Literature Review of Health Effects of Humidity-Influenced Indoor Pollutants

UC Berkeley Indoor Environmental Quality (IEQ)

Title Indoor Humidity and Human Health--Part I: Literature Review of Health Effects of Humidity- Influenced Indoor Pollutants

Permalink https://escholarship.org/uc/item/5kz1z9cg

Authors Baughman, A. Arens, Edward A

Publication Date 1996

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California 3951 lndeer Humidity and HumanHealth Part Literature Reviewof Health Effects

Anne V. Baughman Edward A. Arens, Ph.D. Student MemberASHRAE Member ASHRAE

ABSTRACT under elevated humidities, although humidityinteractions with nonbiotic pollutants, such as formaldehyde, mayalso cause Standards for indoor thermal conclitions and ventilation include upper limits for relative humidio,(RH) that O~pically adverse effects. Existing limits appear to be based on engi- are hz the range of 60%to 80%RH. Although the reasons for neering experience with such humidityproblems in buildings. the limits are often not explicitly stated, it is generallyknown The position of any upper humidity limit has great that they were set out of concernfor the health effects that economicsignificance, particularly in hot and arid parts of the might occur shouMthe humidio, becometoo high. The pri- country, where evaporative cooling is an energy-conserving mal3, health effects o~ high humidio, are caused by the g~vwth option. In the West,it affects the needfor billions of dollars of and spread of biotic agents, although humidiO, interactions newpeak electrical generatingcapacity that could be offset by with nonbiotic pollutants, such as formaldehyde, may also noncompressor-based cooling. It also directly affects a causeadverse ~ffects. This literature review ident~’es the most substantial fraction of the cooling load in hot, humidclimates. important health issues associated with high humidities and Undersuch economicimperatives, it is desirable to carefully presents humidio, requirements, ~,pical contamination sites examinethe position of any upper humiditylimit. Ideally, one within buiMings, and remediation measures for each pollut- wouldbe able to assess the health risks against the economic ant. Part ~vo of the paper addresses the physical causes of benefits for any given humiditylimit. At present, there is not moisture-relatedptvblems in buildings. enoughinformation on this subject to even begin such an analysis. INTRODUCTION Standards for indoor thermal conditions and for ventila- This reviewof the literature identifies a numberof health- tion have traditionally put upper limits on the amount of related agents that are affected by indoor humidity.All of them humiditypermissible in interior spaces becauseof concernfor affect humanhealth primarily throughtheir inhalation fiom the the health effects that mightoccur should the humiditybecome air, althoughsome of themhave lesser effects throughthe skin. too high. Such limits are found in past versions of ASHRAE Biological agents require appropriate conditions in the building Standards 55-1992 (ASHRAE1992) and 62-1989 (ASHRAE lbr their germination,growth, release to the air’, andtransport to 1989)and in most international standards. The values set for the humanhost. Airborne levels of nonbiological pollutants, the upper limits have typically ranged from 60%to 80%RH, such as formaldehyde and ozone, may also be affected by although boundaries of absolute humidityhave also been used. humiditythrough influences on offgassing and surface reaction To date, the relationship of high humidityto the full spectrum rates. Finally, the occupants’susceptibility to these agents may of air quality issues and to the relevant characteristics of also be a function of humidity, although this appears to be a building envelopes and conditioning systems has not yet been problem primarily at low humidities, when respiratory addressed in a comprehensivemanner’. This situation affects ailments result from dry mucousmembranes (Green 1985). our ability to set rational standardsand buildingspecifications. The health implications of low humidities are not addressed in Humanhealth is not affected by high levels of humidity this paper. Part two of this paper addressesthe relationships of per sen Knownhealth effects related to high humidity are the environmentswithin buildings and conditioning systems to primarily caused by the growth and spread of biotic agents the growthof biological pollutants.

AnneV. Baughmanis a graduatestudent researcherand EdwardA. Arensis a professorin the Departmentof Architectureat the University of California,Berkeley~

ASHRAETransactions: Research 193 OVERVIEW OF HUMIDITY-RELATED Nonallergic immunologicresponses, characterized by HEALTH CONCERNS recurrent flu-like symptoms(e.g., hypersensitivity pneu- The primary influences of humidityon health are through monitis, farmer’s lung, and humidifierfever), seemto be unre- biological pollutants. The following outline describes the lated to the IgE antibody. Theyoccur as a result of repeated health issues most commonlyassociated with biological pollutant exposures that trigger other antibody-dependent pollutants. mechanismsas well as cellular immuneresponses. Although there seemsto be no genetic predisposition, only a fraction of Infectious disease (pathogens) those exposed develop overt symptoms(Burge 1988). bacteria (e.g., Streptococcus,Legionella) viruses (e.g., commoncold, flu) Mycotoxinsare producedby fungi and can lead to respira- fungi (e.g., Aspergillusfumigatus) tory irritation, interference with pulmonarymacrophage cells, and/or higher risks of cancer (Flanniganand Miller n.d.). Many Allergic reactions(e.g., asthma,rhinitis) fungi also produce volatile organic compounds(VOCs) that dust mites (dried bodyparts and fecal excreta) maybe respiratory irritants and have been suggested as a fungi contributing factor to sick-building-type symptomsin microbi- Nonallergic immunologicreactions (e.g., hypersensitivity ally contaminated buildings (Bjurman1993; Sorenson1989). pneumonitis) Nonbiological pollutants, such as formaldehyde,ozone, fungi oxides of nitrogen, and sulfur, affect humansprimarily through bacteria chemicalirritation of the mucousmembranes. Formaldehyde is Myctoxicosis released into the indoorair frombuilding materials in waysthat fungi are dependenton atmospherichumidity. Surface reactions, and Infectious disease can occur when viable pathogenic consequently the amountand toxicity of ozone and nitrogen organismsenter (usually through inhalation) and colonize oxides (NOx)and sulfur oxides (SOx) in the air, maybe influ- the body of a susceptible host. The most commonlyfound enced by humidity levels. The extent to which humidity pathogensare bacteria or viruses, although fungal pathogens, increases or’ decreases the health impactsof these pollutants, such as Aspergillus fimffgatus, also exist (Flannigan 1992). however’,is relatively small comparedto other environmental Most pathogens are transmitted through human-to-human factors, such as air changerates and outdoorpollutant levels. contact whendroplet nuclei form as a result of sneezing or For exarnple, the use of direct evaporative cooling leads to a coughingand are subsequently inhaled by a humanreceptor: A rise in indoor humidity levels, which mayreduce ozone by few pathogens, most notably the bacterium Legionella, can increasing surface reactions. However,this effect is relatively colonize abundantly within moist environments outside the insignificant comparedto the increased influx of outdoorair; humanbody and becomeairborne given proper conditions. whichtends to increase indoor ozone concentrations to levels Noninfectioushealth conditionsrelated to biological pollut- near those of the outdoor’air’ (Stocket aL 1993). ants include allergic, immunologic,and toxic responses, The primarysources of these adversehealth effects are the byproducts DUST MITES of organismsrather than the viable organismitself. The term Introduc~ion allergyis usedspecifically to refer to illnesses that take placeas a result of the formationof IgEantibodies in affectedpersons. All Mites are considered one of the most important allergens humanbeings have someIgE antibodies, but only a fraction of the in housedust, particularly in regions with high humiditiesand population respondsreadily to allergen exposureand produces temperate climates. The most commongenus of mites found in enoughIgE antibodies to cause an allergic reaction. Oncethe house dust in North America and Europe is Dermatopha- antibodies form, the person becomessensitized and re-exposure goides, of whichthere are two species, D. pteronyssinusand D. to the allergencan then trigger larger’ immunereactions resulting farinae. It is estimatedthat 10%of the populationin the U.S. is in allergic symptoms.This IgE-mediatedreaction develops in allergic to housedust and 70%of these people are specifically 20%to 30%of the peoplein the UnitedStates (Seltzer 1995). The allergic to mite allergens (Bates et al. 1993). The actual allergic diseases with well-documentedlinks to indoorair quality allergen is not the mites themselves, whichare approximately (IAQ)include allergic rhinitis (rhino conjunctivitis), primarily 1/3mmin length at maturity, but the dried fragmentsof their affecting the nasal area, and allergic asthmaand bronchopulmo- bodyparts and fecal excreta. Theseby-products are initially 10 nary aspergillosis (ABPA),both of whichaffect the lowerairways to .50/.tm in diameter but break downinto smaller fragments and alveolL The majority of patients suffering fromasthma are that becomeairborne whendust is disturbed. Accordingto one allergic to dust mites, mold,and/or animaldander. Theestimated study, morethan half of the weightof mite allergens within a overall prevalenceof asthmaand rhinitis maybe as high as 20% homewere found to be less than 5 p.m in length (Reed et al. of the population (Berglundet al. 1992)and the AmericanLung 1986). These particles are the primary health concern since Association estimates that the numberof people with reported they can be inhaled into the lower airwaysof the lungs and, if asthmain the U.S. has greatly increased in recent years, with a quantities are significant, IgE antibodies can form, leading to 49%increase since 1982. allergic reactions in the susceptible portion of the population.

194 ASHRAETransactions: Research A numberof studies have demonstrateda high prevalence The regional diversity of mite studies in the literature of sensitization to mite allergens amongpatients with asthma suggests that mites occurindoors all over the world, fromarctic and nonspecific respiratory symptoms(Voorhorst et al. 1964; Greenland to tropic Africa (Anderson and Korsgaard 1986). Korsgaard1983b; Platts-Mills et alo 1989; Smith et al. 1985; The study by Arlian et al. (1992) included eight different Arlian et al. 1992). In mostof these studies the patients were geographicregions within the U.S. Theyfound that D. pteron- referred to the researchers by clinics and comparedwith yssinus and D. farinae were by far the most commonspecies, control subjects randomlyselected from the sameor a similar with D. pteronyssinus predominating in humid regions with population base. Sensitization to mite allergens was demon- moderateclimates and D. farinae predominatingin areas with strated by a positive skin prick test. Thestudy of Danishhomes prolonged periods of dry weather. This finding was supported by Voorhorstet al. (1964)was the first to establish a definitive by Lang et al. (1977), whoexamined mite populations in four link betweenthe presence of mite allergens and respiratory different climatic zones of southern California and found symptoms.Korsgaard (1983a, 1983b) confirmed this finding significant numbersof D. pteronyssinusand D. farinae in 14 of whenhe found significantly higher concentrations of dust 15 of the coastal homesand in 9 of 15 of the inland valley mites in the homesof 25 asthmatic patients comparedto 75 homes. D. pteronyssinus was the predominant species in the randomlyselected homes.Arlian et al. (1992) conducteda five- coastal region, while D. farinae predominatedin the inland year study of 252 homes inhabited by dust-mite-sensitive regions. people in eight different regions of the UoS.They found that Mites are relatively sparse in regions with low outdoor 83%of the homeshad average mite densities greater than the humidity, such as at high elevations and in desert areas (Brun- estimated sensitivity threshold of 100 mites/gmof dust. drett 1990; Murrayet aim 1985; Langet al. 1977). However,if the indoor humidityis allowed to rise due to internal sources Studies in the literature also provideevidence to support a such as direct evaporative cooling, mite populations, particu- connection betweendamp housing and sensitivity to dust mites larly D. farinae, can becomesignificant even whenoutdoor and childhood respiratory symptoms.For example, Murrayet humidities are low. For example,O’Rourke et aim (1993) evalu- al. (1985) studied 774 homesinhabited by children with respi- ated 190 evaporatively cooled homesin Tucson, Arizona, and ratory symptomsin British Columbiaand found that morethan detected mites in morethan half of the homes,with D. farinae 90%of these children lived in areas defined as "humid"(i.e., being the overwhelminglypredominant species (greater than indoor humidity estimated to be 50%or greater for four or 98%of all mites recovered). moremonths out of the year). In addition, there was a significant difference in the numberof mite-sensitive children in the Environmental Requirements "humid"areas (skin prick test positive for D. farinae in 31% and D. pteronyssinus in 40%)as comparedto those living in Mites contain about 70%to 75%water by weight and areas defined as "dry," with an indoor RHof 50%or higher for must maintain this in order to reproduce (Arlian 1992). Their no morethan two monthsper year (skin prick test positive for primarysource of water is ambientwater vapor, whichthey are D. farinae and D. pteronyssinus in 3%and 2%, respectively). able to extract directly from unsaturated air by meansof a Verhoeffetal. (1995) conducteda study that included 259 chil- hygroscopicsalt solution in the supracoxalgland (Fernandez- dren with chronic respiratory symptomsand 257 control chil- Caldaset al. 1994). The amountof water gained throughinges- dren. Therewere morecases of mite and moldsensitization in tion of moistfood is relatively small. Laboratorystudies of D. the children with respiratory symptoms.These children were pteronyssinus suggest that optimal conditions for growthand also slightly morelikely than the controls to havebeen living in development occur between 70%and 80%RH at 25°C, with homeswhere mold or dampwas reported or observed. acceptable ranges of 55% to 80% RH and 17°C to 32°C (Andersonand Korsgaard 1986). The upper humidity limit Alongwith respiratory symptoms,high levels of dust mite constrained by the possibility of mold growth, particularly allergens have also been correlated with atopic dermatitis above 88%, which can inhibit mite development (Brundrett (AD),characterized by itchy, irritated skin (Harvinget al. 1990; 1990). August 1984)mIn general, these studies suggest that those Arlian (1992) performed laboratory studies of both susceptible to mites (i.e., those likely to formIgE antibodies) farinae and D. pteronyssinusand foundthat the critical equilib- are also likely to developskin sensitization if exposedto high rium humidity (CEH)for fasting mites, defined as the lowest concentrations of mite allergens. For example,Colloff (1992) RHat whichmites are able to maintain their water balance, was examinedthe density of dust mite populations in mattresses in 73%and 70%RH at 25°C for D. pteronyssinus and Dfarinae, the homesof 23 people with ADwho were mite-sensitive and respectively. The CEHwas found to be influenced by tempera- found that counts were significantly higher than in the ture and ranged from 55%at 15°C to 75%at 35°C for D. mattresses of the nonatopiccontrol group°Colloff also cites a farinae. Accordingto Arlian, this temperature relationship, numberof references that link atopic dermatitis to high dust together with the fact that feeding mites do gain small amounts mite exposure, including IgE antibody responses to mite aller- of water fromfood, mayexplain whysignificant populations of gens, amongpatients with ADand marked clinical improve- mites are found in environmentswith relative humidities below mentfollowing intensive eradication of mite allergens. 70%.Survival of mites for prolongedperiods at lower humidi-

ASHRAETransactions: Research 195 ties mayalso be explainedin part by the crystallization of salts Absolute vs. Relative Humidity within the supracoxal gland, which mayslow downthe rate of Someresearchers have suggested that absolute humidity dehydration(Fernandez-Caldes et al. 1994). rather than relative humidityis the limiting factor controlling Underoptimal conditions, mites live for three monthswith mite metabolism(Korsgaard 1983a, 1983b; Platts-Mills et al. three different larval stages. Thesurvival of active adult mites 1987b)~ However,the predominant evidence from laboratory (both maleand female) is limited to 4 to 11 days at humidities studies and field worksuggests that relative humidityis the below 50%RH at 25°C (Arlian et al. 1982). The protonymph, controlling factor. Mites have a high surface-to-volumeratio however,which is one of the dormantlarval forms, can survive and are poikilothermal(i.e., theh" bodytemperature is identical for monthsat low humiditiesand then evolve to the moreactive to that of the surrounding environment) (Andersonand Kors- forms whenoptimal conditions return (Arlian 1992). This gaard 1986). Since there is no temperature gradient between the mite and the surroundingenvironment, the relative differ- observationis supportedby the field study by Langet al. (1978) ence betweenthe air vapor pressure and the mite’s internal in whichthe different stages of mite developmentwere quanti~ saturation vapor pressure is proportional to the relative fled. In this study a higher numberof protonymphswere found humidityrather than the absolute humidity. Arlian (1992) has whenthe RHfell belowcritical levels (50%to 65%RH). These demonstratedthat the driving force in the uptakeof waterfrotn protonymphsare particularly difficult to removewith normal unsaturated air is the numberof water moleculesimpinging on vacuumingsince they can bury themselves within surfaces the mite’s uptake surface. Arlian also performedlaboratory (Arlian 1992). studies that suggest that mites are able to maintain a water As one might expect, most mite allergens are fot~ned by balance at 20°Cand 79%RH but die at 27°C and 56%RH (i.e., adults during their active phase° Thus, for a given numberof the sameabsolute humidity). mites, the highest levels of allergens found in the environment Field Studies: Residential usually correspond to optimal humidity conditions. Arlian (1992) examined the effect of RHon mite metabolism for Field studies within homesgenerally support the labora- range of relative humidities between 22% and 95% and tory findings that indoorrelative humidityis the mostsignifi- observed that feeding rates, and consequently the amountof cant environmental condition associated with high mite fecal matter produced, increased with increasing RH~The populations and allergen concentrations. However,it is not effect was pm~icularly significant between75% and 85%RH, clear fromthese studies whatspecific level of humidityis crit- for whichthere was a fivefold increase in the weight of food ical. Significant concentrations of mites and allergens have consumedfor’ both D. pteronyssbms and Dfarinae. Belowthe been found at indoor humidities as low as 40%RH (O’Rourke CEH,Arlian found that mites fed sparingly and producedlittle et al. 1993) but moreoften at indoor humidities above50% RH. fecal matter. Theseresults suggest that significant reductionsin For example,in a study of homesin Vancouver;Murray et al. the level of mite allergens, whichconsist primarily of meta- (1979) detected significant numbersof mites only whenthe bolic by-products, mayoccur ifRH is reduced below the CEH. was greater than 50%for at least part of every day during the (For more detailed information on the mite life-cycle and monthof collection. Smith et al. (1985) also found a direct correlation betweenmite population and indoor RHin a study metabolism,see Arlian [1992].) of 20 homesof mite-sensitive children, with mite populations Laboratory studies suggest that temperatures within the peaking at RHof 50%or greater; Hart and Whitehead(1990) rangetypically foundin occupiedspaces have little direct effect evaluated 30 homesin the United Kingdomand found that mite on the length of the mite’s life-cycle and that mites are able to populations were most strongly correlated with indoor RHand survive extremetemperature conditions for limited periods. For that bedrooms with humidities above 64% RH contained example, under laboratory conditions, more than half of D. significantly moremites in mattresses than those with humidi- pteronyssinus survive after 12 days when continuously ties belowthis level. exposed to 34°C and 75% RH, while at 2°C and 75% RH Regionalstudies suggest that dust samplesfrom different approximately64% of D. farinae adults survive after 72 hours homeswithin the same region can exhibit wide differences in (Langet al. 1977). Mites are also able to reproduceat temper- mite concentrationdue to differences in indoor humidityalone. atures as low as 17°C, albeit mote slowly than at 25°C(Murray For example,Lintner et al. (1993) evaluated 424 homesacross et al. 1979). the U.S. and foundthat the greatest variations in mite population occurred as a result of differences in the indoor relative Mites subsist primarily on shed humanand animal skin humidity amonghomes rather than regional climatic differ- scales. It is believed that mites cannotdigest lipids withi~ the ences. Korsgaard (1983a, 1983b) conducted a four-season skin scales themselvesand require the aid of xerophilic fungi, study of 50 Danishapartments, all within the sameregion, and of the genusAspergillus, to dissolve the lipids for them(Flan- found that seasonal variation in dust mite populations in nigan 1992; Hart and Whitehead1990; Platts-Mills et al. mattresses con’elated with the indoor humidity, while homes 1989). This suggeststhat conditionsfor’ mite survival mustalso with the lowest indoor humidities did not contain detectable be suitable for these fungi. levels of mite populations. Ellingson et al. (1995) studied the

196 ASHRAETransactions: Research effect of direct evaporative cooling on the prevalenceof mite site for mites becausethere is an amplesupply of food and the allergen in Coloradohomes. Their results showthat during the humiditywithin an occupiedbed is higher than that of the air of peak cooling season 48 of 95 samples from homeswith evapo- the surroundingspace. This is also true of furniture upholstered rative coolers (average interior RHof 51%or greater) had with permeablefabric that can absorb and retain the moisture levels of Der p 1 and Der f 1 of 2 gm/gmdust or greater, but given off by an occupant. Field studies have found less varia- only 5 of 95 control samples(average interior RHof less than tion over time in the number of mites within bedding as 45%)had levels of 2 gm/gmdust or greater. comparedto other sites (Smith et al. 1985). Murray et Long-termstudies suggest that seasonal trends in mite (1979) also foundsignificant numbersof mites in mattress dust populations correlate with seasonal variations in indoor during the winter even whenthe indoor RHfell below 50%. humidity. For example, in a 19-monthstudy of six homesin southern California, Langet al. (1978) foundseasonal differ- Carpeting can also be a localized site of increased ences in species composition, with D. pteronyssinus more humidity and consequently maybe an important reservoir for abundant from July through Novemberand D. farinae more allergens in both homesand schools. Studies conducted in abundant from August through December.Both species were schools have demonstratedthat carpets contain high levels of a found to be less prevalent from late spring to July. Although variety of allergens including pollen, cat and dog dander, and monthlypopulation fluctuations correlated with indoor relative mite and moldallergens (Fernandez-Caldaset aL 1994). This humidity, the population increases lagged one to two months maybe the primarysource of exposurefor youngchildren, who behind the time that conditions first becamefavorable, while generallylive closer to the floor anddo not havehigh exposures populationdeclines correlated directly to the time that relative in bedding since they usually sleep on plastic-covered humiditylevels fell belowcritical levels (47%to 50%RH for mattresses. Arlian (1992) studied the microenvironmentof D.farinae and 60%to 65%RH for D. ptero~o,ssinus). Arlian et carpeted floor and found that the relative humiditywithin the al. (1982) also foundsignificant seasonal fluctuations in the two-year study of 19 homesin Ohio, with highest densities of carpeting was 9.6%higher than that of the ambientair 1 to 2 meters above the floor. This was attributed to the decreased mites occurring during the humid summermonths and lowest densities occurringduring the drier late heating season. In the temperature of the floor (3.7°C lower on average than the study by Korsgaard(1983a, 1983b), those apartments that had ambient air), whichdrove the relative humidity up. In this low absolute indoor humidities in the winter did not contain study, Arlian also found that long-pile carpeting contained noticeable concentrations of mites in the summerand autumn significantly more mites than short-pile carpets and tile or despite the fact that the humidityconditions increased to levels woodenfloors (i.e., short-pile carpets did not contain signifi- that were high enough to support peak populations. This cantly moremites than floors without carpets). This finding is suggests that, in this case, winter conditions mayhave been also supported by a study by O’Rourkeet al. (1993) in which severe and long enoughto kill off even the dormant proto- house mites were found four times more frequently in homes nymph,assuming other eradication steps were not taken. with wall-to-wall carpets than in homeswith other floor types. Basedon a numberof field studies it is also apparentthat allergen levels correlate with seasonal variations and that Field Studies: Office Buildings changes in allergen levels lag behind both increases and decreases in indoor RH.Lintner and Brame(1993) studied 424 The few studies of mite populations that have been homesacross the U.S. and found a distinct seasonal fluctuation conducted within commercialbuildings have shownthat mite of mite allergens for both D. pteronyssinus and D. farinae levels are generally low (Menzieset alo 1992). In a study species, with the D. pteronyssinusallergens peakingin July and office buildings in the mid-Atlanticstates, Hunget al. (1992) the D. farinae allergens peaking in September.In a study by found moderate to high levels of mite allergens within Friedmanet al. (1992) of homesin the upper ConnecticutRiver carpeting and chairs of one of the five buildings studied. In a Valley, there wasa markedseasonal increase in total D. ptero- study of buildings in the NewEngland area by Friendmanet al. nyssinus allergens from June to September.In a one-year study of 12 homesin central Virginia, Platts-Mills et aL (1987a) (1992), very low population levels of dust mites were found found that increases in both mite populations and allergen within the carpets of workplaces.The observation of low mite levels lagged approximately one month behind increases in levels within commercialbuildings maynot be surprising since indoor humidityand that several monthspassed before a fall in these buildings tend to be less humidthan residences due to the allergen levels was detected after a drop in indoorhumidity. frequent use of air conditioning and fewer internal sources of In all of the field studies cited, the highest concentrations moisture (e.g, cooking, showering,etc.). In addition, commer- of mites were found in mattresses, thick carpeting, and/or cial buildings do not usually contain bedding and thick heavily used fabric-upholstered furniture. This suggests that carpeting, the mostcommon sites for mites in residences. It has mites thrive best within micr0environmentsthat contain a also been suggested that mites tend to be more commonon source of food (shed skin scales) and have a relatively high and groundfloors than on upper stories and are rare in hotels (Reed consistent moisture level. For example,bedding is a common et al. 1986).

ASHRAETransactions: Research 197 Remediation the greatest health significance (< 2 gm)easily pass through The strong correlation betweenindoor relative humidity the filter bags of conventionalvacuums, causing a significant and dust mite population has led to recommendationsto reduce increase in the concentration of airborne allergens during and indoor humidity. However’, exactly where the upper limit shortly after vacuuming(Kalra et al. 1990). In a two-yearstudy shouldlie is not obvious.Most of the field studies suggest that of ~nites in 19 homesin Ohio, Arlian et al. (1982) found whenindoor humidity is kept below50% RH, mite populations significant correlation betweenthe numberof mites and the do not growto significant levels. Laboratorystudies, on the frequency or thoroughnessof cleaning, amountof dust, or age other hand, in which the microenvironmentof the mite is in of furnishings or dwelling. Arlian (1992) suggested that the equilibriumwith the surroundingair, suggest that mite popula- ineffectivenessof cleaning mayalso be related to the difficulty tion growthand metabolism(related to the amountof allergen in removinglarval forms of mites adhering to surfaces. produced)can be significantly reducedif relative humidityis Acaricides are nowavailable that have been specially kept below 70%RH at 25°C (Arlian 1992). designed to eliminate mites from carpeting. Onesuch product Onereason for the discrepancy betweenfield and labora- uses benzyl benzoateas the active ingredient. It is formulated tory studies maybe the difference in relative humiditybetween as a moist powderwith a waxto bind mite fragmentsand excre- the mite’s microhabitat, whichcan be consideredto be within a ment so that they can be vacuumed.It is designedto be reap- few millimeters of the horizontal surfaces on whichthey lie, plied every six to eight months. Results from a numberof and that of the surroundingair’ due to differences in tempera- studies suggest that this product has been successful in ture as well as the ability of certain types of surfaces to retain reducing mite populations (Htut 1994). Benzyl benzoate was moisture. The humidity measurementsfor the field studies initially marketedin Europeand has been approvedfor use in cited weretaken fromthe air within the core of the room,which all states in the U.S. except California (eL 1994). Fungicides maynot correlate with the RHwithin the rnicroenvironments such as natamycinkill the fungi required by mites to digest fromwhich the mite sa~nples weretaken. Thetime fi’ame of the lipids in the skin scales and have also been used with some RHmeasurements was also not indicated for mostof these field success (Flannigan1992; Platts-Mills et al. 1987b;Htut 1994); studies---instantaneous measurementstaken at the time of however, in one study in which a double-blind, placebo- samplingmay not be representative of long-termconditions. In controlled methodwas used, no significant improvementwas addition, seasonal changesin indoor humidityhave a signifi- observed(Reiser et al. 1990). Other surface treatments that cant effect on mite populations and allergen levels, as have been used include liquid nitrogen, benzyl benzoate in suggestedby the long-termfield studies; however,the specific combinationwith tannic acid, and benzoic acid (Htut 1994). time constraints have not yet been resolved. Other possible methodsof reducing mite levels include the ’Iqaere are a numberof effective remedial methodsdirected use of electric blankets, whichcan reduce the local humidity at reducing allergens and mites within their microhabitat in within bedding (Hart and Whitehead1990), and dehumidifica- addition to control of relative humidity.These include special- tion/air conditioning(Lintner et al. 1993). Accordingto a study ized vacuumingprocedures, removalof long-pile carpeting and by de Boer and van der Geest (1990), a reduction in dust mite heavily used upholsteredfur:niture, regular hot-water cleaning populations of 19%to 84%can be achieved by heating the of bedding, encasementof mattresses and pillows, and applica- mattresses with electric blankets whenthe beds are not in use. tion of acaricides (Htut 1994). For example, in a study In a field study by Cabrera et al. (1995), dust mite allergens laundry procedures, McDonaldand Tovey(1992) found that all were reduced by morethan 50%with the use of a dehumidifier. mites were killed by water at 55°C or higher. In a study by Improvedventilation systems within homescan also help Platts-Mills et al. (1989),a tenfold or greater reductionin mite reduce mite levels by counteracting internal sources of allergen levels wasachieved in manyhouses by hot-washingall humidity, such as cooking and showering, in climates where beddingat least every 10 days and removingcarpets and uphol- outdoor humidity is not the major source of moisture. stered furniture. Wickmanet al. (1994a) found a significant Wickmanet al. (1991) suggest that housedust mite infestation decreasein mite allergen levels on mattress surfaces six months used to be rare in Stockholm;however, mite-sensitive children after they had been encased with a semipermeablepolyure- are nowfrequently observed, whichmay indicate an increased thane cover’. Based on the observed seasonal effects for infestation rate. Theauthors attribute this to a reductionin the temperateclimates, it seemsthat late winter and early spring ventilation rate resulting from the energy conservation are the best ti~nes to clean mattresses and carpets aggressively programs.In a follow-up study, Wickmanet al. (1994b) looked to kill the fewmites that survivedthe winter. Theoretically,this at the concentration of dust mites in 70 homesin Stockholm should reduce the chancesof havinga large infestation in the belonging to two major house types--4.hose with crawlspace summermonths. basements and those with concrete floor slabs--and deter- Vacuumingis effective only if central vacuumingsystems, minedthat the highest risk factors for allergen concentration HEPAfilters, or systems that entrain dust in a liquid medium exceeding the median were unimprovednatural ventilation are used. Conventionalvacuuming does not help to reduce mite (i.e., no mechanicalexhaust), concrete floor slabs, and conden- populations and allergens within carpets and can actually sation on windows.In a study of Danishhomes, Harving et al. aggravate the problem.Allergen particles in the size range of (1994) found that decreases in indoor humiditylevels through

198 ASHRAETransactions: Research the use of supply-and-exhaustventilation systemssignificantly counts reduceddue to settling and filtration within air-condi- reduceddust mite levels. tioning equipment. In a study of Canadianoffice buildings, Miller (1992) found that those buildings not associated with FUNGI microbial problems had micofloral counts that were qualita- Introduction tively similar and quantitatively lowerthan those of outdoorair, Fungi (via airborne fungal spores, fragments of hyphal while contaminatedbuildings tended to have a higher propor- mat, and metabolic by-products) have been linked to a number tion of nonphylloplane fungi, particularly Penicillium and of adverse health effects, including allergic reactions, hyper- AspergillusoIn a study of fungal concentrationswithin daycare sensitivity pneumonitis, mycotoxicosis, and pathogenic centers and dwellings, Hyvarinenet al. (1993) found that the disease. In general, however,fewer peopleare allergic to fungi total concentration of airborne fungal spores was higher in than to dust mites and animal dander (Flannigan et aL 1991). moldybuildings. In addition, the concentrationsof Aspergillus Beaumontet al. (1985) demonstratedthat manymore respira- and Oidoidendronin the fall and Aspergillus and Penicilliumin tory patients with suspected allergies react to animal dander the winter were higher in the buildings with mold problems (34%) and house dust (44%) than to molds (3%). The than in the reference buildings. The presence of wet-habitat commongenera knownto cause asthma and rhinitis include fungi, such as Phoma,Stachybotrys, Trichoderma,and Ulocla- Alternaria, Aspergillus, Cladosporium,and Penicillium (Flan- dium, in significant quantities suggests the existence of either rotting vegetation near the air intake or an extremely damp nigan and Miller n.do). A few genera, such as Aspergillus niger and A. fi#nigatus), Histoplasma,and Cryptococcus,are amplification site within the building (Flannigan1992). Xero- pathogenicand can infect the lungs, ears, or eyes in susceptible philic species, including the toxigenic species Penicillium hosts; however,reported cases are relatively rare (Gravesen auranteogriseum and Aspergillus versicolor, can form an 1979; Flannigan 1992; Miller 1992). Metabolicgases produced appreciable percentage of the population within indoor dust by fungi contain volatile organic compounds(VOCs) that are samples. These had not been widely detected until the recent responsible for the mildew odor. These VOCsmay be a use of new sampling methodsdesigned for detection of xero- contributing factor to sick-building-type symptoms,including philic species (Miller 1992). eye, nose, and throat irritation; headache;and fatigue (Bjurman 1993). In a study of microbially contaminatedbuildings, VOCs Environmental Requirements of the type commonlyassociated with indoor man-mademate- Fungineed water, carbon, and nitrogen for growth, as well rials were actually found to be the metabolic by-products of as minute amounts of other nutrients normally present in fungi growingin the buildings (Bayeret al. 1993). Highindoor natural environments~ Typical construction materials spore levels of fungi such as Cladosporiumand the dry rot containing nutrients used by fungi include wood, cellulose, fungus Sepula have been associated with cases of hypersensi- wallpaper, organic insulation materials, textiles (especially tivity pneumonitis. Fungal spores and vegetative myceliumare natural fibers), and glues and paints containing carbohydrates also knownto contain toxic substances (mycotoxins)that can or proteins. Althoughmaterials such as metal, concrete, plas- lead to respiratory symptomsunrelated to allergic mechanisms tics, fiberglass, and other synthetic products cannot be used (Flannigan and Miller n.d.). Flannigan et al. (1991) list directly by fungi, they can collect organicdebris that serve as a numberof toxigenic species isolated from the indoor air of nutrient source for fungi. For example,despite air filtration, houses. somedust containing living microorganismspasses through Mostfungi originate outdoors and are saprophytic (i.e., air-handling units and settles on porousinsulation within ducts. growon substrates of dead or dying plant and animal matter). If this insulation material then becomeswet (e.g., due to Outdoorconcentrations vary with the season, the time of day, condensation), fungi will grow and release spores into the local weatherconditions, and whetherthe site is rural or urban. ventilation air (Moreyet al. 1991; Pasanenet al. 1993). For example, phylloplane (leaf-loving) fungi, which include Fungi acquire most of their nutrients through a solvent Cladosporiumand Alternaria, are morecommon in rural areas process (Griffin, 1981). Thus the moisture on and within and showa strong seasonal activity with peak concentrationsin substrate is the important factor determining fungal growth the summer.Penicilliu~n and Aspergillus are the most common rather than the moistureof the ambientair (Block1953)o Labo- soil fungi found in urban environments, and airborne spore ratory studies support this observation. For example,Pasanen concentrations of these species remain relatively constant et al. (1991) measuredcolony diameters for both Penicillium throughout the year (Brundrett 1990). Fungal spores are typi- sp. and Aspergillus.fitmigatusas a function of RHin the range cally in the rangeof 3 to 30 p.min diameterand, once they are of 11%to 92%and found that fungal colonies grew on wet released into the air, can travel intercontinental distances. substrates even at low levels of atmospheric humidity. The Airbornespores enter buildings through ventilation equipment authors conclude that growthis dependenton substrate mois- and can set up colonies on surfaces wheremoisture and nutrient ture and is not directly affected by atmospheric moisture. conditions are favorable. Systems containing water, such as the water reservoirs of Buildingswith no internal sources of fungi have nearly the humidifiers, favor the growthof bacteria, algae, protozoa, and sameproportion of fungal species as outdoor air, with total certain types of fungi, especially yeasts. Mostfungi, however,

ASHRAETransactions: Research 199 prefer surfaces of moistmaterials to liquid water(Pasanen et al. as the ratio of the water vapor pressure at the surface of the 1992). Thus, since nutrients and airborne fungal spores are moist material to that of a pure liquid watersurface at the same abundantwithin buildings, the availability of moisture on and temperature and pressure (Ayerst 1969; Griffin 1981; Flare within surfaces appearsto be the limiting factor for growth. nigan1992). This is also r’efe~x’edto as the equivalentrelative Fungi are able to withstand dry periods to someextent by humidity (ERH)when written in the form of a percentage(i.e, becomingdormant or by utilizing metabolically generated an aw of 0.80 is the same as an ERHof 80%). ERHis equal to water that they add to the substrate. For’ example,wood will not RHat the surface of the material only whenthe system is decayif the moisturecontent is less than 20%to 25%of its dry confined to the extent that the atmosphere above a moist weight, except whenit has been invaded by a dry-rot fungus surface is at the samevapor pressure and temperature as that such as Meruliuslacrymans, which is able to translocate water directly at the moist surface. In actual environments,however, (Moore-Landecker1982). Fungi that inhabit soil and wood there is usually a gradient of vapor pressure from the surface grow better in moderate rather than high moisture contents into the air above or vice versa. In this case, the relative since soil aeration (and therefore oxygensupply) is limited humidity of the air has only an indirect influence through whenthe moisture content is high (Moore-Landecker1982). drying and moisteningof the materials it contacts. In general, fungi can grow at temperatures between 0°C In general, favorable conditions for fungal growthdepend and 40°C. Belowtemperatures of 0°C the fungi maysurvive on the species and the type of substrate on whichit is growing. but will not continue to grow, and above temperaturesof 40°C In addition, fungalgrowth does not occur"in isolation but rather fungi cannot survive for long periods (TenWoldeand Rose within a complexmicrobiological system that includes yeasts 1993)~Temperature variations within the range found in most and bacteria as well as molds. The following excerpt from conditionedbuildings do not appear to be a limiting factor but Flannigan (1992) presents his ecological classification mayaffect growthrates. For example,the study by Pasanenet moldsin terms of their moisture requirements, and describes al. (1992) demonstrated that both Penicillium sp. and the process by whichdifferent types of molds take hold and Aspergillus fi~migatus grew at all temperatures from 10°Cto grow. 30°C, with Aspergillus growingfastest at 30°C and Penicil- Althoughall moldsgrowing on surfaces in buildings grow lium growingfastest around 20°C. mostrapidly in pure culture at a high ~w.... individual species The potential for fungal growthon a substrate has often can be classified as: been attributed to moisture content (MC),defined as the ratio o Primarycolonizers, whichare able to growat an ~ of of "fi’ee water"in the material to the material’s dry weightafter less than 0.80 and also are [eferred to as xerophilic being dried in an oven (free water refers to water held in fungi becausethey are able to growat lower ~ than other molds,e.g., species in the Aspergillusglaucus porous material by van der Waals forces [i.e., hydrogen group(A. amstelodami,A. repens,etc.), A versicolot; bonding]or’ capillary attraction, as opposedto water of hydra- and Penicilliumbrevicompactum. tion, which is chemically bound to the materials). For wood o Secondarycolonizers, whichare able to growat an aw of products, percent moisture content is defined as the weightof 0.80-0.90, e.g, Cladosporium cladosporoidesand C. removed water divided by the weight of oven-dried wood. sphaerospermum. Woodconsists of hygroscopic cell walls surrounding cellular o Tertiary colonizers,which are only able to growat an wa spaces filled with water and/or air~ Belowfiber saturation, morethan 0.90, e.g., Alternariaalternata, Phomaher- whenthe cell walls are fully hydrated yet have no water barum,Ulocladium consortiale, andStachybotrys atra. contained in the cellular’ spaces, fungal growthis inhibited Wherethere is ingress of water over a restricted area, e.g., since the fungi are not able to readily extract the waterheld by as a result of rain water’penetrating via a structural fault in a van der Waalsforces (Wilcox1995). Thefiber saturation point wall, tertiary colonizers rnay be found at or near the site of for woodoccurs at MCsof 25%to 30%. ingress and primarycolonizers, such as A. versicolor and Peni- Althoughmoisture content is commonlymentioned, it is cillium, at the less wet marginsof the affected area (Grantet al. not the most appropriate measureof a substrate’s potential to 1989). Pasanen et al. (1992) found that Aspergilli and Peni- support fungus. This is becausematerials differ in howtightly celia (primary colonizers) grew under all conditions when they hold fi’ee water; and the measurementof moisture content samples of timber, plywood, gypsumboard, fiberboard, and mayvar3~ depending on the procedure used. For example, wallpaper were incubated in atmospheres of 75-76%RH and Block (1953) evaluated fungal growthon a numberof different above, but species of Cladosporium(secondary colonizers) and materials, including leather’, wood,cheese, wool, and cotton, Stachybotrysand Trichoderma(tertiary colonizers) only devel- and found a commonmold growth threshold value of about oped at the highest RH,where the substrate aw wouldbe 0.96- MC= 0.1, whichhas often been quotedin the literature. This is 0.98. The degree of dampnessdetermines what species are able significantly belowthe fiber’ saturation point for’ wood.More to grow and sporulate, and therefore strongly influences the recently, Foar’de et al. (1993) demonstrateda critical MC compositionof the spora in indoor air. 0.055 to 0.065 for Penicilliumgrowing in porousceiling tiles. This statement agrees with the findings of Kalliokoski et It has been suggestedthat for biological purposesthe more al. (1993), whocarried out a controlled-chamber study physically meaningfulparameter is water activity, a~ defined fungal growth on a number of moisture-damaged building

200 ASHRAETransactions: Research materials. Basedon this study, the authors suggest that fungal The most extensive fungal contamination problemsoccur growth is dependenton temperature, composition, and hygro- in hot, humidclimates; control of indoor humidity in these scopicity of materials and fungal species and is likely whenthe regions is an important factor. Bayer and Downing(1992) ERHexceeds 76%to 96%. The growth rate of a xerophilic observed fungal contamination in schools in a climate where fungi from a numberof different ecological sites and contami- outdoor humidities ranged from 75%to 90%for most of the nated materials was studied in the laboratory by Avari and year. High indoor humiditiesresulted in visible moldgrowth on Allsopp (t983). Optimumranges for growth were found to ceiling tiles, fan-coil units, papers, and books. In one case, betweena~,, levels of 0.97 and 0.90. For all of the species carpeting adhesive was not able to cure in the highly humid studied, no growthwas observed after one monthat au, levels of conditions and provided a mediumfor microbial growth. Effec- approximately0.80. This study is in agreementwith the recom- tive remediation procedures focused on removingand cleaning mendationby the International EnergyAgency (Annex 14) that contaminatedmaterials and controlling indoor humiditylevels. the monthlyaverage water activity of a material surface should Withincommercial buildings, microbiologicalcontamination not exceed 0.8 (IEA 1991). This recommendationrecognizes is frequentlya result of the absenceof’ or inadequatepreventive the importanceof the surface microclimateand was developed maintenanceof conditioning systems.Morey (1988) evaluatedthe in response to the request from building professionals for a occupiedspace and heating, ventilating, and air-conditioning simple criterion by whichto .judge the likelihood for mold (HVAC)systems of 21 commercialbuildings for the presence growth. microbiologicalreservoirs and amplification sites. Microbiolog- ical growthwas detected in 18 of the 21 buildings. In nine of the Field Studies buildings, contaminationwas found in the porousduct insulation. Othersignificant sourcesincluded stagnant water in drain pans(10 The most commoncause of fungal spore contamination buildings), excessiverelative humidity(6 buildings),flood damage within residences is condensationon surfaces and reoccurring (6 buildings),and the locationof outdoorair’ intakesnear external spills or leaks (Seltzer 1995). Besidessuperficial condensation, bioaerosol sources (6 buildings). Ezeonuet al. (1994) demon- interstitial condensationwithin porousbuilding materials (such strated that the fiberglass ductliners andboards from eight build- as concrete, brick, and/or plaster) mayprovide a reservoir for ings whoseoccupants complainedof unacceptableor moldyodors fungal growth. Interior dampnessproblems are usually related wereheavily colonized by fungi, particularly by Aspergillusversi- to construction faults, such as inadequate insulation and color. thermal bridges, in combinationwith inadequate ventilation and/or the pattern of use within homes.For example,in a study Remediation of 86 newly built energy-efficient residences in the Pacific Northwest, Tsongas (1991) found that one-third had mold The most effective remediation procedures depend on the growing on indoor wall surfaces and one-third had mold on source of the contaminationand regional climatic conditions. windowframes and/or sills. Althoughhomes were well insu- Control of indoor humidity is an important factor in hot, lated, condensationstill occurred due to internal sources of humid climates; however, for other climatic types other humiditythat were not properly ventilated. Nevalainenet al. means of controlling moisture, such as insulation to keep (1991) studied residences in Finland, where microbial prob- interior surfaces above the dew point, proper placement of lems in buildings are relatively uncommondue to the heating vapor barriers to control vapor and airflow betweenindoors and insulation requirementsof a cold climate. Theyfound that and outdoors, and control of external rain and groundwater most of the houses with mold problemshad improperly weath- penetration into the building, maybe more critical Proper erproofedoutside walls, whichallowed rainfall into the insula- maintenance of HVACequipment is also an important factor, tion material, and/or inadequatedrainage that allowedmoisture particularly in commercial buildings. This may require to penetrate in through the floor. Becker (1984) conducted design innovations focused on improving accessibility and post-occupancyevaluation of 200 homesin a coastal region of maintenance procedures. These issues are discussed in Israel, whichhas mild winters. Heconcluded that condensation greater detail in part two of this paper. was the mainsource of fungal growthand that the majorfactors Once a material has becomecontaminated, it is almost contributing to moldproblems were location and orientation of impossible to completely eliminate the fungi and removal is the dwelling (affecting wall surface temperatures), occupant often the only option. In a recent study of microbial growthin density, cookinghabits, and the type of paint or wall covering. chipboard, Thogersenet al. (1993) found that water damage Abe (1993) developeda biosensor method, using a xerophilic resulted in massive growthand that, even after drying, the fungus, to study the potential for growth within different material still contained spores. Use of biocides is usually microenvironmentsof a Japaneseapartment. A significant vari- discouraged, since most are toxic and continuous use may ation amongand within rooms was evident, with lower poten- increase corrosion and encourage the developmentof resis- tials for growth observed in spaces with walls adjacent to tant strains (Nevalainen 1993). Bleach treatment has been internal spaces and highest potentials observedat cold corners unsuccessful in cleaning contaminated duct liners (Morey of northern and eastern walls adjacent to the outside. 1988).

ASHRAETransactions: Research 201 BACTERIA AND VIRUSES virus wasinversely proportional to the level of RHand temper- Introduction ature. Mostviral and bacterial respiratory infections are trans- Expelled droplets and skin flakes that settle out may mitted amonghuman hosts. This mayoccur by touching an survive in dust and transmit disease if re-entrained when infected person or an object they have infected or by inhaling surfaces are disturbed. There is evidence that outbreaks of contaminated airborne droplets expelled from the nose and bacterial infections in hospitals have been associated with mouthduring sneezing, coughing, and talking. Most of these cleaning processes (Brundrett 1990). Studies of viability airborne droplets are large enoughto fall to the groundwithin a bacteria in dust suggest that there is a trend of decreasing meter. Smaller"droplets quickly shrink through desiccation to viability with increasing relative humidity(Brundrett 1990). form "droplet nuclei," which are small enough(between 0.5 addition, dust is less likely to becomeairborne at higher"humid- and 5 gmin diameter’) to remainsuspended in the air for" long ities. periods (LaForce1986). Droplets within this range are of the Field studies of airborne pathogensurvival at high humid- greatest important whenconsidering health effects since they ities at’e limited. Studies at lower humiditiessuggest a higher’ are small enoughto penetrate deep into the lungs. If these drop- survival rate for airborne viruses at humidities below 30% lets contain viable infectious organismsand are in sufficient (Green1985). In general, consideringthe results fromin vitro number’s,they will cause infection in susceptible hosts. Rela- studies, there is little evidenceto suggestthat for humiditiesin tive humiditycan affect the desiccation process of droplets, the upper range (> 50%RH) one specific level is better than which,in turn, affects droplet size and the viability and infec- any other’ in reducing the viability or numberof suspended tivity of the airborne pathogens(Burge et al. 1991). For’ more microorganisms~ informationconcerning specific types of infectious disease, see Burge (1989, 1995). Building-Related Sources A few microbes, pathogenic to humans, are able to AirborneViability flourish in nonhumanenvironments. These can be introduced Most of the information concerning the viability of into building systems from outside sources and proliferate if airborne pathogens has been determined through in vitro conditions are favorable. The most important exampleof such studies. Evidencefrom these studies suggests that the humidity a contaminantis Legionella, whichcan lead to fatal pneumonia level has complexeffects on the viability and virulence of in susceptible hosts. Aside from Legionnaire’s disease, no airborne pathogensthat vary from organismto organism, while specific infections have been documentedto be of great clinical the effect of temperatureis not significant within the range of importance in commercialbuildings (Hodgson1989). interest for’ conditioned environments.In manycases certain Hypersensitivity pneumonitishas been directly associated bacteria were found to have a windowof relative humidityat with microorganisms (pat’ticularly thermophilic actino- whichthey died morequickly (Andersonet al, 1968; Cox1966; mycetes)cultured fi’om poorly maintained humidification and Brundrett 1990). For’ example,Cox (1966) found that Escheri- air-conditioning systems(Hodges et al. 1974; Fink et al. 1976; chia coli strain jepp had minimalviability in the rangeof 70%to Burgeet al. 1980). Patients often report a relief of symptoms 80%RH and that the viability increased at RHvalues aboveand upon avoidance of the environmentcontaining the offending belowthis window.The results of a study by Hambletonet al. contaminant(LaForce 1986). Outbreaksof bacterial infections (1983) suggest that Legionella pneumophila has minimal in hospitals and flare-ups of asthmahave also been associated viability at two humiditylevels---between .50%and 60%RH and with humidificationsystems (Covelli et al. 1973; Airoldi et al. at 30%RH. Hambletonet al. also demonstrated that at the 1972; Smith et al. 1977; Solomon1974; Benckoet al. 1993). In optimumhumidity of 65%RH, only about 20%of the cells are all these cases the offendinghumidifiers have been of the spray viable after’ one hour; Experimentson the bacterium Pneu- or mist types that formaerosols in the airstream. moccussuggest a shatp decreasein viability in a narrowband at Use of direct evaporative cooling is a potential concern .50%RH (Brundrett 1990). In a study of the survival of Strepto- because poor’ maintenance,which is not uncommonin residen- coccus, Flynnet al. (1971) foundthat the changein the viable tial systems, can result in microbial growthwithin sumpwater count was insignificant for the RHrange of 0%to 92%. (Macher and Girman1990; Stetzenbach et al. 1990; Macher hz vitro studies of viruses suggest that particular strains, 1994). However,it is not apparent that this could lead to an including mengovims37A, polio virus, foot-and-mouth outbreak of disease. Onestudy of homesin the Las Vegasarea disease virus, and encophalomyocarditisvirus, are unstable as cooled by direct evaporative systems traced the presence of aerosols in atmospheres below about 70%RH (Cox 1989). gram-negativebacteria to a fouled sumpin one of the homes, contrast to these viruses, other’ strains, including vesicular although noneof the dweller’s was infected (Stetzenbachet aL stomatitis, vaccinia, and influenza, are least stable at high RH. 1990). In a laboratory study using tracer’ organisms, Macher Coxsuggests that this difference can be attributed to specific (1994) found"minimal transfer" from the fouled sumpinto the structural differencesin the viruses~Mbithi et al. (1991)studied air under normaloperating conditions. Conversely,direct evap- the survival of hepatitis A virus on surfaces at RHlevels of orative cooling mayhelp to reduce human-to-humanspread of 25%, 55%, 80%, and 95%and found that the survival of the infectious disease becauseof the relatively high supply rate of

202 ASHRAE Transactions: Research outside air required. Increased ventilation has been shownto m3 when the relative humidity increased from 34%to 70% lead to decreasedrates of viral respiratory infections (Burgeet duringa period of 24 hours. In addition, they foundthat short- al. 1991) and is often a recommendedmeans of reducing term increases in relative humiditywithin a period of one to indoor air contaminants (ASHRAE1989). five hours, as occur whenusing a sauna or drying clothes, also causedincreases in formaldehydelevels. The offgassing rate of NONBIOLOGICAL AIR POLLUTANTS chipboard within a chamberwas also examinedin this study Formaldehyde and it wasfound that sealing chipboardon all sides with "reac- Formaldehydeis found in numerousbuilding materials, tive paint" significantly reducedoffgassing rates. including plywood, particleboard, and other pressed wood Ozone products;fi~rnishings; carpets; andtextiles. It is also a ma.jor componentof urea-formaldehyde foam insulation, which is Ozoneis well recognized as a respiratory irritant and a nowbanned in the U.S. Formaldehydeis highly water soluble significant problemin urban areas of southern California, and causes irritation of the mucousmembranes within the eyes where the EPAstandard for outdoor ozone concentration is and upperrespiratory tract at concentrationsstarting at 0.1 ppm often exceeded. Ozoneis formed as a result of reactions but is most frequently reported at or above 1 ppm(Berglund et betweenphotochemically reactive hydrocarbonsand oxides of al. 1992). Formaldehydeis also classified by the U.S. Environ~ nitrogenunder the influenceof sunlight. Becauseof its strongly mental Protection Agency(EPA) as a probable carcinogen. oxidizing properties and low water solubility, ozonecan pene- Both ASHRAEand the World Health Organization (WHO) trate deepinto the alveoli of the lungs and affect lung function. have set maximumguidelines of 0.1 mg/m3 to ensure suffi- It is also an irritant to the mucousmembranes of the eyes. ciently low formaldehyde concentrations in indoor air The primarysource of ozoneindoors is infiltration/venti- (Puhakka and Karkkainen 1993; Gammage1990). The effect lation of polluted outdoorair. Indoorsources of ozone, such as of these standards, along with the ban on urea-formaldehyde photocopiers and air-cleaning equipment, exist; however, foaminsulation, has been an overall decrease in indoor formal- under normal living and working conditions there is no dehyde concentrations within the last decade (Marburyand evidence that these would reach levels high enoughto be of Krieger 1991). concern. Onceindoors, ozone decomposesthrough heteroge- The rate of formaldehydeoffgassing from pressed-wood neousreaction with indoor surfaces. Muelleret al. (1973) esti- productsdecreases exponentiallywith age and is sensitive to a matethe half-life of ozonein a typical bedroomto be less than numberof factors, including the initial properties of the mate- 6 minutes,while Weschler et al. (1991)suggest a half-life esti- rial, temperature, and humidity (Gammage1990; Meyer1986). mate of 11.1 minutes for a typical office environmentwith a Laboratoryand field studies agree that temperatureis the most surface-to-volumeratio of 2.9 m-1. Using this general knowl- significant environmental effect (van Netten et al. 1989; edge, authorities have typically advised the public to remain Godishand Rouch1986; Arundelet al. 1992). In general, since indoors and reduce infiltration as muchas possible during formaldehydewithin binding resins is water soluble, higher episodes of high outdoor ozone levels. However,this recom- humiditylevels also tend to increase the offgassing rate. In a mendationis not usefifl for those buildings that depend on controlled environmentstudy within mobile homes,a decrease natural ventilation or evaporative cooling as a means of in humidity from 70%to 30%resulted in an approximate 40% cooling. This condition is mademore problematic by the fact reduction of formaldehydelevels (Godish and Rouch1986). that ozone episodes often correspond with high outdoor a study of formaldehydeoffgassing from chipboard within a temperatures. controlled chamber, Andersonet al. (1975) found that the In the study by Weschleret al. (1991), ozone concentra- formaldehydeconcentration within the air was directly propor- tions were measuredfor three office buildings with different tional to temperature and air humidity. A change in relative ventilation rates. Theindoor values closely tracked those of the humidity from 30%to 70%doubled the equilibrium formalde- outdoor values and, dependingon the ventilation rate, were hyde concentration, while a 7°C rise in temperature in the 20%to 80%of those outdoors. Weschler et al. also cite a range of 14°Cto 35°Ccaused the formaldehydeconcentration numberof previous studies in whichthe concentration of ozone to double. In a study of 20 homesreferred to by Arundelet al. in the indoor environment was measuredto be 10%to 80%of (1992), a significant correlation was found betweenthe indoor outdoor concentrations. Considering this information, along relative humidity and the formaldehydeconcentration in the with the fact that people spend morethan 90%of their time air. indoors, Weschler suggests that exposure (concentration Seasonal fluctuations have also been observed, with the time) to ozoneindoors is a moresignificant issue than outdoor highest rates of offgassing occurring in summermonths, when exposures. markedincreases in temperature, solar gains (which can cause Thereis little informationregarding the effect of humidity localized increases in wall temperature), and humidity occur on ambient ozone concentrations within indoor environments. (Marburyand Krieger 1991). Puhakkaet al. (1993) studied One controlled-chamber study found that the rate of ozone apartments in Finland and found that the concentration of decay increased as either temperature or humidity was formaldehydein the air increased from 0.08 mg/m3 to 0.20 mg/ increased (Muelleret al. 1973). However,in the case of direct

ASHRAETransactions: Research 203 evaporative cooling, this may not be a significant effect humidity impact zones using bars that decrease in width, comparedto that of the higher ventilation rate. The impactof suggestinga decreasein the effect for each of the eight environ- direct evaporativecooling vs. refrigerated air conditioningwas mentalhealth factors addressed. Thesebars convergefor all of studied by Stock et al. (1993) in homesin El Paso and Houston, the eight categories into a nat~’ow recommended"optimum" Texas. They found that the average indoor/outdoor ratio of zone between 40%and 60%RH; both low- and high-humidity ozone concentrations was muchhigher in homeswith evapora- effects are addressed. There is clearly a need for’ such a tive cooling comparedto those with refrigerated air-condi- summarybecause it has appeared in numerousjournals and tioning systems(0.7 vs. 0.1). conferences(Arundel et al. 1986, 1992). It is also arrestingly The most promising remediation methodavailable is the drawnand easy to grasp, whichadds to its appeal. This figure is use of activated carbon filters, whichhave been shownto be the basis of the recommendationin ASHRAEStandard 62- successfulin significantly reducingozone levels (Muelleret al. 1989 that humidity in the occupied space should be between 1973;Weschler et al. 1991). It maybe possible that these could 30% and 60% RH. be incorporatedinto evaporativecooler’s. However,such factors There are issues that maybe raised with a figure that as engineeringpracticality, cost, efficiency, and service life attempts to combinethe influences of so manyfactors. One wouldneed to be tested under actual conditions to determineif concernis that it does not assign relative weights(severities) this optionis viable. for the different health factors. Anotheris that the practicality of the recommendedhumidity limits in various climates and Nitrogen and Sulfur Oxides building types is not assessed. In practice, these issues need to As with ozone, nitrogen and sulfur’ oxides are produced be addressed. As an example, ASHRAEStandard 62 ove~’ides primarily from outdoor sources~ However,nitrogen dioxide (without explanation) the recommendedlower limit of 40%, and nitrous and nitric acids are also combustionby-products of loweringit to 30%although, based on the figure, the impactof gas cookingstoves and heaters and can accumulateindoors as this action is not muchdifferent than that of raising the upper a result of improper ventilation. Nitrogen and sulfur oxides limit from 60%to 70%. This may have represented a value react with water on indoor surfaces to form acid aerosols, judgment by ASHRAEabout the relative severity of the which are generally found in higher concentrations indoors different health effects--differences that are not expressedin (Leadereret al. 1993). Althoughnitrogen and sulfur oxidesare the figure. Conversely, the change mayhave been forced by commonpollutants, surprisingly little workhas been donethus practical reality in conditioning the indoor environmentsof far to determinethe respiratory toxicity of their acid aerosols. buildings. Their acidic nature, reactivity, and aqueoussolubility, however, In defense of the figure, it is ultimately necessary for suggest that respiratory damageis possible (Brauer et al. 1993) designers to make decisions combining disparate elements and increased indoor humiditydoes seemto increase the heter- even without completejustification. Someoneneeds to draw a ogeneous reaction on surfaces. In one chamberstudy it was line in the sand. Thefigure doesthis for them. foundthat at the highest relative humiditytested (80%),nitrous A moresevere criticism is based on the factual substantia- acid (HONO)concentrations were approximately 8%of the tion of the health impact zones presented. As the figure is observed NO2 level, while 30%and 4.5% relative humidities drawn and captioned, these zones imply linear relationships resulted in HONO/NO2 ratios of 0.9% and 2o7%,respectively betweenhumidity and health effects that are not supported by (Brauer et al. 1993). In te~xns of direct evaporative cooling, the literature. In addition, the focus on ambient RHto the again the issue of high ventilation rates maybe a significant exclusion of other’ environmentalconditions is misleading in factor to consider if outdoorair levels of nitrogen and sulfur that it suggests that RHis the controlling environmentalfactor oxides are high. for all of the pollutants listed withoutregard to climate and the conditions of the building system. DISCUSSION In addition, the specific points at whichthese zones begin To the maximumextent possible, building standards and end are not consistently supported by the references should reflect the knowledgeavailable whenthey are written. provided. The recommendationsfor bacteria (RHbelow 60%), The subject of humidity limits is a complex one and many viruses (RH below 70%), and fungi (RH below 60%) are questions remain at this time. Nonetheless, there is a wide supportedby the discussion or the referencesgiven in the orig- range of research under way through numerousdisciplines, inal paper or its morerecent versions. Theliterature cited for muchof which can inform the preparation of standards. This the mite recommendation(below 50%RH) is incomplete, and paper summarizes this research and suggests a format for seasonal effects, whichmay be a significant factor, are not putting newinformation into a building standards context. addressedlThe limits for allergic rhinitis are presumablybased A paper’by Sterling et al. (1985)also addressesthe topic on the potential for mites and fungal growth (for upper humidityand health in buildingsand is the only cited reference humidity limits), although most of the discussion concerns pertinent to humidity in ASHRAEStandard 62-1989. This problemsof mist humidifiers and low humidities. The chem- paper includes a figure that has received wide circulation ical interactions category (RHbelow 30%) appears to be based within the HVACengineering profession. It graphically depicts entirely on two studies of formaldehydeoffgassing rates (the

204 ASHRAETransactions: Research only other chemicalinteraction mentionedin the text is the effective in reducing mite levels. In terms of seasonal conversionof sulfur and nitrogen dioxides to acids, for which effects, it is not yet knownwhether use of evaporative no specific limits werecited). In light of the recent reductions cooling a few months out of the year, sometimes only in the formaldehydelevels within materials, this limit maybe during the day, leads to increased mites. In general, since outdated. The use of the term "ozoneproduction" as a pollutant mite contamination occurs primarily in residences and category is confusing and possibly based on the incorrect affects only a subset of the population, it maybe that, assumption that ozone production from office lighting and when necessary, they should be controlled by other equipmentis a significant pollutant source. meanssuch as cleaning and covering bedding, treatment For the purposeof setting humiditystandards, the figure is or removalof carpets, and insulation of cooled floor slabs clearly inadequate. To promote good building design it is under carpets. importantto identify the specific physical causes and solutions ¯ Fungal contamination occurs primarily as a result of to health hazardsand to regulate design practice to avoid them. condensation on surfaces and/or water damage.Field and Table 1 summarizesthe results of this review. It addresses laboratory studies suggest that fungal growth does not the humidity requirements, the actual site of contamination, becomean issue below 70%or even 80%RH unless there and the meansof control for each of the biological pollutants. are other factors influencing their growth on building CONCLUSIONS surfaces. Studies that reported problemsat lower RHvalues appearedto have problemsthat could be corrected other- ¯ Mostof the identified biological health agents growon the wise. In setting a maximumlimit to air humidityin the surfaces of the building, its systems,and its furnishings,or space,there is little, if any, evidencefrom field studies that in standing waterwithin or outside the building. Noneof the provides a reason for distinguishing 60%relative humidity agents growsin the air of the occupiedspace or the mechan- from 70%. ical system.Their growthis therefore only indirectly related ¯ to the atmospheric humidity measured in the occupied The health impactsof nonbiologicalhealth agents are hard space or the ducts of its mechanicalsystem. To control to assess at this time. Formaldehydegeneration is exacer- these, one needs to ensure that the surfaces remain dry. bated in somematerials by higher humidity. Becauseof the Thereare a numberof waysto achieve this in the design, greater awarenessof the adversehealth effects, newbuild- furnishing,and operation of buildings.It is also necessaryto ing productsand furnishings generate far less formaldehyde avoid producing aerosols of water from the mechanical than before. Theeffect of this changewill needto be eval- system or humidifiers. Howthis is doneis independentof uated. For a given ventilation rate, indoorozone concentra- the level of indoorair humidity. tions may decrease as humidity increases due to an increased rate of surface reactions. However,at the high ¯ The single exception to this is dust mite contamination ventilation rates associatedwith direct evaporativecooling, (particularly D. farinae), which appears to be directly the level of ozonewill not be significantly offset by the related to ambient RH. Controlled laboratory studies higher humiditylevels. Oxidesof nitrogen and sulfur are suggest that optimal conditions for growthare 70%to 80% primarilyof outdoororigin, but the severity of their effects RHat 25°C. A numberof field studies have found mite on health mayincrease with higher humiditylevels. At this contamination in residences with ambient RHsas low as point there is little evidencein the literature to suggestthat 50%. Rather than ambient RH, however, the more rele- this is a significanthealth effect. vant factor maybe the RHwithin the microhabitat of the mite (within a few millimeters of the horizontal surfaces Theother significant source of biological health agents is on which they lie). In the laboratory, surface and air humansharboring infectious diseases. This source (prima- temperatures can be controlled to provide equilibrium rily the respiratorytract but also the skin) is largely inde- conditions; however,in actual environmentsequilibrium pendent of the humidity level in the space. However,the conditions seldomexist. In one field study, the RHwithin spread of infectious disease agents dependssomewhat on carpets was found to be more than 9%higher than the atmospherichumidity in the space, in that aerosol evapora- space RH, suggesting lower temperatures or local mois- tion rates anddeposition rates mayaffect viability of anti- ture production that maybenefit mites. This difference gens, bacteria, and viruses enclosed in water droplets. between surface and ambient RH maypossibly explain Space humiditymay also affect the settling rate of dust the discrepancy betweenthe laboratory and field find- particles to whichbacteria are attached. Theviability of ings. Remedialmethods are designed to address contam- these dustborne organismsalso varies with humidity, with ination at the source. For example,one of the most widely viability optimaoccurring throughoutthe range of RH.For recommendedremedial methods is to encase mattresses eachof the aboveconsiderations, there is little evidenceto with a semipermeablepolyurethane cover. This discour- suggest that any humiditybetween 50% and 90%is signif- ages moisture from getting into the bedding where the icantly better than any other in reducing the viability or mites live. Researchershave also suggestedthat electric numberof suspendedinfectious disease organisms,as well blankets, which can lower the RHwithin bedding, are as the susceptibility of the humanreceptor.

ASHRAETransactions: Research 205 o Direct evaporative cooling through porous mediaappears to Arlian, L.G., D. Bernstein, I.L. Bernstein, S. Friedman, A. be benignin that field and laboratory studies suggest that Grant, E Lieberman, et al. 1992. Prevalence of dust biological organismsin the cooling water appear not to be mites in the homesof people with asthmaliving in eight aerosolized or transmitted downstream.The wet pads may different geographicareas in the United States. Journal have benefits over dry filters in removingincoming pollut- of Allergy and Clinical Immunology90(3): 292-300. ants. However,this needsto be experimentallyinvestigated. Arundel, A.V., E.M. Sterling, J.H. Biggin, and T.D. Sterling. In addition, the once-throughventilation requh’edby such 1986. Indirect health effects of relative humidity in systemsshould act to dissipate the concentrationof infec- indoor environments. Environmental Health tious organismsin the ai~; since such organismsare almost Perspectives 65:351-361. alwaysinternally generated. This process also needs to be Arundel, A.V., E.M. Sterling, J.H. Biggin, and T.D. Sterling. systematicallyevaluated. 1992. Indirect health effects of relative humidity in ¯ Finally, very little of the literature on health effects is indoor environments. Desiccant Cooling and expressedin termsof risk to the occupant:first, the likeli- Dehumidification, pp. 3-12. Atlanta: AmericanSociety hood of humidity-influencedpollutants occur~’ing in the of Heating, Refrigerating and Air-Conditioning building and then the likelihood of the pollutant affecting Engineers, Inc. the occupant. ASHRAEo 1989~ ANSI/ASHRAE Standard 62-1989, Ventilation.for acceptable indoor air. Atlanta: American ACKNOWLEDGMENTS Society of Heating, Refl’igerating and Air-Conditioning Engineers, Inc. The authors would like to thank Alison Kwok,a Ph.D. candidate at the Universityof California at Berkeley, for help ASHRAE.1992. ANSI/ASHRAEStandard 55-1992, Thermal environmental conditions3’br humanoccupancy. Atlanta: with the literature search. Wealso wish to thank WilliamFisk AmericanSociety of Heating, Refi’igerating and Air- of the Indoor’ EnvironmentProgram at Lawrence Berkeley Laboratory and Professor WayneWilcox of the University of Conditioning Engineers, Inc. California Forest Products Laboratory for their detailed August, P.J. 1984. Housedust mite causes atopic eczema. A reviews of the manuscript. Tile research reported here was preliminary study. British Journal of Dermatology funded by the California Institute for Energy Efficiency 3(suppl 26): 10-11. (CIEE),a research unit of the Universityof California. Publi- Avari, G.E, and D. Allsopp. 1983. The combinedeffect of cation of the research does not imply CIEEendorsement of or pH, solutes, and water activity on the growth of some agreementwith these findings, nor that of any CIEEsponsor. xerophilic Aspergillus species. Biodeterioration 5, T.A. Oxley and S~ Barry, eds, pp. 548-556. NewYork: John REFERENCES Wiley & Sons Ltd. Ayerst, G. 1969. The effects of moisture and temperature on Abe, K. 1993. A method for’ numerical characterization of growth and spore germination in some fungi. Journal of indoor climates by a biosensor using xerophilic fungus. Stored Produce Research 5: 127-141. lndoorAir 3: 344-348. Bates, J.M., D.A. Rorek, and M.H. Ballantye. 1993. Dust Airoldi, T., and W. Litsky. 1972. Factors contributing to the mite counts and mite allergens in family homesbefore microbial contamination of cold-water humidifiers. and after dry extraction carpet cleaning. Indoor Air ’93, American Journal of Medical Technology 38(12): 491- Proceedings of the 6th International Conferenceon Air 495. QualiO, and Climate, July 4-8, Helsinki, Finland, vol. 2, Anderson,I., and J. Korsgaard. 1986. Asthmaand the indoor pp. 33-38. environment: Assessmentof the health implications of Bayer, C.W., and C.C. Downing. 1992. Indoor humidity in higl~ indoor’ air’ humidity. EnvironmentInternational 12: schools with insufficient humiditycontrol. Environments pp. 121-127. .¢br People: Proceeding of IAQ ’92, pp. 197-200, San Anderson, J.D., EA. Dark, and S. Peto. 1968. The effect of Francisco. Atlanta: American Society of Heating, aerosolation upon survival and potassium retention by Refrigerating and Air-ConditioningEngineers, Inc. various bacteria. Journal of General Microbiology 52: Bayer’, C.W, and S. Crow. 1993. Detection and pp. 99-105. characterization of microbially producedvolatile organic Arlian, L.G. 1992. Water balance and humidity requirements compounds. Indoor Air ’93: Proceedings of the 6th of house dust mites. Experimental and Applied International Conference on Air QualiO, and Climate, Acarology 16: 15~35. July 4-8, Helsinki, Finland, voL6, pp. 297-302. Arlian, L.G, I.L. Bernstein, J.S. and Gallagher~ 1982. The Beaumont,E, H.E Kauffman,H.J. Sluiter, and K. de Vries. prevalence of house dust mites, dermatophagoids, and 1985. Volumetricaerobiological survey of conidial fungi associated environmental conditions in homesin Ohio. in the North-East Netherlands. II. Comparison of Journal of Allergy and Clinical Immunology69(6): 527- aerobiological data and skin tests with mouldextracts in 532. an asthmatic population. Allergy 40:181-186.

206 ASHRAETransactions: Research TABLE 1 BIOLOGICAL POLLUTANTSIN INDOOR AiR Summaryof Environmental Requirements and RecommendedRemediation Procedures

HealthImplications Environmental CommonSites of Frequently Recommended Requirements Contamination RemediationProcedures Dust Allergic reactions: asthma,Sourceof nutrients:shed ¯ Mattresses/bedding ¯ Removalof contaminatedcarpet- Mites rhinitis, dermatitis humanand animal skin ¯ Thick carpeting ing, bedding,and/or furniture Dustmite allergens: mite scales ¯ Heavily used upholstered " Frequenthot-water cleaning of bodyparts and fecal matter Sourceof water: furniture bedding ¯ are the most commonsource ¯ Uptakewater vapor (Dust mites are more Encasementof mattresses with of allergic reactionswithin directly fromair commonlyassociated with semi-permeablevinyl casing house dust RHof 70-80%at 25°Cis residences rather than Special vacuumingprocedures Species most commonly optimalfor growth commercialbuildings) (e.g., HEPAfilters, central Critical relative humidity: vacuumsystem with outside associatedwith disease: equipment) 55%at 15°Cto 75%at 35°C Dermatophagoides Surfacetreatments (e.g., benzyl Dopteronyssinus- more(For moredetailed informa- benzoate) commonin humid tion on the humidityrequire- Reductionof ambienthumidity or regions mentsfor mites see Arlian, specifically within the microhabi- D. farinae - more 1992) tat of mites(e.g, throughthe use of commonin relatively electric blankets,radiant heating dry regions of carpetedfloor surfaces,etc.) Euroglyphus- generally rare, foundonly in humidregions Fungi Allergic reactions: asthma,Source of nutrients - Moisturedamaged build- ¯ Removalof damagedmaterial rhinitis, dermatitis(most organicdebris, dirt, organic ing materials(walls, wherepossible (i.e. carpeting,duct common:Alternaria, buildingmaterials carpeting,books, etc.) liners, wallpaper,etc.) Aspergillus, Cladosporium,Sourceofwater:moisture Within fiberglass duct ¯ Cleaningof waterresistant materi- and Penicillium) on and within surfaces lining in whichcondensa- als with chlorinebleach tion has occurred ¯ HypersensitivityPneu- ¯ Specific water require- Proper maintenanceand operation monitis(e.g., Cladospo- mentvaries fromspecies Onwall surfaces with of conditioningsystems (eog., rium, Sepula) to species. high ERHor on which cleaningand disinfecting cooling condensationhas coils and drain pans, continual Mycotoxicosis ¯ Growthof xerophilic occurred operationof forcedair systemsto species such as Aspergil- (e.g., Aspergillus,Penicil- Withinpoorly main- avoid condensation) lus is likely to beginat an o lium, Stachybotrysatra, ERHof 75%to 80%. tained conditioning Properconstruction techniques to Trichodermavirde ) systemscontaining water avoidwater damage(e.g., proper (For moredetailed informa- Infectious disease (e.g., humidifiers,cool- placementof vaporbarriers to tion on humidityrequire- ing coil drip pans) avoidcondensation within walls, (e.g.,Aspergillusfi~migams,ments see Flannigan1992 designof drainagesystems to Cryptococcus) and IEAAnnex 14 guide- avoidflooding and water incur- lines.) sion) Adequateventilation to reduce internal humidityloads Bacteria Infectiousdisease: person- ¯ Virulentbacteria and ¯ Infected humans Person-to-personspread of and to-person(e.g., common viruses are transferredto ¯ Improperly maintained airborneinfection: Viruses cold, flu, measles,TB, etco) hosts throughdroplet building systemsthat ¯ Isolation of contaminatedpersons Infectiousdisease: build- nuclei expelled when have the potential to form¯ Increasefresh air exchangerate coughing/sneezingor aerosols:(e.g., spray-type ing related(e.g., Legionella formedwithin contami- Growthwithin building systems: pneumophila) humidifies,cooling nated building systems towersetco) ¯ Removesource of contamination Hypersensitivitypneu- that formaerosols (e.g., replacesystem, biocide treat- monltis (most commonly ¯ Bacteriaand viruses that ment,etco) associated with thermo. growoutside of the ¯ Routinecleaning of water systems philic actinontycetes) humanhost require liquid and/orfilters water for growth ¯ Relocateintake vents (in the case (For detailedinformation, of cooling tower contamination) see Cox1989.)

ASHRAETransactions: Research 207 Becket’, R. 1984. Condensation and mould growth in solutions of protecting agents, as a function of relative dwellings--Parametric and field study. BuiMings and humidity. Journal of General Microbiology43: 383-399. Environment 19(4): 243-250. Cox, C.S. 1989. Airborne bacteria and viruses. Science Bencko,V., J. Melichercik, V. Melichercikova,and Z. Wirth. Progress 73: 469-500. 1993. Microbial growthcontrol in spray humidifier’s of deBoer, R., and L.P. van der Geest. 1990. House dust mite health facilities. IndoorAir3: 20-25. (Progylphidae) populations in mattresses, and their Berglund, B., B. Brunekreep, H. Knopple, T. Lindvall, L. control by electric blankets. Experimental and Applied Maroni, et al. 1992. Effects of indoor air pollution on Acarology 9(1-2): 113-122. humanhealth. IndoorAir 2: 2-25. Ellingson, A.R., R.A. LeDoux, P.K. Vedanthan, and R.W. Bjurman, J. 1993. Thermal insulation materials, Weber. 1995. The prevalence of Detmatophagoldesmite microorganismsand the sick building syndrome. Indoor allergen in Coloradohomes utilizing central evaporative Air ’93: Proceedingsof the 6th International Conference coolers. Journal of Allergy and Clinical Immunology on Air Quality and Climate, July 4-8, Helsinki, Finland, 96(4): 473-479. vol. 4, pp. 339-344. Ezeonu, LM., J.A. Noble, R.B. Simmons,D.L. Price, S.A. Block, S.S. 1953. Humidityrequirements for mold growth. Crow, D.G. Ahearn, et al. 1994. Effect of relative Applied Microbiology 1: 287-293. humidityon fungal colonization of fiberglass insulation. Brauer, M., T.R. Rasmussen, and S.K Kjaergaard. 1993. Applied and Environmental Microbiology 60(6): 2149- Indoor nitrous acid, formation and human exposure 2151. effects. Indoor Air ’93: Proceedings of the 6th Fernandez-Caldas, E., W.L. Trudeau, and D.K. Ledford. h, ternational Conference on Indoor Air Quality and 1994. Environmentalcontrol of indoor’ biologic agents. Climate, July 4-8, Helsinki, Finland, vol. 3, pp. 141-146. Journal of Allergy and Clinical hnmunology94(2): 404- Brundrett, G.W.1990. Criteria for moisture control. London: 412. Butterworth & Co~ Fink, J.N., G.T. Hensley, D.P. Schleuter, and G.E Unger. Burge, H.A. 1988. Environmentalallergy, definition, causes, 1976. Interstitial lung disease due to contamination of control. Proceedings of IAQ ’88. Atlanta: American forced air systems. Annals oflnternal Medicine84: 406- Society of Heating, Refrigerating and Air-Conditioning 413. Engineers, Inc. Flannigan, B. 1992. Approaches to assessment of the Burge, H.A. 1989. Indoor air and infectious disease. microbial flora of buildings. Environmentsfor People: Occupational Medicine: State of the Art Reviews 4(4): Proceedings oflAQ ’92, September, San Francisco, pp. 713-721. 139-145. Atlanta: American Society of Heating, Burge, H.A. 1995~ Aerobiology of the indoor environment. Refrigerating and Air-ConditioningEngineers, Inc. OccupationalMedicine: State of the Art Reviews 19(1): Flannigan, B, and D. Miller. n.d. Health implications of 27-40. fungi in indoor environments---An overview. Burge, H.A., W.R.Solomon, and J.R. Boise. 1980. Microbial Unpublished. prevalence in domestic humidifiers. Applied and Flannigan, B., E.M. McCabe, and E McGarry. 1991. Environmental Microbiology 39(4): 840-844. Allergenic and toxigenic micro-organisms in houses. Burge, H.A., LC. Feeley et al. 1991. Indoor air pollution and Journal of Applied Bacteriology, Symposium infectious diseases. Indoor Air’ Pollution, J.M. Samet Supplement,vol. 70, pp. 61S-73S. and JD. Spengler; eds., pp. 273-284~ Baltimore, Md.: Flynn, D.D., and L.J. Goldberg. 1971. Effect of relative Johns HopkinsUniversity Press. humidity on aerosol persistence of Streptococcus CL 1994. Personal communications.Berkeley, Calif.: Center salivarius. Archives of EnvironmentalHealth 23: 40-42. Laboratories. Foarde, K., R Dulaney, E. Cole, D. VanOsdell,D. Ensor, and Cabrera, P., G. Julia-Serda, EC. Rodriguez, J.B.D. J. Chang. 1993. Assessmentof fungal growth on ceiling Camminero,and T. Carrillo. 1995. Reduction of house tiles under environmentally characterized conditions. dust mite allergens after dehumidifier use. Journal of Indoor Air ’93: Proceedings of’ the 6th International Allergy and Clinical hnmunology95(2): 635-636. Conferenceon Indoor Air Quality and Climate, July 4-8, Colloff, M.J. 1992. Exposureto house dust mites in homesof Helsinki, Finland, vol. 4, pp. 357-362. people with atopic dermatitis. British Journal of Friedman, RM., H.M. Friedman, and G.T. O’Connor. 1992. Dermatology127(4): 322-327. Prevalence of dust mite allergens in homes and Covelli, H.D., J. Kleeman,J.E. Martin, W.L. Landau, R.L. workplaces of the Upper Connecticut River Valley of Hugheset al. 1973. Bacterial emission fi’om both vapor NewEngland. Allergy Proceedings 13(5): 259-262. and aerosol humidifier’s. American Review of Gammage,R.B. 1990. Exposure to formaldehyde in indoor Respirator), Disease 108: 698-701. air. Risk Analysis 10(1): 77-83. Cox, C.S. 1966. The survival ofEscherichia coli sprayed into Godish, T., and J. Rouch. 1986. Mitigation of residential air and into nitrogen from distilled water and frorn formaldedydecontamination by indoor climate control.

208 ASHRAETransactions,: Research American h~dustrial I~ygiene Association Journal 47: materials. Indoor Air ’9.3: Proceedings of the 6th 792-797. bzternational Conference on Air Quality and Climate, Grant, C., C.A. Hunter’, B. Flannigan, and A.E Bravery. 1989. July 4-8, Helsinki, Finland, vol. 4, pp. 299-303. International biodeterioration 25: 259-284. Kalra, So, S.J. Owen,J. Hepworth, and A. Woodcock.1990. Gravesen, S. 1979. Fungi as a cause of allergic disease. Airbornehouse dust mite antigen after vacuumcleaning. Allergy 34: 135-154. Lancet 336: 449. Green, G.H. 1985. Indoor relative humidities in winter and Korsgaard, J. 1983a. Mite asthma and residency, a case- the related absenteeism. ASHRAETransactions 91 (1). control study on the impact of exposure to house-dust Griffin, D.H. 1981. Fungal physiology. NewYork: Wiley and mites in dwellings. American Review of Respiratory Sons. Disease 128:231-235. Hambleton, E, M.G. Broster, EJ. Dennis, R.E Henstridge, Korsgaard, Jo 1983b. House-dust mites and absolute indoor JoW.Conlan et al. 1983. Survival of virulent Legionella humidity. Allergy 38(2): 85-92. pneumophilain aerosols. J. Hyg. Camb.90:451-460. LaForce, EM. 1986. Airborne infections and modern Hart, B.J., and L. Whitehead. 1990. Ecology of house dust building technology. EnvironmentInternational 12:137- mites in Oxfordshire. Clinical and ExperimentalAllergy 146, 20: 203-209. Lang, J.D., and M.S. Mulla. 1977. Distribution and Harving, H., J. Korsgaard,R. Dahl, H.I. Beck, E Bjerring et abundance of house dust mites, Dermataphagoides, in al. 1990. Housedust mites and atopic dermatitis. A case- different climatic zones of southern California. control study on the significance of house dust mites as Environmental Entomology6(2): 213-216. etiolofic allergens in atopic dermatitis. Annalsof Allergy Lang, JoD., and M.S. Mulla. 1978. Seasonal dynamics of 65: 25-31. house dust mites, Dermatophagoides, in homes in Harving, H, J. Korsgaard, and R. Dahl. 1994. House-dust southern California. Environmental Entomology 7(2): mite exposure reduction in specially designed, 281-286. mechanicallyventilated "healthy" homes.Allergy 49(9): Leaderer, B.E, J. Sullivan, E Koutrakis, and J.M. Wolfson~ 713-718. 1993. A passive monitor for the measurementof nitrous Hens, H., and E. Senave. 1990. Condensation and energy, acid and sulfur dioxide, h~doorAir ’9.3: Proceedingsof Zolder case study. Journal Ventilation System the 6th hzternational Conferenceon Indoor Air Quality Performance, pp. 81-99, 11 AIVC Conference, and Climate, July 4-8, Helsinki, Finland, vol. 2, pp. 233- September18-21, Belgirate, Italy. 238. Hodges, G.R, J.N. Fink, and D.E Schlueter. 1974. I, intner, T.J., and K.A. Brameo1993. The effects of season, Hypersensitivity pneumonitis caused by a contaminated climate, and air-conditioning on the prevalence of cool-mist vaporizer. Annals of Internal Medicine 80: Dermatophagoides mite allergens in household dust. 501-504. Journal of Allergy and Clinical hnmunology91(4): 862- Hodgson, M.J. 1989. Clinical diagnosis and managementof 867. building-related illness and the sick building syndrome. Macher, J.M. 1994. Personal communication. Occupational Medicine: State of the Art Reviews 4(4): Macher, J.M., and J.R. Girman. 1990. Multiplication of 593-606. microorganisms in an evaporative cooler and possible Htut, T. 1994. Airbornehouse dust mite allergen: Are short- indoor air contamination. EnvironmentInternational 16: duration avoidance measures helpful? Indoor 203-211. Environment3: 48-53. Marbury, M.C., and R.A. Krieger. 1991. Formaldehyde. Hung, L.I_,., C.So Yang, EJ. Dougherty, EA. Lewis, EA. Indoor Air Pollution: A Health Perspective, J.M. Samet Zampiello, and L. Magniaracina. 1992. Dust mite and and J.D. Spengler, eds., pp. 223-251. Baltimore, Md.: cat dander allergens in office buildings in the mid- Johns HopkinsUniversity Press. Atlantic region. Environmentsfor People: Proceedings Mbithi, J.N., V.S. Springthorpe, and S.A. Sattar. 1991. The oflAQ ’92, October 18-21, San Francisco, pp. 87-93. effect of relative humidity and air temperature on Hyvarinen, A., T. Reponen,T. Husman,J. Ruuskanen,and A. survival of Hepatitis A virus on environmentalsurfaces. Nevalainen. 1993. Characterizing mold problem Applied and Environmental Microbiology 57(5): 1394- buildings~Concentrations and flora of viable fungi. 1399. IndoorAir 3: 337-343. McDonald, L.G., and E. Tovey. 1992. The role of water lEA. 1991. lEA Annex 14 (Condensation and Energy), temperature and laundry procedures in reducing house Guidelines and practice: Energy conservation in dust mite populations and allergen content of bedding. building and communitysystems. Paris: International Jou~7~alof Allergy and Clinical hnmunology90(4, pt 1): Energy Agency. 599-608. Kalliokoski, E, Z Juutinen, E Pasanen, and A. Pasanen. Menzies, R., R. Tamblyn,E Comtois, C. Reed, J. Pasztor, et 1993. Fungal growth and activity on moist building at. 1992. Case-control study of microenvironmental

ASHRAETransactions: Research 209 exposures to aero-allergens as a cause of respiratory temperature and humidity. Environment International symptoms---Part of the sick building syndrome (SBS) 17: 225-228. symptom complex. Environments for People: Pasanen, A.L., H. Heinonen-Tanski,E Kalliokoski, and M.J. Proceedings of IAQ ’92, Octobex; San Francisco. Jantunen. 1992. Fungal microcolonies on indoor Atlanta: AmericanSociety of Heating, Refrigerating and surfaces--An explanation for the base-level fungal spore Air-Conditioning Engineers, Inc. counts in indoor air. Atmospheric Environment 26B: Meyer, B. 1986. Formaldehyde exposure from building 117-120. products. EnvironmentInternational 12: 283-288. Pasanen, E, A. Pasanen, and M. Jantunen. 1993. Water Miller, J.D. 1992. Fungi and the building engineer. condensation promotes fungal growth in ventilation Environments for People: Proceedings of IAQ ’92, ducts. IndoorAir 3:106-112. October, San Francisco, pp. 147-159. Atlanta: American Platts-Mills, T.A.E., M.L. Hayden, M.D. Chapman,and S.R. Society of Heating, Refrigerating and Air~Conditioning Wilkins. 1987. Seasonal variation in dust mites and Engineers, Inc. grass-pollen allergens in dust fromthe housesof patients Moore-Landecker, E. 1982. Fundamentalsof the fungi, 2d with asthma. Journal of Allergy and Clinical ed. EnglewoodCliffs, N.J.: Prentice-Hall. Immunology79(5): 781-791. Morey, RR~1988. Microorganisms in buildings and HVAC Platts-Mills, T.A.E., A.L. de Wecket al. 1989. Dust mite systems, a summary of 21 environmental studies. allergens and asthma--A worldwide problem. Journal Proceedings of IAQ ’88, pp. 10-24. Atlanta: American of Allergy and Clinical Immunology83(1-3): 416-427. Society of Heating, Refrigerating and Air-Conditioning Puhakka, E., and J. Karkkainen. 1993. Formaldehyde Engineers, Inc. concentrationsin indoor air in newbuildings; factors that Morey,RR., C.M.Williams et al. 1991. Is porous insulation affect the prevalence of formaldehyde.Indoor Air ’93: inside an HVACsystem compatible with a healthy Proceedings of the 6th International Conference on building? Health), Buildings: Proceeding03" IAQ ’91, Indoor Air Quality and Climate, July 4-8, Helsinki, September, Washington, D.C. Atlanta: American Finland, vol. 2, pp. 147-152. Society of Heating, Refrigerating and Air-Conditioning Reed, C.E., and M.C. Swanson. 1986~ Indoor allergens, Engineers, Inc. identification and quantification. Environment Mueller, EX., L. Loeb, and W.H. Mapes. 1973. International 12:115-120. Decomposition rates of ozone in living areas. Reise~; J., D. Ingram, E.B. Mitchell, and J.O. Warner.1990. EnvironmentalScience and Technology 7(4): 342-346. Housedust mite allergen levels and an anti-tnite mattress Murray, A~B., and R Zuk. 1979. The seasonal variation in a spray (natamycin) in the treatment of childhood asthma. population of house dust mites in a North Americancity. Clinical and Experimental Allergy 20(5): 561-567. Journal of Allergy and Clinical Immunology64(4): 266- Seltzer, J.M. 1995. Biologic contaminants. Occupational 269. Medicine: State-@the-Art Reviews 10(1): 1-25. Murray, A.B., A.C. Ferguson, and B.J. Morrison. 1985. Smith, RW.,and R.M. Massanari. 1977. Roomhumidifiers as Sensitization to house dust mites in different climatic the source of Actinetobacter infections. JAMA237(8): areas. Journal of Allergy and Clinical Immunology 795-797. 76(1): 108-112. Smith, T.E, LoB. Kelly, RW.Heymann, S.R. Wilkins, and Nevalainen, A. 1993. Microbial contamination of buildings. T.A.E. Platts-Mills. 1985. Natural exposure and serum Indoor Air ’93: Proceedings of the 6th h~ternational antibodies to house dust mite-allergic children with Conferenceon Indoor Air Quality and Climate, July 4-8, asthma in Atlanta. Journal of Allergy and Clinical Helsinki, Finland, vol. 4, pp. 3-12. Immunology76(6): 782-788. Nevalainen, A., A.L. Pasanen, M. Nininen, T. Reponen, R Solomon, W.R1974. Fungus aerosols arising from cold-mist Kalliokoski, and M.J. Jantunen. 1991~The indoor air vaporizers. Journal of Allergy and Clinical Immunology quality in Finnish homes with mold problems. 54(4): 222-228. EnvironmentInternational 17: 299-302. Sorenson, W.G. 1989. Health impact of mycotoxins in the O’Rourke, M.K., L. Fiorentino, D. Clark, M. Ladd, and S. homeand workplace: An overview. In Biodeterioration Rogan. 1993. Building characteristics and importance of Research 2, C.E. O’Rear and G.C. Llewellyn, eds., pp. house dust mite exposures in the Sonoran desert, USA. 201-215. NewYork: PlenumPress. Indoor Air ’93: Proceed#~gsof the 6th h~ternational Sterling, E.M., A. Arundel,and T.D. Sterling. 1985. Criteria Conferenceon b~doorAir Quality and Climate, July 4-8, for humanexposure to humidity in occupied buildings. Helsinki, Finland, vol. 4, pp. 155-160. ASHRAETransactions 91 (1). Pasanen, A.L., E Kallioloski, E Pasanen, M.J. Jantunen, and Stock, T.H., and E.A. Venso.1993~ The impact of residential A~ Nevalainen. 1991. Laboratory studies on the evaporative air cooling on indoor exposure to ozone. relationship between fungal growth and atmospheric h~door Air ’93: Proceedings of the 6th International

210 ASHRAETransactions: Research Conference on b~door Air and Climate, July 4-8, respiratory symptoms:The role of sensitization to dust Helsinki, Finland, vol. 4, pp. 251-256. mites and molds. American Journal of Epidemiolog), TenWolde,A., and W.B.Rose. 1993. Criteria for humidity in 141(2): 103-110. the building and the building envelope. Bugs, Moldand Voorhorst, R., M.I.A. Spieksma-Boezeman, and ETH.M. Rot II: Proceedings of Workshopon Control of Humidity Spieksma. 1964. Is a mite (Dermatophagoides) the for Health, Artifacts and Buildings, November16-17, producer of the house-dust allergen? Allergy und Asthma pp. 63-65. Oak Ridge, Tenn.: Oak Ridge National Band 6: 329-334. Laboratory. Weschler, C.J., V.N. Datta, and H.C. Shields. 1991. Indoor Thogersen, K., 1,. Gunnarsen, and P.A. Nielsen. 1993. The ozone exposures from the infiltration of outdoor ozone. effects on indoor air quality by water-damaged Indoor Air Pollution: Radon, Bioaerosols and VOC’s, chipboards, hzdoor Air ’9.3: Proceedings of the 6th J.G. Kay, G.E. Keller, and J.E Miller, eds. pp. 83-99. International Conference on Indoor Air and Climate, Wickman,M., S.L. Nordvall, G. Pershagen, J. Sundell, and July 4-8, Helsinki, Finland, vol. 2, pp. 537-542. B. Schwartz. 1991. House dust mites sensitization in Tsongas, G. 1991. A field study of indoor moisture problems cihildren and residential characteristics in a temperate and damagein new Pacific Northwest homes. Health), region. Journal of Allergy and Clinical Immunology Buildings: Proceeding of IAQ ’91, September, 88(1): 89-95. Washington, D.C., pp. 202-209. Atlanta: American Wickman,M., S.L. Nordvall, G. Pershagen, J. Korsggard, N. Society of Heating, Refrigerating and Air~Conditioning Johansen, and J. Sundell. 1994a. Mite allergens during Engineers, Inc. 18 monthsof intervention. Allergy 49(2): 114-119. van Netten, C., C. Shirtleffe, and J. Svec. 1989. Temperature Wickman,M., G. Emenius, A.C. Egmar, G. Axelsson, and G. and humidity dependenceof formaldehyde release from Pershahen. 1994b. Reduced mite allergen levels in selected building materials. Bulletin of Environmental dwellings with mechanical exhaust and supply Contaminationand Toxicology 42: 558-565. ventilation. Clinical and Experimental Allerg), 24(2): Verhoeff, A.P., R.T. van Strien, J.H. van Wijnen, and B. 109-114. Brunekreef. 1995. Damp housing and childhood Wilcox, W. 1995. Personal communication.

ASHRAETransactions: Research 211