Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
10 IMPACT OF VENTILATION AND AIR CLEANING ON AS th ma Indoor exposures to pollutants associated with the incidence or symptoms of asthma are affected by many aspects of building design, maintenance, and operation. Building features modify the indoor sources of pollutants, the rates of pollutant entry from out- doors, and the rates of pollutant removal from indoors. Building ventilation and air cleaning are the two primary processes used intentionally within buildings to remove pollutants from the in- door air and maintain acceptable indoor environmental condi- tions. This chapter provides an overview of the relationship of building ventilation and particle air cleaning to exposures to in- door-generated pollutants that are associated with asthma. The findings from experimental assessments of the effects of air clean- ing on allergy and asthma symptoms are also summarized. Be- cause the association of asthma with pollutants from outdoor air is not a primary focus of this report, even though the exposures may occur primarily indoors, the dependence of these exposures on ventilation and air cleaning is not addressed in this chapter. THEORETICAL BACKGROUND This section provides a very brief overview of theoretical con- siderations that are necessary to understand the influence of building ventilation and air cleaning on indoor pollutant concen 327
328 CLEARING THE AIR "rations. Emphasis is placed on indoor particles because the in- door-generated pollutants most clearly associated with asthma are particles. Appendix A provides a more detailed technical dis- cussion of this topic along with the equations and parameter val- ues used for the theoretical predictions later in this chapter. From conservation of mass, the steady-state indoor air con- centration~ of a pollutant that is emitted indoors and absent from outdoor air equals the indoor pollutant generation rate divided by the sum of all pollutant removal rates. In the present context, the most important pollutant removal processes are (1) ventila- tion (i.e., the flow of indoor air containing pollutants to outdoors); (2) pollutant depositional losses on indoor surfaces; and (3) air cleaning (i.e., intentional removal of pollutants from indoor air by air filters and other types of air cleaners). The influence of changes in ventilation or air-cleaning rates on the indoor pollut- ant concentration depends on the magnitude of the other two pol- lutant removal processes. Many of the indoor-generated pollutants important for asthma are particles with diameters ranging from a fraction of a micrometer (1 ,um equals one-millionth of a meter) to approxi- mately 20 ,um. Table 10-1 provides information of the sizes (aero- dynamic diameters)2 of these particles. The available data are lim- ited and sometimes contradictory. Many of the bioaerosols associated with asthma, particularly dust mite allergens, whole pollens, cockroach allergen, and many fungal spores are large par- ticles greater than a few micrometers in diameter. There are con- tradictions among available data on the size of particles with cat allergen; however, a significant fraction of airborne cat allergen appears to be associated with particles smaller than a few mi- crometers. Environmental tobacco smoke is composed almost en- tirely of submicron-size particles (i.e., particles smaller than 1 rim). Droplet nuclei from coughs and sneezes, which often contain vi- rus, are included in Table 10-1 because viral infections are strongly iFor this discussion, we have assumed perfect mixing of the indoor air. See Appendix A for more information. 2Except that the sources of the data for pollens and fungal spores do not indi- cate whether the sizes are physical or aerodynamic diameters.
329 Cal In ~5 Cal . _ Cal o cn cn cn Cal . _ Cal o cn o . _ Q . _ cn . _ C] . _ CD ~5 o Q tr to - m U' C' Or ~7 o Q Or U' Cal Or .N CD .O Cal o Q O) Q a) _ ._~ ~ ~ ~t O) U'~ ~ _ ~ ~ ~ it - ~ o - ~ · ~C/~ Us ~ ~ Us Cry oo - Q (D C,,) ~ TIC ~ .= ~- a' _ ~cn s ~ ~ ~ . =, ~o ~ o =7 ~CD ° s~ ~tc: ~ c, ti, O E '~ ' v E s W c, ~=' ° ° ~ ~ E ~, ~ v 5 E = e E s, E E _ o o ~D ~ V ae :~ V ae E ~ ~ ~ co ae . CO ~C~ ~1 ~1 1 U' _ ~ llJ U' _ =7 ~C~ o U' U' U' ° ° o o ~C ~ ~ E ° o ~`,, E ~ c ~ ._ Q ~O C ~O ~0 C] ~C~ C ~C~ 113 C]
330 linked to exacerbation of asthma, at least in children Johnston et al., 1995~. There is evidence that rates of building ventilation and occupant density modify the rates of respiratory illness experi- enced by building occupants (Fisk, 1999; Fisk and Rosenfeld, 1997), presumably by changing exposures to infectious droplet nuclei. Data on the size distribution of droplet nuclei are ex- tremely limited and the methods employed to obtain the data may have resulted in an undercounting of the larger particles. The available data indicate that most of these particles are submicron in size but most of the particle volume is associated with particles larger than 1 ,um. It is not clear whether the number concentration or volume concentration of infectious droplet nuclei is more rel- evant for disease transmission. The magnitude of two of the particle removal processes- deposition on surfaces and air cleaning can vary dramatically with particle size. Particles deposit on indoor surfaces when in- door air motion, gravitational settling, electrostatic forces, and other phenomena cause them to collide with indoor surfaces. For particles, larger than a few micrometers in diameter, depositional losses are dominated by rates of gravitational settling. A 20-,um particle falls a distance of 1 m in about 80 seconds so it remains suspended indoors for only a short period. The deposition losses of such large particles tend to overwhelm normal rates of particle removal by ventilation or air cleaning. In contrast, a 0.2-,um par- ticle falls a distance of 1 m in about five days. The rate of deposi- tional removal of 0.2-,um particle from the indoor air, which is controlled by the indoor air motion, indoor surface roughness, and other factors, is almost a factor of 100 lower than the rate of depositional removal of a 20-,um particle. Some gaseous pollutants such as nitrogen dioxide and ozone are also removed from indoor air at a significant rate by deposi- tion (often called sorption) on or reaction with indoor surfaces. Rates of depositional removal depend on the chemical nature of the pollutant, the intensity of indoor air motion, and other fac- tors. Gravitational settling is unimportant for gaseous pollutants. Particles deposited on indoor surfaces can be resuspended in indoor air when the surfaces are disturbed by human activities (e.g., walking, vacuuming) or by high air velocities (e.g., air exit- ing a fan). Theory (Hinds, 1982) and limited empirical data
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 331 (Thatcher and Layton, 1995) indicate that resuspension occurs predominantly for particles larger than approximately 2 ,um. Based on our current knowledge of the behavior of particles, ex- posures to some of the larger particles associated with asthma may be substantially influenced by the localized resuspension of particles that results from occupant activities. BUILDING VENTILATION Background In this document, the term "ventilation" refers to the flow of outside air indoors, which is always accompanied by an equal flow of indoor air outdoors. Ventilation removes and dilutes in- door airborne pollutants, brings outdoor air pollutants into build- ings, and often removes or supplies heat and water vapor. Venti- lation is also needed to maintain oxygen concentrations inside buildings, although the quantity of ventilation needed to supply oxygen is very small relative to other ventilation requirements. Increasing the rate of ventilation generally leads to overall improvements in indoor air quality; however, the indoor concen- trations of some pollutants from outdoors, such as outdoor par- ticles and ozone, can increase with the ventilation rate. Indoor humidity can increase or decrease with ventilation rate. When it is cold and dry outdoors, increased ventilation usually reduces the indoor humidity. While increased ventilation rates are usually considered ben- eficial for health and for improving perceived air quality (e.g., odors), ventilation air must often be heated (and sometimes hu- midified) or cooled and dehumidified. Consequently, the ventila- tion rates selected for buildings must strike a balance between the benefits of energy savings with reduced ventilation and the known or suspected benefits to health with increased ventilation. Several metrics are used to specify the rates of building venti- lation. Generally, these metrics are flow rates of outside air nor- malized by the number of occupants, floor area, or indoor vol- ume. Corresponding units of ventilation rates include the following: liters per second per person (L so per person); liters
332 CLEARING THE AIR per second per square meter of floor area (L so per square meter); and air changes per hour (h-~. Municipalities typically adopt one of the several building de- sign codes used in the United States. These codes, or state energy codes, include building design provisions intended to maintain ventilation rates above a minimum rate that varies with the build- ing type. The American Society of Heating, Refrigerating, and Air Conditioning Engineers (ASHRAE) publishes a minimum venti- lation standard that is the basis for the ventilation specifications in many codes. The current version of the ASHRAE standard is Standard 62-1999 Ventilation for Acceptable Indoor Air Quality (ASHRAE, 1999~. Standard 62-1999 lists 0.35 ho as a minimum ventilation rate in residences,3 10 L so per person (20 cubic feet per minute [cfm] per person) as a minimum ventilation rate in offices, and 8 L so per person (15 cfm per person) as a minimum ventilation rate in schools. Due to a paucity of scientific data on the relationship of building ventilation rates with the health and well-being of occupants (Seppanen et al., 1999), the minimum ventilation rates in the ASHRAE standard are based substantially on professional judgment and on studies performed in laborato- ries with conditions quite different from those encountered in real buildings. Building design codes and ASHRAE's minimum ventilation standard do not ensure that all buildings maintain the specified minimum ventilation rates. In most states and municipalities, there are no legal requirements to actually maintain ventilation rates at or above the levels in building design codes. Additionally, building ventilation rates are difficult to measure accurately, in- frequently measured, and as discussed later, poorly controlled. Ventilation systems, although intended to remove indoor pol- Jutants, can also become sources of pollutants. Portions of venti- ration systems, particularly components that become wet, can be- come colonized by microorganisms and produce bioaerosols that 3Standard 62-1999 states that the 0.35 h-i of ventilation is normally satisfied by infiltration and natural ventilation but includes no technical specifications for the building to ensure that this ventilation rate is met continuously or on average. Standard 62-1999 also specifies installed mechanical exhaust capacities of 50 L s-i (100 cfm) per kitchen and 25 L s-i (50 cfm) per bathroom.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 333 are transported by the airflow to the occupied space. In addition, particles, fibers, and odorous and potentially irritating volatile organic compounds (VOCs) may be emitted from synthetic mate- rials, including fibrous insulation materials, from residual oils used in component production, from deposited dusts, and from microorganisms. Ventilation also affects the indoor humidity which in turn influences the growth or survival of microorgan- isms within buildings. Heating, ventilating, and air conditioning (HVAC) is the more general process of thermally conditioning and ventilating build- ings. In commercial buildings, these functions are usually inte- grated. The HVAC process employed in commercial buildings is reviewed here because HVAC features may influence exposure to pollutants that are known or thought to be associated with asthma. Methods and Rates of Ventilation in U.S. Single-Family Residences Diamond (1999) has summarized many of the basic physical characteristics of the U.S. residential building stock. In 1997, de- tached single-family units and row houses constituted 73°/O of the U.S. housing stock, 6% of the housing stock was mobile homes, and the remainder was apartments. The average heated floor space in all U.S. housing stock was 181 m2 (1,950 square feet) and air conditioning was installed in 70°/O of these dwellings. The av- erage conditioned floor area of mobile homes was 87 m2 (940 square feet) and 70°/O of mobile homes had air conditioners. Four- teen percent of all housing units used humidifiers, and nine per- cent had dehumidifiers. When windows are closed, the ventilation of single-family residences in the United States is almost exclusively an uncon- trolled process. In air infiltration (or infiltration and exfiltration), air leaks through unintentional cracks and holes in the building envelope. The infiltration rate is driven by small pressure differ- ences across the building envelope that are typically less than a few pascals in magnitude. These pressure differences arise due to the differences between the indoor and outdoor air temperatures, resulting in different indoor and outdoor air densities, and also as
334 CLEARING THE AIR a consequence of wind. Unintentional air leakage in the ductwork of forced-air heating and air-conditioning systems located in at- tics and craw! spaces, also causes large increases in air infiltra- tion. Even if the ducts do not leak, forced-air systems can pressur- ize or Repressurize specific rooms relative to the outdoor pressure, forcing air leakage through the building envelope. U.S. homes often have intermittently-operated exhaust fans in bathrooms and kitchens. When operated, these fans draw out- door air into the building. Window and door opening by occu- pants, predominantly during mild weather, also has a large influ- ence on residential ventilation rates. A very small portion of single-family dwellings in the United States have mechanical ventilation systems (i.e., fans operating continuously or intermittently to provide ventilation). Mechani- cal ventilation is most common in the State of Washington be- cause the state energy code now requires mechanical ventilation. The technologies used to mechanically ventilate residences are described in Roberson et al. (1998~. Ventilation rates in residences vary considerably over time. The lowest ventilation rates occur during mild weather with win- dows and doors closed. When weather is more severe, windows remain closed but ventilation rates are higher due to increased indoor-to-outdoor temperature differences and increased use of forced-air heating and air conditioning. The highest ventilation rates generally occur when windows or doors are open. Present data on ventilation rates in U.S. single-family resi- dences are limited and possibly not representative of the building stock. One source of information is measurements of the airtight- ness of building envelopes with windows and doors closed. Ven- tilation rates are predicted with semiempirical models, using mea- sured values of building airtightness4 combined with climate data and indicators of a building's shielding from wind, as mode] in 4Airtightness is used here as a general term understandable to a broad audi- ence. The actual measured parameter is the effective leakage area (ELA) at a ref- erence pressure, usually 25 or 50 Pa across the building envelope. The ELA is the area of an orifice that would leak air at the same rate as all the leakage paths in the building envelope. The ELA is usually normalized with building floor area and height to produce a normalized leakage.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 335 puts. When annual average ventilation rates are desired, the pre- dictions may also include terms to account for natural ventilation via windows; however, the current knowledge of window use and effects on ventilation is cursory. The second source of infor- mation on residential ventilation rates is measurements made us- ing a tracer-gas procedure. Although considered more accurate than predictions based on airtightness, the measured data are more sparse than airtightness data. Therefore, we presently have only crude estimates of residential ventilation rates. Based on airtightness and climate data for about 12,000 houses, Sherman and Matson (1997) estimate that the arithmetic average effective ventilation rate of houses in the United States is 1.1 hot. This average reflects ventilation rates when windows are closed and also the higher ventilation rates that occur with open windows during mild weather. Airtightness normalized by house size is highly variable (Sherman and Dickerhoff, 1994), with a standard deviation that is approximately 50% of the mean. The mean of the airtightness data from individual states varies among states by more than a factor of three. In the available data, there is no trend in airtightness with severity of climate. The available data indicated that houses constructed after 1980 are more air- tight (by ~50%) than older houses (Sherman and Dickerhoff, 1994~; however, there was no trend evident in airtightness with age for houses constructed after 1980. A set of 2,844 measurements of residential ventilation rates in U.S. houses was analyzed by Murry and Burmaster (1995~. The measured data from 66 research projects are not from a represen- tative sample of residences; however, this analysis is probably the best available information on the distribution of ventilation rates in U.S. houses. When considering all climate zones and seasons, the arithmetic and geometric mean ventilation rates were 0.76 ho and 0.53 hot, with a geometric standard deviation of 2.3. There are large variations in ventilation rates with season and climate zone. The winter and summer arithmetic means, for all climate zones, are 0.55 and 1.50 hot. Approximately one-third of the measure- ments in the winter season are less than the 0.35 hot, the rate in the current ASHRAE ventilation standard. In the coldest climate zone, approximately 55% of the measured ventilation rates, from all seasons, are less than 0.35 hot.
336 CLEARING THE AIR Methods, Patterns, and Rates of Ventilation in U.S. Multifamily Apartment Buildings In 1997, 21% of U.S. housing units were apartments. The aver- age conditioned floor area of apartments was 85 m2 (920 square feet) and air conditioning was installed in 65% of apartments (Dia- mond, 1999~. Published information on the methods and rates of ventilation in multifamily apartment buildings are extremely sparse. Based on the limited information available,5 older low- rise (i.e., less than Three stories) apartment buildings usually have no mechanical supply of ventilation air. Much like single- family dwellings, these buildings are ventilated primarily by un- controlled infiltration and natural ventilation windows that can be opened. Intermittently operated bathroom and kitchen exhaust fans cause temporary increases in ventilation rates. Leakage in the ductwork of forced-air heating and air-conditioning systems and pressurization or Repressurization of individual rooms can drive infiltration and exfiltration in apartments, just as it does in single-family houses. Newer low-rise apartment buildings are ventilated similarly to older low-rise buildings; however, a larger portion of these buildings have continuous mechanical exhaust ventilation from the bathrooms and/or kitchens of each apart- ment. Older apartment buildings with more than approximately three stories typically have no mechanical air supply or some mechanical supply to the interior hallways. The air supply sys- tem, when present, is frequently not functional (Shapiro-Baruch, 1993~. Apartments within these buildings sometimes have a sys- tem for continuous exhaust ventilation from bathrooms and kitch- ens, although it is not always operational. Some portion of these older high-rise buildings have a vertical ventilation shaft that functions much like a chimney and passively draws air from the apartments. In new apartment buildings with more than three stories, 5The information in this section is based primarily on case studies, on two gen- eral guidance documents (Diamond et al., 1999; Liddament, 1996) and on discus- sions with Dr. Rick Diamond of Lawrence Berkeley National Laboratory. who conducts research on energy use and ventilation in apartment buildings.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 337 exhaust ventilation is usually drawn continuously from the kitchen and bathrooms of each apartment. The exhaust fans may serve groups of apartments or individual apartments. Outside air enters either from unintentional leaks and vents at windows or via ventilation systems that supply air continuously to each apart- ment. When a mechanical air supply is present, often this air is supplied to a single room of each apartment from a duct system in the building's interior hallway. The airflow in heated multistory apartment buildings with- out mechanical ventilation often occurs in an upward direction from lower-level to upper-level apartments (e.g., Diamond et al., 1986; Modera et al., 1986~. Coo] outdoor air leaks into the lower apartments; flows upward, picking up moisture and pollutants; and exfiltrates through the walls and ceilings of upper-level apart- ments. Due to this airflow pattern, the lower-level apartments tend to have more fresh air supply, lower humidity, and more cold drafts. Humidity and pollutant levels are often increased in upper-level apartments because a portion of the air entering these apartments comes from lower levels of the building. As moist air exfiltrates out of the upper-level apartments, water vapor may condense within cold walls and ceilings. Possibly, the higher pol- lutant levels and humidity in upper-level apartments could con- tribute to asthma symptoms. This same upward-flow phenomenon occurs to a variable de- gree in all heated multistory buildings. When the building is air conditioned (i.e., cooled), the airflow direction can reverse; how- ever, the downward airflow in air-conditioned buildings will be less pronounced because the indoor-to-outdoor temperature dif- ferences are typically much smaller during air conditioning than during heating of buildings. By reducing the openings between floors, the vertical airflow between floors can be reduced. Me- chanical ventilation can also reduce or overwhelm the upward buoyancy-driven airflow. In addition to the buoyancy-driven upward airflow in apart- ments, other unintentional flows of air between adjacent apart- ments are reported commonly from case studies. These flows oc- cur through unintentional openings in walls and floors, and may be driven by mechanical ventilation systems, buoyancy, and wind.
338 CLEARING THE AIR The vertical and horizontal airflows between apartments transport pollutants; therefore, occupants may be exposed to pol- lutants such as tobacco smoke, cooking odors, and pet allergens from the apartments of neighbors. These airflows also transport indoor moisture, potentially causing moisture problems. The ASHRAE ventilation standard has the same minimum ventilation requirements for multifamily and single-family resi- dential buildings, 0.35 hat, with exhaust capacity required in kitchens and bathrooms. Measurements of ventilation rates in apartments are extremely limited. Based on case studies in a small number of buildings, the reported ventilation rates are 0.5-1.5 ho for low rise apartments with frame or brick construction and 0.2- 1.0 ho for high-rise apartments with a more airtight concrete con- struction (Diamond et al., 1999~. Based on the available informa- tion, we can expect ventilation rates to vary a great deal with time, among apartment buildings, and among apartments within a building. Heating, Ventilating, and Air Conditioning in U.S. Commercial and Institutional Buildings Building and HVAC Characteristics Approximately 40°/O of commercial floor space has windows that open and approximately one-half of this floor space is in buildings with a floor area less than 4650 m2 (50,000 square feet) (DOE-EIA, 1994, 1995~. Most U.S. commercial buildings have HVAC systems that thermally condition and ventilate the occu- pied spaces. Larger buildings, as well as many smaller buildings, typically use HVAC systems that thermally condition air in me- chanical rooms or in equipment located at the rooftop. The condi- tioned air, usually a mixture of outdoor air and recirculated in- door air known as the supply airstream, is supplied throughout the building or building section. The supply airstream typically passes through a fibrous particle filter, heat exchangers (called heating and cooling coils) that add or remove heat, a supply fan, a system of air ducts, dampers used to regulate airflow rates, and supply registers located at or near the ceilings. The return air- stream is drawn from the occupied spaces usually at or near the
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 339 ceiling and typically flows back to the mechanical equipment through return ducts or through the plenum between a sus- pended ceiling and the floor of the next-higher story. Return air fans are used in many HVAC systems, particularly larger systems. A portion of the return airstream is exhausted to the outdoors, and the remainder is mixed with outdoor air and resupplied to the space after filtering and thermal conditioning. Dampers lo- cated in the outside, recirculated, and exhaust airstreams are au- tomatically or manually adjusted to controls the airflow. In addi- tion to these central HVAC systems, some buildings have distributed systems located throughout the building with com- ponents for adding and removing heat, particularly at the build- ing perimeter that is most affected by heat loss and gains through the building envelope. Many variations in HVAC system designs are described elsewhere (e.g., ASHRAE, 1996~. The supply airflow rates of HVAC systems serving offices or schools are typically 3.5-5 L so per square meter of floor area (0.7-1.0 cfm per square foot) which is equivalent to approximately four to five indoor air volumes per hour. In some smaller build- ings, the supply air fan stops when there is no demand for heat- ing or cooling. The supply airstream usually contains more recir- culated indoor air than outdoor air. The ratio of flow of outside air to total supply air is called the percentage of outside air. Some HVAC systems are designed and adjusted to maintain an approxi- mately fixed rate of outside air intake, or a fixed percentage of outside air, consistent with code requirements. To save energy, many HVAC systems employ an economizer cycle that increases the rate of outside air intake during mild weather to reduce the need for mechanical cooling. While the above-described approach for controlling outside air supply in commercial buildings appears straightforward, in practice it often works poorly. The flow rate of outside air is con- trolled by a small pressure drop across the outside air dampers, and this pressure difference is rarely monitored or controlled. HVAC systems almost never incorporate instruments for directly monitoring the rate of outside airflow. Air-balance professionals may monitor and adjust HVAC system airflows after initial build- ing construction and occasionally thereafter; however, accurate measurements of the rates of outside air intake are very difficult
340 CLEARING THE AIR even for these personnel. Often, the damper systems used to regu- late airflows are nonfunctional, disconnected from the damper actuators, or casually adjusted by building operators. Occasion- ally, building operators will fully close outside air dampers, as- suming that leakage provides adequate outside air. It is not un- common to find inside air flowing out of the building through outside air dampers. In HVAC systems serving smaller buildings with supply air provided only during heating or cooling, there may be no outside air supply during mild weather. The poor per- formance of control systems for outside air is reflected in the mea- sured ventilation data. Measured minimum ventilation rates in commercial buildings vary a great deal among buildings and of- ten deviate substantially from code requirements (Dols and Persily, 1994; Fisk and Faulkner, 1992; Lagus Applied Technolo- gies, 1995; Persily and Grot, 1985; Teijonsalo et al., 1996; Turk et al., 1989~. Outside air drawn into HVAC systems and mixed with recir- culated air is distributed to the various rooms of a building with a complex duct system. In many buildings the rate of air supply to each section of the building is modulated automatically over time to maintain a comfortable indoor temperature. Sometimes out- side air is distributed very unevenly to the rooms or floors within a building so that effective ventilation rates within rooms of the same building vary by a factor of two or three (e.g., Dols and Persily, 1994; Fisk and Faulkner, 1992; Teijonsalo et al., 1996~. In other buildings, the effective ventilation rate is spatially uniform. In addition to the ventilation provided by HVAC systems, natural ventilation through windows that open occurs in the 40°/O of commercial floor space having such windows. Air infiltration also occurs in all commercial buildings, just as it does in resi- dences; however, the mechanical ventilation systems of commer- cial buildings can drive air infiltration and exfiltration through the building envelope, modifying or sometimes reversing the air- flows that would occur due to wind and buoyancy. For many large commercial buildings, particularly those in warm humid climates, the design intent is to pressurize the building slightly with the mechanical ventilation system in order to prevent unde- sirable infiltration of unconditioned air, moisture, and outdoor air pollutants. In cold climates, some buildings are designed to be
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 341 slightly Repressurized to prevent warm humid indoor air from flowing outward through the building envelope. In practice, in- door-outdoor pressure differences are often poorly controlled, and many buildings are not pressurized (Persily and Norford, 1987~. Ventilation Rates in Commercial and Institutional Buildings Commercial and institutional buildings are ventilated mechani- cally with the HVAC system, via air infiltration and through the openable windows present in 40°/O of the commercial floor space. The ventilation provided through each of these mechanisms varies over time. In warm climates, the mechanical supply of outside air is minimized when the outside air exceeds approximately 20-24°C, hence, the building may be operated with a minimum ventilation rate for much of the cooling season. Minimum ventilation is also provided during cold weather when the outside air temperature fails below a set point that varies among buildings. During periods of minimum ventilation, concentrations of indoor-generated pollut- ants will tend to be highest. During mild weather in the spring and fall, the indoor-to-outdoor concentration ratio for some outdoor air pollutants, such as outdoor particles (e.g., molds) and ozone, will tend to be at a maximum. Ventilation rate data from United States commercial and in- stitutional buildings are very limited and come primarily from convenience samples of buildings and modest-size research projects. Consequently, only very rough estimates can be pro- vided on the distribution of ventilation rates in U.S. commercial buildings. Table 10-2 provides summary information from the largest data sets. The measured ventilation rates vary over a large range and usually exceed the minimum ventilation rates speci- fied in ASHRAE Standard 62-1989. However, the ventilation rates measured when the HVAC systems provided the minimum amount of outside air were less than the corresponding minimum rate specified in the current ASHRAE standards in a majority of 6Some buildings were constructed when the applicable ventilation standard specified lower minimum ventilation rates than ASHRAE Standard 62-1999 (ASHRAE, 1999).
342 CLEARING THE AIR buildings (i.e., in 8 of 13 buildings studied by Turk et al., 1989; 20 of 49 buildings [including 13 out of 14 schools] studied by Lagus Applied Technologies, 1995; and 8 of 8 buildings studied by Persily and Grot, 1985~. In commercial and institutional buildings that are not occu- pied continuously, HVAC systems are often not operated when the building is vacant. The concentrations of indoor-generated air pollutants may increase when the building is vacant and there is no mechanical ventilation, and then decrease over a period of a few hours after mechanical ventilation starts. Therefore, occu pants' exposures to indoor air pollutants depend on the operat- ing schedule for the HVAC system. Even less is known about the proportion of commercial build- ing ventilation that results from air infiltration; however, the ex- isting data suggest that infiltration is appreciable, particularly in the smaller buildings (Lagus Applied Technologies, 1995; Persily, 1999~. The ratio of infiltration to total ventilation may be impor- tant for asthma because the air that infiltrates a building is not filtered to remove outdoor particles before it enters the building. Pollutant Sources in HVAC Systems HVAC systems can become sources of indoor air pollutants, possibly increasing asthma symptoms. Pollutant emissions from certain types of HVAC systems constitute one of several potential explanations for the consistent association between the type of HVAC system and the prevalence of nonspecific health symptoms (called sick building syndrome) experienced by office workers (e.g., Mendell, 1993~. In particular, almost all studies have found a statistically significant increase in symptom prevalence among occupants of office buildings with mechanical ventilation and air conditioning (Mendell, 1993) relative to occupants in naturally ventilated offices. There is also some evidence that humidifiers are associated with increased symptoms (Mendell,1993~. The sub- jectively assessed (i.e., via questionnaire) nonspecific health symptoms in these studies usually include a few symptoms po- tentially indicative of asthma, such as wheezing, tight chest, and difficulty breathing. Therefore, the evidence that HVAC systems can become sources of pollutants is relevant for asthma. The avail
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 343 able information about pollutant sources in HVAC systems is quite limited and comes predominately from commercial build- ings; however, similar pollutant sources, or risk factors for sources, may be present in HVAC systems of other types of build- ings. This section provides a brief review of the issue. Liquid water is often present at several locations in or near commercial building HVAC systems, facilitating the growth of microorganisms that may contribute to asthma. The outdoor air is often drawn from the rooftop or from a below-grade "well" where water (and organic debris) may accumulate. Raindrops, snow, or fog can be drawn into HVAC systems with the incoming outside air, although the systems are usually designed to prevent or limit this moisture penetration. Moving along the supply air- stream in the direction of airflow leads to the cooling cod] where moisture condenses and ideally drips from the surfaces of the cod! into a drain pan with a drainage pipe connected to the sewer or to outdoors. Frequently, drain pans contain stagnant water because they do not slope toward the drain line. The drain also may be plugged or nonfunctional because air pressure differences pre- vent drainage, sometimes causing the drain pan to overflow with water. If the velocity of air passing through the cooling coils is too high, drops of water on the surface of the cooling cod! can become entrained in the supply airstream and deposit in the HVAC sys- tem downstream of the cooling coil. Air exiting cooling coils is frequently nearly saturated with water vapor, and the high hu- midity of this air increases the risk of microbiological growth. The HVAC system may have a humidifier that uses steam or some evaporation process to add moisture. Humidifiers, used predomi- nantly in the colder climates, may have reservoirs of water, sur- faces that are frequently wetted, or water drops that do not evapo- rate. Thus, there are many sources of liquid water and very high humidity in HVAC systems. Microbiological contamination of HVAC systems has been reported in many case studies and investigated in a few multi- building research efforts (e.g., Batterman and Burge, 1995; Bencko et al., 1993; Martiny et al., 1994; Morey, 1994; Morey and Williams, 1991; Shaughnessy et al., 1998~. Sites of reported microbiological contamination include the outside air louvers, the mixing box where outside air mixes with recirculated air, filters, cooling coils,
344 CLEARING THE AIR TABLE 10-2 Summary Information from the Three Largest Surveys of Ventilation Rates in U.S. Commercial Builclings Reporte' Type of Number of Operating Study Building Buildings Condition Mean Turk et al., 1989 Offices, schools 38 As found 28 and libraries 38 As found 1.5 combined 13 Min. vent.a 10 14 Min. vent.a 1.6 M M M M Lagus Applied Tech., 1 995 Persily and G rot, 1985 Small office 17 Large office Retail School 13 14 Offices 9 in. vent.a in. vent.a in. vent.a in. vent.a Yearly avg. Min. vent.a 1.4 0.8 2.2 2.4 0.7 0.4 aWith the ventilation system providing the minimum rate of supply of outside air, a normal condition during cold weather and also when outside temperatures exceed the indoor temperature. cooling cod] drain pans, humidifiers, and the surfaces of ducts. The porous insulating and sound-adsorbing material called duct liner, which is used inside some HVAC systems may be particu- larly prone to contamination (Morey, 1988; Morey and Williams, 1991~. Supporting the evidence of microbiological contamination from field studies are numerous laboratory-based studies dem- onstrating that fungi can grow on various HVAC materials at a wide range of temperatures and humidities. The flow of air through HVAC systems to the occupied spaces is a means of trans- porting bioaerosols from contaminated sites inside HVAC sys- tems to occupants. Thus, microbiological contamination of HVAC systems is theoretically a risk factor for asthma. However, rela- tively little is known about the extent to which microbiological contaminants within HVAC systems actually increase bioaeroso! exposures or affect asthma symptoms. HVAC systems can also be sources of other pollutants such as fibers, nonbiological particles, and VOCs. Sources of these pollut
IMPACT OF VENTI~TION AND AIR CLEANING ON ASTHMA 345 of Reported Ventilation Rates Estimated ASH RAE Mean SD Min Max Unitsb Std. 62-1999 Equivalentb 28 18 4.5 84 Ls-4pe~4 10Ls-4pe~4 1.5 0.9 0.3 3.6 h-4 10 9 1.5 37 Ls-4pe~4 10Ls-1pe~4 1.6 0.7 0.4 3.0 h-4 1.4 0.7 0.3 2.7 h-4 0.8 h-4 0.8 0.6 0.7 2.7 h-4 0.8 h-4 2.2 1.6 0.5 7.0 h-4 1.2 h-4 2.4 1.6 1.2 2.9 h-4 3.0 h-4 0.7 0.2 0.3 1.1 h-4 0.7 h-4 0.4 0.2 0.1 0.6 h-4 0.7h-4 bLegend for ventilation rate units: L s-4 peck = liters per second per person; h-4 = air changes per hour; Min. vent. = minimum ventilation; SD = standard deviation; and Yearly avg. = yearly average. ants inside HVAC systems include duct liners, gaskets, of] left on surfaces after manufacturing, dust deposited on surfaces includ- ing dusts generated during building construction and renovation, and the wear of fan belts. It is clear that HVAC systems and com- ponents, particularly dirty filters and the oily surfaces of ducts, can release odorous compounds that significantly degrade the perceived acceptability of the air supplied to the occupied space (e.g., Pasanen et al., 1995; Pejtersen, 1996; Seppanen, 1998~; how- ever, the relevance of these pollutants for asthma is not known. Causes of HVAC Problems The rates of ventilation provided by HVAC systems and the contamination of HVAC systems with pollutants depend as much on system construction, maintenance, and operation practices as on system design. Proper operation and maintenance of HVAC systems is hindered by system complexity, inaccessibility of com
346 CLEARING THE AIR portents, and lack of training of construction and maintenance staff. The absence of economic incentives for proper HVAC op- eration and maintenance may be even more important. When buildings are leased, their owners and operators are usually un- affected economically by the adverse health of building occu- pants, unless the health problems are severe and clearly linked to the building. In speculative new construction, the designer is similarly unaffected. However, these designers, owners, and op- erators have an incentive to reduce the highly tangible costs of building construction, maintenance, and operation. Costs of Ventilation and Associated Carbon Dioxide Emissions Ventilation is one of the most energy-intensive methods of reducing indoor pollutant concentrations primarily because of the need to thermally condition ventilation air. For several reasons, including great uncertainty about average ventilation rates in buildings, the available estimates of the energy used for ventila- tion are relatively crude. Orme (1998) estimates that ventilation accounts for about 30°/O of the energy used in U.S. residential buildings. A rough estimate of the average annual cost per resi- dence of energy for ventilation is $400.7 In the U.S. service sector (e.g., commercial, institutional, and government buildings), the estimated energy consumed for ventilation (Orme, 1998) is ap- proximately one-quarter of total service-sector building energy use. The cost of energy for ventilating commercial buildings de- pends highly on climate, building size, occupant density, and type of HVAC system. If the average ventilation rate is 10 L so per person, Emmerich and Persily (1998) have estimated that 46% of the total heating and cooling load in the stock of office buildings is attributable to ventilation. With nearly 50°/O of the U.S. work force, or 57 million workers, in offices (U.S. Department of Com 7This estimate is based on the 30% of total residential energy used for ventila- tion (Orme, 1998) and the average U.S. household's total energy cost of $1,355 (Diamond, 1999~.
IMPACT OF VENT TON AND AIR CLEANING ON ASTHMA 347 merce, 1997; Table 645) and average energy prices, the annual cost of ventilation per office worker is roughly $25.8 Carbon dioxide, a greenhouse gas, and other air pollutants are generated in the production of the energy used for ventila- tion. The annual CO2 emissions attributed to ventilation are ap- proximately 1000 and 800 million tons for the residential and ser- vice sectors, respectively (Orme, 1998~. Influence of Ventilation Rates on Indoor Concentrations of Pollutants Direct Impacts Measured data quantifying the influence of ventilation rates on indoor concentrations of indoor-generated pollutants are sur- prisingly limited, particularly for pollutants associated with asthma. Data from surveys with ventilation rate and pollutant concentration measurements have emphasized measurements of pollutants that are generated indoors and also present in outdoor air reducing the influence of ventilation rates on indoor concen- trations. None of the large cross-sectional surveys have monitored concentrations of indoor-generated particles associated with asthma, such as environmental tobacco smoke (ETS) particles and indoor-generated bioaerosols. Indoor particle mass concentra- tions have been measured in some surveys, but these measure- ments reflect both indoor-generated particles and particles from outdoors. In a study of 150 residences located in Riverside, Cali- fornia (Ozkaynak et al., 1996), there were statistically significant, but small, increases in indoor particle concentrations with in- creased ventilation rates. The values of PM and PM2 5 increased about 12 and 5 ,ug m-3, respectively,9 for a 1 ho increase in venti- lation rate. Estimated heating and cooling energy are 110 and 28 PI, respectively (Emmerich and Persily, 1998~. Average prices of natural gas (~0.69 per therm) and electricity (~0.098 per kilowatt-hour) from U.S. Department of Commerce (1997, Table 768) have been used to convert energy use to energy costs, which are then divided by the size of the office work force. 9PM and PM2 5 are particle mass concentrations for particles smaller than 10 and 2.5 ,um, respectively.
348 CLEARING THE AIR In data from surveys, there is sometimes no clear association of indoor pollutant concentrations with ventilation rates. In a sur- vey of 26 houses near Spokane, Washington, and 35 houses near Portland, Oregon, the indoor formaldehyde concentration de- creased 40 and 300 parts per billion (ppb), respectively, per 1 ho increase in ventilation rate (Turk et al., 1987b). However, in sur- veys of 38 commercial buildings in the U.S. Pacific Northwest, there were no clear associations of ventilation rates with indoor concentration of respirable particles, nitrogen dioxide, or formal- dehyde (Turk et al., 1987a, 1989~. Apparently, variations in indoor pollutant emission rates and other factors are large enough to ob- scure the effects of ventilation in a survey of this size. Few experiments have been performed with changes in ven- tilation rates as the only intervention. Again, most of the experi- mental data involve pollutants generated indoors but also present in outdoor air. Offermann et al. (1982) used mechanical ventila- tion systems to increase wintertime ventilation rates in nine houses from an average of 0.35 to 0.83 hot. The average indoor formaldehyde~° concentration decreased 21% and average indoor relative humidity decreased 13%. Indoor nitrogen dioxide con- centrations increased slightly because concentrations were higher outdoors than indoors. In a set of two houses, inhalable particle concentrations decreased 30°/O. Using a mechanical ventilation system with a filter, Turk et al. (1997) increased ventilation rates in two classrooms from about 0.7 to 2-3 ho and indoor particle concentrations decreased by approximately 50°/O. Berglund et al. (1982) increased the percentage of outside air in the air supply of an office building from 20 to 50°/O, and then to 80%, and their mea- sure of total indoor-generated VOCs in detector-response units decreased from 120 to 50 and then to ~20. Menzies et al. (1993) increased ventilation rates in a set of four office buildings and found that formaldehyde and total VOC concentrations were 40 and 60% lower, respectively, during periods with a higher venti- lation rate. Hodgson et al. (1999) measured concentrations of 24 VOCs in a house with ventilation rates of 0.14 and 0.32 hot. At the i°Indoor concentrations of formaldehyde are usually much larger than outdoor concentrations due to indoor formaldehyde sources.
IMPACT OF VENTI~TION AND AIR CLEANING ON ASTHMA 349 higher ventilation rate, on average for the 24 compounds the in- door concentrations was 42% lower, with a standard deviation of 11%. Some experimental studies did not detect clear reductions in indoor pollutant concentrations when ventilation rates were in- creased. Shaughnessy et al. (1997) increased ventilation rates from ~ 0.2 ho to 3 ho in one elementary classroom and from ~ 0.7 ho to 3 ho in another classroom. There were no clear changes in indoor PM concentrations reportedly because of large natural temporal fluctuations. Indoor total volatile organic compound (TVOC) con- centrations decreased by about a factor of two, but natural tem- poral variations precluded firm conclusions about the effects of ventilation. Nagda et al. (1990) increased ventilation rates in an office building from 0.84 ho to 1.08 ho and indoor concentrations of formaldehyde, nicotine, respirable particles, and carbon mon- oxide did not change significantly. Neither the cross-sectional nor experimental data provide us with a clear indication of the influence of ventilation rates on in- door concentrations of indoor-generated pollutants. The cross- sectional findings are subject to many confounders. Experimental data are complicated by temporal variations in indoor pollutant emissions. Neither type of study has focused on pollutants with only indoor sources. Therefore, Equation A1 in Appendix A has been used to provide theoretical predictions for particles of vari- ous sizes as well as for some types of gaseous pollutants. For predictions, ventilation rates have been varied from 0.2 to 4 hot, and 0.75 ho has been used as a reference point for which the con- centration is arbitrarily set equal to one. The low end of this range (0.2 h-~) is equivalent to the lowest ventilation rates commonly reported in buildings. Because of energy costs and equipment re- quirements, increasing ventilation rates to greater than 4 ho dur- ing periods of heating or cooling seems unlikely except in spaces with a high occupant density such as classrooms. A ventilation rate of 4 ho is a factor of five greater than the arithmetic mean ventilation rate of residences reported by Murry and Burmaster iiThe same equation can be used for many gaseous pollutants, as long as the indoor pollutant emission rate is unaffected by the indoor concentration.
350 CLEARING THE AIR (1995) and corresponds approximately to the total rate of supply of recirculated plus outdoor air in U.S. commercial buildings. The reference ventilation rate of 0.75 ho is typical of the ventilation rates in residences (Murry and Burmaster, 1995) and also roughly equivalent to the minimum ventilation rates in schools and com- mercial buildings (Table 10-2~. The influence of ventilation rate on pollutant concentrations depends on the magnitude of pollutant removal by air cleaning. Figure 10-1 provides some predicted relationships for a space with no air cleaning. In this figure, relative concentrations are used, which are defined as the actual concentration divided by the con- centration at the reference ventilation rate of 0.75 hot. One clear observation is that the relative concentrations of some indoor-gen- erated pollutants increase dramatically as ventilation rates be- come unusually low (e.g., below a few tenths of an air change per hour). Because ventilation rates are poorly controlled in build- ings, a portion of the building stock will have these low ventila- tion rates. The airflow requirements and associated energy penal- ties of avoiding these particularly low ventilation rates are o ~ 2.5 ~ Q) c) Q) A,, 1.5 ~ Q) 1.0 0.5 O Gas, no deposit loss O NO2 0 0.1- and 0.5-,um particles 1.0-,um particle .)K 2.0-,um particle 5-,um particle 1 O-,um particle WOW ~ + + + ~ ~ D1 [1 ~ ~ E] - 20-,um particle 30-,um particle _~ 0 0.5 1 1.5 3 3.5 4 4.5 Air Exchange Rate (h-1) FIGURE 10-1 Predicted trends in the relative concentrations of indoor- generated pollutants with ventilation rate.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 351 modest, but low ventilation rates cannot be prevented with cur- rent ventilation technologies and processes (e.g., natural infiltra- tion) that control ventilation rates poorly. AS ~ustratea In Figure ~ u- ', tne predicted cnange In pollut- ant concentrations with ventilation rate is greatest for an "ideal" gaseous pollutant that is not removed by deposition or sorption on surfaces and that has an emission rate unchanged by ventila- tion rate. In this highly ideal case, often used to illustrate the ef- fects of ventilation, the relative concentration in Figure 10-1 de- creases from 3.0 to 0.2 as the ventilation rate increases from 0.2 to 4 hot. For real pollutants, ventilation rates have a smaller pre- dicted impact on concentrations in many cases, a much smaller impact. The range in predicted relative concentrations of nitrogen dioxides is 1.6 to 0.3 much less than the range for the ideal pol- lutant because of losses of NO2 via sorption on surfaces. If out- door NO2 concentrations are significant relative to those indoors, ventilation rates will cause an even smaller change in indoor NO2 concentrations. For indoor-generated VOCs, the dependence of indoor concentration on ventilation rates is more difficult to pre- dict and will vary among compounds, but concentrations will usually be less affected by ventilation rate than the ideal gaseous pollutant. For indoor-generated particles, the predicted effects of venti- lation rate depend highly on particle size because the depositional losses of particles increase dramatically with particle size. Practi- cal changes in ventilation rates have a substantial predicted im- pact on indoor concentrations of small particles, such as ETS par- ticles and the portion of airborne cat allergens associated with particles smaller than a couple micrometers. Ventilation rates are not likely to have an appreciable directly impact on indoor con- centrations of the larger particles associated with dust mite and · · .. . . . · ~ · ~ ~ ~ .. . - . . . i2To simulate the change in nitrogen dioxide concentrations with ventilation rate, a deposition velocity (see Appendix A) of 7.4 x 10-5 m s-i was used (Nazaroff et al., 1993), and for illustrative purposes, the outdoor NO2 concentration was neglected. i3However, as discussed later, ventilation rates can affect indoor humidity, which in turn modifies the indoor sources of bioaerosols.
352 CLEARING THE AIR cockroach allergens and many indoor-generated fungal spores. For example, in Figure 10-1 the predicted range in relative con- centrations of 10-,um indoor-generated particles is only 1.14 to 0.57 as the ventilation rates varies from 0.2 to 4 hot. If there is some form of air cleaning within the building, ventilation rates will change indoor pollutant concentrations by a smaller amount. The previous discussion has focused on indoor-generated air pollutants. Outdoor pollutants are not substantially addressed in this document; however, the influence of ventilation rate on in- door concentrations of outdoor pollutants is a relevant issue for asthma. Increasing ventilation rates will increase the rates of en- try of outdoor pollutants into buildings. For pollutants that are not removed indoors by deposition on surfaces or air cleaning (e.g., outdoor carbon monoxide), a change in ventilation rate will not substantially affect the time-average indoor concentration, al- though peak concentrations may be affected. However, an in- crease in ventilation rate will increase the indoor concentrations of other outdoor pollutants such as particles and ozone from out- door air. Indirect Impacts Ventilation rates may have a very significant indirect impact on indoor concentrations of some pollutants because they affect indoor humidities, which in turn modify indoor pollutant sources. Potentially most important for asthma is the influence of indoor humidity on indoor dust mites, molds, and bacteria. Two humidity parameters are necessary to characterize the relationship between indoor and outdoor humidity. The humid- ity ratio is the mass of water vapor divided by the mass of dry air. The relative humidity (RH) is the partial pressure of water vapor divided by the partial pressure of water vapor in air at the same temperature that is saturated with water vapor. The amount of water vapor required to saturate air decreases with decreasing air temperature; hence, 0°C outdoor air with a 100% RH will have a low humidity ratio relative to room-temperature indoor air with aRHof50°/O. Due to indoor moisture generation, the humidity ratio in
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 353 doors will exceed that outdoors,~4 unless the indoor air or incom- ing outside air is dehumidified. In cold winter climates, outdoor air is very dry and ventilation rates have a pronounced impact on humidities indoors, with the humidity decreasing as the ventila- tion rate increases. In warm humid climates, air-conditioning sys- tems must remove moisture from the indoor air to maintain ac- ceptable indoor relative humidities. Increasing the ventilation rate in warm humid climates will increase indoor humidity ratios and relative humidities unless the rate of water removal by the air- conditioning system increases accordingly. This relationship be- tween ventilation rate and humidity places constraints on the ven- tilation rates in buildings. Particularly high ventilation rates during cold weather causes indoor RHs to decrease to a point (less than ~25%) that causes discomfort and dryness symptoms. Particularly high ventilation rates in warm humid weather over- load the moisture-removal capabilities of typical HVAC systems, leading to excessive indoor relative humidities. Recognizing the relationship of ventilation rate to RH, a few studies have investigated whether indoor dust mite levels or al- lergy symptoms in dwellings can be controlled by increasing ven- tilation rates with mechanical ventilation systems (Table 10-3~. In cold climates, mechanical ventilation appears to be associated with decreases in indoor humidity and dust mite levels (Emenius et al., 1998; Harving et al., 1994) or allergic symptoms (Aberg et al., 1996~. In more humid climates, the results are mixed. Mechani- cal ventilation was associated with significant reductions of dust mite levels in the small study by McIntyre (1992) in the United Kingdom. However, Fletcher et al. (1996) did not measure signifi- cant reductions in dust mites in England and Aberg et al. (1996) did not find significant reductions in allergic symptoms in the more humid areas of Sweden. Summary Ventilation rates are poorly controlled in residential and com- mercial buildings. Measured data indicate that ventilation rates i4At equilibrium.
354 CLEARING THE AIR TABLE 10-3 Investigations of the Association of Resiclential Mechanical Ventilation Systems with Indoor Humiclities, Dust Mite Levels, and Allergy Symptoms Reference (Study Type) Location (Climate) No. of Buildings Increase Fletcher et al., 1996 Northwest England 18 (9 experimental andNot ass (controlled experiment) (mild, humid) 9 controls) Aberg et al., 1996 Sweden 1,694 questionnaires andNot ass (survey) (cold, dry winters) 400 skin prick allergy tests Emenius et al., 1998 Stockholm area 59Median (cross-sectional) (cold, dry winters) mecha ~0.26 ventila Harving et al., 1994 Aarhus, Denmark 32 families moved to houses0.4 h-4 i (experimental study with (mild) with mechanical ventilation;before control group) 29 controls1.5 h-4 mocha For the medial unchain Wickman etal., 1994 Suburban Stockholm 70Not ass (cold, dry winters) McIntyre, 1992 South Wales, UK 8 houses (excluding 4 flats)Not ass (cross-sectional) (moderate humid climate) NOTES: Cl = confidence interval; OR = odds ratio.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 355 nical I ncrease in Ventilation Rate Findings Not assessed Not assessed tests ruses ilation; lats) Median ~0.6 h-4 with mechanical ventilation and ~0.26 h-4 with natural ventilation 0.4 h-4 in experimental group before move, versus 1.0 and 1.5 h-4 in the post-move mechanically ventilated homes. For the control group, the median ventilation rate was unchanged at 0.29 ha. Not assessed Not assessed Indoor RHs during autumn and winter were about 10% lower with increased ventilation, but still generally above 50%. No significant effects of increased ventilation on mite counts or mite allergen concentrations (p>.1~. Reported severity of water condensation indoors diminished. Mechanical ventilation of the home was associated with a reduced prevalence of allergic disease with relative risk for one or more symptoms of 0.68 (95% Cl 0.46-1.00), but only in northern Sweden. Median indoor RH ~9% lower (30 vs. 40%) with mechanical ventilation (p < .001~; median dust mite allergen in mattress dust ~70% lower with mechanical ventilation (p < .0001~; median TVOC levels ~60% lower with mechanical ventilation (p < .0001~. Move to mechanically ventilated house associated with ~30% reduction in dust mites in mattress dust of experimental group (p < 0.05~. The post-move absolute humidity was also significantly decreased in one of two measurement periods. For the control group, humidity did not change significantly and dust mite levels in mattress dust increased nonsignificantly by a factor of ~2. Results of studies of bedroom floor dust were similar. Presence of mechanical ventilation system was associated with significantly lower indoor humidity ratio (OR = 0.1; 95% Cl 0.0-0.2) and significantly lower levels of dust mite allergen in mattress dust (OR = 0.1; 95% Cl 0.0-0.5~. Significantly lower humidity (p < .05) and ~90% lower dust mite levels in dust (p < .01 ) in 4 houses with continuous mechanical ventilation relative to 4 houses that did not continuously use mechanical ventilation.
356 CLEARING THE AIR vary widely among buildings of the same class (e.g., residences, schools, offices) and that rates of ventilation are frequently below or well above the rates specified in the current ASHRAE ventila- tion standard. Insufficient data are available for conclusions about changes in building ventilation rates during the past few decades; however, there are some indications that ventilation rates have decreased. Concentrations of indoor-generated air pollutants associated with asthma may be influenced by ventilation systems via three primary mechanisms. First, changes in the rate of ventilation modify indoor concentrations of indoor-generated air pollutants by changing the rates of pollutant removal and dilution with out- door air. Measured data and mode] predictions indicate that the changes in indoor pollutant concentrations associated with the typical range of indoor ventilation rates vary dramatically among pollutants. Because of the sparseness of empirical data, mode! predictions are the best available means of estimating the typical changes in indoor pollutant concentrations that result from changes in ventilation rates. However, there are considerable un- certainties in the applicability of these predictions to actual com- plex indoor environments, with rates of pollutant loss by deposi- tion on indoor surfaces being one of the largest sources of uncertainty. Based on these predictions, the indoor concentrations of some pollutants increase dramatically as ventilation rates become un- usually low (e.g., <0.25 hub. The energy costs associated with avoiding these particularly low ventilation rates are modest; how- ever, current methods and technologies of ventilation do not con- sistently prevent these low ventilation rates. Increases in ventilation rates are one obvious approach for reducing exposures to some pollutants associated with asthma. Increasing ventilation rates from a typical value (0.75 h-~) to a high value (4 h-~) should decrease concentrations of indoor-gen- erated particles 1 ,um or smaller by up to 80%. Indoor-generated particles in this size range that are relevant for asthma include those associated with ETS, portions of airborne cat allergen, and most of the droplet nuclei produced during coughing and sneez- ing. Some grass and birch allergens from outdoors are also smaller than 1 ,um. As particle size increases, changes in ventilation rates
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 357 have a diminishing predicted impact on indoor concentrations. For particles 10 ,um in size, increasing the ventilation rate from 0.75 to 4 ho would be expected to decrease indoor concentrations by about 50°/O. For 20- and 30-,um particles, an even smaller change in concentration with ventilation rate is expected. The limited data available indicates that much of the airborne dust mite and cock- roach allergen is associated with particles that are 10 ,um or larger . . In size. The impact of ventilation rates on concentrations of gaseous pollutants associated with asthma will generally be modest. With- out a strong indoor source of nitrogen dioxide, outdoor NO2 con- centrations are often higher than indoor concentrations, and in- creases in ventilation rates will increase the concentrations indoors. Even when strong indoor sources are present, such that the outdoor concentration is negligible, practical changes in ven- tilation rates should change indoor nitrogen dioxide concentra- tions by 50°/O less. The influence of ventilation rates on indoor concentrations of indoor-generated VOCs will vary among com- pounds and, for many compounds, is not easily modeled with existing data. In general, concentrations of the more volatile com- pounds are more strongly affected by ventilation rates. several, tnese prealcuons ana evade aata suggest that large increases in ventilation rates would be most effective in re- ducing exposures to ETS, cat allergens, infectious droplet nuclei, and some volatile organic compounds. Such large increases in ventilation rates, applied broadly, would substantially increase building energy use and energy costs and would significantly in- crease emissions of the greenhouse gas carbon dioxide. Unless ventilation systems with heat recovery were used, increasing ven- tilation rates to ~4 ho in U.S. residences would typically increase annual energy costs from several hundred to more than a thou- sand dollars. When possible, eliminating or reducing indoor sources of these pollutants would be a much more effective method of reducing exposures and generally would not increase building energy use. Microorganisms and other pollutants that are potentially rel- evant for asthma can contaminate HVAC systems and be trans- ported via the system to occupied spaces. There are many sources of liquid water in HVAC systems, as well as regions with very ~.. .. .. .. . .. .. . .
358 CLEARING THE AIR humid air, that facilitate the growth of microorganisms. The con- tamination of HVAC systems with microorganisms and other pol- lutants is one hypothesized explanation for the consistent find- ings that office workers in buildings served by HVAC systems with air conditioning have more nonspecific health symptoms than occupants of naturally ventilated office buildings. However, the relevance of this association to asthma remains unclear. Pre- venting HVAC system contamination requires improvements in system design, construction, and maintenance. Ventilation rates also affect indoor humidities, in turn, influ- encing indoor levels of dust mites. Higher indoor humidities may also increase indoor molds and bacteria. When it is cold and dry outdoors, increases in ventilation decrease the indoor humidity. When it is warm and humid outdoors, increases in ventilation rate tend to increase the humidity in air-conditioned buildings. The limited data available suggest that using mechanical ventila- tion systems to increase ventilation rates in residences can result in significantly lower dust mite levels only in cold climates. Most U.S. houses and many schools rely for ventilation on air leakage through unplanned openings in the building envelope plus natural ventilation through windows and doors. Increased window opening will increase ventilation rates, but not in a con- trolled or predictable manner. Mechanical ventilation systems would have to be installed in these buildings to achieve controlled increases in ventilation rate. Mechanical ventilation systems are commercially available, but they are often expensive (hundreds of dollars to more than a thousand dollars) and unfamiliar to many homeowners. Several health effects other than asthma, including nonspe- cific irritation symptoms, allergies, and communicable respiratory illnesses, are potentially influenced by ventilation rates and ven- tilation system contamination. All of these health outcomes have to be considered when policies and education programs about ventilation are established. Conclusions Regarding the Relationship of Ventilation to Asthma Existing data are inadequate for conclusions regarding the
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 359 association between ventilation rates or ventilation system mi- crobiological contamination and either the exacerbation of asthma symptoms or asthma development. However, there are both theo- retical evidence and limited empirical data indicating that fea- sible modifications in ventilation rates can decrease or increase indoor concentrations of some of the indoor-generated pollutants associated with asthma by up to 75°/0. Research Needs Additional research is needed on ventilation, but only a small proportion of these research needs are critical to advancing our understanding of the relationship of ventilation to asthma. At the present time, our understanding of the influence of changes in ventilation rates on concentrations of (or exposures to) indoor- generated pollutants associated with asthma is very limited ac- cordingly, mode! predictions that have not been adequately evalu- ated are the best source of information. Experiments in actual buildings, with manipulation of ventilation rates, are the pre- ferred approach for quantifying the direct (i.e., via pollutant re- moval) influence of changes in ventilation rates on the indoor con- centrations of these pollutants. Because indoor pollutant source strengths can vary temporally, experiments should be repeated several times. To assess how changes in ventilation rates affect indoor humidities and, in turn, the proliferation of dust mites and molds in buildings will require either long-term experiments last- ing a year or more or large cross-sectional studies with control for confounding factors. Airtight building envelopes and low rates of ventilation have been cited as factors that may contribute to asthma incidence or symptoms or may explain recent increases in asthma; however, very few relevant data are available. The evidence of a linkage of ventilation rates with asthma is not sufficient to justify large stud- ies intended to resolve only this issue. However, measurements of ventilation rates should be included, when possible, in future asthma case-control studies or cross-sectional surveys. Ventilation measurements in houses can be performed using nonobtrusive tracer-gas methods with passive tracer-gas emitters and samplers (Dietz and Cote, 1982; Stymne and Eliasson, 1991~.
360 CLEARING THE AIR Finally, research is needed to advance our very limited cur- rent knowledge about microbiological contamination of HVAC systems, its influence on microbial exposures, and its influence on asthma development or asthma symptoms. PARTICLE AIR CLEANING: INTRODUCTION AND REVIEW OF CONVENTIONAL PRACTICE Background Because many of the indoor pollutants associated with asthma are airborne particles, particle air cleaning is considered a potentially beneficial technology for the prevention of asthma or asthma symptoms. Particle air cleaning is any process used inten- tionally to remove particles from the indoor air. Filtration and electronic air cleaning are the two most common examples. Natu- ral deposition of particles on indoor surfaces, ventilation, and measures that reduce indoor "article emission rates are not con- sidered particle air cleaning. The magnitude of the reduction in indoor particle concentra- tions accomplished with particle air cleaning depends on the air cleaner's particle removal rate relative to the particle removal rate by all other processes. The rate of particle removal by an air cleaner varies with particle size and equals the flow rate of air through the air cleaner Sac multiplied by the air cleaner's par- ticle removal efficiency £. The product of Qac and £ integrated over particle size for a specific type of particle source is sometimes called the clean air delivery rate (CADR). Based on standard test protocols, the American Home Appliance Manufacturers pro- vides CADRs for tobacco smoke, dust, and pollen for many por- table air cleaners. Although particle air cleaning reduces indoor particle con- centrations, microorganisms can grow on some air-cleaning equipment such as filter media; thus, air cleaners are also a poten- tial source of indoor pollutants. Particle Air-Cleaning Technologies By far, the most common method of air cleaning is to circulate
IMPACT OF VENTI~TION AND AIR CLEANING ON ASTHMA 361 air through a fibrous filter. The filtration media is usually a mat of thin fibers, often glass fibers. The efficiency of the filter is a func- tion of fiber diameter, fiber packing density (e.g., distance be- tween fibers), thickness of the media, velocity of the air as it passes through the media, particle size, and other factors. A pleated (i.e., folded) filter media is often employed to increase the media sur- face area, reducing the air velocity in the media, and the resis- tance to airflow through the media. A wide range of filter prod- ucts are available, with a correspondingly wide range of efficiencies and prices. Based primarily on data from Haniey et al. (1994) and data from manufacturers, Figure 10-2 provides ex- amples of particle removal efficiency versus particle size for fil- ters with a range of efficiency ratings, using an ASHRAE rating method (ASHRAE, 1992~. At one extreme are the coarse pane] filters, usually called furnace filters, commonly used in residen- tial forced-air furnace systems. This filter has a negligible effi- ciency for particles smaller than ~0.5 ,um and a low efficiency (e.g., < 20%) for particles smaller than 10 ,um. The other extreme is a Furnace filter (not oiled) ASHRAE 65% efficiency O Electrostatic precipitator · ASHRAE 40% efficiency OK ASHRAE 95% efficiency ~ HERA filter 80 100 - 5t 90t x0 0 0 0 70 60 50 40 30 20 10 O ~ ~K t~l ~ 01~. 0 ~OK . . · ~ - · t ~ Is-.. 0.01 0.1 1 10 Particle Size (,um) FIGURE 10-2 Efficiency curves for filters and an electrostatic precipitator.
362 CLEARING THE AIR high-efficiency particle air (HEPA) filter, which has a minimum efficiency of 99.97 °/O for 0.3-,um particles and a higher efficiency for smaller and larger particles. For indoor applications, the effi- ciency of a HEPA filter is effectively 100% at all particle sizes if all air that flows through the air cleaner actually passes through the filter. In addition to fibrous filters, a wide range of electronic air cleaners are available. Electrostatic precipitators first produce ions that attach to and electrically charge particles entering the air cleaner. These charged particles then pass through an electric field where they migrate to a surface and attach. The collection sur- faces must be cleaned or replaced periodically. An example of an efficiency curve for an electrostatic precipitator is also provided in Figure 10-2. An ion generator is another device that produces ions that attach to and charge particles, but ion generators often have no particle collection surfaces. Instead, ion generators may increase the rate of particle deposition on normal indoor surfaces. A variety of hybrid technologies employ both fibrous filters and particle charging or fibrous filters and electric fields. Electronic air cleaners can produce ozone, sometimes in significant quanti- ties (U.S. EPA, 1999; Viner et al., 1989~. Many units have a char- coal filter to remove the ozone produced by the air cleaner. The efficiency data in Figure 10-2 are from tests with all air passing through a previously unused filter. In practical installa- tions, a portion of the air bypasses the filter, flowing between ad- jacent filters or between filters and their housing, causing the air cleaner's particle removal efficiency to be lower than the effi- ciency of the filter media. In commercial installations, gaps of a few centimeters between adjacent filters, even missing filters, are common. Systems of higher-efficiency filters usually have fewer gaps and more gaskets to reduce the quantity of air that bypasses the filter media. Almost no information is available on typical by- pass rates. As filters accumulate deposited particles, their effi- ciency generally increases (sometimes markedly); however, the resistance to airflow through the filter also increases. The resistance to airflow as air passes through an air cleaner is a common concern of engineers, building operators, and HVAC equipment providers. If airflow rates are constant, more power- ful fans and increased fan energy are required when the air pres
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 363 sure drop increases. More efficient fibrous air filters tend to have a greater pressure drop. However, the pressure drop of high-effi- ciency filters can be diminished by increasing the degree of pleat- ing of the filter malt, which usually increases both the thickness of the filter in the direction of airflow and the filter cost. Addition- ally, a larger number of filters in parallel can be used to reduce the air velocity and pressure drop. Thus, pressure drop does not nec- essarily increase with increased filter efficiency; however, the cost of, and space required for, the filters increases unless a greater pressure drop is accepted. Electronic air cleaners tend to have smaller pressure drops than fibrous filters with the same particle removal efficiency. There are two basic installation options for particle air clean- ers. First, air cleaners can be installed within existing HVAC sys- tems and rely on the normal fan-driven airflow through the HVAC system to force air containing particles through the air cleaner. Air drawn from the occupied indoor space flows though a duct system, through the air cleaner, and then back to the occu- pied space. With this type of installation, often called "in-duct" air cleaning, the rate of air cleaning is limited by the rate of air- flow through the HVAC system, and air cleaning occurs only when the HVAC fan operates. For example, an air cleaner in a residential furnace system cleans air only when the residence is heated, unless the furnace fan is forced to operate at other times. One option for improving filtration in buildings is to increase the efficiency of the in-duct filters or air cleaners. The particle removal rate increases only for particles that are removed with a higher efficiency after the filtration upgrade. Hence, substitution of a high-efficiency in-duct air cleaner for a lower-efficiency device may dramatically increase the particle removal rate for small par- ticles but have a small or negligible influence on the removal rate for large particles. The second option is a supplemental air cleaner with an inte- gral fan that forces airflow through the air-cleaning devices. Nor- mally, these are portable devices placed on the floor, designed to clean the air predominantly in a single room. The addition of por- table air cleaners to a building with an in-duct air cleaner in- creases the total flow rate of filtered air. Portable air cleaners are often a significant source of noise. In some studies, occupants
364 CLEARING THE AIR have turned air cleaners off because this noise is annoying. The air flow rate of some portable air cleaners is too small for mean- ingful air cleaning within a room or house. Typical Particle Air-Cleaning Practices in U.S. Buildings In U.S. single-family homes with forced-air heating or air-con- ditioning systems, a filter designed to remove coarse particles is generally installed in the stream of air circulated through the heat- ing or air-conditioning system. Usually this is a coarse pane! fil- ter, such as the furnace filter with an efficiency versus particle size as depicted in Figure 10-2. Recently, filters and electronic air cleaners with a higher removal efficiency for small particles have been designed to replace the standard furnace filter. Standard fur- nace filters cost only a few dollars or less and should be replaced a few times per year. Higher-efficiency replacements for furnace filters often cost between $10 and $20 per filter. An unknown proportion of residential heating and cooling systems have an additional in-duct particle air cleaner within the recirculated airstream, often an electronic air cleaner. The price of supplemental in-duct air cleaners for residential applications var- ies widely (e.g., $500 to $800) among products. The cost of elec- tricity used to operate the electronic components of the air-clean- ing system is usually insignificantly (e.g., $25 per year). If the recirculation fan is operated only during heating or air condition- ing, there is no significant incremental cost for fan energy. Occa- sionally, home owners will operate the HVAC fan continuously in order to have continuous air cleaning. If 4,380 incremental hours of fan operation (50°/O of the year), a fan power of 300 W. and an electricity prices of $ 0.098 per kilowatt-hour (kWh) are assumed, the annual incremental fan energy cost would be $130. The rates of airflow through these air cleaners correspond to the rates of airflow through the residential heating and air-condition- ing system, with about four indoor air volumes per hour being typical. i5This excludes the cost of energy used by fans that drive airflow through the air cleaner and associated in-duct heating or air-conditioning system.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 365 In U.S. commercial and institutional buildings with HVAC systems, the system invariably contains a filter in the supply air- stream (i.e., in the mixture of outside air and recirculated indoor air. A typical supply air filter has an efficiency rating of roughly 40°/O (Figure 10-2~;~7 however, the efficiency rating for supply air filters varies widely among buildings. Historically, the rationale for these supply filters has been to reduce the deposition of large particles on the heat-transfer equipment inside HVAC systems. In the past few years, more attention has been directed toward filtration to protect human health. The cost of filters in commer- cial buildings varies with the product used. Using cost estimates from Burroughs (1997) for filters with a 30°/O ASHRAE efficiency rating, annual costs per person for a typical level of air filtration in a commercial building are on $4 to $8.~8 Fisk and Rosenfeld (1997) estimated that the annual incremental cost of using very high efficiency filters in an office building was $24 per person. 19 In a typical commercial HVAC system, the filter is upstream of many of the HVAC components, including the cooling coils, cod] drain pans, humidifiers (when present), and sections of duct work that can become contaminated with microorganisms. Thus, the supply filters do not prevent particles released from these components from entering the occupied spaces of the building. Supplemental portable air cleaners are not standard equip- ment in any particular type of building; however, they are widely available and used in many buildings, particularly residences. Although a very large range of products is available, many of the heavily marketed products incorporate HEPA filters and a multi- speed fan, with airflow rates at maximum fan speed ranging from 50 to 200 L so-. This range of flow rates corresponds to 8 to 32 room air volumes per hour in a small 22-m3 bedroom and to 0.4 to i6Electronic air cleaners may also be used in commercial buildings, but they are much less common than filters. i7The most common efficiency rating is the "dust spot" rating in ASHRAE Stan- dard 52.1-1992 (ASHRAE, 1992). i8A total supply flow rate per person of 70 L s-i (150 cfm) was assumed. Esti- mate includes cost of filters, labor, and energy. i995% efficiency for 0.3 ,um particles.
366 CLEARING THE AIR 1.6 house volumes per hour in a 450-m3 house. In practice, units may often be operated with less than maximum airflow or they may be unused, for example, because they are noisy. These por- table air cleaners are sized primarily for cleaning the air in single rooms. When the door to the room is open or a forced-air heating or cooling system operates, room air will often mix with air throughout the building and the air-cleaning system must have a larger capacity to obtain the same reduction of particle concentra- tions within the room. The retail cost of portable air cleaners var- ies widely. As an example, the retail cost of one of the most com- monly used HEPA room filter units with a maximum flow rate of 165 L so is about $250 ($1.50 per L so maximum airflow).20 Peri- odic replacement of pre-filters and HEPA filters in this unit would cost about $70 annually. At maximum fan speed, this unit con- sumes 350 W of electrical power, with an annual electrical cost of ~$300 if operated continuously. If a product life of four years is assumed, the annualized cost is roughly $400 ($2.4 per L so of maximum airflow), with 75°/O of this cost associated with electric- ity to operate the fan. The annualized costs per unit airflow of other commercially available products could be considerably higher or lower. Higher-capacity supplemental air cleaning units, sometimes mounted at ceiling level, are also readily available. These devices are intended primarily for use in health care facilities, smoking rooms, and restaurants, but they could be used in residences. Prices vary widely, and in many instances, there is no obvious relationship between product price and published specifications. Air Cleaner Standards Standard test procedures for assessing the efficiency of par- ticle air cleaners are available from ASHRAE (1992~. The most commonly cited current test methods yields an "ASHRAE dust- spot efficiency," hereafter called an ASHRAE efficiency, but this standard does not provide an efficiency rating versus particle size. 2020 room volumes per hour for a 30-m3 bedroom; 1.3 room volumes per hour for a 450-m3 house.
IMPACT OF VENT TON AND AIR CLEANING ON ASTHMA 367 HanIey et al. (1994) have provided efficiency curves for typical products. Air cleaner manufacturers also often provide size-de- pendent efficiency data upon request. At present, there are no standards that specify the minimum allowable filter efficiency in HVAC systems in U.S. buildings. A proposed revision of the ASHRAE minimum ventilation standard includes a minimum efficiency specification. . Predicted and Measured Influence of Particle Air Cleaning on Indoor Concentrations of Indoor-Generated Particles of Various Sizes Measurements Figure 10-3 presents the results of experimental studies of air cleaners that specify both the rate of airflow through the air cleaner and the reduction in indoor pollutant concentration. Some of the studies summarized later in Tables 10-4 and 10-5 also pro- vide a measured reduction in indoor pollutant concentration but no rate of airflow through the air cleaner. The studies specifying the rate of air cleaning are listed at the top of each table in order of decreasing air cleaner flow rate. Most experimental studies have used portable air cleaners with HEPA filters in rooms of homes. Ten studies provide some information on the decrease in airborne allergen or particle concentrations, usually within the room containing the air cleaner, associated with air cleaner opera- tion.22 Six of these studies reported large or statistically signifi- cant decreases in airborne particles or allergens. The reported per- centage reductions in particles or allergens in five studies ranged from 30 to 90°/O and averaged 60%. Overall, these data indicate that air cleaners can significantly reduce airborne allergens or par- ticles in some applications. However, these findings should be interpreted with caution. Temporal variations in indoor allergen Artless than 50% of the studies provide information on the rate of air cleaning. 220ne study checked for changes in dust mite concentrations in dust samples, a parameter unlikely to change as a consequence of short-term air cleaner opera- tion.
368 CLEARING THE AIR O Cat allergen Pollens and mold spores All particles > 0.3,um · Generated particles, 0.7,um mass median diameter ETS particles, room with 9 air changes per hour ventilation rate 0 100- O ~ 90 a.) 80 C' 70- ~O .~60 .~ 50- ~ ~O 40 30 ~ 20 0, 10 O O- 1 1 1 1 1 1 1 1 0 2 4 6 8 10 12 14 1 to l6 18 Flow Through Air Cleaner (room volumes per hour) FIGURE 10-3 Measured reductions in indoor pollutant concentrations with air cleaner operation. NOTE: Solid data points are for pollutants that are generated only indoors. SOURCES: Cat allergen Wood et al. (1998) and Blay et al. (1991~; all pollens and mold spores-Chang et al. (1998) and Nelson et al. (1933~: all particles > 0.3 ,um Reisman et al. (1990~; generated particles Miller-Leiden et al. (1996~; and ETS particles Bascom et al. (1996~. production or resuspension rates and in outdoor allergen concen- trations make accurate experimental assessments of the effects of air cleaning quite difficult. Most of the experimental studies have been short term, thereby increasing the potential errors from natu- ral fluctuations in allergen sources. The studies of de Blay et al. (1991) on the effects of air cleaning on airborne cat allergen con- centrations serve as an example of changes in allergen sources. The influence of air cleaner use on airborne cat allergen concen- tration varied widely (Figure 10-3), with large reductions occur- ring only if the room was previously vacuum cleaned, presum- ably because the resuspension of allergen from indoor surfaces
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 369 was enhanced by the higher indoor air velocities that occur with . . alr c .eamng. A few of the experimental studies provide information on the type of air cleaner needed to reduce concentrations of various types of pollutants. Van der Heide et al. (1997) used multistage filter systems and found that most of the dust mite antigen was removed by the coarse filters indicating that HEPA filters are unnecessary for the large particles associated with dust mite anti- gen. Miller-Leiden et al. (1996) determined that HEPA filter units did not perform significantly differently from non-HEPA units for particles with a mass median diameter of 0.7 ,um. Offermann et al. (1991) evaluated the effectiveness of six types of air cleaners installed in a forced-air furnace duct system for removal of envi- ronmental tobacco smoke. The highest rates of ETS removal oc- curred using a filter with an ASHRAE efficiency rating of 95°/O, followed by an electrostatic precipitator, and then by a HEPA fil- ter. The HEPA filter had a higher airflow resistance and thus, a lower airflow rate and smaller particle removal rate. Predictions The influence of various particle air-cleaning options on con- centration of indoor-generated particles was predicted using the steady-state mass balance equation and assumptions provided in Appendix A. Figure 10-4 provides predictions of the percentage reductions in indoor particle concentrations versus particle size for various air filtration systems installed within the forced-air heating and air-conditioning system of a house. Figure 10-5 pro- vides predictions of the effects of using portable filter units in isolated bedrooms or entire houses. All predictions are for spaces with perfectly mixed air. Several observations based on these predictions follow. First, to obtain a substantial reduction in indoor concentrations of 10- ,um particles, the rate of airflow through the air cleaner per unit of indoor air volume must be high. Even with 10 room volumes per hour through a filter with a 100% efficiency for 10-,um particles, the predicted reduction in concentration is only 70°/O. The effec- tiveness of air cleaning would diminish rapidly with increases in particle size above 10 ,um because of increases in gravitational
370 CLEARING THE AIR TABLE 10-4 Influence of Air Cleaner Use on Indoor Allergens and Allergy or Asthma Symptoms of Subjects with Perennial Allergic Disease Air Cleaner Flow Rate Change in Air Cleaner (room volumes Allergen Reference Setting Type per hour) Concentra Studies with Information on Air Cleaning Rate Criep and Green, Hospital rooms Electrostatic 25 (assuming Separate 1936 and homes precipitator in 30-m3 room); demons patient's room or 1.6 assuming reductio house 450-m3 house Reisman et al., 1990 Homes HEPA 17 (assuming 70% redly 30-m3 room); particle ~ 1.1 assuming for partial a 450-m3 house Wood et al., 1998 Homes with cats HEPA in bedrooms 15 bedroom Airborne c volumes; reduced 1.3 (assuming (p = .04! 450-m3 house) Mitchell and Elliott, Homes Electrostatic 9 (assuming Not asses' 1980 precipitator in 30-m3 room); bedroom 0.6 assuming 450-m3 house Warburton et al., 1994 Homes HEPA in bedroom 2-4 in bedroom Nonsignifi decrease gram-pc bacteria Studies Without Information on Air Cleaning Rate van der Heide, 1997 Homes (smokers Portable unit with Not specified in 15 of 45) coarse filter, cyclone, high efficiency filter Nonsignifi decrease . . positive and fund
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 371 llergy caner Rate ~ volumes our) Change in Allergen Concentrations Blinded or Subjects Placebo? Reported Change in Health Outcomes ssumlng me room); assuming ~-m3 house ssumlng me room); assuming '0-m3 house droom limes; (assuming ~-m3 house) ,umlng me room); assuming -m3 house bedroom necified Separate experiment demonstrated large reductions 70% reduction in particle concentration for particles >0.3 ,um Airborne cat allergen reduced 30% (p = .045) Not assessed Nonsignificant decrease in . . gram-posltlve bacteria and fungi Nonsignificant 45 non decrease in gram positive bacteria and fungi 61 patients No with some with asthma 32 allergic patients (children and adults) 35 Yes cat-allergic subjects Allergy symptoms improved moderately or completely in 47 of 61 subjects. No statistical test. Total study: no significant change in symptom or medication scores. Weeks without respiratory illnesses: several symptom and medication scores improved significantly but no change in asthma. No significant changes in symptom scores, peak flow rates, sleep disturbance, or medication use, although a few of these parameters improved nonsignificantly. 10 children Yes No significant change in peak flow. with asthma 12 atopic Placebo air adult cleaners, asthmatics with foam filtered smoking allergic asthmatic patients No significant changes in symptom scores, spirometry, or bronchial hyperreactivity. Nonsignificant trend toward higher peak flow with air cleaner use Placebo air Significant improvement in airway cleaners with hyperresponsiveness and coarse filters eosinophils, but only with both and cyclones air cleaners and impermeable mattress covers. No significant change in peak flow. Table continued on next page
372 TABLE 10-4 Continuecl CLEARING THE AIR Air Cleaner Flow Rate Change in Air Cleaner (room volumes Allergen Reference Setting Type per hour) Concentra Bowleretal., 1985 Homes Electrostatic Not specified, Not asses' precipitator and not operated HERA in bedroom continuously Friedlaender and Homes Electrostatic 6 (in 56-m3 Not asses' Friedlaender, 1954 precipitator room) Antonicelli et al., 1991 Homes HERA in bedroom Not specified No signifier in dust r in bedro samples Warneretal., 1993 Homes ionizer Not applicable Airborne r decrees' (p <.OC Studies with Fresh Air Delivery from or near the Headboard of the Bed Verall et al., 1988 Homes HEPA-filtered air Not applicable from bed's headboard Zwemer and Karibo, Homes 1 973 Villaveces et al., 1977 Homes Studies with Air Conditioners Trasoff and Hospital rooms Blumstein, 1936 Not asses' HEPA-filtered air Not applicable from bed's headboard HEPA-filtered air Not applicable directed over pillow Air conditioner 8 with filters Not asses' Not asses' Not asses' aPlacebo air cleaner may have removed some larger allergen particles.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 373 caner Rate Change in ~ volumes Allergen Blinded or Reported Change in our) Concentrations Subjects Placebo? Health Outcomes pacified, operated sinuously Not assessed 9 subjects Yes with asthma, age 14-31 No significant change in symptom score or peak expiratory flow. 56-m3 Not assessed 18 subjects No Excellent or moderate improvement m) (16 with in symptoms in 40% of subjects. asthma) No controls, no statistical tests. pacified No significant change 9 patients No No significant improvement in in dust mite allergen allergic to forced expiratory volume or in bedroom dust dust mites symptom score. samples pplicable Airborne mite allergen 20 children Yes No change in peak flow, symptoms, decreased by ~50% with or medication use. (p <.0001) asthma pplicable Not assessed 13 dust Yes Significantly less medication mite-allergic (p < .05) and significant asthmatics improvement in airway responsiveness (p=.05~. Crossover study. pplicable Not assessed 12 asthmatic Yes (but Total symptom score reduced 75%; children notfully) 300% increase in days of uninterrupted sleep, Lost school days decreased from 15 to 0. No statistical test. pplicable Not assessed 13 asthmatic Yes No significant improvement in peak children flow. Significantimprovementin symptom scores (p < .05~. Not assessed 22 (all with No No clear benefits, relative to asthma) patients in other hospital rooms. No statistical test.
374 CLEARING THE AIR TABLE 10-5 Influence of Air Cleaner Use on Indoor Allergens and Allergy or Asthma Symptoms in Subjects with Seasonal Allergic Disease Air Cleaner Flow Rate Change in Air Cleaner (room volumes Allergen Reference Setting Type per hour) Concentra Studies with Information on Air Cleaning Rate Criep and Green, 1936 Hospital rooms ESP 25 (assuming In separat and homes 30-m3 patient large ret room in artific introduc' Rappaport et al., 1932 Hospital ward Air filter unit, 10 Not asses' efficiency not specified Nelson et al., 1933 Hospital wards Air filter units, no 12 efficiency data Relative to wards, F reduced Scherr and Peck, Converted railroad One or two HERA 1.5 room Not asses' 1977 bunkcars units volumes per hour per air cleaner Studies Without Information on Air-Cleaning Rate Cohen, 1927 Home and work Filtered air 9 (assuming pressurizes room, 30-m3 rooms) no efficiency rating Not asses'
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 375 llergy or caner Rate ~ volumes our) Change in Allergen Concentrations Blinded or Subjects Placebo? Reported Change in Health Outcomes ssumlng me patient m tom limes per r per air Iner ,umlng me rooms) In separate experiments, 61 subjects No large reductions with in artificially allergies, introduced allergens some with asthma 105 patients No? with pollen allergy, some with asthma Not assessed Relative to control 51 patients No wards, pollen counts with hay reduced about 90% fever, 35 patients with pollen asthma 130 children Yes with asthma at summer camp Not assessed Not assessed 4 case No studies Symptoms improved in 77% of subjects. No statistical test The 35 patients with hay fever but no asthma experienced complete or moderate relief on 100 of 115 patient nights. Asthma symptoms and medication use persisted in many patients. However, data from 39 patients were excluded because they left after one night or had infections. 50% of hay fever patients reported great or complete symptom relief. Symptoms in 17 of 35 asthma patients improved or disappeared. No statistical tests. Borderline significant improvement in score for nighttime asthma attacks (p= .07~. Asthma and allergy symptoms improved in three subjects. Hay fever symptoms improved in one subject. No controls, self- assessment of symptoms, no statistical test. Table continued on next page
376 TABLE 10-5 Continuecl CLEARING THE AIR Air Cleaner Flow RateChange in Air Cleaner (room volumesAllergen Reference Setting Type per hour)Concentral Friedlaender and 9 homes and ESP 6 in 56-m3>90% red Friedlaender, 1954 3 in hospital roompollen, ~ measure with sub Kooistra et al., 1978 Homes with central Air cleaner in return Not specifiedAllergen c air conditioning duct of heating nonsign system, 99% lower wi efficiency at 6 ,um cleaner Studies with Air Conditioners Vaughan and Cooley, Hospitals Air conditioner 1799% reduce 1933 without filter, with pollen c' closed windows room wi window. air cond Trasoff and Blumstein, Hospital rooms Air conditioner with 8 Not asses' 1 936 filters NOTE: ESP = electrostatic precipitator. settling rates with particle size. Thus, air cleaning does not ap- pear to be a very attractive option for reducing exposures to dust mite allergens, which predominantly contain particles larger than 10 ,um. The second observation is that increasing the filter effi- ciency rating from a furnace filter to an ASHRAE 95% efficiency filter, while maintaining a constant rate of airflow through the filter, decreases the predicted indoor concentrations by approxi- mately 40°/0 or less for particles 5 ,um or larger. An equivalent re- duction in concentrations of particles within this size range is ob- tained using an ASHRAE 65% filter in place of the ASHRAE 95% filter. Thus, for many of the bioaerosols associated with asthma, high-efficiency filters, which are more expensive and often require larger and noisier fans, are not likely to be superior to moderate
IMPACT OF VENTI~TION AND AIR CLEANING ON ASTHMA 377 caner late volumes fur) Change in Allergen Blinded or Concentrations Subjects Placebo? Reported Change in Health Outcomes 16-m3 m necified >90% reduction in pollen, but not measured in rooms with subjects 12 (6 with No asthma) Allergen concentrations 20 adults nonsignificantly with hay lower with air fever, 6 of cleaner use these with asthma Yes 99% reduction in 2 No pollen compared to room with open windows and no air conditioner Not assessed 10 (all with No asthma) Excellent relief in 75% of subjects and moderate relief in 25%. No controls, no statistical tests. 14% decrease in nocturnal symptoms, significant only at nighttime. Significant daytime improvement in 10 most sensitive subjects. Moderate improvement in one of two subjects. No controls, no statistical test. Marked improvement in symptoms in 90% of subjects. No controls, no statistical test. efficiency filters. Third, it appears feasible to reduce concentra- tions of particles smaller than 2 ,um, such as ETS particles, droplet nuclei, and the smaller particles of cat allergen, by 70°/O or more using air cleaners with a moderate to high efficiency rating and a flow rate of several indoor air volumes per hour. Figure 10-6 provides predicted reductions in indoor-gener- ated particle concentrations from the use of typical filters and higher-efficiency filters in a commercial and institutional build- ing HVAC system. These upgrades could substantially reduce concentrations of submicron particles but have virtually no effect on particles larger than a few micrometers. Portable air cleaners could be used to reduce concentrations of the larger-size indoor-generated particles in the air within com
378 ~ 100 o . _ . _ .o tL 90 80 70 60 50 40 30 20 10 O CLEARING THE AIR O Furnace filter with continuous furnace fan operation · Furnace filter with half-time furnace fan operation vs. no filter ASHRAE 65% efficiency filter with half-time furnace operation vs. no filter ASHRAE 95% efficiency filter with half-time furnace fan operation vs. no filter O ASHRAE 95% efficiency filter with continuous furnace fan operation vs. no filter | $ 1 o 00 0 Y ~ `. ~0 ox . o 1 1 2 4 o 1 6 8 10 12 Particle Size (,um) FIGURE 10-4 Predicted reduction in indoor-generated particles in a house with use of various filters in the forced-air heating system if a recirculation flow of four house volumes per hour and a ventilation rate of 0.75 h-1 are assumed. mercial buildings. As in residences, high rates of airflow through the filters would be necessary to obtain substantial percentage reductions in indoor concentrations. Use of portable air cleaners in isolated individual offices of asthmatics is not likely to be effec- tive unless the office door is kept closed, and even with closed doors the reductions in indoor particle concentrations may be quite small. Effects of Particle Air Cleaning on Allergy and Asthma Symptoms The influence of air cleaner use on asthma and allergy out- comes has been evaluated in many experimental studies, and was the subject of a 1997 review by the American Lung Association (ALA, 1997~. Building on the review articles of Nelson et al. (1988)
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 379 · Portable HERA filter with 10 room volumes per hour air flow in isolated bedroom with no other air cleaning OK Portable ASHRAE 65% efficiency filter in an isolated bedroom with no other air cleaning · Portable HERA filter with air flow of 2 house volumes per hour in house with a furnace filter Portable ASHRAE 65% efficiency filter with air flow of 5 house volumes per hour in a house with a furnace filter 100 o In 90 ~ 80 c) o 70- C~ .~ 60 .°~ 50 c) ' 40 30 20 10 O ~ · ~ _ 1 1 1 1 1 1 0 2 4 6 8 10 12 Particle Size (,um) FIGURE 10-5 Predicted reduction in indoor-generated particles from op- eration of portable air cleaners in an isolated bedroom or an entire house, where a ventilation rate of 0.75 h-i is assumed. NOTE: Isolated bedroom: bedroom door closed and no air circulation between the bedroom and other rooms. Furnacefilter: furnace recirculates air through- out the house. and de Blay et al. (1997a), these studies are summarized in Tables 10-4 and 10-5 for subjects with perennial and seasonal symptoms, respectively.23 Many of these experimental studies have important limita- tions, including a very small number of subjects (e.g., <15), lack of blinding, low or unspecified rates of air cleaning, no informa- tion on building ventilation rates, virtually no specification of rel 23The committee is using the perennial versus seasonal categorization of Nelson et al. (1988) and de Blay et al. (1997a), but the proper categorization of individual studies is sometimes ambiguous.
380 o ._ ~ o ~ o c) CLEARING THE AIR 100 90 80 70 - 60 ' 50 - 40 ~ 30 20 10 O · ASHRAE 40% filter (base-case) vs. no filter ASHRAE 95% filter vs. no filter ASHRAE 65% filter vs. no filter . 0 2 1 1 1 1 ~ 4 6 8 10 12 Particle Size (,um) FIGURE 10-6 Predicted reduction in indoor-generated particles from air filtration in a commercial building HVAC system with a recirculation rate of three indoor volumes per hour and a ventilation rate of 0.75 hat. event buildings characteristics,24 short-term study periods, and reliance on self-reports of symptoms. Many studies assessed changes in allergy symptoms rather than clear indications of asthma symptoms. Some studies selected subjects who were al- lergic to dust mites, and as discussed above, air cleaning is un- likely to be highly effective in reducing exposures to the large particles that carry dust mite allergens. A minority of the studies quantified the reduction in airborne allergen levels. Many studies reported whether changes in outcomes were significant but did not provide relative risks or odds ratios. Thus, the results of these studies are suggestive, rather than a basis for firm conclusions. Most of the studies involving subjects with perennial allergic or asthma symptoms (Table 10-4) used portable air cleaners, most often portable HEPA filter units, in bedrooms. Large reductions 24Important building and building operation characteristics not reported in the literature include the normal position of doors (open or closed) in bedrooms with air cleaners, the presence or absence of carpets, and the type of building heating system (e.g., forced air recirculating system with a filter).
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 381 in concentrations of allergens within these bedrooms would not be expected unless bedroom doors were kept closed and forced- air heating or cooling systems did not mix air between the bed- rooms and the remainder of the house. Excluding three studies with filtered air delivery over the bed, only 4 of 11 studies re- ported improvements in symptoms or reduced use of medica- tion, and in 3 of these studies the subjects were not blinded. In one of the four studies (Reisman et al., 1990), symptoms and medi- cation use improved significantly only during the period when patients had no respiratory illnesses. In another of the four stud- ies, only air cleaners used in combination with impermeable mat- tress covers were associated with significant improvements in air- way hyperresponsiveness and eosinophils.25 Three experimental studies involving subjects with perennial allergic or asthma symptoms (bottom of Table 10-4) used air- cleaning devices that supplied filtered air over the bed. All three studies reported improvements in outcomes, suggesting that sup- plying cleaned air to the breathing zone may be more effective than attempting to clean the air in entire rooms or buildings. The results of studies involving subjects with seasonal aller- gic or asthma symptoms are much more encouraging. Excluding the two studies using only air conditioning as a form of air clean- ing, six of seven studies reported improvements in symptoms, and the seventh study showed a borderline significant improve- ment (p = 0.07~. However, subjects were blinded in only two of these studies, and most studies were old and without formal sta- tistical tests. One additional experimental study (Bascom et al., 1996) per- formed in an experimental chamber (not included in the tables) assessed the reduction in acute symptoms from ETS exposure when particle air cleaners are used to reduce concentrations of ETS particles. With a 50°/O reduction in particle concentrations, some symptoms (headache, eye irritation, rhinorrhea) and the minimum cross-sectional area of the nasal passage improved. The relevance of these findings for asthma is not clear. Overall, these data suggest that air cleaners are probably help 25Impermeable mattress covers are intended to reduce exposures to dust mite allergens.
382 CLEARING THE AIR fu] in some situations in reducing allergy or asthma symptoms, but air cleaning, as applied in the studies, is not consistently and highly effective in reducing symptoms. It is conceivable that a more consistent or larger reduction in symptoms could be ob- tained using higher rates of air cleaning. The available data pro- vide no information regarding the effects of air cleaning on the development of asthma or the development of sensitization to allergens. Air Conditioning as ~ Substitute for Air Cleaning Previous reviews of air cleaning and asthma have included studies of air conditioning. Air conditioners can remove some particles from the indoor air because they often contain coarse particle filters and because some particles may be removed from the air along with water in the air conditioner's cooling coil. Con- sequently, Table 10-5 also summarizes the results of two investi- gations in the 1930s that evaluated the effects of air conditioning on symptoms. One study demonstrated that the use of an air con- ditioner and closing windows resulted in a large reduction in in- door pollens. Symptoms improved in one of two subjects. In the second study the air conditioners contained filters, and marked improvements in asthma symptoms were reported for 9 of 10 sub- jects. Neither study was blind, and neither performed a statistical test to assess the significance of the change in outcomes. Air conditioners are not designed to remove particles from air. Air conditioning will, in general, be less effective than con- ventional air cleaning in reducing exposures to indoor-generated particles. However, air conditioners enable the occupants of build- ings to keep windows and doors closed during warm weather, greatly reducing the rate of entry of pollens and other outdoor pollutants into buildings. Additionally, air conditioners reduce indoor temperatures and humidities that may influence asthma symptoms. Home owners and medical doctors sometimes consider resi dential air conditioning a form of ventilation. In the general mode of operation, residential air conditioners do not provide ventila- tion they simply remove heat and moisture (and, to some de- gree, particles) from a stream of recirculated indoor air. However,
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 383 to save energy, some modern air conditioners will stop the recir- culation and cooling of indoor air and instead provide ventilation (i.e., supply outdoor air to the building) when the outdoor air is relatively coo! and suitable for cooling the house. Summary and Discussion of Limitations of Assessment Many of the indoor pollutants associated with asthma are air- borne particles; thus, particle air cleaning has been considered a potentially beneficial technology for the prevention of asthma or asthma symptoms. Technologies for particle air cleaning are well developed. Air filters with a moderate to high efficiency for par- ticles larger than approximately 2 ,um are used routinely in the heating and air-conditioning systems of buildings. The magnitude of the reduction in indoor-generated particle concentrations accomplished with particle air cleaning depends on the air cleaner's particle removal rate relative to the particle removal rate by all other processes including ventilation and par- ticle deposition on surfaces. The rate of particle removal by an air cleaner varies with particle size and is proportional to the flow rate of air through the air cleaner multiplied by the air cleaner's size-dependent particle removal efficiency. The two primary air cleaning options for reducing indoor particle concentrations are to replace the existing filters in heating and air-conditioning sys- tems with higher-efficiency filters and to operate supplemental air cleaners with integral fans in the occupied space. In field studies, enhanced air cleaning has been associated with reductions in airborne particle concentrations that range from negligible to more than 90°/0. For the airborne particles asso- ciated with asthma, the published data are very limited. Simple mode] predictions indicate that substantial reductions in indoor concentrations of 10-,um particles can be obtained only when the rate of airflow through the air cleaner per unit of indoor air vol- ume is large, for example, 10 room volumes per hour or more. The predicted effectiveness of air cleaning diminishes rapidly with increases in particle size above 10 ,um because gravitational settling rates increase with particle size. Thus, air cleaning does not appear to be an attractive option for reducing exposures to dust mite allergen, which predominantly involves particles larger
384 CLEARING THE AIR than 10 ,um. However, based on predictions it is feasible to reduce concentrations of particles smaller than 2 ,um, such as ETS par i] ticles, droplet nuclei, and smaller particles with cat allergen, by 70°/O or more using air cleaners with a moderate to high efficiency rating and a flow rate of several indoor air volumes per hour. Both the available experimental data and mode] predictions ndicate that HEPA filters, which are more expensive and often require larger and noisier fans, are not likely to be superior to lower-efficiency filters in reducing concentrations of many of the bioaerosols associated with asthma. Even for submicron-size ETS particles, available data indicate that HEPA filters are not neces- sarily the preferred option. Thus, the very common recommenda- tion that HEPA filtration, in contrast to lower-efficiency air clean- ing, be used by allergic and asthmatic individuals when they choose to employ air cleaning, is not supported by either experi- ments or theoretical predictions. Unfortunately, the limited per- formance data available for many non-HEPA residential air clean- ers make it difficult to provide alternate recommendations. The influence of air cleaner use on asthma and allergy out- comes has been evaluated in numerous experimental studies; however, most of these studies have important limitations. Over- all, the data suggest that air cleaners are helpful in some situa- tions in reducing allergy or asthma symptoms, particularly sea- sonal symptoms, but it is clear that air cleaning, as applied in the studies, is not consistently and highly effective in reducing symp- toms. The available data provide no information regarding the effects of air cleaning on the development of asthma or the devel- opment of sensitization to allergens. Conclusion Regarding Air Cleaning and Asthma There is limited or suggestive evidence that particle air clean- ing is associated with a reduction in the exacerbation of asthma symptoms. There is insufficient evidence to determine whether or not the use of particle air cleaners is associated with decreased asthma development. Theoretical and limited empirical data sug- gest that air cleaners are most likely to be effective in reducing the indoor concentrations of particles smaller than approximately 2 ,um. Much of the airborne allergen appears to be within larger
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 385 particles. Relevant particles smaller than 2 ,um include environ- mental tobacco smoke particles, significant portions of airborne cat, grass, and birch allergen, and virus-containing droplet nuclei from coughs and sneezes. Research Needs Related to Air Cleaning and Asthma The results of existing experimental studies are inadequate to draw firm conclusions regarding the benefits of air cleaning for asthmatic and allergic individuals. Many of the existing studies have important limitations, such as small study size, lack of blind- ing, a small or undefined rate of air cleaning, placebo air cleaners that may significantly remove the larger particles associated with asthma, and no exposure assessment or inadequate assessment. Additional research to assess the benefits of air cleaning is clearly warranted, but future studies must overcome as many of these limitations as possible. Because air cleaning is most promising for reducing indoor concentrations of particles smaller than a couple of micrometers, future research should emphasize these agents. Sensitization to allergens a critical step in the development of allergic asthma often occurs early in life. No information is available to indicate whether air cleaning of spaces occupied early in life can reduce the rate of allergic sensitization. Research is needed to address this issue. As described in Appendix A, particles larger than a few mi- crometers have a complex and inadequately understood behav- ior in the indoor environment, including rapid rates of gravita- tional settling, resuspension from surfaces, and possibly incomplete mixing with the indoor air. Consequently, the influ- ence of air cleaning systems on exposures to particles in this size range is not well understood and the associated benefits from air- cleaning cannot be predicted with a high degree of confidence. A combination of aerosol science and air-cleaning research is needed to fill this gap in our knowledge. The limited data on the size distribution of many of the bioaerosols and allergens associated with asthma limit our un- derstanding of the benefits of air cleaning. Additional data are needed particularly for pet allergens and pollens. As stated earlier, HEPA filter units have been widely recom
386 CLEARING THE AIR mended for allergy and asthma patients who desire to use air cleaners. Air cleaner manufacturers have responded by aggres- sively marketing air cleaners with HEPA filters and offering few other products. However, experimental data and theoretical pre- dictions indicate that air cleaners with a lower efficiency rating are likely to be equally effective in reducing the concentrations of most, and perhaps all, of the indoor-generated particles associ- ated with allergies and asthma. These lower-efficiency air clean- ers could have a lower product cost, less powerful or noisy fans, higher rates of airflow and particle removal, and reduced energy consumption. The scientific and medical community should de- velop revised recommendations regarding the selection of air cleaners by allergic and asthmatic individuals, and air cleaner manufacturers should respond by providing new air-cleaning products. REFERENCES Aberg N. Sundell J. Eriksson B. Hesselmar B. Aberg B. 1996. Prevalence of allergic diseases in schoolchildren in relation to family history, upper respiratory infections, and residential characteristics. Allergy 51~4~:232-237. American Lung Association. 1997. Residential air cleaning devices: types, effectiveness, and health impact. American Lung Association, Washington, DC. American Thoracic Society. 1997. American Thoracic Society Workshop, Achieving Healthy Indoor Air. American Journal of Respiratory and Critical Care Medicine 156(Suppl 3~:534-564. Antonicelli L, Bilo MB, Pucci S. Schou C, Bonifazi F. 1991. Efficacy of an air cleaning device equipped with a high efficiency particulate air filter in house dust mite allergy. Allergy 46~8~:594-600. ASHRAE (American Society of Heating, Refrigerating, and Air-Conditioning Engineers). 1992. ASHRAE Standard 52.1-1992-Gravimetric and Dust-Spot Procedures for Testing Air-Cleaning Devices Used in General Ventilation for Removing Particulate Matter. Atlanta, GA: American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. ASHRAE. 1996. 1996 ASHRAE Handbook: HVAC Systems and Equipment. Atlanta, Georgia: American Society of Heating, Refrigerating, and Air- Conditioning Engineers, Inc. ASHRAE. 1999. ASHRAE Standard 62-1999. Ventilation for Acceptable Indoor Air Quality. Atlanta, GA: American Society of Heating, Refrigerating, and Air- Conditioning Engineers, Inc.
IMPACT OF VENTI~TION AND AIR CLEANING ON ASTHMA 387 Bascom R. Fitzgerald TK, Kesavanathan J. Swift DL. 1996. A portable air cleaner partially reduces the upper respiratory response to sidestream tobacco smoke. Applied Occupational and Environmental Hygiene 11~6~:553-559. Batterman SA, Burge H. 1995. HVAC systems as emission sources affecting indoor air quality a critical review. International Journal of HVAC and Research 1~1~:61-80. Bencko V, Maelichercik J. Melichercikova V, Wirth Z. 1993. Microbial growth in spray humidifiers of health facilities. Indoor Air 3~1~:20-25. Berglund B. Johansson I, Lindvall T. 1982. The influence of ventilation on indoor/ outdoor air contaminants in an office building. Environment International 8:395-399. Bowler SD, Mitchell CA, Miles J. 1985. House dust control and asthma: A placebo- controlled trial of cleaning air filtration. Annals of Allergy 55~3~:498-500. Burroughs HE. 1997. Filtration: an investment in IAQ. Heating, Piping, and Air Conditioning (August):55-65. Cohen MB. 1927. The prophylaxis and treatment of hay fever and asthma in rooms made pollen and dust free by means of mechanical filters. Journal of Laboratory and Clinical Medicine 13:59-63. Criep LH, Green MA. 1936. Air cleaning as an aid in the treatment of hay fever and bronchial asthma. Journal of Allergy 7:120-133. Custovic A, Simpson A, Pahdi H. Green RM, Chapman M, Woodcock A. 1998. Distribution, aerodynamic characteristics, and removal of the major cat allergen Fel d 1 in British homes. Thorax 53~1~:33-38. de Blay F. Chapman MD, Platts-Mills TAP. 1991. Airborne cat allergen (Fel D 1~: Environmental control with the cat in-situ. American Review of Respiratory Disease 143~6~:1334-1339. de Blay F. Sanchez J. Hedelin G. Perez-Infante A, Verot A, Chapman M, Pauli G. 1997a. Dust and airborne exposure to allergens derived from cockroach (Blattella germanica) in low-cost public housing in Strasbourg (France). Journal of Allergy and Clinical Immunology 99~1 Pt 1~:107-112. de Blay F. Colas F. Richard MC, Ott M, Verot A. 1997b. Air cleaners and airborne allergens. Journal of Investigational Allergology and Clinical Immunology 7~5~:335-337. Diamond RC. 1999. An Overview of the U.S. Building Stock. In: Indoor Air Quality Handbook. Spengler JD, Samet JM, McCarthy JF, Eds. McGraw Hill, ~n-press. Diamond RC, Modera MP, Feustel HE. 1986. Ventilation and occupant behavior in two apartment buildings. In: Proceedings of the 7th Air Infiltration and Ventilation Centre Conference. Coventry, Great Britain: Air Infiltration and Ventilation Centre, pp. 6.1-6.18. Diamond RC, Feustel HE, Matson NE. 1999. Energy Efficient Ventilation for Apartment Buildings. U.S. Department of Energy, Rebuild America technical guide series. July. Dietz RN, Cote EA. 1982. Air infiltration measurements in a home using a convenient perfluorocarbon tracer technique. Environment International 8~1- 6~:419-433.
388 CLEARING THE AIR DOE-EIA. 1994. Commercial Building Characteristics 1992. Report DOE/EIA- 0246~92), (U.S. Department of Energy, Energy Information Administration), Washington DC: U.S. Government Printing Office. DOE-EIA. 1995. Commercial Buildings Energy Consumption and Expenditures 1992. Report DOE/EIA-0318~92), (U.S. Department of Energy, Energy Information Administration), Washington DC: U.S. Government Printing Office. Dots SW, Persily AK. 1994. Measurements of Outdoor Air Distribution in an Office Building. Report NISTIR 5320. Gaithersburg, MD: National Institute of Standards and Technology. Emenius G. Egmar A, Wickman M. 1998. Mechanical ventilation protects one- story single dwelling houses against increased air humidity, domestic mite allergens and indoor pollutants in a cold climate region. Clinical and Experimental Allergy 28~11~:1389-1396. Emmerich S. Persily A. 1998. Energy impacts of infiltration and ventilation in U.S. office buildings using multi-zone airflow simulation. Proceedings of IAQ & Energy 98. Atlanta, GA: American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc., pp. 191-203. Fisk WJ. 1999. Estimates of potential nationwide productivity and health benefits from better indoor environments: an update. Lawrence Berkeley National Laboratory Report, LBNL-42123, To be published as a chapter in Indoor Air Quality Handbook. Spengler J. Samet JM, McCarthy IF, Eds. McGraw Hill, in press. Fisk WJ, Faulkner D. 1992. Air exchange effectiveness in office buildings: measurement techniques and results. In: Proceedings of the 1992 International Symposium on Room Air Convection and Ventilation Effectiveness, July 22- 24, Tokyo, pp. 213-223, ASHRAE, Atlanta. Fisk WJ, Rosenfeld AH. 1997. Estimates of improved productivity and health from better indoor environments. Indoor Air 7~3~:158-172. Fletcher AM, Pickering CAC, Custovic A, Simpsom J. Kennaugh J. Woodcock A. 1996. Reduction in humidity as a method of controlling mites and mite allergens: the use of mechanical ventilation in British domestic dwellings. Clinical and Experimental Allergy 26~9~:1051-1056. Friedlaender S. Friedlaender AS. 1954. Effectiveness of portable electrostatic precipitator in elimination of environmental allergens and control of asthma symptoms. Annals of Allergy 12:419-428. Gerone PI, Couch RB, Keefer GV, Douglas RG, Derrenbacher ED, Knight V. 1966. Assessment of experimental and natural viral aerosols. Biological Reviews 30~3~:576-588. Hanley IT, Ensor DS, Smith DD, Sparks LE. 1994. Fractional aerosol filtration efficiency of in-duct ventilation air cleaners. Indoor Air 4~3~:169-178. Harving H. Korsgaard J. Dahl R. 1994. House-dust mite exposure reduction in specially-designed, mechanically ventilated "healthy" homes. Allergy 49~9~:713-718. Hinds WC. 1982. Aerosol Technology. New York: John Wiley & Sons.
IMPACT OF VENT TON AND AIR CLEANING ON ASTHMA 389 Hodgson AT, Rudd AF, Beat D, Chandra S. 1999. Volatile organic compound concentrations and emission rates in new manufactured and site built houses. LBNL-43519, Lawrence Berkeley National Laboratory Report, Berkeley, California. (in draft). Holmquist L, Vesterberg O. 1999. Quantification of birch and grass pollen allergens in indoor air. Indoor Air 9~2~:85-91. Johnston SL, Pattemore PK, Sanderson G. Smith S. Lampe F. Josephs L, Symington P. O'Toole S. Myint SH, Tyrrell DA, Holgate ST. 1995. Community study of role of viral infections in exacerbations of asthma in 9-11 year old children. British Medical Journal 310~6989~:1225-1229. Kooistra JB, Pasch R. Reed CE. 1978. The effects of air cleaners on hay fever symptoms in air-conditioned homes. Journal of Allergy and Clinical Immunology 61~5~:315-319. Lagus Applied Technologies. 1995. Air Change Rates in Non-Residential Buildings in California. Report P400-91-034BCN, Prepared for the California Energy Commission by Lagus Applied Technology, Inc., San Diego, CA. Liddament MW. 1996. A Guide to Energy Efficient Ventilation. Document AIC- TN-VENTGUIDE-1996. Coventry, Great Britain: Air Infiltration and Ventilation Centre, International Energy Agency. Luczynska CM, Li Y. Chapman MD, Platts-Mills T. 1990. Airborne concentrations and particle size distribution of allergen derived from domestic cats (Felis domesticus). Measurement using cascade impactor, liquid impinger, and two- site monoclonal antibody assay for Fel d 1. American Review of Respiratory Disease 141~2~:361-367. Martiny H. Moritz M, Ruden H. 1994. Occurrence of microorganisms in different filter media of heating, ventilation and air conditioning (HVAC) systems. In: Proceedings of IAQ '94 Engineering Indoor Environments, Atlanta, GA: ASHRAE. pp. 131-137. McIntyre DA. 1992. The control of house dust mites by ventilation: a pilot study. In: Proceedings of the 13th AIVC Conference Ventilation for Energy Efficiency and Optimum Indoor Air Quality. Air Infiltration and Ventilation Centre, Coventry, Great Britain. pp. 497-507. Mendell MJ. 1993. Non-specific symptoms in office workers: a review and summary of the epidemiologic literature. Indoor Air 3~4~:227-236. Menzies R. Tamblyn R. Farant JP, Hanley J. Nunes F. Tamblyn R. 1993. The effect of varying levels of outdoor-air supply on the symptoms of sick building syndrome. New England Journal of Medicine 328~12~:821-827. Miller SL. 1996. Characterization and Control of Exposure to Indoor Air Pollutants Generated by Occupants. Ph.D. Dissertation, Civil and Environmental Engineering, University of California, Berkeley. Miller-Leiden S. Lobascio C, Nazaroff WW, Macher JM. 1996. Effectiveness of in- room air filtration and dilution ventilation for tuberculosis infection control. Journal of the Air and Waste Management Association 46:869-882. Mitchell EA, Elliott RB. 1980. Controlled trial of an electrostatic precipitator in childhood asthma. Lancet 2~8194~:559-561.
390 CLEARING THE AIR Modera MP, Brunsell IT, Diamond RC. 1986. Improving diagnostics and energy analysis for multi-family buildings: a case study. In: Proceedings, Thermal Performance of the Exterior Envelopes of Buildings III, SP 49. Atlanta, GA: ASHRAE. pp. 689-706. Morey PR.1988. Microorganisms in buildings and HVAC systems: a summary of 21 environmental studies. In: Proceedings of IAQ'88 Engineering Solutions to Indoor Air Problems. Atlanta, GA: ASHRAE. pp. 10-24. Morey PR. 1994. Suggested guidance on prevention of microbial contamination for the next revision of ASHRAE Standard 62, In: Proceedings of IAQ'94- Engineering Indoor Environments. Atlanta, GA: ASHRAE, pp. 139-148. Morey PR, Williams CM. 1991. Is porous insulation inside an HVAC system compatible with a healthy building? In: Proceedings of IAQ'91 Healthy Buildings. Atlanta, GA: ASHRAE, pp. 128-135. Murry DM, Burmaster DE. 1995. Residential air exchange rates in the United States: empirical and estimated parametric distributions by seasonal and climatic region. Risk Analysis 15~4~:459-465. Nagda N. Koontz M, Lumby D, Albrecht R. Rizzuto J. 1990. Impact of increased ventilation rates on office building air quality. Proceedings of Indoor Air 90(4):281-286. Nazaroff WW, Gadgil AG, Weschler CJ. 1993. Critique of the Use of Deposition Velocity in Modeling Indoor Air Quality. ASMM Standard Technical Publication 1205. Philadelphia, PA: American Society for Testing and Materials, pp. 81-104. Nelson HS, Hirsch SR, Ohman JL, Platts-Mills TAE, Reed CE, Solomon WR. 1988. Recommendations for the use of residential air-cleaning devices in the treatment of allergic respiratory diseases. Journal of Allergy and Clinical Immunology 82~4~:661-669. Nelson T. Rappaport BZ, Walker WH.1933. The effect of air filtration in hay fever and pollen asthma: further studies. Journal of the American Medical Association 100:1385. Offermann FJ, Hollowell CD, Nazaroff WW, Roseme GD, Rizzuto JR. 1982. Low infiltration housing in Rochester New York: a study of air exchange rates and indoor air quality. Environment International 8:435-445. Offermann FJ, Loiselle SA, Sextro RG. 1991. Performance comparisons of six different air cleaners installed in a residential forced air ventilation system. Proceedings of IAQ'91. Atlanta, GA: ASHRAE. pp. 342-350 Orme M. 1998. Energy impact of ventilation. Technical Note 49. Coventry, Great Britain: International Energy Agency Air Infiltration and Ventilation Centre. Ozkaynak H. Xue J. Spengler J. Wallace L, Pellizzari E, Jenkins P. 1996. Personal exposure to airborne particles and metals: results from the Particle TEAM Study in Riverside, California. Journal of Exposure Analysis and Environmental Epidemiology 6~1~:57-78. Pasanen PO, Pasanen AL, Kalliokoski P. 1995. Hygienic aspects of processing oil residues in ventilation ducts. Indoor Air 5~1~:62-68.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 391 Pehkonen E, Rantio-Lehtimaki A. 1994. Variations in airborne pollen antigenic particles caused by meteorologic factors. Allergy 49~6~:472-477. Pejtersen J. 1996. Sensory pollution and microbial contamination of ventilation filters. Indoor Air 6~4~:239-248. Persily AK. 1999. Myths about building envelopes. ASHRAE Journal 41~3~:39-47. Persily AK, Grot RA. 1985. Ventilation measurements in large office buildings. ASHRAE Transactions 91~2A):488-502. Persily A, Norford L. 1987. Simultaneous measurements of infiltration and intake in an office building. ASHRAE Transactions 93~2~:42-56. Platts-Mills TA, Heymann PW, Longbottom JL, Wilkins SR. 1986. Airborne allergen associated with asthma: particle sizes carrying dust mite and rat allergens measured with a cascade impactor. Journal of Allergy and Clinical Immunology 77~6~:850-857. Rantio-Lehtimaki A, Viander M, Koivikko A. 1994. Airborne birch pollen antigens in different particle sizes. Clinical and Experimental Allergy 24~1~:23-28. Rappaport BZ, Nelson T. Walker WH. 1932. Effect of air filtration in hay fever and pollen asthma. Journal of the American Medical Association 98:1861. Reisman RE, Mauriello PM, Davis GB, Georgitis JW, DeMasi JM. 1990. A double blind study of the effectiveness of a high-efficiency particulate air (HEPA) filter in the treatment of patients with potential allergic rhinitis and asthma. Journal of Allergy and Clinical Immunology 85~6~:1050-1057. Roberson JA, Brown RE, Koomey JG, Greenberg SE. 1998. Recommended Ventilation Strategies for Energy Efficient Homes. LBNL-40398, Lawrence Berkeley National Laboratory, Berkeley, CA. Schappi GF, Suphioglu C, Taylor PE, Knox RB. 1997. Concentrations of the major birch tree allergen Bet v 1 in pollen and respirable fine particles in the atmosphere. Journal of Allergy and Clinical Immunology 100~5~:656-661. Scherr MS, Peck LW. 1977. The effects of high efficiency air filtration system on nighttime asthma attacks. West Virginia Medical Journal 73~7~:144-148. Seppanen O. 1998. Ventilation strategies for good indoor air quality and energy efficiency. In: Proceedings of IAQ and Energy 98 Using ASHRAE Standards 62 and 90.1. Atlanta, GA: ASHRAE, pp. 257-276. Seppanen O. Fisk WJ, Mendell MJ. 1999. Association of ventilation rates and CO2- concentrations with health and other responses in commercial and institutional buildings. Indoor Air 9~4~:226-252. Shapiro-Baruch I. 1993. Evaluation of Ventilation in Multifamily Dwellings. Report 93-5. Albany, NY: New York State Energy Research and Development Authority. Shaughnessy RJ, Turk B. Casey M, Harrison J. Levetin E. 1997. Use of energy recovery ventilators to provide ventilation in schools and the impact on indoor contaminants. Proceedings of Health Buildings'97 1:161-166. Shaughnessy RJ, Levetin E, Rogers C. 1998. The effects of UV-C on biological contamination of AHUs in a commercial office building: Preliminary Results. In: Proceedings of IAQ and Energy 98 Using ASHRAE Standards 62 and 90.1. Atlanta, GA: ASHRAE, pp. 229-236.
392 CLEARING THE AIR Sherman MH, Dickerhoff D. 1994. Air-tightness of U.S. dwellings. Proceedings of the 15th AIVC Conference The Role of Ventilation. Coventry, Great Britain.: Air Infiltration and Ventilation Centre, pp. 226-234. Sherman MH, Matson N. 1997. Residential ventilation and energy characteristics. ASHRAE Transactions 103~1~:717-730. Spieksma FT, Kramps JA, van der Linden AC, Nikkels BH, Plomp A, Koerten HK, Dijkman JH. 1990. Evidence of grass-pollen allergenic activity in the smaller micronic atmospheric aerosol fraction. Clinical and Experimental Allergy 20~3~:273-280. Spieksma FT, Kramps JA, Plomp A, Koerten HK. 1991. Grass pollen allergen carried by the smaller micronic aerosol fraction. Grana 30:98-101. Stymne H. Eliasson A. 1991. A new passive tracer gas technique for ventilation measurements. Proceedings of the 12th AIVC Conference: Air Movement and Ventilation Control Within Buildings 3:1-18. Teijonsalo J. Jaakkola JJK, Sepannen O. 1996. The Helsinki Office Environment Study: air change in mechanically ventilated buildings. Indoor Air 6~2~:111- 117. Thatcher TL, Layton DW. 1995. Deposition, resuspension, and penetration of particles within a residence. Atmospheric Environment 29~13~:1487-1497. Trasoff A, Blumstein G. 1936. The value of air-conditioned room in the treatment of seasonal and perennial asthma. Journal of Laboratory and Clinical Medicine 22:147-150. Tsubata R. Sakaguchi M, Yoshizawa S. 1996. Particle size of indoor airborne mite allergens (Der p 1 and Derf 1~. Proceedings of Indoor Air '96 3:155-160. Turk BH, Brown JT, Geisling-Sobotka K, Froelich DA, Grimsrud DT, Harrison J. Koonce JF, Prill RJ, Revzan KL. 1987a. Indoor Air Quality and Ventilation Measurements in 38 Pacific Northwest Commercial Buildings, Volume 1: Measurement Results and Interpretation, LBL-22315 i/:, Lawrence Berkeley National Laboratory, Berkeley, CA. Turk BH, Grimsrud DT, Harrison J. Prill RJ. 1987b. A Comparison of Indoor Air Quality in Conventional and Model Conservation Standard New Homes in the Pacific Northwest: Final Report. LBNL-23429, Lawrence Berkeley National Laboratory, Berkeley, CA. Turk BH, Grimsrud DT, Brown JT, Geisling-Sobotka KL, Harrison J. Prill RJ. 1989. Commercial building ventilation rates and particle concentrations. ASHRAE Transactions 95~1~:422-433. Turk B. Powell G. Casey M, Fisher E, Ligman B. Marquez A, Harrison J. Hopper R. Brennan T. Shaughnessy R. 1997. Impact of ventilation modifications on indoor air quality characteristics at an elementary school. Proceedings of Healthy Buildings '971:155-160. U.S. Department of Commerce. 1997. Statistical Abstract of the United States 1997, Washington, DC. U.S. EPA. 1999. Ozone Generators That Are Sold as Air Cleaners. URL: http:// www.epa.gov/iaq/pubs/ozonegen.html. Accessed September 6,1999.
IMPACT OF VENTILATION AND AIR CLEANING ON ASTHMA 393 van der Heide S. Kauffman HE, Dubois AEJ, de Monchy JGR. 1997. Allergen reduction measures in houses of allergic asthmatic patients: effect of air- cleaners and allergen-impermeable mattress covers. European Respiratory Journal 10~6~:1217-1223. Vaughan WT, Cooley LE. 1933. Air conditioning as a means of cleaning pollen and other particulate matter and of relieving pollinosis. Journal of Allergy 5:37-44. Verall B. Muir DC, Wilson WM, Milner R. Johnston M, Dolovich J. 1988. Laminar flow air cleaner bed attachment: a controlled trial. Annals of Allergy 61~2~:117- 122. Villaveces JW, Rosengren H. Evans J. 1977. Use of laminar flow portable filter in asthmatic children. Annals of Allergy 38~6~:400-404. Viner AS, Lawless PA, Ensor DS, Sparks LE. 1989. Ozone emissions from electronic air cleaners. Paper 89-84.1, presented at the 82nd Annual Meeting of the Air and Waste Management Association, Anaheim, CA. Warburton CJ, Niven RM, Pickering CAC, Fletcher AM, Hepworth J. Francis HC. 1994. Domiciliary air filtration units, symptoms and lung function in atopic asthmatics. Respiratory Medicine 88~10~:771-776. Warner JA, Marchant JL, Warner JO. 1993. Double-blind trial of ionizers in children with asthma sensitive to the house dust mite. Thorax 48~4~:330-333. Wickman M, Emenius G. Egmar AC, Axelsson G. Pershagen G. 1994. Reduced mite allergen levels in dwellings with mechanical exhaust and supply ventilation. Clinical and Experimental Allergy 24~2~:109-114. Wood RA, Johnson EF, Van Natta ML, Chen PH, Eggleston PA. 1998. A placebo- controlled trial of a HEPA air cleaner in the treatment of cat allergy. American Journal of Respiratory and Critical Care Medicine 158~1~:115-120. Zwemer RJ, Karibo J. 1973. Use of a laminar flow device as adjunct to standard environmental control measures in symptomatic asthmatic children. Annals of Allergy 31~6~:284-290.