu and Yao 2011; Tan et al. 2011, respectively).

In the first paper, six bioaerosol samplers are compared to evaluate their collection efficiency and culturable biological activity in both indoor and outdoor environments. Sampling collection mechanisms included filtration, impaction, liquid impingement, and electrostatic. Factors impacting the biological collection efficiency included the sampling mechanism, impaction velocity, relevant cutoff size of the sampler, and degree of particle embedding in the filter or the collection medium. Collected samples were either directly cultured, or washed or extracted from the collection media. All culture samples were incubated for 2 days for bacteria and 3 days for fungi. Biological collection efficiency for bacteria was determined based on the number of colony forming units observed through manual counting. Bioaerosol diversity for bacteria was obtained by denaturing gradient gel electrophoresis and examining the observed number of bands. Identification of fungal colonies was obtained by washing the agar plates, applying a portion to a microscope slide, and manually identifying and counting the fungi. Results indicated that bioaerosol concentrations and diversity varied among the samplers with differences due to different stresses impacting the bioaerosol activity as a result of the different sampler designs and sampling environments. Results also differed among samplers in terms of which reported the highest bioaerosol concentration and which had the greatest diversity. Culturing method also impacted activity and diversity. Tan et al. (2011) describe the development and evaluation of an automated electrostatic sampler (AES) for the collection and transport of biologically active aerosol to a variety of biosensors and the reporting of results in real time. The physical collection efficiency of the sampler was obtained by sampling outdoor and indoor particles of different sizes and monitoring particle counts with an optical particle counter. Different flow rates and voltages applied to the AES were tested to optimize the physical collection efficiency. Indoor and outdoor air also was used to obtain biological collection efficiency as compared to a reference sampler. The biological collection efficiency was evaluated at two flow rates. These tests were conducted using the larger central electrode and the voltage set to 20 kv as optimized in the physical collection efficiency study. Biological collection efficiencies within a factor of 3 and closest to the reference were obtained at the lower flow rate tested. Particle charging at the inlet to the AES, also increased collection efficiencies.

ACKNOWLEDGEMENTS

The 2010 AAAR Air Pollution and Health Conference would not have been possible if not for the generous support by a number of sponsors, including the U.S. EPA, Health Effects Institute, American Chemistry Council, American Petroleum Institute, California Air Resources Board–Research Division, Electric Power Research Institute, NARSTO, National Aeronautics and Space Administration, National Institute for Public Health and the Environment (RIVM), National Oceanic and Atmospheric Administration, South Coast Air Quality Management District, Southern Company, Air & Waste Management Association, International Society of Exposure Sciences, and Springer. The AAAR (the professional society sponsoring the meeting), along with Association Headquarters, AAAR's management company, also were pivotal in making the conference a huge success. I especially thank the conference co-chair, Maria Costantini, conference committee members, and those who attended and participated in the meeting.

Preparation of this preface was supported by the U.S. EPA through its Office of Research and Development. The preface has been subjected to the agency's administrative review and approved for publication.

Paul A. Solomon

U.S. Environmental Protection Agency

Office of Research and Development

Las Vegas, Nevada

E-mail: [email protected]

Paul A. Solomon, a senior environmental scientist with the National Exposure Research Laboratory, Office of Research and Development, at the U.S. EPA in Las Vegas, Nevada, organized and chaired the AAAR Air Pollution and Health International Specialty Conference along with Dr. Maria Costantini of HEI. His research interests include developing and evaluating air quality measurement methods and conducting measurement projects in support of regulatory and research programs in the atmospheric sciences, and in air pollution exposure and health effects programs with an emphasis on particulate matter.

This article not subject to United States Copyright Law.