Tobacco Documents Online

Indoor Air 1997; 7:143-150 B925 XF~.7$ lad] Printed in Denmark. All rights reserved IND~IOR AIR 97 ~ClffdHKSGAARD INT PU]~L LTD DE Copyright © Munksgaard 1997 ll~DOOR AIR ISSN 0905-6947 Subjective Indoor Air Quality in Schools in Relation to Exposure GRETA SMEDJE1, DAN NORB~CK1 AND CHRISTER EDLING1 Abstract This paper presents data on indoor air quality in schools as perceived by those working in them and relates these data to exposure measurements. Data on subjective air quality, domestic exposures and health aspects were gathered by means of a ques- tionnaire which was sent to all personnel in 38 schools; it was completed by 1410 persons (85% of the total). Data on exposure were gathered by exposure measurements in classrooms. The re- sults indicate that 53% of the personnel perceived the indoor air quality as bad or very bad. It was perceived as worse by those who were younger, those who were dissatisfied with their psy- chosocial work climate and those who were not exposed to to- bacco smoke at home. In older school buildings and buildings with displacement ventilation there was less dissatisfaction with the air quality. There were no significant relations between com- plaints and air exchange rate or concentration of carbon dioxide. The air quality was perceived as worse at higher levels of ex- posure to a number of airborne compounds including volatile organic compounds, moulds, bacteria and respirable dust. It was concluded that exposure to indoor pollutants affects perception even at the low concentrations normally found indoors in nonin- dustrial buildings. Key words Subjective air quality; School personnel; Personal factors; Building characteristics; VOC; Microorganisms. Received 27 June 1995. Accepted for publication 12 November 1996. © Indoor Air (1997) Introduction Complaints concerning poor indoor air quality are common (Burge et al., 1987; Mendell, 1993; Skov et al., 1987; Sundell, 1994). Such complaints are often the starting-point for discussions about the medical rel- evance of indoor air quality, of investigations and measurements in the buildings, as well as of alter- ations. These measures often become both extensive and expensive. The validity of such complaints as an indicator of exposure is potentially of great interest. Most studies relating subjective air quality to exposure have dealt with one single exposure factor such as hu- midity, temperature or volatile organic compounds (VOC) (Andersen et al., 1974; Broder et al., 1993; Hud- nell et al., 1992; M61have et al., 1986; Reinikainen et al., 1992). Most of these studies were designed as experi- ments conducted over a certain period of time. In some epidemiological studies, mainly dealing with sick building symptoms, data on subjective air quality have been included. These studies typically deal with office workers (Zweers et al., 1992; Wallace et al., 1993). In recent years, there has been growing concern about the school environment in Sweden. The vast ma- jority of Swedish schoolchildren attend public elemen- tary and secondary schools from the age of 7 to 16 years. In schools the population density is high and poor ventilation, lack of maintenance and unsatisfac- tory cleaning are all thought to be common. Poor in- door air quality has been suggested as being related to the increase of allergic diseases that has occurred particularly among children and youths (NIPH, 1994) but few studies have been published on aspects of air quality in schools (Thorstensen et al., 1990; Gravesen et al., 1986; Munir et al., 1994). Norback has recently conducted a study dealing with different aspects of subjective air quality such as temperature, dry air and dust levels and has related these to exposures (Nor- back, 1995). The results show an association with the amount of fabrics in the classroom, the concentration of VOC and relative humidity, and the psychosocial work climate. We have undertaken a number of studies of asth- matic and sick building symptoms among school per- sonnel and pupils and have related these to the school environment and the indoor air quality. In this paper we pkesent the results concerning the subjective air quality in schools as perceived by the personnel in re- ~DepartmentofOccupationaland EnviromnentalMedicine, University Hospital, S-751 85Uppsala, Sweden. Fax +4618519978 THIS ARTIOLE IS FOR INDIVIDUAL USE ONLY AND ~AY NOT ~E FURTHER REPRODUCED OR STORED ELECTRONICALLY NITHOUT HRITTEN '" PERHISSION FROH THE COPYRIGHT HOLDER~ UHAUTHDRZZED R~PRGDUGFEO~ ~AY RESULTed.. IN FINANCIAL AND OTHER PENALTIES. ;

Smedje, Norb~ick and Edling lation to degree of exposure. The aims of the study were: to study the prevalence of complaints about the indoor air among school personnel, to relate these complaints to factors in the school environment, and to consider the importance of personal and psychosocial factors and domestic exposures. A number of hypotheses were tested. Subjective in- door air quality was assumed to be related to different characteristics of the school building, indoor environ- ment, pollutants in the classroom air, psychosocial fac- tors of the work environment, personal factors, and ex- posures in the home. Material and Methods The Study Population In the county of Uppsala in mid Sweden, there were in 1992 approximately 130 public schools from which we randomly selected 40. The headmasters were asked if they wished their school to participate and 38 schools agreed to do so. The schools varied in respect of factors such as age, construction and size. The small- est had less than 10 employees and 50 pupils, while the largest had almost 100 employees and more than 500 pupils. One third of the schools were situated in the city of Uppsala (117000 inhabitants), one in the town of Enk6ping (19 000 inhabitants) and the others in minor communities or in the countryside. All public employees working in the school buildings were in- vited to participate in the study, regardless of occur pation or number of hours/week at work. In a few schools, cleaning was performed by private contractors whose employees were not included in the study. Information from the Personnel Symptoms were recorded by means of a self-adminis- tered questionnaire mailed in January-February 1993 to the homes of 1652 employees. The questionnaire requested information on per- sonal factors such as age, smoking habits, present dis- eases and symptoms and included questions about domestic exposure to e.g. environmental tobacco smoke and damp in the home. There were also three questions on different aspects of the psychosocial cli- mate at work; general satisfaction, stress, and climate of cooperation. Each of these questions consisted of an analogue rating scale measuring from 0 (minimum) to 1 (maximum). These questions had previously been used by Norback et al. (1990). One question concerned the subjective air quality by asking "How do you perceive the quality of air inside the school?" The possible answers were "very good", "good", "bad" and "very bad". If the quality of air was felt to vary, the subjects were asked to make an assessment of the average quality. The question con- cerned the indoor air quality during the last three months. Assessment of Exposure Between March and June 1993 exposure measurements were performed in the schools. In each school we chose 2-5 classrooms so that the different buildings were rep- resented; a total of 96 classrooms was investigated. We inspected the buildings and noted details of their con- struction, building materials, equipment such as the type of ventilation system, room size, lighting levels, and the presence of open shelves and fabrics. Odours and signs of damp in the construct-ion were also noted. The cleaning staff were asked about their cleaning rou- tines. In each classroom we measured the temperature, relative humidity and rate of air exchange, and the levels of carbon dioxide (CO2), carbon monoxide (CO), nitrogen dioxide (NO2i, formaldehyde, other volatile organic compounds (VOC), respirable dust, moulds, bacteria, settled dust and mite allergen. Respirable dust and CO2 were recorded during 15 minutes by direct reading instruments, the Sibata P- 5H2 and Riken RI 411-A, respectively. The Sibata was calibrated at the factory (Sibata Scientific Technology Ltd.); the Riken was calibrated at the Department of Occupational and Environmental Medicine. Room temperature and air humidity .were recorded by an Assman psychrometer. Dust, CO2, temperature and humidity were measured twice in each classroom, at the end of a lesson. General and local air exchange rates were measured by a tracer gas decay method using acetone as the tracer gas (Anundi et al., 1992). Based on the air exchange rate and the room volume, the supply air rate was calculated. Formaldehyde was measured with glass fibre filters impregnated with 2.4-dinitro-phenylhydrazine (And- ersson et al., 1981) with a sampling rate of 0.2 L/min for 4 h. The filters were analysed by liquid chromato- graphy. VOC were sampled in parallel on beaded char- coal sorbent tubes (SKC Anasorb 747) and coconut charcoal with the same sampling rate and time as for formaldehyde. The charcoal tubes were desorbed with one ml of carbon disulphide, and analysed by gas chromatography and mass spectrometry. Fourteen common compounds were identified and quantified using an external standard technique and selective ion monitoring (SIM). Airborne microorganisms were sampled on 25 mm nucleopore filter with a pore size of 0.4 ~m and a sampling rate of 1.5 L/min for 4 h. The total concentration of airborne microorganisms 144

Subjective Indoor Air Quality in Schools in Relation to Exposure was determined by the CAMNEA method (Palmgren et al., 1986). Viable moulds and bacteria were deter- mined by incubation on two different media. The de- tection limit of viable organisms was 30 colony form- ing units (cfu) per m3 of air. Nitrogen dioxide was sampled with a passive sam- pling badge obtained from Toyo Roshi Kaisha, Ltd. (Yanagisawa and Nishimura, 1982) and placed in the classroom for 6-7 days after which it was analysed by liquid chromatography. However, following the recom- mendations of Lee et al. (1993), an overall mass trans- fer coefficient of 0.10 cm/s was used. Carbon monox- ide was measured by a passive colorimetric detector tube (Dr~ger 50/a-D) placed in the classroom for the same period as for NO2. Settled dust was collected from desks, chairs and the floor by a 400 W vacuum cleaner provided with a special dust collector from ALK Laboratories, Copen- hagen, containing a Millipore filter (pore size 6 ~m). After passing through a sieve containing a filter with a porosity of 300 gm, the amount of fine dust was de- termined by weighing the filters. The content of major mite allergens in the dust was determined by enzyme immunoassays and by the semi-quantative Acarex test (Bischoff et al., 1992). The measurements were made during normal activi- ties and under representative conditions. If the win- dows were normally kept open during lessons, they were also kept open when the measurements were made. When measuring the rate of air exchange, how- ever, the windows and doors were all closed. Temperature, relative humidity and levels of carbon dioxide, respirable dust and VOC were measured in the outdoor air, using the same methods as those ap- plied indoors. lationships, the relation to subjective air quality was analysed for each degree of room temperature and litre of supply air/person. Relations between different exposure variables were analysed by linear regression. In all the statistical analyses, two-tailed tests and a significance level of 5% were used. Results Questionnaire Data The questionnaire was mailed to 1652 subjects and I410 completed forms were returned, a response rate of 85%. The response rate was 87% for women and 80% for men, a difference which was statistically sigificant. Information about occupation was gathered by the questionnaire. For about two thirds of the study popu- lation, occupation was stated on address lists obtained from the schools, and it was therefore possible to esti- mate that the response rate was significantly higher among teachers than among those with other occu- pations. Those who had filled in the questionnaire but had not been working during the previous three months (49 subjects) were excluded, as were 58 sub- jects who had not answered the question about subjec- tive indoor air quality. The mean age of the subjects was 45 years, and the age range was from 16 to 64 years. Data on personal factors and domestic exposures are given in Table 1. On the analogue rating scale, general satisfaction with work was rated as 0.67, stress at work as 0.55 and climate of cooperation as 0.66. The subjective air quality was rated as bad, or very bad, by 53% of the subjects; the remainder considered it to be good, or very good. There were no significant Statistical Methods Analysis of relations between subjective air quality, questionnaire data and exposures were undertaken with multivariate statistical methods. Multiple logistic regression analysis was performed in several steps using the SPIDA statistical package (Gebski et al., 1992). Regression diagnostics available in the SPIDA package were used to test for collinearity. As a first step, all personal factors and domestic ex- posures were forced into the model. Secondly, non-sig- nificant factors were excluded. Thirdly, all school ex- posure variables were forced into the model one by one, keeping also the significant personal factors and domestic exposures in the statistical models. For total number of moulds and bacteria, logarithmic trans- formations of the raw data were used. In order to detect nonlinear exl~osure-response re- Table 1 Personal factors, domestic exposures and other character- istics of 1 303 school employees Gender, female 76% male 24% Occupation, teacher 54% other 46% Atopy 29%a Hay fever 16% Allergy to pets 9% Childhood eczema 14% Nickel allergy 22% Asthma 8% Smoker, present 19% former 29% Detached/semi detached domestic house 64% • Repainting indoors last year 24% Building dampness 15% Tobacco smoke at home 34% House pet 42% a Hay fever, pet allergy or childhood eczema. 145

Smedie, Norb~ick and Edling Tabel 2 Subiective air quality in schools as perceived by 1 303 school employees Judgement Women Men Total % % % Very bad 15 13 14 Bad38 39 38 Good 43 45 44 Very good 4 3 4 differences in the ratings of women and men (Table 2). Among those working in schools with natural venti- lation, 49% were dissatisfied, as were 61% of those working in schools with a mechanical exhaust air sys- tem only, 56% of those in mechanically ventilated schools with a mixing flow, and 48% of those in dis- placement ventilated schools. Building Characteristics The mean age of the school buildings was 33 years; the oldest was built around the year 1900, the newest in 1992. The majority of the buildings had 1-2 storeys (82%) and 38% had a basement. Most (63%) were mainly built of stone. Mechanical supply and exhaust air systems, without air-conditioning, were found in 61%, while 27% had natural ventilation. The mean air exchange rate in the classrooms was 5.5 L/s • person, with a range from 0.1 to 22.4 L/s • person. The lowest air exchange rate was in buildings with natural venti- lation only. All the classrooms had hard floor cover- ings, almost all of PVC or linoleum. Eighty-four per- cent had walls of painted plaster. In 19% of the class- rooms there were visible signs of damp or a mouldy odour. All the classrooms had fluorescent strip light- ing; the mean lighting level was 14.8 W/m2. None of the schools had kerosene heating or other sources of indoor combustion. In about 70% of the schools the floors were cleaned once a day with a moistened mop, while 30% were cleaned every second day. The desks were usually wiped once a week. In nearly half the schools there were no routines for washing the cur- rains, while in the others they were usually washed once a year. Exposure Measurements The mean room temperature was 23.5°C but in ap- proximately a quarter of the measurements the tem- perature was 25°C or higher. The concentration of CO2 was above 1 000 ppm in 41% of the measurements. In 55% of the classrooms the concentration of form'alde- hyde was below the detection limit of 5 ~tg/m3. The highest concentrations of VOCs were of (x-pinene, lim- onene and 5-carene, n-un~tecane and n-decane, and toluene and xylene. The highest concentrations of NO2 and CO were found in schools in the city of Uppsala. Smoking was not allowed in any school building apart from in special smoking rooms. The most common microorganisms were Cladosporium, Mycelia sterilia, Penicillium, Yeasts and Pseudomonas. Using monoclonal antibodies, allergens from house dust mites Der p I were found in one sample only and Der f I in none. The Acarex method showed the presence of mites in three schools (Table 3). Data on exposure outdoors by the school is given in Table 4. Subjective Air Quality in Relation to Personal Factors and Domestic Exposure Air quality was perceived as being worse by those who were young, those who had an atopic disposition (hay fever, an allergy to pets or childhood eczema) or an allergy to nickel, those who were teachers, or those who were dissatisfied with the psychosocial climate at work. Air quality was perceived as being better by those who were exposed to tobacco smoke at home or those who had pets (Table 5). There were no significant relations between subjective air quality and gender, Table 3 Exposures in 96 classrooms Exposure factor Arithmetic Min-Max mean Temperature °C) Relative humidity (%) Carbon dioxide (ppm) Formaldehyde (lag/m3) Volatile organic compounds, coconut charcoal (l~g/m3)a Volatile organic compounds, Anasorb 747 (gg/m3)a Nitrogen dioxide (pg/m3) Carbon monoxide (pg/m3) Respirable dust (gg/m3) Total bacteria (103/m3) Viable bacteria (103/m3) Total moulds (103/m3) Viable moulds (103/m3) Settled dust (rag/classroom) 23.5 19.5-27.5 38 16-75 990 425-2 800 6 <5-72 30 1-280 35 2-302 6 1-11 0.2 <0.1-0.9 19 6-60 52 8-290 0.9 0.1-18 40 7-360 0.5 0,1-4.5 172 26-370 Sum of 14 identified VOC. Table 4 Exposures outdoors by 38 schools Exposure factor Arithmetic Min-Max mean Temperature (°C) 12 2-21 Relative humidity (%) 62 39-92 Carbon dioxide (ppm) 425 375-525 Volatile organic compounds, 5 1-25 coconut charcoal (l~g/m3)a Volatile organic compounds, 6 1-24 Anasorb 747 (gg/m3)a Respirable dust (l~g/m3) 10 5-22 a Sum of 14 identified VOC. 146

Table 5 Personal factors, domestic exposures and psychosocial work climate sigificantly related to subjective air quality in schoolsa Factor Odds ratio CI (95%) Age 0.7b 0.6-0.8b Atopy 1.6 1.2-2.0 Nickel allergy 1.6 1.2-2-1 Environmental tobacco smoke at home 0.9 0.8-0.99 House pet 0.7 0.6-0.9 Stress at work 3.0 1.7-5.3 Cooperation at work 0.3 0.2-0.6 Being a teacher 2.1 1.6-2.6 Analysed by multiple logistic regression. Odds ratio expressed as change of coefficient per 10 years of age. Table 6 School exposures significantly related to subjective air quality in schoolsa Factor Odds ratio CI (95%) Age of building 0.9b 0.9-0.99b Exhaust ventilation 1.8 1.2-2.8 Displacement ventilation 0.7 0.5-0.9 Total moulds 1.9~ 1.3-2.8¢ Total bacteria 1.6¢ 1.1-2.3c Total VOC, coconut charcoal 1.8d 1.1-3.0d Total VOC, Anasorb 747 1.6d 1.1-2.4d Respirable dust 1.3e 1.1-1.5e Settled dust 1.5f 1.1-1.8f Analysed by multiple logistic regression and controlled for per- sonal and psychosocial factors and domestic exposure. Odds ratio expressed as change of coefficient per 10 years. Odds ratio expressed as change of coefficient per 10-fold in- crease of organisms. Odds ratio expressed as change of coefficient per 100 pg/m3. Odds ratio expressed as change of coefficient per 10 ~g/m3. Odds ratio expressed as change of coefficient per 100 mg. smoking habits, wearing of contact lenses, allergy to- wards furry animals, proneness to infection as a child, number of persons in the home, number of small children at home, age of the home, recent repainting, damp or wall-to-wall carpets at home. Subjective Air Quality in Schools in Relation to Characteristics of the School Buildings The air quality was perceived to be better in older buildings and in buildings with a mechanical air sup- ply and exhaust air system operating with the dis- placement principle. It was perceived as being worse in buildings with a mechanical exhaust system (with- out a mechanical air supply) (Table 6). No significant relations were found between subjective air quality and natural ventilation system, mechanical ventilation system with a mixing flow, air exchange rate, visible damp, lighting level, daylight factor, shelf-factor or fleece-factor. Subjective Indoor Air Quality in Schools in Relation to Exposure Subjective Air Quality in Schools in Relation to School Exposures The air quality in schools was perceived to be worse when the concentration of respirable dust, settled dust, total moulds, total bacteria or total VOC was higher (Table 6). When forcing these exposure factors into the statistical model at the same time, only total moulds and total VOC remained significantly related to subjective air quality. Thus, several of the ex- posure factors were correlated with each other. No significant relations were found between subjective air quality and CO2, room temperature, relative air humidity, CO, NO2, formaldehyde, viable moulds, or viable bacteria. Discussion Subjective air quality was significantl~..related to ex- posure to such factors as microorganisms, VOC and dust in the school environment, and to personal fac- tors, including age, at.opic disposition and perceived social climate at work. In all epidemiological studies, problems concerning the validity of the results will arise. One such problem concerns the representativeness of those who have completed the questionnaire compared with those who did not. In this study there was a response rate of 85%, which should be satisfactory, but there were some sig- nificant differences between those who participated and those who did not. Men, and those who were not teachers, were more frequently represented among the non-responders. Since there was no difference between men and women concerning the dependent variable, the difference in response rate between the sexes does not sigificantly affect the results of the study. Our re- sults showed that teachers were those most negative towards the indoor environment; therefore the lower response rate among those in other occupations may have resulted in the proportion of all employees who were dissatisfied with the air quality being slightly overestimated. The question used to characterize subjective air quality was more general than those used by certain other investigator's who asked about several different aspects such as "dry air", "stuffy air", "draught", and/or "hot/cold" (Nordstr6m et al., 1995; Norb~ick, 1995). It may be possible that by formulating the ques- tion as we did, the answers we received did not give us such detailed information, but our question corre- sponded more to people's perceptions since in problem b. uildings the complaints are often expressed as "poor ~ir", "dry air" or even "no air". Our question also cor- responded to the decipol unit for subjective air quality, 147

Smedje, Norb~ick and Edling which is based on a two-level (acceptable, not accept- able) comprehensive judgement (Fanger, 1988). It may be argued that the subjects" judgement of air quality is really an accumulation of air quality percep- tions over a period of time, and that exposure measure- ments, conducted mainly in the course of one day, are not representative of that period. However, several of the exposure values, including temperature, CO2, res- pirable dust and VOC, are comparable with those re- corded in an earlier study of schools (Norb~ick, 1990), indicating that the environment is relatively stable. More important, since the exposure measurements were performed in randomly chosen classrooms on randomly chosen days, the potential lack of stabilily of the exposures would most probably lead to a non- differential misclassification which would, of course, result in less significant relations between the different exposures and subjective indoor air quality. So the re- lationships found should not have been caused by such potential variations of the exposures. It could also be said that the results can be explained by the subjects drawing a conclusion concerning the quality of air on the basis of visible defects such as obvious damp. There was, however, no significant cor- relation between subjective air quality and visible damp; this explanation thus seems unlikely. The results show that subjects with an atopic dispo- sition or an allergy to nickel were more negative about the indoor air quality in schools than subjects without such allergies. Younger subjects were also more prone to describe the indoor air quality as poor, which may reflect that younger persons have a lower odour threshold (Cometto-Muftiz and Cain, 1991). In this study we were able to confirm that the psychosocial work climate affects the perception of the indoor en- vironment (Norb~ick, 1995; NordstrOm et al., 1995). Those who were exposed to tobacco smoke (ETS) at home were more satisfied with the quality of the air at school; the demands for air quality made by these sub- jects may thus be fairly low. Although there was no significant relation between subjective air quality and smoking, many of those exposed to ETS were smokers themselves (46%). Those who had pets also perceived the subjective air quality as better. It should be noted that this result cannot be attributed to recall bias, since we controlled for allergic disposition. It has been reported from office environments (Stenberg, 1994, Skov et al., 1989) that women are more sensitive than men to poor indoor air. This has been assumed to be partly related to the fact that women and men often have different jobs and different work- ing conditions but we found no significant relatiorr'be- tween subjective air quality and gender. This may re- flect that the working conditions of men and women in schools are relatively similar. In this stud~ age of the building, exhaust ventilation system, and mechanical ventilation with supply air by displacement, were significantly related to subjective air quality. Ventilation by displacement has become widely used in Scandinavia during the last few years. In this type of system the air is supplied through a low velocity diffuser located at floor level, and extracted at ceiling level (SkSret and Mathi~en, 1983). Our results indicate that air quality may vary not only between buildings with mechanical ventilation and buildings with only natural ventilation, but also between build- ings with different types of mechanical ventilation system. Significant relations were found between subjective air quality and exposures to settled and airborne dust, total number of moulds, total bacteria and total VOC. It would thus seem that, even at low concentrations, these pollutants affect the perception of air quality. We have not traced the sources of the microorgan- isms. Visible ~igns of damp or smell were observed in 19% of the classrooms but the common presence of Cla- dosporium implies that there could be outdoor sources as well. Volatile organic compounds were sampled onto two different media (coconut and synthetic charcoal), and significant relations were found with both methods. The concept of total VOC has been questioned, one rea- son being that it includes different compounds with different effects (M61have and Damgaard Nielsen, 1992). Others claim that at these low level concen- trations, different VOCs form together a "'pattern", and that it is this pattern that can be perceived by humans (Berglund et al., 1982; Baird et al., 1987). It is important to investigate in future whether subjective air quality is related to any particular compound or group of com- pounds. Both settled dust and respirable airborne dust were related to perception of poor air quality. We have con- sidered the amount of settled dust collected by stan- dardized vacuum cleaning to be a measure of the stan- dard of cleaning. It is obvious that the cleaning rou- tines in s~hools are less comprehensive than those in, e.g., offices. Since dust may contain allergens or other compounds (Gyntelberg et al., 1994; Munir, 1994), cleaning routines should include not only the floor, but also the desks, chairs and fabrics. Raw et al. (1993) have shown that cleaning of furniture and fabrics sig- nificantly reduced sick building symptoms in an office building. Complaints about poor indoor air quality were not related to the concentration of CO2 or the air exchange 148

rate. The significance of ventilation is still not known satisfactorily. It seems that at the same level of ex- posure to airborne pollutants, there are fewer com- plaints about poor air quality in naturally ventilated buildings compared to mechanically ventilated build- ings. Since mechanically ventilated buildings have higher air exchange rates, this indicates that the emis- sion of pollutants in these buildings differs consider- ably and is higher. This should be an important area for further research. Conclusions Our results show that subjective air quality in schools, as perceived by the occupants, is related to certain per- sonal, psychosocial and domestic factors. Controlling for these factors, it was found that subjective air qual- ity in schools is still significantly related to measured exposure levels of several airborne compounds. This indicates that complaints about poor indoor air quality are related to deficiences in the indoor environment. Acknowledgements This study was supported by grants from the Swedish Council for Work Life Research. References Andersen, I., Lundqvist, G.R., Jensen, P.L. and Proctor D.F. (1974) "Human response to 78-hour exposure to dry air", Archives of Environmental Health, 29, 319-324. Andersson, K., Hallgren, C., Levin, 5.0. and Nilsson, C.A. (1981) "Chemosorption sampling and analysis of formalde- hyde in air: influence on recovery during the simultaneous sampling of formaldehyde, phenol, furfural and furfuryl alchohol", Scandinavian fournal of Work, Environment and Health, 7, 282-289. Anundi, H., Norb~lck, D., Kinigalakis, G. and Johanson G. (1992) "Simplified measurement of air change rate using acetone as tracer gas". In: Proceedings of 41th Nordic Meeting on Work Environment, Reykjavik, Statens ArbetsmiljOstyrel- se, pp. 227-228 (in Swedish). Baird, J.C., Berglund, B., Berglund, U., Nicander-Bredberg, H. and Noma, E. (1987) "Distinguishing between healthy and sick preschools by chemical classification", Environment In- ternational, 13, 167-174. Berglund, B., Berglund, U., Lindvall, T. and Nicander-Bred- berg, H. (1982) "Olfactory and chemical characterization of indoor air. Towards a psychophysical model for air qual- ity", Environment h~ternationaI, 8, 327-332. Bischoff, E.R.C., Kniest, F.M. and Shirmacher, W. (1992) "Re- sults in dust samples with a new quantitative guanine as- sessment and evaluating of risk groups", Environmental Technology, 13, 377-382. Broder, I., Corey, P. and Pilger, C. (1993) "Influence of volatile organic compounds on the well-being of workers in office buildings". In: Volatile Organic Compounds in the Environ- ment, Indoor Air International, London: Lonsdale Press, pp. 595-604. Subjective Indoor Air Quality in Schools in Relation to Exposure Burge, S., Hedge, A., Wilson, S., Harris-Bass, J. and Robert- son, A.S. (1987) "Sick building syndrome: A study of 4373 office workers", Annals of Occupational Hygiene, 31, 493-504. Cometto-Mufiiz, J.E. and Cain, S.C. (1991) "Influence of air- borne contaminants on olfation and the common chemical sense", In: Getchell, T.V. et al. (eds), Smell and Taste in Health and Disease. New York: Raven Press, pp. 765-785. Fanger, EO. (1988) "Introdution of the olf and the decipol units to quantify air pollution perceived by humans in- doors and outdoors", Energy and Buildings, 12, 1-6. Gebski, V., Leung, O., McNeil, D. and Lunn D. (1992) SPIDA users manual, Version 6, Eastwood, Australia, Statistical Computing Laboratory. Gravesen, S., Larsen, L., Gyntelberg, F. and Skov, P. (1986) "Demonstration of microorganisms and dust in schools and offices", Allergy, 41, 520-525. Gyntelberg, F., Suadicani, P., Wohlfahrt Nielsen, J., Skov, P., ValbjOrn, O., Nielsen, P.A., Schneider, T., JOrgensen, O., Wol- koff, P., Wilkins, C.K., Gravesen, S. and Norn, S. (1994) "Dust and the sick building syndrome", Indoor Air, 4, 223-238. Hudnell, H.K., Otto, D.A., House, D.E. and MOlhave, L. (1992) "Exposure of humans to a volatile organic mixture. II. Sensory", Archives of Environmental Health, 47, 31-38. Lee, K., Yanagisawa, Y., Spengler, J.D., Ozkaynak, H. and Billick LH. (1993) "Sampling rate evaluation of NO2 Badge: (I) in indoor environments", Indoor Air, 3, 124-130. Mendell, M.J. (1993) "Non-specific ~ymptoms in office workers: A review and summary of "the epidemiologic literature", IndoorAir, 3, 227-236. Munir, A.K.M. (1994) Exposure to Indoor Allergens and Relation to Sensitization and Asthma in Children, Sweden, LinkOping University Medical Dissertations, No. 412. (Thesis). MOlhave, L., Bach, B. and Pedersen, O.F. (1986) "Human reac- tions to low concentrations of volatile organic com- pounds", Environment International, 12, 167-175. MOlhave, L. and Damgaard Nielsen, G. (1992) "Interpretation and limitations of the concept "total volatile organic com- pounds" (TVOC) as an indicator of human responses to exposures of volatile organic compounds (VOC) in indoor air", Indoor Air, 2, 65-77. National Institute of Public Health (NIPH) (1994) "The air that we breathe indoors", 1994:16 (in Swedish). Norb/ick, D. (1995) "Subjective indoor air quality in schools - the influence of high room temperature, carpeting, fleecy wall materials and volatile organic compounds (VOC)", In- &or Air, 5(4), 237-246. Norb~ick, D., TorgOn, M. and Edling, C. (1990) "Volatile or-~ ganic compounds, respirable dust, and personal factors re- lated to prevalence and incidence of sick building syn- drome in primary schools", British Journal of Industrial Medicine, 47, 733-741. NordstrOm, K., Norb~ick, D. and Akselsson, R. (1995) "Subjec- tive indoor air quality in hospitals - the influence of build- ing age, ventilation flow, and personal factors", Indoor En- vironment, 4, 37-44. Palmgren, U., StrOm, G., Blomqvist, G. and Malmberg, P. (1986) "Collection of airborne microorganisms on nucleo- pore filters, estimation and anlysis - CAMNEA method", Journal of Applied Bacteriology, 61, 401-406. Raw, G.J., Roys, M.S. and Whitehead, C. (1993), "Sick build- ing syndrome: Cleanliness is next to healthiness", Indoor Air, 3, 237-245. Reinikainen, L., Jaakkola, J.J.K. and Sepp~inen, O. (1992) "The effect of air humidification on symptoms and perception of indoor air quality in office workers: A six-period cross-over trial", Archives of Environmental Health, 47, 8-15. Skov, P., ValbjOrn, O. and DISG (1987) "The "sick" building syndrome in the office environment: The Danish town hall study", Environment hlternational, 13, 339-349. 149

Smedje, Norb~ck and Edling Skov, P., Valbjorn, O., Pedersen, B.V. and DISG (1989) "Influ- ence of personal characteristics, job-related factors and psy- chosocial factors on the sick building syndrome", Scandi- navian Journal of Work, Enviromnent and Health, 15, 286-295. SkAret, E. and Mathisen, H.M. (1983) "Ventilation efficiency - a guide to efficient ventilation", ASHRAE Transactions, 89, 2B, 490-495. Stenberg, B. (1994) Office Illness - The Worker, The Work and The Workplace, UmeA University Medical Dissertations, New Series No 399. (Thesis) Sundell, J. (1994) On the Association between Building Venti- lation Characteristics, some Indoor Environmental Exposures, some Allergic Manifestations and Subjective Symptom Reports, Indoor Air, Supplement 2. Thorstensen, E., Hansen, C., Pejtersen, J., Clausen, G.H. and Fanger, P.O. (1990) "Air pollution sources and indoor air quality in schools". In: Proceedings of Indoor Air "90, Ottawa, Canada Mortgage and Housing Corporation, Vol. 1, pp. 531-536. Wallace, L.A., Nelson, C.J., Highsmith, R. and Dunteman, G. (1993) "Association of personal and workplace character- istics with health, comfort and odor: A survey of 3948 of- rice workers in three buildings", Indoor Air, 3, 193-205. Yanagisawa, Y. and Nishimura, H. (1982) "A badge-type per- sonal sampler for measurements of personal exposure to NO2 and NO in ambient air", Environment International, 8, 235-242. Zweers, T., Preller, L., Brunekreef, B. and Boleij, J.S.M. (1992) "Health and indoor climate complaints of 7043 office workers in 61 buildings in the Netherlands", Indoor Air, 2, 127-136. 150