There are many physical and chemical processes that affect the accumulation of outdoor pollutants. In recent years some of the information and concepts previously ascribed to outdoor pollution has been found to be useful in examining indoor dynamic and chemical processes. Further, becau se of the confining nature of the indoor environment, processes such as the “grasshopper effect” can lead to sustained higher levels of semivolatile chemicals indoors and affect multiroute (inhalation, dermal, incidental dietary, and nondietary ingestion) exposures. Such processes can also lead to a complex mixture of both semivolatile and volatile compounds in indoor air and on surfaces or within objects. This article specifically examines the above in combination with another indoor issue, indoor chemistry, and places the results into a context that can be used to evaluate (1) multipollutant cumulative or aggregate exposures and risks indoors, (2) exposure reduction strategies that can create healthy indoor environments. It is not a review of the entire field of the indoor environment or indoor air or the indoor environment, which has been covered in numerous volumes and reports. The complexities of the scientific issues are discussed by also placing them into our traditional approaches outdoor and indoor to pollution management, to indicate the difficulty in establishing the exposures that require mitigation or prevention. Further, some emerging issues are discussed as well as how to specifically address long-term single or multiroute exposures to semivolatile compounds within the “Total Indoor Environment.”
Indoor locations can be viewed as dynamic environments where even chemical reactions can occur and lead to build-up of secondary pollutants (Fan et al., 2003; Nazaroff and Cass, 1986; Weschler, 1999). The multimedia sources of concern for individual compounds are not just points of direct release but also include the retention or re-emission of semivolatile and volatile compounds that are present within and on the surfaces of objects (Tucker, 2001; Lioy et al., 2002; Nazaroff and Weschler, 2004). If viewed in the context of ambient environmental pollution, the physical dynamics and chemistry that can occur indoors are similar to outdoor processes, although characteristically different in terms of photochemistry and the boundary conditions that affect accumulation (the build-up of materials in air or in/on surfaces) and deposition. As a result, the indoor environment contains a very complex mixture of materials that can affect multiple routes of human exposure and human health. Thus, there is a need to consider total exposure indoors (Lioy, 1990b, 1999; National Research Council, 1991; Duan and Mage, 1997; Seifert et al., 2000; OECD, 2003). For this reason, the indoor environment will require strategies for reduction and control based upon ideas currently being evaluated or re-evaluated for ambient pollution, as well as those specific to the structures that define “indoors.”
The following attempts to provide a context for understanding the evolving knowledge about the indoor environment using traditional and current information on the nature of pollution dynamics and chemistry, and the methods that are employed to control such pollution. It also discusses some of the challenges that must be met to ensure that we can reduce or prevent exposure that may have adverse effects on human health.
Background: protection of the ambient environment
Traditionally, the approaches and tools used to address and mitigate environmental health problems (e. g., cancer, developmental and/or neurological effects) are employed after the identification of individual pollutants of concern. Initially, efforts are made to more fully characterize the toxicity and measure environmental quality. More recently, approaches have been employed to characterize exposure (Foley et al., 2003; USEPA, 2000a, 2001, 2003). Evaluation of the results obtained from many types of studies lead, when necessary, to the implementation of mitigation strategies to decrease risk, including those required to control or eliminates sources. The selected strategies could affect sources located in a single community or a much larger area. After implementation, systems are needed to track changes in pollutant levels in the ambient environment. The results will also yield the information required to set, and evaluate compliance to national standards associated with individual contaminants found in the air, soil, water or food. A recent summarization and evaluation on each of the above is provided within Management of Air Pollution (National Research Council, 2004b).
The systematic evaluation of individual contaminants occurs in spite of the fact that most environments and media usually contain a complex mixture of many different types of pollutants. The occurrence of pollutant mixtures in the environment (at times in the 100s to 1000s of individual compounds), and their collective potential to yield health impacts, are not new to the environmental health sciences (Pellizzari et al., 1995; Suk et al., 2002). This point has also led to the consideration of aggregate exposure (same pollutant from multiple routes of contact) and cumulative exposure (all pollutants and all routes) with similar mechanisms of action (USEPA, 2001, 2003). For example, the California Air Resources Board now has regulations to control cumulative exposure by reducing atmospheric exposures from a large number of consumer products (CARB, 2005). This, in effect, will change local outdoor/indoor air issues in California.
As a result of the above, it is important to focus on improving our understanding of additive synergistic or antagonistic effects on health (USEPA, 2000b). This is especially true today since the types of pollutants encountered during our daily lives continue to change and possibly increase, and can enter the body by multiple routes. In addition, more information is needed on how such mixtures affect susceptible individuals and subgroups of the general population that may or may not have susceptibilities enhanced by one or more response modifiers. The fetus, neonate and developing child are members of subgroups that are susceptible to environmental agents that can be associated with developmental health effects or childhood disease. Some diseases of concern for current and future analysis include leukemia, asthma, Parkinson's disease, autism and developmental defects.
Historically, that is, from before 1950 to ∼1980, the emissions of individual pollutants (e.g., sulfur and lead in air) in the United States and many developed countries were much higher than the emissions of the same pollutants released today from similar sources. Major progress has been made to reduce the levels of many contaminants released by sources to the air, water, etc. (e.g., CO and SO2 into air, organics to the water). Recently, new approaches for identifying and addressing complex mixture problems, such as those associated with air pollution, are being considered under one or more umbrella frameworks. For air pollution, one proposed conceptual framework is called “One Atmosphere” (IJC, 1998). It encourages the simultaneous consideration of all pollutants present in a local or regional atmosphere as part of the etiology of detected or suspected ecological or health outcomes. The “one atmosphere” framework provides an opportunity to conduct better environmental science, and environmental health studies and analyses. It can also encourage development of new approaches to obtain the types of information required for prioritization and then effective implementation of strategies to mitigate multipollutant problems. The NRC review of air pollution management also discussed the need to expand the approaches for air quality management to encompass mixtures of ambient contaminants (National Research Council, 2004b). Similarly, such approaches should be considered as ways to examine and mitigate problems in what could be described as the “Total Indoor Environment.”
The “One Atmosphere Framework” has also led to new ideas for augmenting current research on outdoor air pollution to reduce environmental health problems within the general population. For example, results from recent air pollution research on fine particulate matter (PM2.5) has begun to direct attention toward understanding relationships between excess cardiac and pulmonary morbidity or mortality, and the ambient air levels or exposures to multiple pollutants that occur in many cities and other locations around the world. These include the chemical components of PM2.5 and co-occurring gaseous species. Recommendations made in the fourth report of the NRC committee on Particulate Matter Research, completed in 2004, strongly supported the need for research to seriously address the issue of particulate matter health effects in context of the levels of other pollutants that can co-occur in the atmosphere (National Research Council, 2004b; Wu, 2004).
In the future, approaches that improve the characterization of exposure by using “one atmosphere” for single or multiroute exposures (such as aggregate or cumulative exposures) can also be used to develop new exposure scenarios for use “in vitro” or “in vivo” toxicological methods. These are necessary studies for evaluating additive synergistic and antagonistic effects caused by mixtures in different types of animal models. They could augment the toxicological studies normally completed to identify the effects caused by individual pollutants (Suk et al., 2002). However, as noted by the National Occupational Research Agenda (NORA), mixed exposure time of the study mixtures is not simple, and a good reference is NIOSH (2004).
Background: exposure reduction in the indoor environment
Indoors, there have been and continue to be major exposure issues in the United States, Europe and other countries that are readily identified with a single pollutant, for example, radon or asbestos, or with a single pollutant source, for example, cigarettes. (Spengler et al., 2001; Joint Research Commission, February, 2005). Further, one must consider viable and nonviable biological materials, such as bacteria and mold, as continuing indoor environmental problems (Spengler et al., 2001; Wu, 2004). An excellent summary of the exposure and toxicological information available for many indoor air pollutants is found in the Indoor Air Quality Handbook (Spengler et al., 2001).
Similar to the way in which we have considered and addressed outdoor environmental issues, the resolution of indoor environmental problems has been pollutant by pollutant or by pollutant class (e.g., leads, pesticides), and reduction of exposure. Included are product replacement or elimination, mitigation and general or case-specific ventilation practices. In some cases, the issue of indoor air quality has been considered a multipollutant problem, and some of the current ideas are summarized in recent work by Weschler (2003). Over the past 20 years, a major feature of indoor mixtures and the health effects evaluations has been associated with the “sick building syndrome” (Cone and Hodgson, 1989; Spengler et al., 2001).
The prevalence of both single and multiple pollutants indoors has increased because of changes in lifestyle, building construction and the continued accumulation of new sources of pollutants. (Lioy et al., 1999; Rudel et al., 2001; Spengler et al., 2001; Wilson et al., 2003; Dearry, 2004; Fan et al., 2003; Hites, 2004; Wu, 2004) However, it should be recognized that the resulting exposures can come from multiple routes of contact with the toxicant of concern. Typically, the route of exposure considered for indoor problems has been inhalation. However, significant concerns can be derived from dietary (contaminated surfaces from surface c ontact) and nondietary ingestion (e.g., lead) and dermal transfer and absorption (e.g., organic compounds) including pesticides. Aggregate exposure characterization has begun to address the above through its initial application to residential pesticide exposure (USEPA, 2001), and this approach is being used in the development of the Total Human Exposure Assessment Study (THEXAS-CHEM) for Europe through the JRC Project Knowledge System- FP6-WP2005-Action no 1312 (http://projects-2005.jrc.cec.eu.int).
Similar to outdoor pollution, current analyses of indoor and personal environmental health issues recognize that the emission, transformation and accumulation (e.g., build-up in air or on surfaces) of indoor pollutants have changed in character over the years. For example, as a consequence of the energy crises in the 1970s, buildings in America and many other parts of the world saw a decrease in interior ventilation rates that are represented in terms of number of air changes per hour (e.g., 0.5 means approximately half the air replaced in an hour). The decline in indoor ventilation to <0.5 air changes per hour provided a clear indication that such conditions will be conducive for the accumulation of high levels of contaminants with indoor sources (Spengler et al., 2001). Thus, with low ventilation rates and tighter buildings, the levels of air pollutants emitted by similar indoor sources can be higher today than they were 30 years ago in many buildings. In parallel, more synthetic materials are being used to make and finish indoor furnishings, products and materials, which at a minimum, increases the number of contaminants that can be emitted simultaneously or independently indoors. Many of these “new” sources have been found to release a large variety of trapped volatile and semivolatile compounds, an issue recently discussed and summarized in a review by Nazaroff and Weschler (2004). As a result, there can be the release or accumulation of such materials from media or objects (e.g., plush furniture or porous materials) that can lead to aggregate or cumulative exposures from inhalation, dermal absorption and/or nondietary ingestion.
Over the past 15 years, the ventilation recommendations made by the American Society of Heating, Refrigeration and Air Conditioning Engineers (ASHRAE) for new buildings have changed, and increased the desirable number of building air changes per hour each day (Spengler et al., 2001). However, it will take years to rectify the problems created by the design of tight and poorly ventilated commercial buildings and even longer for residential structures. Further, since many designers and architects generally continue to prefer “nonoperable” windows in new commercial construction, the situation will remain ripe for indoor pollutant accumulation in the air or on surfaces, or within objects.
Concerns about indoor mixtures have expanded from just air pollutants to include house dust, water pollutants associated with bathing and washing, and materials trapped in or adsorbed on surfaces of objects These situations can be caused by the addition of “new” or unanticipated sources with a low or high likelihood of occurrence; such as, hazardous waste infiltration, or dust mites, or mold in wet buildings. In addition, new materials or products can contaminate indoor environments. Each of these sources can emit chemical, physical or biological contaminants, and may distribute them on or in other surfaces or objects. Typical sources include home furnishings, construction products, as well as consumer products (e.g., candles, air fresheners), synthetic fibers and personal products, Finally, our attempts to sterilize places indoors, for example, bathroom and kitchen, and even toys, may result in overuse or misuse of products redefining the indoor environment to the point where the natural immunological processes have been altered by the over use of bactericides. As a result, people may be more susceptible to certain types of environmental insults that traditionally may have been less obvious problems (e.g., allergic responses) (Riedler et al., 2002).
Scientists who have examined indoor pollution have attempted to place it in the context of multipollutant source and concentration build-up (accumulation) issues, but primarily focused on the consequences of inhalation exposures. For example, the evaluation of the health effects associated with Molhave mixture of typical indoor air volatile organic compounds during the 1980s was one of the first attempts to examine symptoms associated with the “sick building syndrome”. The controlled exposure response studies used a mixture of over 20 compounds commonly found indoors (Molhave, 2001). Recent indoor air research has expanded the Molhave mixture to include some of the more reactive volatile organic compounds emitted indoors by various home products (Wainman et al., 2000). Thus, multipollutant research strategies analogous to those being recommended for outdoor particulate matter pollution and clinical evaluation strategies similar to the Molhave approach should be considered for more extensive use to effectively address current total indoor exposures issues. Included would be toxicants trapped in objects or materials, and outdoor pollutants that infiltrate indoors and react with indoor air pollutants or with indoor surfaces; the result will be primary or secondary pollutants that are available for contacts leading to ingestion, dermal or inhalation exposure. (Nazaroff and Cass, 1986; Colombo et al., 1993; Zhang and Lioy, 1994a, 1994b; Weschler and Shields, 1997; Wainman et al., 2000; Wainman et al., 2001; Fan et al., 2003; Weschler, 2003).
Pollutant dynamics: outdoor and indoor similarities
One major outdoor environmental process that can be linked to indoor processes, and help explain multiroute exposure issues, is the “grasshopper effect”. Outdoors the “grasshopper effect” has been used extensively to describe the movement of many persistent organic pollutants (POPs) from location to location around the world, Figure 1. For example, it has been shown to be a primary driver for the deposition and evaporation of POPs in the Great Lakes (Fernandez and Grimalt, 2003). It involves physical and chemical processes in an atmosphere that lead to volatilization and/or condensation of a chemical. Then, in combination with the diffusion and convective processes, the “grasshopper effect” augments transport of POPs to and from surfaces. In the end, a semivolatile pollutant emitted in one region, state or country can eventually deposit in multiple locations at great distances from the original release site.
The process occurs regularly outdoors, but the concepts can also be adapted to describe indoor pollution processes and accumulation in the air or in/on surfaces. However, the impact of the “grasshopper effect” on potential human exposure will be much different indoors than outdoors. Outdoors, low levels of POPs (e.g., PCBs, Dioxin, chlordane, DDT) spread to many distant locations. Conversely, the physical barriers (e.g., walls) will prevent the rapid transport of indoor POPs and other organics to the outdoors and can lead to continuous and higher indoor inhalation, dermal contact or nondietary ingestion exposures. (Zhang and Lioy, 1994b; Zhang et al., 1994; Chuang et al., 1995; Seifert et al., 2000; Wainman et al., 2000, 2001; Rudel et al., 2003; Olson and Corsi, 2004; Nazaroff and Weschler, 2004).
Grasshopper Effect and Indoor Chlorpyrifos: An Example
As an indoor example, consider the work of Gurunathan et al. (1998), which examined the mechanisms of release and accumulation of the semivolatile pesticide, chlorpyrifos within a residential indoor environment. An important observation made by the investigators was that similar to the processes that govern the movement of semivolatile organics outdoors, indoor air levels of chlorpyrifos change via volatilization and recondensation. Therefore, after a routine pesticide application, the chlorpyrifos that had been deposited directly on indoor surfaces and within objects did not stay within or on that one surface. The implication is that after application of any semivolatile pesticide, etc., the contaminants cannot be considered exclusively as a pesticide “residue”—a significant conclusion, and their semivolatility increases the availability of such compounds for multiroute contacts and aggregate exposures via inhalation, nondietary ingestion, incidental dietary ingestion (adsorption on food product) and/or dermal transfer. Further, chlorpyrifos, other pesticides and other semivolatile compounds can be translocated among various locations within a residence or other indoor spaces. This is not just due to pesticide drift from outdoor pesticide applications (e.g., farms); it is associated with volatilization that occur post-indoor application, or tracking of the pesticide indoors (Lu et al., 2000; Lioy et al., 2002).
Work completed by Bukowski et al. (1996) on the pesticide chlorpyrifos has also demonstrated that after a pesticide application, the concentrations found in the air within a home did not quickly drop to background within the ∼2 h previously assumed and recommended as pesticide re-entry criteria. In fact, depending upon the room temperature, or changes in temperature, the indoor levels will stabilize, and then change (increase or decrease) over time. These secondary peaks resulted partially from the absorption/adsorption and the re-emission of the compound as it tries to re-establish quasi-equilibrium within the home based upon the ventilation rate and ambient (indoor) air temperature.
The systematic release and movement of chemicals via the “grasshopper effect” can be used to explain the patterns of chlorpyrifos accumulation in the indoor air after a professional pesticide application as found by Gurunathan et al. (1998). After chlorpyrifos was applied, the initial air concentrations of chlorpyrifos did not exponentially decay immediately to a relatively stable background concentration range or level. The peak is delayed and the decline was relatively slow with secondary peaks occurring at later time. The latter was indicative of the presence of a secondary source of emission or re-emission. This behavior is illustrated in Figure 2, since the initial peak in chlorpyrifos concentration occurred 50 h post-application, not 2 h post-application. Then after a slight increase at 75 h post-application, there was a slow decline that was followed by second increase in the airborne chlorpyrifos concentration 4 days post-application. The changes in chlorpyrifos shown in Figure 2 were consistent with an iterative pattern of re-emission from, and re-absorption/adsorption on, active surfaces. More importantly, the “grasshopper effect” helped to explain the time course of the adsorption and absorption on objects and surfaces of pesticides and other semivolatile materials in the home during a long period of time after application.
This point was also illustrated in the same study by Gurunathan et al. (1998), which also found that plastic and felt toys, frequently present indoors, absorbed and retained the semivolatile pesticide. The objects selected by Guranathan et al. were found to have adsorbed or absorbed the chlorpyrifos during the first 3–5 days after a pesticide application, and also provided a reservoir for re-emission. Subsequently, the pesticide was slowly released or continued to be absorbed by the toys from day 7 through day 14 post-application, as illustrated in Figure 3. Thus, the toys acted as both a secondary source and receptor and continued the potential for inhalation and nondietary ingestion exposure over an extended period of time.
In addition to the above, the physical and chemical processes that control the accumulation (on or in objects) and release of the chlorpyrifos can be expanded to include all plush objects that have an affinity for pesticides or other semivolatile materials. Included are objects and products present in buildings and homes that contain poly-foam materials, for example, sofa, chair cushions. The nature of such materials must be evaluated or re-evaluated for emissions of other semivolatile compound to better define the sources, the duration of contact and the types of exposures that can occur. For example, Hore et al. (2005) has shown that if chlorpyrifos applications are made using a crack and crevice technique instead of the broadcast method, the primary and secondary loadings are reduced, and the bioaccumulation of chlorpyrifos was minimal in children after a routine and professional application.
Ultimately, efforts to understand the redistribution of semivolatiles indoors can be used to define and/or reduce the risk posed by indoor contact with one or more semivolatile materials over time via inhalation or nondietary ingestion. The result would be aggregate or cumulative exposures for single or multiple pollutants. It can also be concluded that to better characterize potential exposures, for example, indoor aggregate or cumulative exposures, one must expand the indoor contaminant list to non-pesticide chemicals and other agents deposited or redistributed on or within surfaces.
Indoor Surfaces as a Source and a Reservoir of Indoor Pollutants
Surfaces play an important role in assessing the significance of outdoor environmental transport and deposition processes. However, indoor surfaces have an even more important role in characterizing indoor environmental health problems; given that indoor surface to volume ratios are substantially higher than outdoors. These surfaces are usually located well within the range of influence of the boundaries that define size (volume) of the indoor environment, and can (1) increase or decrease air indoor concentrations (and exposures), (2) cause surface reactions among chemicals and (3) increase direct exposure to materials on/within the surface (Freeman et al., 1995; Edwards et al., 1998; Roberts, 1998; Jorgensen et al., 1999; Lioy et al., 2000, 2002; Weschler, 2003; Nazaroff and Weschler, 2004; Colt et al., 2004).
The surfaces present at the boundary of a room or building can be active absorbers/adsorbers for some pollutants. The absorbed/adsorbed material may be re-released from the surfaces (e.g., formaldehyde) or volatile products released by biological materials (e.g., mold or bacteria). Thus, the same surfaces become emitters, which can yield inhalation exposures. The surfaces can become a reservoir of absorbed/adsorbed chemicals, and act as both a source and a receptor of environmental pollutants. Eventually these can lead to contacts that cause significant dermal or ingestion-related exposures (Zhang and Lioy, 1994b; Zhang et al., 1994; Chuang et al., 1995; Rudel et al., 2003).
Weschler (2003) has shown that under most circumstances the amount of a semivolatile material adsorbed by a room surface would be much higher than on the surface of particles deposited on a surface (because of the much greater room surface area than particle surface area). The above concepts provide a foundation for examining, and addressing the environmental health aspects indoor mixtures of pollutants within the framework of aggregate or cumulative exposures (USEPA, 2001, 2003). Our indoor environment contains many types of organic and inorganic mixtures that can be found simultaneously in the air or deposited on surfaces. Some may collectively or individually be of concern, while others may not be of concern to human health (Spengler et al., 2001). In addition to the surface of household objects and furnishings being a reservoir of semivolatile compounds, the actual dust and soil that has settled on or within surfaces found indoors provides more surface area for initiating or promoting various physical and chemical processes. These can occur while on a material is a surface or when resuspended into the air. (Lioy et al., 2002; USEPA, 2001, 2003; Nazaroff and Weschler, 2004).
A review of indoor dust and dust sampling by Lioy et al. (2002) showed that dust is composed of many materials that include man made and natural products (human skin and hair). The very large surface area to mass of dust makes it a very efficient medium to absorb compounds. Thus, surfaces that collect dust deposited in the home can serve as reservoirs of the materials that were both transported and emitted within a home, and such reservoirs can also act as a “new” source for multiroute exposure (e.g., inhalation, dermal, and nondietary ingestion) over many years (Farfel et al., 1994a, 1994b; Molhave et al., 2000; Lioy et al., 2002; Ilacqua et al., 2003) Dust analysis results reported by Millette et al. (2004) found that the composition of indoor dust can differ from room to room, and from city to city. The data in Table 1a and 1b provide a summary that illustrates differences in composition that may occur between rooms or between cities.
The review by Lioy et al. (2002) also indicated that surfaces will release materials into the air as well as adsorb or absorb chemicals and other materials from the air. The data presented in Table 2 show that many chemicals, including phthalates, PAHs and Hg, can be present in dust. The processes that control deposition and release of contaminants are similar for both the surface of objects, and the surface of deposited particles; however, the emission rates of semivolatile or volatile components will be different for specific materials or surfaces. Nonvolatile species will require other mechanisms for release or resuspension into the indoor air, for example, vacuuming (Lioy et al., 1999). The particle dispersion (and size distribution) of the resuspension processes will be influenced by the temperature and humidity in a room or building (Wang et al., 1995). In addition, activities completed within a room by children or adults or vacuum cleaners (Lioy et al, 1999) can lead to particle resuspension in the room or around the person that could be characterized by a process which is most easily visualized in the Pig-Pen Effect® (a Shultz Comics registered trademark).
Dynamic Processes and Steady-State Conditions
An examination of Figure 4 shows that there can be transport into and within indoor spaces caused by primary emissions; absorption and evaporation; adsorption; infiltration transformations and decay (secondary emission of the same chemicals). If one removed the walls of a structure, the most significant changes would be the addition of direct sunlight to fuel photochemistry, dilution of the concentrations and removal of pollutants absorbed or adsorbed on the wall.
The indoor conceptual model illustrated in Figure 4 can be used to demonstrate the movement and accumulation of many compounds indoors. These include toxicants emitted from a variety of materials (e.g., house furnishings), and the infiltration of outdoor pollutants. The variables and processes illustrated can also lead to the consideration of scenarios for quasi-steady-state conditions that can minimize or maximize indoor levels in the air or on surfaces (Lioy, 2000). Such analyses can help define the magnitude of indoor exposures and, identify the variables needed to develop source-to-dose computational/models (Georgopoulos and Lioy, 2006 (in press)).
The dynamics of accumulation for individual materials could include some or all of the identified processes. In each case, the contributions will depend upon the nature and amount of the contaminants present or released indoors. Many of the processes identified in the conceptual model or for use in the source-to-dose model will govern how, or if, a semivolatile compound may be able to achieve quasi-steady-state levels indoors. In addition, these processes must be considered in attempts to define the dynamics and kinetics for the passive release of volatile organic compounds (high Henry's law constant). These are initially trapped in materials, containers and objects commonly used indoors or permanently placed indoors or semivolatile and other chemicals that can be transported indoors from outdoor sources. Thus, many sources present indoors or outdoors will be available to yield inhalation, ingestion and/or dermal exposure. These need to be measured or modeled to assist in identifying the routes of primary concern (National Research Council, 1991). Each, however, may require different types of strategies to mitigate or prevent the single route or aggregate or cumulative human exposure (OECD, 2003; USEPA, 2001, 2003).
In the home, the rate of sorption/volatilization of individual compounds will be controlled by the gas to particle partition coefficients, or the gas to surface partition coefficient; as well as, temperature, relative humidity and air speed. The actual process of sorption/volatilization, however, would be affected by the composition surface type (e.g., rugs, plush objects, hardwood), surface loading, object loading and particle loading. Therefore, it becomes apparent that there is probably a source or group of sources that are present in a typical home or building, or outside that provide baseline (quasi-steady-state) and possibly higher concentrations of semivolatile compound mixtures after being released or transported indoors. The nature of some of these chemicals will probably lead to transformation or depletion over time (aging). Replenishment of the chemical or chemicals would be dependent on the application of more material to the indoor environment and its adsorption or absorption on or in surfaces. These processes can be augmented by the addition of pollutants emitted by other sources; for example, PAHs during combustion (Chuang et al., 1995; Tucker, 2001; Wilson et al., 2003).
Weschler reviewed the volatility of a number of air pollutants, and identified a number of compounds that would be affected by the processes identified above (Weschler, 2003). In the list published by Weschler (2003), Table 3, some of these common chemicals are found in the indoor environment. Another class of compounds not listed by Weschler, but of concern indoors, is polybrominated diphenol ethers (PBDEs) (Rudel et al., 2003).
An important point made by Weschler's work is that the partitioning between the vapor and particle phase will be dependent upon the volatility which in turn is affected by many of the variables shown in Figure 4. As a result, the levels in air, on surfaces and/or dust will be affected by volatility. This affects the intensity of exposure or risk that can result from inhalation, ingestion or dermal contact on a given day because of changes various indoor or outdoor environmental variables (e.g., temperature, humidity, ventilation rate or wind speed) or lack of changes (quasi-steady-state) in these variables.
As particles are deposited on a surface, higher loadings can result from multiple applications or the general use of a semivolatile material, and higher surface loadings could enhance the partitioning between phases. However, the competing processes of emission and adsorption can become more significant for individual compounds depending upon whether or not there is a large surface area available for adsorption/absorption. A further issue is the number of materials that can deposit on or be present in a material or object. Each can act as an independent source within an active reservoir. Thus, a rug or carpet can absorb, adsorb or release one or more volatile or semivolatile species, and in some situations can become an active reservoir for multiroute exposures to many toxic substances and other materials, for example, fibers. Further, materials deeply embedded in a carpet will not be easily removed by a vacuum cleaner and these trapped semivolatile materials can be re-emitted to the indoor air or available for contact with a hand or other parts of the body for years.
The retention of the materials will also be affected by the amount of water attached to the surface, as well the hydrophilic nature of the materials and the strength of the Van der Waals sources (electrical) that hold a compound to a surface. Absorption will also be affected by diffusivity of a compound within materials found within a residence. Again, a good example is the absorption of pesticides.
It has been demonstrated that poly foam is an excellent absorbent for organic phosphate pesticides, and it has been used as a collection medium for pesticides when obtaining air samples. In homes, there are many materials and objects that contain or are primarily composed of poly-foam, for example, toys and plush pillows, and these present a set of surfaces and volumes for preferential absorption and redistribution in the indoor environment (Gurunathan et al., 1998). Other pollutants that fit into the category of low vapor pressure compounds that will be partitioned between the gas phase and the sorption phase include those that are added to objects and surfaces, consumer products (e.g., polybrominated diphenol ethers), and combustion products. For example PAHs are emitted by a variety of combustion sources (indoors or outdoors), and each PAH species varies in volatility and absorption, which are dependent upon each compound's molecular weight and structure. As a result, some PAHs act primarily as gases, while others can act either as a particle or gas, and the distribution of levels sampled in the air or from a surface will be dependent upon the source of the emissions (indoors or outdoors) and the time post-emission. Weschler (2003) estimated the partitioning of several PAH's between the gas phase and the particle phase, and on surfaces within a hypothetical typical room. For naphthalene, the mass in the room in the gas phase was estimated as 200 μg, its mass in the particle phase was 0.012 μg and the mass on a 10m2 carpets was 7400 μg. In the case of chrysene, the mass in the gas phase was 20 μg, the particle phase was 3.6 μg, and the mass on the carpet was 3 × 106 μg. Thus, a higher molecular weight PAH compound would preferentially attach to a surface, and the lower molecular weight PAH compound (e.g., naphthalene) would be more likely found in the air (Weschler, 2003).
The room temperature, the method and rate of room ventilation, and outdoor infiltration rate will affect the gas to particle partitioning of each volatile or semivolatile compound. Owing to the above processes and the potential distribution of many contaminants in homes and offices, a compound with low volatility can achieve pseudo-steady-state air levels indoor over the course of a day. Thus, it is necessary to define or characterize the nature of the temperature humidity and ventilation cycles in a variety of buildings. (Spengler et al., 2001; Wallace et al., 2002; Hood, 2004; Johnson et al., 2004). For example, if a home has radiant heating under the floor in the home and the floor has been finished with a polyurethane type material, during the first few months post completion there can be significant quantities of petroleum distillates released into the indoor air (Case 1). In contrast, if the floor is not heated, the emission rate will be much lower (Case 2). While volatile petroleum distillates will be found in the air for both cases the concentration will vary, because of the different emission rates for similar ventilation rates. For Case 1, a significant indoor air concentration and acute exposures can occur over a short period of time because of the higher temperature on the floor. In contrast, for Case 2, the exposures would be the same, but the effect could be of no concern, because much lower concentrations are present over a much longer period of time. Thus, occupants in the second home are not expected to experience potential health effects associated with acute exposures.
Indoor air chemistry
Recent studies have identified or hypothesized the contributions of secondary pollutants formed by gas phase or surface chemistry to the levels and types of air pollutants that accumulate indoors (Nazaroff and Cass, 1986; Zhang and Lioy, 1994b; Zhang et al., 1994; Weschler, 1999, 2000; Fan et al., 2003; Lee et al., 2002; Nazaroff and Weschler, 2004). This is in contrast to other studies that have examined the fraction of chemicals released and then sorbed on surfaces (Gurunathan et al., 1998; Tucker, 2001; Lioy et al., 2002; Rudel et al., 2003; Weschler, 2003).
The composition of the secondary compounds produced by gas phase chemistry needs to be characterized for various settings and conditions. The source of some indoor reactants, particularly oxidative species, is the outdoor environment. These contaminants migrate indoors by active and passive transport mechanisms. Some of the mechanisms for the movement of precursors within or into the indoors are also found in Figure 4. The products yielded by indoor chemical reactions have important implications, since reduction of the levels of the outdoor reactants migrating indoors will compliment or support the development of effective outdoor strategies that can also address potential indoor health issues. Actually, EPA's efforts to control precursors of some outdoor pollutants, for example, ozone, are needed to reduce the reactivity of the indoor air as well as outdoor air, and has serendipitous consequence of reducing total inhalation exposure to ozone. The products of the chemical reactions of the outdoor pollutant ozone and radicals, HO, HO2, etc., with volatile organics and nitrogen oxides emitted indoors can also lead to significant accumulations of both secondary particles (e.g., organics) and gases (e.g., nitric acid, aldehydes) in the indoor air (Zhang and Lioy, 1994b; Zhang et al., 1994; Fan et al., 2003). Thus, maintaining the regulations for control of volatile organics and nitrogen oxides outdoors will reduce exposures to secondary products indoors.
For characterization of inhalation exposure, it is apparent that when outdoor sources and emissions of precursor pollutants are coupled with indoor air precursor sources and chemical processes, the types and the concentrations of pollutants inhaled by an individual can change over time. Research by Wainman et al. (2000, 2001) and Fan et al. (2003) has clearly demonstrated the potential for formation of a variety of gas phase (aldehydes) and secondary particles phase toxicants indoors. A chamber study completed by Fan et al. (2003) examined the kinetics of the ozone reacting with volatile compounds indoors. Fan et al. found, as shown in Figure 5, that reactions between ozone used in the experiments and VOCs produce both secondary particles and gases. The concentration of ozone was only 40 ppb (1/3 of the previous 1-h National Ambient Air Quality Standard for the United States), and the concentration of the VOCs was 6 ppm. These indoor concentrations occurred at a moderate air exchange rate (1 air change per hour) but yielded significant increases in the concentrations of fine particles that were caused by reacting ozone with the VOCs. Clearly, other conditions may lead to more particle formation due to likelihood of finding higher ozone concentrations in non-air conditioned or one pass air conditioned indoor homes/buildings during the summertime. In a study conducted by Zhang and Lioy (1994a) the indoor concentrations ranged from 20 to 140 ppb, depending upon the outdoor levels and the method of ventilation, that is, air conditioning, no air conditioning.
It is important to note that the accumulation of primary or secondary air pollutants indoors will depend not only on changes in the source emission rate but also, on the physical and chemical processes that take place within a structure (building); for example, wall adsorption, humidity and reactions caused by intermediate compounds. Published results have shown that reactions between ozone and organic and inorganic species can lead to the formation of significant quantities of secondary particles. This is an observation that is beginning to be assessed by the scientific community as part of total inhalation exposure (USEPA, 2004a). Part of this assessment will be augmenting the current indoor source and ventilation models with indoor chemistry and the physical processes of evaporation and condensation. Such indoor chemistry models will be very useful for examining the issue of indoor mixtures, including those that affect source-to-dose relationships (Georgopoulos and Lioy, 2006; Georgopoulos et al., 1997; Georgopoulos et al., 2005).
Indoor environment: As complex mixture of pollutants
In the outdoor environment, the absorption, adsorption, condensation, emission and evaporation of pollutants can occur throughout the day, and the “grasshopper effect”, is one of many processes that affect the atmospheric accumulation of outdoor air pollutants and their deposition. Theoretical, laboratory and field studies have been completed that quantify the absorption and emission of materials from sources and surfaces and their reactions during nonphotochemical outdoor events (Kamens et al., 1999; Finlayson-Pitts and Pitts, 2000; Fernandez and Grimalt, 2003; Lohmann and Lammel, 2004). In contrast, there is little information available that provides a detailed understanding of the complex nature of indoor accumulation and evaporation processes associated with the presence of semivolatile compounds indoors. Research reporting on the emission of compounds from surfaces, operations (e.g., photocopying), consumer products, and building products was summarized in an excellent review article by Tucker (2001). Much of this information has been used to identify the types of emissions that can occur inside residential and commercial structures.
Specific studies have examined the absorption or adsorption of semivolatile compounds by various commercial materials (e.g., cotton, steel) and the interaction of semivolatile compounds with surfaces. For example, Hodgson et al. (1993) examined the release of volatile organics from carpets. However, the results provided little information to assist in understanding the variability in the concentration patterns of semivolatile compounds found routinely on and released from indoor surfaces. Further, even less is known to help couple these results to the levels of multiple pollutants that might be expected to accumulate in the indoor air. Eventually, the suite of semivolatile compounds that cause mixture interactions via similar modes of action, either additively or synergistically, needs to be examined in conjunction with the magnitude and extent of multipollutant (mixtures) cumulative exposures that can be found in a variety of indoor situations and lead to “total indoor environment” exposure/health concerns.
Until recently, few studies were designed to systematically examine the nature of the complex mixtures associated with indoor air and/or the total indoor environment. A study by Rudel et al. (2003) found many estrogenic materials present simultaneously in the indoor air and on indoor surfaces. A summary of selected compounds detected by Rudel et al. are presented in Table 4, and include many current and banned materials. They found that the indoor air levels of many of the detected compounds were higher than the levels of the same contaminants present outdoors (Rudel et al., 2003). The observation would be consistent with the “grasshopper effect” by augmenting the accumulation of compounds on surfaces indoors, and their subsequent release over time. Thus, these estrogenic compounds would be available for single as well as multiroute exposure and for longer periods of time. These data support the body of information that has been provided by the NHEXAS study and other studies on pollutants emitted outdoors or indoors and accumulate within the indoor environment (Jantunen et al., 1998; Bonanno et al., 2001; Pellizzari et al., 2001; Roy et al., 2003; Weisel et al., 2005). However, the question remains as to how these compounds vary over time and space in a building?
The processes illustrated in Figure 4 provided a schematic of the complex series physical–chemical processes that govern: (1) the initial and iterative release of a semi-volatile compound within the indoor environment, (2) the penetration of that compound from ambient sources and (3) the build-up of quasi-steady-state levels in the indoor environment (Lioy, 2000). The full range of the complex mixture of semivolatile and volatile materials within individual residences and other buildings that can be released indoors and different types of material need to be characterized for their potential impact on total (multiroute) exposure. The results can then be used to design multipollutant mechanistic or epidemiological studies.
The net result of partitioning between surfaces and the air and chemical reactions will be a complex mixture that would contain a number confounding and possibly synergistic or antagonistic chemicals. Depending upon the toxicity of components of the mixture, each event or general categories of events could lead to a variety of risks associated with exposures from one or more routes of entry into the body. This can be caused by a single contaminant, multiple contaminants with the same toxic effect, or the synergism or antagonism among a number of copollutants. Further, knowledge about each of the above will improve our ability to develop source-to-dose models (Georgopoulos and Lioy, 2006 (in press)).
The migration of volatile materials to and from objects and surfaces will yield variations in concentrations within a dynamic indoor environment. Volatile compounds that adsorb on to surfaces will accumulate; however, their concentration in the air (potential for inhalation) or on a surface (potential for non-ingestion or dermal exposure) change within a day, from day-to-day, and from season to season. The fluctuating air or surface concentrations will depend upon many of the factors illustrated in Figure 4. But, a major contributor to indoor problems is the continued addition of sources into the home, which increases the variety and level of semivolatile and volatile compounds available for release (Tucker, 2001).
The time frame for sustained releases into the indoor environment can become much longer (month to years) when objects are placed indoors by consumers that contain materials that are continually released because of their volatility. The persistence of these reservoirs will be dependent upon both volatility of the chemical and environmental conditions. In some cases, the pollutants are residual chemicals associated with a production process. Examples include formaldehyde, petroleum distillates, polybrominated diphenol ethers and phthalates. In situations where there are a series of semivolatile materials in the home or other building, and there are sequestered volatile materials, changes in temperature and humidity levels in the home will increase or decrease the rate of release of these compounds in the home. Thus, in any indoor environment, the day to day changes in concentration will vary depending upon the dynamic and kinetic processes conceptualized in Figure 4.
An example of the components of such a process was recently published by Clausen et al. (2004) for the release and adsorption of di-2-ethylhexyl phthalate (DEHP) from PVC flooring into two different types of controlled chambers. The experiment yielded a record of the daily release of the phthalates, and the concentration patterns in the chambers were recorded over a series of months. The DEHP concentrations first showed a build-up period and then a period of quasi-equilibrium in a chamber. This was due to a balancing between the emission of the phthalate and its adsorption and re-emission of DEHP by dust on the surface of the PVC in the chambers. Obviously, there will be much more variability in concentration of a phthalate or other material in a home or other structure since the doors and windows can be opened and closed daily. In addition, the DEHP would have to compete with or would be affected by the emissions from other sources and even other components of dust. Within home and between home variability, person and interpersonal variability, lifestyles and source strength will affect the magnitude of the releases, and the indoor concentrations achieved over time. What is clear, however, is that the proximate or ultimate sources of these multiple pollutants must be mitigated or replaced when there is build-up indoors. The application of aggregate and cumulative exposure analyses and then risk assessment tools can help identify the worst source situations (OECD, 2003; USEPA, 2001, 2003).
An example of an evolving set of potential “new” pollutants is engineered nano-particles (<100 nm in diameter) since they may become candidates for replacing materials currently used in “foam” for elasticity. We, at a minimum, will have to consider the potential for the release of these materials from surfaces over time, and the potential exposures that can result among the general public (Hood, 2004). Since nano-particles are only beginning to be introduced as part of commerce, it would be prudent to assess any biological mechanism of action or human exposure issues before large scale applications are attempted for consumer products. Alternately, we will do it after the fact, which has been the case for polybrominated diphenol ethers and many environmental chemicals, for example, PBDEs and MTBE (USEPA, 1999; Hites, 2004). Some recent examples of nano-particle applications are sporting goods, tires, stain-resistant clothing, and sun-screen.
Work by Oberdorster et al. (2004) indicate that these particles will deposit in the nasal passages, lung and transmitted to the brain. Thus, the single or multi-route exposures and potential health outcomes to nano-particles must be considered as a new issue for environmental health and indoor environmental research; but, before they become a pervasive component of household products. The status of knowledge on nano-particles has been reviewed and a good start will be information associated with the National Nanotechnology Initiative, which officially began in 2000. (www.nano.gov) At the present, time NIH receives approximately 80 million dollars for research but EPA receives only about 5 million. Thus, although health or mechanism of action studies are being conducted, there are few resources being studied for examining human exposure.
Implications for Exposure Research and Mitigation of Health Effects
The preceding provides a conundrum to the field of environmental health sciences and the companion field of exposure science. Human health effects can occur because of indoor pollution and such information has been summarized in numerous reviews and reports. Important examples are pesticides (National Research Council (1993a) and www.epa.gov/pesticides/science/models_db.htm, lead (ATSDR, 1999; National Research Council, 1993b)), and indoor air (Spengler et al., 2001). However, because of the many types of materials released or formed or transported indoors, it can become difficult to define the compound or the chemical of concern. Of course, in cases where the health outcome is very specific and serious (e.g., death), and can be associated with the acute toxicity of a specific compound, the problem is relatively straight forward (e.g., carbon monoxide). In contrast, complex mixtures of the indoor environment have not been explored for the potential in additive, synergic and/or antagonistic modes of actions that occur because of the presence of multiple compound and compound classes. This is still primarily a hypothesis, and requires toxicological studies to validate the multichemical and or synergisms that can occur indoors, and epidemiological investigations to establish single or multiroute exposure–response relationships. Problems in developing such studies are: What to select for measurement of exposure and response, and what to select as the best animal models? Since some of these chemicals are semivolatile and volatile, they can be released simultaneously while released indoors because of many environmental factors including temperature and ventilation rates . At first this just may be construed as the release of a series of confounding chemical species. Since we do not understand the toxicology of complex mixtures or reactivity of many complex mixtures and their products, individuals and populations at risk actually may be affected adversely by more than a single toxic agent in a specific situation. Finally, one must also consider that the toxicity of the mixture may be reduced or enhanced after undergoing a series of chemical reactions on surfaces or in the air indoors.
The “total indoor environment” can be considered a chemical world, and the general public and sensitive or susceptible subgroups live in this world. Notable examples would be pregnant women, young children, <2 years of age, and asthmatics. The issues of single and multiroute exposure to such individuals can occur in many different places and multiple times (Lioy, 1990a; National Research Council, 1991). In addition to the home, single or multiroute exposures occur in schools, day care centers, offices and public buildings, arenas or outdoors. Thus, examining the total exposures of susceptible people in such natural settings can be useful in evaluating public health policies to control the indoor and outdoor sources that contribute to single, or aggregate or cumulative exposures (USEPA, 2001, 2003).
Environmental protection provides opportunities to eliminate many problems but it still compartmentalizes problems. Regulators and managers tend to try to explain an apparent effect, without considering the variables that must be examined in conjunction with the agent they consider to be most important. Alternatively, if scientists find a new compound in the environment and try to identify effects, many times they complete studies and analyses without the detection of the distribution of their presence or looking for obvious confounding variables. The work by Rudel et al. (2003) measured many compounds with estrogenic activities in 120 homes, which is an interesting start that can lead to many hypotheses. If a single estrogen in these homes was being considered for health effects, the background concentrations they measured for the multitude of species present (from indoor/outdoor sources) could confound any results. Further, the air and dust levels of estrogenic compounds need to be measured to identify the contacts that can lead to significant exposures. Subsequently, it is incumbent upon exposure science to define the “real world” for those who would conduct mechanistic studies needed to define the mode of actions caused by the presence of multiple estrogenic compounds found indoors. Toxicological studies on these mixtures will also need to consider the levels of the pollutant mixtures and temporal relationships associated with interactions at or near the NOEL or NOAEL (National Research Council, 1994; Skipp and Allen, 1994).
The evaluation and characterization of multipollutant indoor contamination does not end with semivolatile, volatile compounds and suspended particles or resuspendable dust. The multipollutant indoor issues must include bacteria (e.g., Legionella bacteria, and Mycobacterium tuberculosis), mold, fungi and other forms of viable material, as well as any volatile byproducts or toxins (e.g., endotoxins, mycotoxins and dust mite allergen) that are released by biological agents in a building (Spengler et al., 2001). In such cases, however, the “in vivo” biological processes may or may not mirror the same physical-chemical processes that govern the dynamics of the “grasshopper effect.” For example, dust mites and their allergenic byproducts are found in high use areas of the home, their density is significantly affected by relative humidity, their distribution is affected by cleaning frequency and intensity, and it is difficult to determine the minimum amount of material that can cause an effect (Arlian, 1989; Hirsch et al., 1998; Lewis et al., 1998).
Similar to the outdoor environment, the processes that govern the release of indoor air pollution through (1) combustion, (2) direct application of consumer products, (3) resuspension of particles on the floor by vacuuming, walking etc., “pigpen effect” (i.e., ®Schultz Comics registered trademark) and (4) chemical transformations caused by infiltrating outdoor pollutants would not follow the same dynamics as the semivolatile or volatile compounds (Farfel et al., 1994a, 1994b; Adgate et al., 1995; Roberts, 1998; Lioy, 1999; Roberts et al., 1999) They all, however, lead to the chemical world indoors.
The above points to the need to provide good baseline information on the levels of various chemical and physical agents that can be present in homes and other indoor environments. This was supposed to be the goal of the NHEXAS program (Pellizzari et al., 1995; Sexton et al., 1995). The research design of NHEXAS provided an excellent opportunity to characterize long-term exposure to many toxicants (although not bacteria and mold), especially those associated with the indoor and personal environment. However, after the initial studies were completed, the funding of the program faded into a series of data analysis opportunities, and no studies are planned for the future. There is some potential for large scale exposure characterization for one segment of the population, children, within the National Children's Study.
The advances we have made, and will continue to make, for the purposes of improving the quality of life for each American and members of the world community needs a series of scientific protocols for examining exposure that can identify problems early, and intervene quickly. The agencies that deal with environmental health issues must then prioritize research based upon the significance of acute or chronic exposures that affect each of us and the general public. The study of Air Quality and Water Quality do not address the many issues that currently affect most Americans and people living in developed countries.
Our built indoor environment has always offered humans more protection against the elements, and other dangers. It must, however, be designed
to minimize the additional burden from use or presence of toxic materials in the home that can lead to unanticipated acute or chronic health outcomes
to minimize “total indoor environment” exposures to toxicants”
to include a comprehensive evaluation of exposure and health interactions among physical, chemicals and biological agents
to include exposure prevention strategies in the design of new or green buildings. This can include both chemicals and biological agents.
The indoor environment has taken a central role in the lives of all Americans and people living in developed countries. This impact is partly due to the advent of air conditioning, the proliferation of the television, and the impact computers and the internet have on our daily personal and professional lives. Coincidentally, there has been changes in emissions and reservoirs of pollutants indoors due to the presence or use of a variety or commercial, residential and personal products; however, outdoor pollution will always be a factor in determining the quasi-steady-state levels of various pollutants found indoors.
Eventually the above will become a part of the total indoor environment in developing countries around the world. Unfortunately, in contrast to outdoor environmental issues, we do not totally appreciate the dynamics of indoor pollution sources and pollutant accumulation, and how they affect single route, aggregate or cumulative exposures. Part of this problem is the lack of understanding about the role semivolatile and volatile materials play in creating primary and secondary pollutants. These materials and associated processes can lead to complex mixtures that may produce complex environmental health problems among the general population or susceptible subgroup of the general population. Since there are various agents that are present or can be introduced indoors, there needs to be systematically designed field and/or modeling studies of the single route and multiroute exposures that can occur in the indoor environment. These are necessary to determine or estimate the aggregate or cumulative exposures and/or doses that can be of concern to human health. Such data and model estimates should be used to minimize or eliminate the total exposures that may occur, and develop approaches to evaluate the potential exposures that can be associated with new products before their introduction into the indoor environment. One goal will be to reduce or prevent single or multiroute exposures and the risks to single and multiple contaminants that cycle through or are formed indoors. However, the ultimate goal should be to still use commercial and personal products safely within the total indoor environment.
The author thanks Drs. J. Zhang and C. Weisel of EOHSI, and Dr N. Freeman, U. Florida, for their assistance in reviewing the manuscript, and their excellent comments and suggestions for improvement, and Dr. C. Weschler of EOHSI for our continuing discussions on total exposure and the indoor environment. The author thanks the many people who have conducted indoor research over the past 30 years, and helped improve our current understanding of fundamental issues. The research supported by the Center for Exposure and Risk Modeling (CERM) (Grant No.: CR8162501); the Center for Childhood Neurotoxicology and Exposure Assessment (POIES11256-01), and the NIEHS Center at EOHSI (Grant No.: P30 ESO5022).
About this article
Journal of Exposure Science & Environmental Epidemiology (2008)