Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The benefits of whole-house in-duct air cleaning in reducing exposures to fine particulate matter of outdoor origin: A modeling analysis

Abstract

Health risks of fine particle air pollution (PM2.5) are an important public health concern that has the potential to be mitigated in part by interventions such as air cleaning devices that reduce personal exposure to ambient PM2.5. To characterize exposure to ambient PM2.5 indoors as a function of residential air cleaners, a multi-zone indoor air quality model was used to integrate spatially resolved data on housing, meteorology, and ambient PM2.5, with performance testing of residential air cleaners to estimate short-term and annual average PM2.5 of outdoor origin inside residences of three metropolitan areas. The associated public health impacts of reduced ambient PM2.5 exposure were estimated using a standard health impact assessment methodology. Estimated indoor levels of ambient PM2.5 varied substantially among ventilation and air cleaning configurations. The median 24-h average indoor–outdoor ratio of ambient PM2.5 was 0.57 for homes with natural ventilation, 0.35 for homes with central air conditioning (AC) with conventional filtration, and 0.1 for homes with central AC with high efficiency in-duct air cleaner. Median modeled 24-h average indoor concentrations of PM2.5 of outdoor origin for those three configurations were 8.4, 5.3, and 1.5 μg/m3, respectively. The potential public health benefits of reduced exposure to ambient PM2.5 afforded by air cleaning systems were substantial. If the entire population of single-family homes with central AC in the modeling domain converted from conventional filtration to high-efficiency in-duct air cleaning, the change in ambient PM2.5 exposure is estimated to result in an annual reduction of 700 premature deaths, 940 hospital and emergency room visits, and 130,000 asthma attacks in these metropolitan areas. In addition to controlling emissions from sources, high-efficiency whole-house air cleaner are expected to reduce exposure to particles of outdoor origin and are projected to be an effective means of managing public health impacts of ambient particle pollution.

Introduction

Exposure to fine particles (PM2.5) in outdoor air is associated with a variety of adverse effects in humans including numerous forms of cardiovascular and respiratory illness as well as impaired activity, such as school and work loss days (Pope and Dockery, 2006). The public health impacts of exposure to ambient PM2.5 are estimated to be substantial. For instance, using a conventional health impact assessment methodology, the US Environmental Protection Agency estimated that controlling emissions of particle precursors from power plants in the United States would result in 17,000 fewer premature deaths and over 12,000 fewer hospitalizations for respiratory and cardiovascular disease each year, among other health outcomes (EPA, 2005). Because Americans spend approximately 69% of their time indoors at home (Echols et al., 1999; Klepeis et al., 2001) and particles that comprise PM2.5 penetrate buildings easily (Leaderer et al., 1999), residential indoor levels of PM2.5 of outdoor origin are an important source of personal exposure to ambient PM2.5.

Air conditioning has been shown to attenuate levels of PM2.5 of outdoor origin within residences (Suh et al., 1992; Rodes et al., 2001). Air conditioning use is likely associated with lower air exchange rates, and therefore lower infiltration of PM2.5 into homes compared to natural ventilation. The prevalence of air conditioning in a community appears to modify the health risks of ambient PM2.5 on a population scale (Janssen et al., 2002; Franklin et al., 2007). In the study by Janssen et al. (2002), data on cardiovascular disease and PM was analyzed for 14 communities in the US. The authors observed an approximately 50% decrease in the PM-related relative risk of cardiovascular disease for every 20% increase in central air conditioning use in a community. By extension, factors that increase removal rates indoors of PM2.5 should further reduce exposure to PM2.5 of ambient origin and the associated health impacts. For instance, use of high volume indoor air cleaners in apartments in Copenhagen was associated with an approximately 60% reduction of PM2.5 from both indoor and outdoor sources and improvement in a sub-clinical marker of microvascular function (Brauner et al., 2008).

A number of indoor air cleaning systems and technologies intended to remove aerosols from within residences are in commerce. The efficiency, removal rates, and other performance characteristics of some of these devices have been characterized empirically (Chen et al., 2006; Shaughnessy and Sextro, 2006; MacIntosh et al., 2008). The potential for advanced filtration technology to mitigate indoor exposure to PM2.5 of ambient origin in Europe was estimated from results of the EXPOLIS study (Hanninen et al., 2005). A similar modeling analysis was conducted for buildings in Singapore (Sultan, 2007). An assessment of exposure to PM2.5 of outdoor origin as a function of residential building stock and air cleaning methods in the United States has not yet been conducted.

In this paper, we use a multi-zone indoor air quality model to examine annual, 24-h average, and diurnal concentrations of PM2.5 of outdoor origin in residential indoor air afforded by three ventilation and air cleaning configurations in comparison to natural ventilation in three metropolitan areas. Estimation of indoor levels of ambient PM2.5 builds upon our study of whole-house clean air delivery rates measured in a fully instrumented test home and our modeling analysis of selected asthma triggers in indoor air (MacIntosh et al., 2008; Myatt et al., 2008). In addition, we estimate the public health benefits expected to result from a reduction in exposure to PM2.5 from use of high efficiency, whole-house air cleaning systems on a population scale. The results of this analysis provide a reasonable estimate of the potential for residential air cleaning systems to mitigate exposure to fine particles of ambient origin.

Methods

The analysis focused on counties that comprise the metropolitan areas of Cincinnati, Cleveland, and Columbus, Ohio, because there was detailed geographically resolved information available in these areas on single-family building stock, residential heating and cooling systems, ambient PM2.5, and meteorology. These data were combined to estimate concentrations of PM2.5 of outdoor origin inside of representative residential building stock in each county for every day in 2005. Three analyses were conducted for each combination of residence type and county: (1) natural ventilation; (2) forced air heating and cooling with conventional in-duct filtration, and (3) forced air heating and cooling with high-efficiency in-duct air cleaning. Results from the individual building types and counties were analyzed to identify the determinants of ambient PM2.5 in indoor air. Lastly, the potential public health benefits of reduced exposure to ambient PM2.5 provided by residential air cleaning were calculated following a standard methodology for health impact assessments. Details of the approach and implementation are presented below.

Indoor Air Quality Simulation

The CONTAM multi-zone indoor air quality model developed at the National Institute of Standards and Technology was used to estimate indoor concentrations of PM2.5 of ambient origin (Walton and Dols, 2006). CONTAM provides dynamic simulations of inter-zonal airflow, ventilation rates, and concentrations of gaseous and aerosol pollutants. The performance of the model has been evaluated extensively (Lansari et al., 1996; Emmerich and Nabinger, 2000; Emmerich et al., 2003; Howard-Reed et al., 2004, 2008). Of particular relevance to our application, Emmerich et al. (2003) and Emmerich and Nabinger, (2000) found that modeled average fine particle levels were within 30% of the measured values during a study of air cleaner performance in a single-zone test home. Principal inputs to the model include the structure of the residential building, pollutant levels in ambient air, meteorology, and sources and sinks of pollutants indoors.

Residential Buildings

Seven residential building templates developed for use in CONTAM were selected to represent the single-family building stock in the modeling domain. The templates were based on the US Census Bureau American Housing Survey (HUD, 1999) and the US Department of Energy Residential Energy Consumption Survey (US Department of Energy, 1999) and were intended to represent typical US residential building stock (Persily et al., 2006). As shown in Table 1, the templates include home volumes, structure, and leakage rates representative of four building eras – homes built prior to 1940, between 1940 and 1969, between 1970 and 1989, and after 1990 — for four types of detached homes and three types of attached homes. To allow for natural ventilation and leakage through and around windows, the templates were modified to include windows sized to 11.5% of the area of each wall (Enermodal, 2001). The proportion of homes in each census tract assigned to each of the seven templates was based on the age, size, construction (detached or attached home), and type of ventilation system (forced air or natural) of the homes as described for each of the respective counties by the American Housing Survey noted above.

Table 1 CONTAM Housing templates selected to represent single-family residential buildings in the modeling domain

Ambient Fine Particles

Twenty-four hour average levels of PM2.5 for 2005 measured in outdoor air at 34 monitoring locations distributed among the three metropolitan areas were obtained from the US EPA Air Quality System (AQS TTN, 2005) and used as the basis for the ambient PM2.5 exposure estimates. PM2.5 levels for days with missing data were estimated using an autoregressive model based on the concentration for the previous days. A map of the modeling domain and locations of the PM2.5 monitors is presented in Figure 1. A combination of universal kriging and inverse distance-weighting methods was used to interpolate the ambient PM2.5 concentration at each census tract based on the monitoring data (Jerrett et al., 2005). These interpolation methods were chosen for the combination of local detail afforded by inverse distance weighting and the ability of universal kriging to describe larger scale trends across the domain. The average of these two prediction surfaces was extracted at the census tract centroids. Finally, each county in the three metropolitan areas was assigned a population-weighted average of the PM2.5 concentration interpolated for each of its constituent census tracts.

Figure 1
figure1

Locations of PM2.5 monitors used to estimate daily concentrations of PM2.5 for each county.

In addition to the county-level analyses, we also evaluated the potential for ventilation and air cleaning to mitigate levels of short-term (hourly) PM2.5 levels inside of a hypothetical home (template DH-72) located in central Cincinnati. For this purpose, we relied upon 1-hour average PM2.5 levels in ambient air (EPA Monitor ID 39-061-0040-88101-1). We modeled indoor concentrations of ambient PM2.5 within that home for each hour in 2005 under each of the four ventilation–air cleaning configurations described previously. This analysis provides a preliminary assessment of the potential for indoor air cleaning systems to modify hourly and other short-term exposures to PM2.5 that have been associated with sub-clinical and intermediate markers of cardiopulmonary effects.

Ventilation

Dynamic air exchange rates are computed in CONTAM from simulations of force convection and radiant leakage based on properties of the building, meteorology, and window and door openings. The templates described above contain the leakage area for each of the residential buildings that we modeled. Hourly wind direction and speed, dry and wet bulb temperature, relative humidity, and cloud cover data for 2005 were obtained from the National Weather Service station at the principal airport for each metropolitan area. Using a temperature-based probabilistic approach based on data from an EPA analysis (Johnson, 2002), window and door opening schedules were generated that produced total ventilation rates for centrally and naturally ventilated periods consistent with corresponding air exchange rates determined from field campaigns reported elsewhere (Suh et al., 1992; Murray and Burmaster, 1995; Sarnat et al., 2006). During periods in which the windows were open, 40% of the total window area was assumed to be open. The air handling units (AHU) duty schedule and the window schedules were linked so that the AHU was never running when the windows were open. The front door was set to a schedule of opening for 15 min five times each day.

Homes with central forced air heating and cooling systems were assumed to have AHU balanced to provide 0.18 m3/min/m2 (0.6 cfm/ft2) of air to each room in the house. The duty schedule during heating and cooling periods was simulated with 1-h resolution based on output from the EnergyPlus Energy Simulation Software (US DOE, 2007). In general, the fraction of each hour devoted to forced air heating or cooling was proportional to the difference between ambient temperature and a set point of 22°C (72°F). Hourly duty schedules ranged from 4 min per hour during temperate periods to 38 min per hour during extreme summer periods and 52 min during extreme winter periods. The duty schedule and the window schedules were linked so that the AHU did not operate when the windows were open.

Air Cleaning

Two types of air cleaning systems were simulated for homes with central forced air heating and cooling. The first was a conventional air handler equipped with a standard 1-inch media filter. In this system, the fan operated only during periods of heating or cooling demand. The second was an air handler equipped with a variable speed fan and a high-efficiency electrostatic air cleaner with HEPA (high-efficiency particle arrestance) -like removal efficiency for aerosols. In this system, the fan operated at full speed during periods of heating and cooling demand and at half-speed during all other times. Homes without a central forced air system were assumed not to have any indoor air cleaning capacity such as portable air cleaners. PM2.5 removal efficiencies were based on measurements reported for a conventional 1-inch media filter (14%) and a high-efficiency electrostatic air cleaner (90%, Trane CleanEffects Model CAP591) as determined in an assessment of whole-house clean air delivery rates measured in a fully instrumented test home (MacIntosh et al., 2008).

In a subset of the analyses, the effect of a portable air cleaner with HEPA filtration sized for a room in addition to conventional filtration was simulated. A PM2.5 removal efficiency of 70% was used for the portable air cleaner based on results from studies conducted for the National Center of Energy Management and Building Technologies (Chen et al., 2006). The portable air cleaner was located in the center of the living room of the residence and operated at a flow rate of 5.6 m3/min (200 cubic feet per minute).

In addition to removal of fine particles by air cleaning, the rate of PM2.5 deposition to indoor surfaces was modeled as 0.5/h (Thatcher et al., 2002). The deposition rate was assumed to be independent of air exchange rate and AHU operation.

Determinants of PM2.5 of Ambient Origin Indoors

Twenty-four hour average PM2.5 of outdoor origin present in indoor air was estimated for each of seven single-family residential building types representative of housing stock in the selected metropolitan areas as described above. We performed a statistical analysis of output from the simulations to identify the determinants of PM2.5 of ambient origin indoors using statistical software (SAS Institute, Cary, NC, USA). Relationships of indoor PM2.5 of ambient origin with outdoor PM2.5, AHU-operating hours, and type of ventilation/air cleaner were estimated with a generalized linear model, defined by:

where ln (Y) is the natural log-transformed 24-h average concentration of indoor PM2.5 of ambient origin, Outdoor PM2.5 is the 24-h average outdoor PM2.5 concentration, AHU Hours is the number of hours that the air handling unit operated during the day, Ventilation is the ventilation configuration (natural, conventional filter, or high-efficiency electrostatic air cleaner), and ɛ is the within-subject measurement error. Controlling for house template did not change parameter estimates for the explanatory variables and this variable was therefore not included in the final model.

Health Impact Assessment

Following public health benefit-cost assessments for PM2.5 published previously, we calculated mortality and selected morbidity benefits associated with reduced exposure to PM2.5 of ambient origin (EPA, 1999, 2005; Levy and Spengler, 2002; Levy et al., 2003; Sultan, 2007). We focused the analysis on premature mortality and the following morbidity outcomes: hospital admissions for cardiovascular and respiratory disease, emergency room visits for asthma exacerbations, and asthma exacerbation. Inputs required for the present application include the size of the affected population, change in indoor concentrations of PM2.5 of ambient origin, average time spent at home, as well as PM2.5 concentration–response functions and baseline incidence rates for the selected health outcomes.

The metropolitan areas of the three cities contain over 5,000,000 residents distributed among 19 counties that comprise 20,877 km2 of urban and suburban areas. The age distribution of the population is similar to that of the United States with 7% less than 5 years, 58% greater than 30 years, and 12% over 65 years. The size of the affected population within the modeling domain was determined from the prevalence of single-family homes with central forced air systems in each county which ranged from 39% to 63% (US Census Bureau, 1998, 2000, 2002). The number of homes per template in each county was estimated from county-specific age of home information reported in the Year 2000 United States Census (US Census Bureau, 2000) and the American Housing Survey (US Census Bureau, 1998, 2000, 2002).

A critical issue within this analysis is the fact that PM2.5 concentration–response functions are derived from ambient monitoring data, but we are estimating changes to personal exposures. Directly applying concentration–response functions would therefore result in a mismatch of information and a systematic bias in outputs. Although simulating the distributions of personal exposures before and after our hypothetical intervention and interpreting these distributions in relation to the underlying personal exposure characteristics of the epidemiological studies is beyond the scope of this analysis, we make some first-order assumptions to allow us to reasonably approximate health benefits.

We assume that the epidemiological evidence described below involves populations with typical time-activity patterns (Echols et al., 1999; Klepeis et al., 2001) and ambient PM2.5 indoor/outdoor ratios (Wallace and Williams, 2005). Using these data, on average, 69% of time is spent at home, with 13% spent outdoors or in vehicles, 13% in other indoor settings or bars/restaurants, and 5% in offices or factories. We assume that the median indoor/outdoor ratio for sulfur reported by Wallace and Williams (2005) is applicable to all residences and other indoor settings and that exposures outdoors or in vehicles correspond with ambient levels. This would imply that a 1 μg/m3 change in ambient PM2.5 would correspond with an approximate 0.6 μg/m3 change in mean personal PM2.5. Thus, as we are simulating changes in mean personal PM2.5, we need to multiply the functions presented below by approximately 1.6 prior to applying them to our modeled outputs. This is clearly a simple approximation, but it captures the general insight that a 1 μg/m3 change in ambient PM2.5 corresponds with a smaller magnitude change in personal PM2.5.

We determined reasonable central estimates for ambient PM2.5 concentration–response functions following methods published previously (Levy et al., 2002). The concentration–response function estimates were generally based on inverse-variance weighted meta-analyses of the most relevant literature. For premature mortality, central estimates in the relevant literature range from a 0.6% to 1.7% increase in mortality per μg/m3 of annual average PM2.5 levels in ambient air (Dockery et al., 1993; Pope et al., 2002; Jerrett et al., 2005; Laden et al., 2006). A formal elicitation of 12 expert opinions on the magnitude of the concentration–response relationship between long-term PM2.5 exposures and mortality risks yielded a range of median estimates from 0.4% to 2.0%, with an average of the medians of 1% and median of the medians of 1.05% (Industrial Economics, 2006). In consideration of the relevant literature and the expert elicitation, we used a value of a 1% increase in mortality per μg/m3 of annual average PM2.5 as a reasonable central estimate from the current knowledge base.

For hospital admissions for cardiovascular causes, we relied on a recent meta-analysis (COMEAP, 2006) which combined 51 published studies to determine that cardiovascular hospital admissions increased by an estimated 0.9% per 10 μg/m3 increase of PM10. This rate was converted to an estimated 0.16% increase in cardiovascular hospital admissions per μg/m3 increase of PM2.5 following a PM2.5/PM10 ratio developed from monitoring data reported by the EPA (EPA, 2004). For respiratory hospital admissions, we derived a concentration–response function of a 0.2% increase in respiratory hospital admissions per 1 μg/m3 increase of PM2.5 based on a meta-analysis of the most relevant literature (Thurston et al., 1994; Schwartz, 1995, 1996; Schwartz et al., 1996; Burnett et al., 1997; Wordley et al., 1997; Atkinson et al., 1999; Gwynn et al., 2000; Hagen et al., 2000; Anderson et al., 2001; Gwynn and Thurston, 2001). Similar to respiratory hospital admissions, we derived a concentration–response function of a 0.8% increase in asthma-related ER visits per μg/m3 increase of PM2.5 based on a meta-analysis of the most relevant literature (Schwartz et al., 1993; Lipsett et al., 1997; Norris et al., 1999; Tolbert et al., 2000; Peel et al., 2005).

For asthma exacerbations, we derived a concentration–response function of a 2% increase in asthma exacerbations (among asthmatics only) per μg/m3 increase of PM2.5 based on a meta-analysis of studies looking at asthma exacerbation as measured by aggregate symptoms, rather than specifically looking at cough, wheeze, or other defined outcomes (Delfino et al., 1998, 2002; Yu et al., 2000; Desqueyroux et al., 2002; Mortimer et al., 2002).

The baseline incidence rates for the health outcomes were developed from a variety of sources. For mortality, we calculated county-specific mortality incidence rates based on all non-accidental deaths (International Classification of Diseases (ICD). 10th Revision, group codes A00–R99) from the years 1999–2003 (CDC, 2007) and population data (US Census Bureau, 2000). Baseline incidence rates for the hospital admission outcomes and the asthma-related emergency room visits are based on data from the Year 2000 NCHS National Hospital Discharge Survey (NHDS) and the Year 2000 National Hospital Ambulatory Medical Care Survey, respectively. In the NHDS, hospitalization records for all respiratory (ICD, 9th Revision, 460–519) and all cardiovascular-related admissions (ICD, 9th Revision, 390–429) were identified. For asthma exacerbations, we used a baseline rate of 20.08 exacerbations per year per asthmatic (NCHS, 1999) and an estimated prevalence of asthmatics in the United States of 3.86% (American Lung Association, 2002).

Results

Annual and 24-h Average PM2.5 Exposure

We estimated population-weighted annual average ambient PM2.5 concentrations in 2005 for the Cincinnati, Columbus, and Cleveland metropolitan areas as 17.1 μg/m3, 16.4 μg/m3, and 16.2 μg/m3, respectively. County-specific annual average concentration estimates ranged from 15.3 μg/m3 for Delaware County near Columbus to 18.5 μg/m3 in Dearborn County, Indiana outside of Cincinnati.

Primary output from the CONTAM model includes air exchange rates (AER) and the ratios between indoor concentrations of PM2.5 of outdoor origin and outdoor concentrations (referred to henceforth as I/O ratios, which should not be taken to include indoor sources). The distributions of 24-h average AER and PM2.5 I/O ratio for homes with natural, conventional forced air, and high-efficiency forced air ventilation for the entire modeling domain are summarized in Table 2. The median AER for homes with natural ventilation was approximately two times as high as the AER in homes with forced air ventilation systems due to the increased use of windows during warm weather. The meteorology and temperature-dependent window schedule resulted in open windows for 17% of the hours in the year in the mechanically ventilated homes and 40% for homes with natural ventilation.

Table 2 Distribution of simulated 24-h average air exchange rate and ambient PM2.5 indoor–outdoor ratios for homes with and without forced air ventilation systems

I/O ratios of PM2.5 were predictably lower for homes with high-efficiency filtration when compared with homes with conventional or natural ventilation. Although not substantially different, stand-alone homes had slightly lower I/O ratios than attached homes (e.g., townhomes) of the same era, and newer homes had slightly lower I/O ratios than older homes. The AER and I/O ratios did not vary considerably among metropolitan areas or among housing templates, but as anticipated did vary by season. The influence of temperature-based ventilation patterns is illustrated in Figure 2 where I/O ratios are shown to be greatest in warm weather months.

Figure 2
figure2

Simulated monthly average indoor–outdoor ratios of PM2.5 for homes with natural ventilation and homes with central forced air heating and cooling equipment with conventional filtration and high-efficiency whole-house air cleaning. Symbols represent the mean of the monthly mean indoor–outdoor ratios for the seven housing templates. The error bars represent the housing type with the highest and lowest monthly mean indoor–outdoor ratio. Monthly average outdoor air temperatures (°C) across the three cities are shown below the horizontal axis.

The distributions of modeled 24-h average indoor PM2.5 concentrations for the three ventilation and air cleaning configurations are illustrated in Figure 3. When comparing all home types and locations, the median daily average PM2.5 in homes with high-efficiency filtration (1.5 μg/m3) was approximately one-third of the levels in homes with conventional filtration (5.3 μg/m3) and approximately one-fifth of concentrations in homes with natural ventilation (8.4 μg/m3). The average reduction in indoor levels of daily average PM2.5 of ambient origin afforded by the conventional filtration over natural ventilation was 3.7 μg/m3 (SD: 5.6 μg/m3). For high-efficiency air cleaning in comparison to conventional filtration, the average difference was 3.4 μg/m3 (SD: 1.8 μg/m3) and 24-h average differences ranged from 0.3 to 38.0 μg/m3 across both days and home types.

Figure 3
figure3

Daily PM2.5 concentrations for the three ventilation configurations (high efficiency, conventional, and natural). Whiskers represent the 10th and 90th percentiles, circles represent the 5th and 95th percentiles.

Cumulative distributions of modeled 24-h average indoor air concentrations of PM2.5 of outdoor origin reveal differences in upper bound exposures among the four ventilation–air cleaning configurations. As shown in Figure 4, there were 21 days where the PM2.5 levels were greater than 10 μg/m3 for the high efficiency filtration configuration. For the majority of the 21 days, the windows were open for all or nearly all of the days. In comparison, there were 44 days, 57 days, and 129 days where the modeled 24-h average PM2.5 concentration was above 10 μg/m3 for conventional plus portable HEPA configuration, the conventional filtration configuration, and the natural ventilation configuration, respectively. The plots in Figure 4 also show that the continuous use of a room-size portable air cleaner with HEPA filtration had little impact on the home-wide 24-h average indoor PM2.5 concentrations.

Figure 4
figure4

Cumulative percentiles of daily PM2.5 results for template 72 in Hamilton county, OH, comparing four ventilation configurations (High efficiency, conventional plus portable air cleaner, conventional alone, and natural).

Determinants of Indoor PM2.5

Analyses of the simulation results from CONTAM revealed that outdoor PM2.5, AHU-operating hours and ventilation configuration were all significant predictors of indoor PM2.5 concentrations (R2=0.60, P<0.0001) (Table 3). The ventilation–air cleaning configuration explained the greatest proportion of variability (partial R2=0.38). As expected, increased outdoor PM2.5 concentrations were associated with increased indoor PM2.5 concentrations, although increased AHU-operating time was associated with decreased indoor PM2.5. The effect of high-efficiency air cleaning on indoor PM2.5 concentrations was significantly greater than that of conventional filtration or natural ventilation. Significant interaction was also observed between the filter type and time the AHU operated, explaining an additional 2% of indoor PM2.5 concentration variability (not shown); the effect of the number of hours the AHU operated on indoor PM2.5 concentrations was modified by filter type, with high-efficiency filters reducing concentrations at a greater rate than the conventional filter.

Table 3 Multivariate regression model results for determinants of modeled indoor concentrations of PM2.5 of ambient origin (natural log-transformed)

Hourly PM2.5 Levels

Examination of results from the simulation of 1-h average PM2.5 for a hypothetical home in central Cincinnati demonstrates the mitigation of short-term ambient PM2.5 concentrations indoors. A typical week of results during which windows were both open and closed throughout the week, with the windows open more frequently for the natural ventilation configuration, is illustrated in Figure 5. When windows are open, indoor PM2.5 is approximately equal to ambient air concentrations for the natural ventilation and conventional air cleaning simulations and moderately below ambient levels when the whole house, high-efficiency air cleaning is used. Short-term levels of PM2.5 indoors are substantially lower when windows are closed compared with open, an effect that is enhanced substantially by concurrent whole house, high-efficiency air cleaning. The dampening of 1-h average concentrations associated with indoor air cleaning results in lower 24-h and annual average indoor air concentrations of PM2.5 of ambient origin.

Figure 5
figure5

Simulated hourly indoor PM2.5 levels for first week of September, 2005. Template 28 in Cincinnati, OH comparing three ventilation configurations (high efficiency, conventional, and natural).

Public Health Impacts

A reduction of annual average PM2.5 indoors of 3.4 μg/m3 in homes with high-efficiency air cleaning compared with conventional filtration translates to a reduction of 2.3 μg/m3 in annual average personal exposure to PM2.5 assuming an average time indoors of 16.6 h per day. Given the assumed ambient–personal relationship underlying the epidemiological evidence, this change in PM2.5 exposure on a long-term basis is associated with an approximate 3.7% decrease in the annual risk of mortality. Similarly, conversion from natural ventilation to forced air with conventional filtration and natural ventilation to forced air with high-efficiency air cleaning is associated with 4.2%, and 7.8% reductions in the annual risk of mortality, respectively.

For morbidity outcomes, a conversion to high efficiency electrostatic air cleaning from conventional filtration would result in an estimated decrease of 0.6% in the risk for cardiovascular hospital admissions, 0.8% for respiratory hospital admissions, 3.0% for emergency room visits for asthma, and 7.4% for asthma exacerbations among asthmatics. Changes in risk between other pairings of the ventilation and air-cleaning configurations can be inferred directly from the preceding results.

Information from the Year 2000 US Census and American Housing Survey (US Census Bureau, 1998, 2000, 2002) shows that the three metropolitan areas that we considered contain approximately 964,000 single family detached and 77,000 single family attached residential units with central forced air heating and cooling systems that can accommodate high efficiency, in-duct air cleaning systems that are marketed in the United States at this time. Approximately 2.7 million people occupy these residences. We are not aware of data on the prevalence of air cleaning methods or technologies employed in those homes. However, substantial public health benefits are estimated to result if all of those homes had conventional in-duct filtration and switched to high-efficiency in-duct air cleaning (Table 4).

Table 4 Estimated annual reduction in mortality and morbidity associated with reduced PM2.5 exposure as a result of converting between types of residential ventilation/filtration system

Discussion

This modeling analysis shows that exposure to PM2.5 of outdoor origin is modified substantially by residential ventilation and air-cleaning configurations. Reductions in indoor levels of PM2.5 infiltrated from outdoors are expected to be realized on an hourly, daily, and annual average basis. Estimated changes in health risks associated with PM2.5 exposure attained by forced air with conventional filtration over natural ventilation and forced air with high-efficiency air cleaning over other configurations could be substantial on a population scale.

Studies in which the performance of CONTAM has been evaluated demonstrate the accuracy of indoor air concentrations of particles and gases predicted with the model. In simulations of a single-story home, predicted concentrations of tracer gas were within 15% of corresponding measured values (Haghighat and Rao, 1996). Likewise, air exchange rates for a single room building predicted with CONTAM were within 5% of measured levels (Emmerich and Nabinger, 2000). In a tracer gas study conducted in a multi-room occupied townhouse, gas concentrations predicted by the model were within 25% of measured concentrations (Emmerich et al., 2003). Finally, measured and predicted 24-h average concentrations of 0.3 to 5 μm particles in a single room building were within 30% of each other (Emmerich and Nabinger, 2000).

I/O ratios of PM2.5 of outdoor origin simulated with CONTAM are comparable to values measured in field studies, where sulfur is typically measured as a surrogate for ambient PM2.5. For example, measured I/O ratios averaged 0.69 for homes with air conditioning and 0.86 for homes without air conditioning studied during the summer in Pennsylvania (Suh et al., 1992). Similarly, in a study of six homes in Massachusetts, five of which did not have central air conditioning, the average sulfur indoor–outdoor ratio was 0.72 over sampling periods in Spring/Summer and Fall/Winter (Sarnat et al., 2002). In our simulations the modeled I/O ratio in summer months ranged between 0.7 and 0.9 for naturally ventilated homes. In a study that included sampling periods in each of the four seasons, 24-h I/O ratios of sulfur ranged from 0.26 to 0.87 (Wallace and Williams, 2005). The 5th and 95th percentiles of daily I/O ratios produced from the simulations for naturally ventilated homes was 0.23 to 0.97, similar to results from the year-long study.

The simulations of AER and I/O ratios agreed well with relevant benchmark values despite the limitations of the data available for input to the indoor air quality model. For example, there is likely to be more variability in building stock, leakage rates and other determinants of AER in actual residences than is represented by the seven single-family home templates included in our modeling. Our assumption that all windows in a residence are open to 40% of their area simultaneously may also be unrealistic. This assumption lead to days with higher AER values, however, the air exchange rates for periods in which the windows were open and closed did not deviate far from values reported in the literature for similar situations. A study of indoor PM in homes in the Boston area reported a mean AER of over 4.7/h and a maximum one-hour AER of over 20/h for a home in which windows and doors were kept open for much of the six-day monitoring period (Long et al., 2000). Similarly, actual deposition rates of PM2.5 indoors due to settling and removal within forced air systems probably vary among and within residences in contrast to our simplifying assumption that the rates are constant for all locations and time. The reasonableness of I/O ratios calculated with the model may be due in part to our use of contemporaneous meteorological and ambient PM2.5 records along with temperature based duty schedule for utilization of forced air heating and cooling that should reflect correlations among these factors that influence the magnitude of PM2.5 levels indoors that result from infiltration.

Previous modeling studies have demonstrated the utility of advanced filtration for reducing exposure to PM2.5 in other areas of the world. Hänninen and co-workers used data on infiltration factors from the Exposures of Adult Urban Populations in Europe Study (EXPOLIS) as inputs into models to calculate an exposure reduction of ambient PM2.5 of 27% for residences of Helsinki, Finland when advanced filtration was employed (Hanninen et al., 2005). The authors of that study noted that if their results were extrapolated to all of Europe, the use of advanced filtration would result in an estimated reduction of 27,000–100,000 deaths per year. In a case study of indoor exposures in Singapore, adoption of forced air heating and cooling and higher efficiency in-duct filtration of schools, offices, and residences was estimated to result in a 6% to 10% reduction of PM10-related mortality and morbidity from current conditions (Sultan, 2007).

Our results are consistent with the previous studies in that reductions in exposure to ambient PM2.5 through use of high-efficiency, in-duct filtration as well as central air conditioning are associated with substantial public health benefits. When scaling results for individual residential building types to the population of metropolitan Cincinnati, Cleveland, and Columbus, we assumed that the entire population of single-family home dwellers with central air systems has conventional filtration. This assumption results in an overestimate of the potential public health benefits to the extent that homes in the area already have high-efficiency in-duct air cleaning. However, we are not aware of data on the prevalence of in-duct air-cleaning technologies required to support a more refined estimate.

Portable air cleaners could be more accessible to the general public than in-duct, high-efficiency systems because of cost, home ownership, and other considerations. However, our results suggest that reductions in population-based exposure to ambient PM2.5 as a result of widespread use of portable air cleaners would be modest. For instance, operation of a portable air cleaner along with conventional filtration system reduced the median 24-h average concentration of ambient PM2.5 indoors by 5% in comparison to conventional filtration alone (Figure 4). The effect of a portable air cleaner may be smaller in homes with natural ventilation as such homes tend to have relatively high ventilation rates in warm weather months due to opening of windows. In that case, high ventilation rates will make particle removal by a portable filtration unit a small component of the total loss rate for PM2.5.

For our concentration–response functions, we determined reasonable central estimates based on a thorough review of the relevant literature; however, it was beyond the scope of this analysis to quantify the uncertainties surrounding our estimates, in part because many important uncertainties extend beyond the statistical confidence intervals reported in the individual epidemiological studies. Our health impact assessment was based on the generally accepted no-threshold linear concentration–response model (NRC, 2002; EPA, 2005; Schwartz et al., 2008), therefore, additional scientific evidence, such as the existence of a population threshold, could significantly change our estimates of health impacts. Uncertainty could be introduced into our analysis if the concentration–response functions were derived from PM2.5 that differed substantially in compositon from the material for which we evaluated exposure. However, our analysis focused on exposure to ambient PM2.5, the same material from which the concentration–response functions upon which we relied were derived, thereby minimizing this potential uncertainty. More uncertainty could be related to our first-order approximation to convert the central site monitoring data from the published epidemiological studies to mean personal exposure values. However, ignoring this disconnect would have led to biased health impact estimates, and our first-order approximation appears reasonable. In addition, by focusing on single-family homes with the ability to accommodate high-efficiency in-duct air cleaning systems, we have presumably focused on higher-income households which have been shown previously to be at less risk of air pollution-mediated health effects than low-income populations (O'Neill et al., 2003). If the public health effects of PM2.5 are disproportionately found among those living in multi-family dwellings or in smaller homes that cannot accommodate these technologies, our benefits could be substantially overestimated. Finally, our analysis does not reflect changes in ambient air quality and corresponding health risks that would be associated with the generation of electricity required to operate fans or electronics required of portable and high-efficiency in-duct air cleaners. The estimate of mortality and morbidity cases avoided remain a reasonable point-estimate for comparing the effect of filtration. Additional research could use a probabilistic approach (e.g., Monte Carlo) to account for model uncertainties in a quantitative manner. Importantly, all of the filtration/ventilation scenarios were evaluated using the same assumptions, which allows for direct comparisons within this analysis.

The ambient PM2.5 concentrations were modeled using a combination of universal kriging and inverse distance-weighting methods developed by Jerrett et al. (2005) based on the data from the PM2.5-monitoring stations. The inclusion of major point sources (e.g., power plants) and area sources (e.g., highways), would likely increase the accuracy of the spatially resolved and modeled ambient PM2.5. However, our modeling demonstrates that both the type of air cleaning and the run time of the AHU significantly impact the levels of PM2.5 indoors independent of ambient PM2.5 levels. Therefore, a more accurate estimate of ambient PM2.5 would not change our findings on public health benefits. The modeling also shows that there is a 1 μg/m3 decrease in indoor PM2.5 for every hour of AHU operation (β=−0.06; natural log scale). Even greater reductions in ambient PM2.5 exposure and health risks may be achieved in regions less temperate than Ohio where air conditioners may be operated more frequently. Population-based epidemiological studies may benefit from the use of models similar to those described here to estimate personal PM exposure with greater accuracy.

The magnitude of the benefits associated with the use of high-efficiency filtration systems can be contextualized by comparing these benefits with those due to source control. For example, an analysis of reductions in emissions due to the application of Best Available Control Technology for five older power plants in the Washington, D.C area yielded a reduction of ambient PM2.5 ranging from 0.009 to 0.9 μg/m3 (Levy et al., 2002). In another study of two coal-fired power plants in Massachusetts, the addition of advanced control technologies yielded a 0.006 to 0.2 μg/m3 reduction of annual average PM2.5 (Levy and Spengler, 2002). By comparison, adoption of whole-house high-efficiency in-duct air cleaning is anticipated to yield more than a 15-fold larger reduction in exposure to PM2.5. However, the number of people affected by source control can be much larger than the number of people affected by citywide application of high-efficiency in-duct air cleaning. Because of this fact, population scale heath benefits from the citywide institution of in-duct air cleaning may be comparable to regional benefits from emission control technologies. In addition, source controls will provide benefits to low-income populations who may be more susceptible and may not be candidates for air cleaning systems. Source controls could result in other benefits as well, such as reductions in air pollutants that influence acid precipitation and visibility, More comprehensive analyses considering the marginal costs and marginal benefits of source controls versus in-duct air cleaning would be warranted, but this comparison suggests that high-efficiency whole-house particle removal in residences will substantially reduce the public health burden of PM2.5-related air pollution in addition to expansion of emission controls on fossil fuel-fired stationary and mobile sources.

In summary, this modeling analysis builds upon a robust body of empirical studies of fine particle levels in ambient air, infiltration into residences, personal exposure, and resulting health effects to estimate changes in indoor levels of PM2.5 of outdoor origin associated with several common ventilation and air-cleaning configurations. Our detailed assessment of housing and air pollution in three major cities of Ohio indicates that adoption of high efficiency, in-duct, air cleaning would result in meaningful reductions in exposure to PM2.5 of outdoor origin on short- and long-time scales. If our results were extrapolated to residences with forced air ventilation systems throughout the United States, the potential health benefits of reductions in PM2.5-related exposure would be substantial. Appropriately sized, high-efficiency whole-house air-cleaning systems appear to be an important tool for mitigating the public health burden of ambient particle pollution among a subset of the population.

References

  1. American Lung Association. Trends in asthma morbidity and mortality, American Lung Association, Best Practices and Programs Services, Epidemiological and Statistics Unit, 2002.

  2. Anderson H.R., Bremner S.A., Atkinson R.W., Harrison R.M., and Walters S. Particulate matter and daily mortality and hospital admissions in the west midlands conurbation of the United Kingdom: associations with fine and coarse particles, black smoke and sulphate. Occup Environ Med 2001: 58 (8): 504–510.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  3. AQS TTN. 2005 US EPA Air Quality System Technology Transfer Network. Retrieved July 2007, from http://www.epa.gov/ttn/airs/airsaqs.

  4. Atkinson R.W., Bremner S.A., Anderson H.R., Strachan D.P., Bland J.M., and de Leon A.P. Short-term associations between emergency hospital admissions for respiratory and cardiovascular disease and outdoor air pollution in London. Arch Environ Health 1999: 54 (6): 398–411.

    CAS  PubMed  Article  Google Scholar 

  5. Brauner E.V., Forchhammer L., Moller P., Barregard L., Gunnarsen L., Afshari A., et al. Indoor particles affect vascular function in the aged: an air filtration-based intervention study. Am J Respir Crit Care Med 2008: 177 (4): 419–425.

    PubMed  Article  Google Scholar 

  6. Burnett R.T., Cakmak S., Brook J.R., and Krewski D. The role of particulate size and chemistry in the association between summertime ambient air pollution and hospitalization for cardiorespiratory diseases. Environ Health Perspect 1997: 105 (6): 614–620.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  7. CDC. CDC WONDER, Centers for Disease Control and Prevention, 2007.

  8. Chen W., Gao Z., Zhang J., Kosar D., Walker C., and Novosel D. Reduced energy use through reduced indoor contamination in residential settings, Alexandria VA, National Center for Energy Management and Building Technologies, 2006.

  9. COMEAP. Cardiovascular Disease and Air Pollution. United Kingdom, Department of Health, Committee on the Medical Effects of Air Pollutants, 2006.

  10. Delfino R.J., Zeiger R.S., Seltzer J.M., and Street D.H. Symptoms in pediatric asthmatics and air pollution: differences in effects by symptom severity, anti-inflammatory medication use and particulate averaging time. Environ Health Perspect 1998: 106 (11): 751–761.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  11. Delfino R.J., Zeiger R.S., Seltzer J.M., Street D.H., and McLaren C.E. Association of asthma symptoms with peak particulate air pollution and effect modification by anti-inflammatory medication use. Environ Health Perspect 2002: 110 (10): A607–A617.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  12. Desqueyroux H., Pujet J.C., Prosper M., Squinazi F., and Momas I. Short-term effects of low-level air pollution on respiratory health of adults suffering from moderate to severe asthma. Environ Res 2002: 89 (1): 29–37.

    CAS  PubMed  Article  Google Scholar 

  13. Dockery D.W., Pope C.A. III, Xu X., Spengler J.D., Ware J.H., Fay M.E., et al. An association between air pollution and mortality in six US cities. N Engl J Med 1993: 329 (24): 1753–1759.

    CAS  PubMed  Article  Google Scholar 

  14. Echols S.L., MacIntosh D.L., Hammerstrom K.A., and Ryan P.B. Temporal variability of microenvironmental time budgets in Maryland. J Exp Analy Environ Epidemiol 1999: 9 (5): 502–512.

    CAS  Article  Google Scholar 

  15. Emmerich S., and Nabinger S. Measurement and Simulation of the IAQ Impact of Particle Air Cleaners in a Single-Zone Building. National Institute of Standards and Technology, Gaithersberg MD, 2000.

    Google Scholar 

  16. Emmerich S.J., Nabinger S., Gupte A., Howard-Reed C., and Wallace L. Comparison of Measured and Predicted Tracer Gas Concentrations in a Townhouse. National Institute of Standards and Technology, Gaithersburg MD, 2003.

    Google Scholar 

  17. Enermodal. Characterization of Framing Factors for Low-Rise Residential Building Envelopes., final report prepared for ASHRAE, Atlanta, GA (USA), 2001.

  18. EPA. The Benefits and Costs of the Clean Air Act: 1990 to 2010. Office of Air and Radiation, Washington, D.C., 1999.

  19. EPA. Air Quality Criteria for Particulate Matter (October 2004). US Environmental Protection Agency, Washington, D.C., 2004.

  20. EPA. Regulatory Impact Analysis for the Final Clean Air Interstate Rule. US Environmental Protection Agency, Office of Air and Radiation, Washington, DC, 2005.

  21. Franklin M., Zeka A., and Schwartz J. Association between PM2.5 and all-cause and specific-cause mortality in 27 US communities. J Expo Sci Environ Epidemiol 2007: 17 (3): 279–287.

    CAS  PubMed  Article  Google Scholar 

  22. Gwynn R.C., Burnett R.T., and Thurston G.D. A time-series analysis of acidic particulate matter and daily mortality and morbidity in the Buffalo, New York, region. Environ Health Perspect 2000: 108 (2): 125–133.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  23. Gwynn R.C., and Thurston G.D. The burden of air pollution: impacts among racial minorities. Environ Health Perspect 2001: 109 (Suppl 4): 501–506.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  24. Hagen J.A., Nafstad P., Skrondal A., Bjorkly S., and Magnus P. Associations between outdoor air pollutants and hospitalization for respiratory diseases. Epidemiology 2000: 11 (2): 136–140.

    CAS  PubMed  Article  Google Scholar 

  25. Haghighat F., and Rao J. A comprehesive validation of two airflow models - COMIS and CONTAM. Indoor Air 1996: 6 (4): 278–288.

    CAS  Article  Google Scholar 

  26. Hanninen O.O., Palonen J., Tuomisto J.T., Yli-Tuomi T., Seppanen O., and Jantunen M.J. Reduction potential of urban PM2.5 mortality risk using modern ventilation systems in buildings. Indoor Air 2005: 15 (4): 246–256.

    CAS  PubMed  Article  Google Scholar 

  27. Howard-Reed C., Nabinger S., and Emmerich S.J. Predicting the Performance of Non-Industrial Gaseous Air Cleaners: Measurements and Model Simulations from a Pilot Study. National Institute of Standards and Technology, Gaithersburg MD, 2004.

    Google Scholar 

  28. Howard-Reed C., Nabinger S., and Emmerich S.J. Characterizing gaseous air cleaner performance in the field. Build Environ 2008: 43 (3): 368–377.

    Article  Google Scholar 

  29. HUD. The American Housing Survey for the United States. US Department of Housing and Urban Development, US Department of Commerce, Washington, DC, 1999.

  30. Industrial Economics. Expanded Expert Judgment Assessment of the Concentration–Response Relationship Between PM2.5 Exposure and Mortality. Prepared for Office of Air Quality Planning and Standards, US Environmental Protection Agency, Cambridge, MA, 2006.

  31. Janssen N.A., Schwartz J., Zanobetti A., and Suh H.H. Air conditioning and source-specific particles as modifiers of the effect of PM(10) on hospital admissions for heart and lung disease. Environ Health Perspect 2002: 110 (1): 43–49.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  32. Jerrett M., Burnett R.T., Ma R., Pope C.A. III, Krewski D., Newbold K.B., et al. Spatial analysis of air pollution and mortality in Los Angeles. Epidemiology 2005: 16 (6): 727–736.

    PubMed  Article  Google Scholar 

  33. Johnson T A Guide to Selected Algorithms, Distributions, and Databases Used in Exposure Models Developed by The Office Of Air Quality Planning and Standards. US Environmental Protection Agency, Office of Research and Development, Research Triangle Park NC, 2002.

    Google Scholar 

  34. Klepeis N.E., Nelson W.C., Ott W.R., Robinson J.P., Tsang A.M., Switzer P., et al. The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. J Exp Anal Environ Epidemiol 2001: 11 (3): 231–252.

    CAS  Article  Google Scholar 

  35. Laden F., Schwartz J., Speizer F.E., and Dockery D.W. Reduction in fine particulate air pollution and mortality: extended follow-up of the Harvard Six Cities Study. Am J Respir Crit Care Med 2006: 173 (6): 667–672.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  36. Lansari A., Streicher J., Huber A., Crescenti G., Zweidinger R., Duncan J., et al. Dispersion of automotive alternative fuel vapors within a residence and its attached garage. Indoor Air 1996: 6 (2): 118–126.

    CAS  Article  Google Scholar 

  37. Leaderer B.P., Naeher L., Jankun T., Balenger K., Holford T.R., Toth C., et al. Indoor, outdoor, and regional summer and winter concentrations of PM10, PM2.5, SO4(2)-, H+, NH4+, NO3-, NH3, and nitrous acid in homes with and without kerosene space heaters. Environ Health Perspect 1999: 107 (3): 223–231.

    CAS  PubMed  PubMed Central  Google Scholar 

  38. Levy J., and Spengler J. Modeling the benefits of power plant emission controls in Massachusetts. J Air Waste Manag Assoc 2002: 52: 5–18.

    CAS  PubMed  Article  Google Scholar 

  39. Levy J.I., Greco S.L., and Spengler J.D. The importance of population susceptibility for air pollution risk assessment: a case study of power plants near Washington DC. Environ Health Perspect 2002: 110 (12): 1253–1260.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  40. Levy J.I., Nishioka Y., and Spengler J.D. The public health benefits of insulation retrofits in existing housing in the United States. Environ Health 2003: 2 (1): 4.

    PubMed  PubMed Central  Article  Google Scholar 

  41. Lipsett M., Hurley S., and Ostro B. Air pollution and emergency room visits for asthma in Santa Clara County, California. Environ Health Perspect 1997: 105 (2): 216–222.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  42. Long C.M., Suh H.H., and Koutrakis P. Characterization of indoor particle sources using continuous mass and size monitors. J Air Waste Manag Assoc 2000: 50 (7): 1236–1250.

    CAS  PubMed  Article  Google Scholar 

  43. MacIntosh D.L., Myatt T., Ludwig J., Baker B.J., Suh H., and Spenger J.D. Whole house particle removal and clean air delivery rates for in-duct and portable ventilation systems. J Air Waste Manag Assoc 2008: 58 (11): 1474–1482 Accepted for Publication.

    CAS  PubMed  Article  Google Scholar 

  44. Mortimer K.M., Neas L.M., Dockery D.W., Redline S., and Tager I.B. The effect of air pollution on inner-city children with asthma. Eur Respir J 2002: 19 (4): 699–705.

    CAS  PubMed  Article  Google Scholar 

  45. Murray D.M., and Burmaster D.E. Residential air exchange rates in the United States: empirical and estimated parametric distributions by season and climate region. Risk Analysis 1995: 15 (4): 459–465.

    Article  Google Scholar 

  46. Myatt T.A., Minigishi T., Allen J.G., and MacIntosh D.L. Control of asthma triggers in indoor air with air cleaners: a modeling analysis. Environ Health 2008: 7: 43.

    PubMed  PubMed Central  Article  Google Scholar 

  47. NCHS. National Health Interview Survey 1999. National Center for Health Statistics, Atlanta, GA, 1999.

  48. Norris G., Young Pong S.N., Koenig J.Q., Larson T.V., Sheppard L., and Stout J.W. An association between fine particles and asthma emergency department visits for children in Seattle. Environ Health Perspect 1999: 107 (6): 489–493.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  49. NRC. Estimating the Public Health Benefits of Proposed Air Pollution Regulations. National Research Council, Washington, D.C., 2002.

  50. O'Neill M.S., Jerrett M., Kawachi I., Levy J.I., Cohen A.J., Gouveia N., et al. Health, wealth, and air pollution: advancing theory and methods. Environ Health Perspect 2003: 111 (16): 1861–1870.

    PubMed  PubMed Central  Article  Google Scholar 

  51. Peel J.L., Tolbert P.E., Klein M., Metzger K.B., Flanders W.D., and Todd K., et al. Ambient air pollution and respiratory emergency department visits. Epidemiology 2005: 16 (2): 164–174.

    PubMed  Article  Google Scholar 

  52. Persily A., Musser A., and Leber D. A Collection of Homes to Represent US Housing Stocks. National Institute of Standards and Technology, Gaithersberg MD, 2006.

    Google Scholar 

  53. Pope C.A. III, Burnett R.T., Thun M.J., Calle E.E., Krewski D., Ito K., et al. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 2002: 287 (9): 1132–1141.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  54. Pope C.A. III, and Dockery D.W. Health effects of fine particulate air pollution: lines that connect. J Air Waste Manag Assoc 2006: 56 (6): 709–742.

    CAS  PubMed  Article  Google Scholar 

  55. Rodes C.E., Lawless P.A., Evans G.F., Sheldon L.S., Williams R.W., Vette A.F., et al. The relationships between personal PM exposures for elderly populations and indoor and outdoor concentrations for three retirement center scenarios. J Expo Anal Environ Epidemiol 2001: 11 (2): 103–115.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  56. Sarnat J.A., Long C.M., Koutrakis P., Coull B.A., Schwartz J., and Suh H.H. Using sulfur as a tracer of outdoor fine particulate matter. Environ Sci Technol 2002: 36 (24): 5305–5314.

    CAS  PubMed  Article  Google Scholar 

  57. Sarnat S.E., Coull B.A., Schwartz J., Gold D.R., and Suh H.H. Factors affecting the association between ambient concentrations and personal exposures to particles and gases. Environ Health Perspect 2006: 114 (5): 649–654.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  58. Schwartz J Short term fluctuations in air pollution and hospital admissions of the elderly for respiratory disease. Thorax 1995: 50 (5): 531–538.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  59. Schwartz J Air pollution and hospital admissions for respiratory disease. Epidemiology 1996: 7 (1): 20–28.

    CAS  PubMed  Article  Google Scholar 

  60. Schwartz J., Coull B., Laden F., and Ryan L. The effect of dose and timing of dose on the association between airborne particles and survival. Environ Health Perspect 2008: 116 (1): 64–69.

    PubMed  Article  Google Scholar 

  61. Schwartz J., Slater D., Larson T.V., Pierson W.E., and Koenig J.Q. Particulate air pollution and hospital emergency room visits for asthma in Seattle. Am Rev Respir Dis 1993: 147 (4): 826–831.

    CAS  PubMed  Article  Google Scholar 

  62. Schwartz J., Spix C., Touloumi G., Bacharova L., Barumamdzadeh T., le Tertre A., et al. Methodological issues in studies of air pollution and daily counts of deaths or hospital admissions. J Epidemiol Community Health 1996: 50 (Suppl 1): S3–S11.

    PubMed  PubMed Central  Article  Google Scholar 

  63. Shaughnessy R.J., and Sextro R.G. What is an effective portable air cleaning device? A review. J Occup Environ Hyg 2006: 3 (4): 169–181; quiz D45.

    CAS  PubMed  Article  Google Scholar 

  64. Suh H., Spenger J.D., and Koutrakis P. Personal exposures to acid aerosols and ammonia. Environ Sci Technol 1992: 26 (12): 2507–2517.

    CAS  Article  Google Scholar 

  65. Sultan Z.M. Estimates of associated outdoor particulate matter health risk and costs reductions from alternative building, ventilation and filtration scenarios. Sci Total Environ 2007: 377 (1): 1–11.

    CAS  PubMed  Article  Google Scholar 

  66. Thatcher T.L., Lai A.C.K., Moreno-Jackson M., Sextro R.G., and Nazaroff W.W. Effects of room furnishings and air speed on particle deposition rates indoors. Atm Environ 2002: 36 (11): 1811–1819.

    CAS  Article  Google Scholar 

  67. Thurston G.D., Ito K., Hayes C.G., Bates D.V., and Lippmann M. Respiratory hospital admissions and summertime haze air pollution in Toronto, Ontario: consideration of the role of acid aerosols. Environ Res 1994: 65 (2): 271–290.

    CAS  PubMed  Article  Google Scholar 

  68. Tolbert P.E., Mulholland J.A., MacIntosh D.L., Xu F., Daniels D., and Devine O.J., et al. Air quality and pediatric emergency room visits for asthma in Atlanta, Georgia USA. Am J Epidemiol 2000: 151 (8): 798–810.

    CAS  PubMed  Article  Google Scholar 

  69. US Census Bureau. 2000 “Year 2000 Census.” 2007, from http://factfinder.census.gov/home/saff/main.html?_lang=en.

  70. US Census Bureau. Current Housing Reports, Series H170/02-25. American Housing Survey for the Columbus Metropolitan Area, 2002.

  71. US Census Bureau. Current Housing Reports, Series H170/98-23. American Housing Survey for the Cincinnati Metropolitan Area, 1998.

  72. US Department of Energy. A Look at Residential Energy Consumption in 1997 1999.

  73. US DOE. EnergyPlus Version 2.1.0., 2007.

  74. Wallace L., and Williams R. Use of personal-indoor-outdoor sulfur concentrations to estimate the infiltration factor and outdoor exposure factor for individual homes and persons. Environ Sci Technol 2005: 39 (6): 1707–1714.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  75. Walton G., and Dols W.S. CONTAM 2.1 Supplemental user guide and program documentation, National Institute of Standards and Technology, Gaithersberg, MD, 2006.

  76. Wordley J., Walters S., and Ayres J.G. Short term variations in hospital admissions and mortality and particulate air pollution. Occup Environ Med 1997: 54 (2): 108–116.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  77. Yu O., Sheppard L., Lumley T., Koenig J.Q., and Shapiro G.G. Effects of ambient air pollution on symptoms of asthma in Seattle-area children enrolled in the CAMP study. Environ Health Perspect 2000: 108 (12): 1209–1214.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

Download references

Acknowledgements

Funding for this research was provided by Trane Residential Systems Inc., Tyler, Texas and Environmental Health and Engineering Inc., Needham, Massachusetts. We thank Helen H. Suh for her review of the design and inputs to this analysis.

Author information

Affiliations

Authors

Corresponding author

Correspondence to David L MacIntosh.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

MacIntosh, D., Minegishi, T., Kaufman, M. et al. The benefits of whole-house in-duct air cleaning in reducing exposures to fine particulate matter of outdoor origin: A modeling analysis. J Expo Sci Environ Epidemiol 20, 213–224 (2010). https://doi.org/10.1038/jes.2009.16

Download citation

Keywords

  • indoor air
  • air cleaners
  • filtration
  • PM2.5

Further reading

Search

Quick links