The sick building syndrome revisited - CiteSeerX

7 downloads 0 Views 3MB Size Report
standardised Danish questionnaire on symptoms and perceived indoor ... interested in studying whether awareness affects ... attention was drawn to asbestos and radon in buildings due ... World Health Organization (WHO) in 1979 to evaluate ...
Department of Occupational Medicine Copenhagen University Hospital Glostrup, Denmark

The sick building syndrome revisited

PhD thesis Charlotte Brauer, MD

University of Copenhagen 2005

2

PhD thesis

The sick building syndrome revisited By Charlotte Brauer, MD Department of Occupational Medicine Copenhagen University Hospital Glostrup, Denmark

Supervisors: Sigurd Mikkelsen, MD, DMSc Department of Occupational Medicine Copenhagen University Hospital DK-2600 Glostrup, Denmark Henrik Kolstad, MD, PhD Department of Occupational Medicine Aarhus University Hospital DK-8000 Aarhus C, Denmark Palle Ørbæk, MD, DMSc National Institute of Occupational Health DK-2100 Copenhagen Ø, Denmark Opponents: Finn Gyntelberg, Professor, MD, DMSc Clinis of Occupational and Environmental medicine Bispebjerg Hospital DK-2400 Copenhagen NV, Denmark Jens Peter Bonde, Professor, MD, DMSc Department of Occupational Medicine Aarhus University Hospital DK-8000 Aarhus C, Denmark Åke Thörn, MD, DMSc The Norrbotten County Council S-971 89 Luleå, Sweden

3

4

PREFACE

This PhD thesis is based on three manuscripts: •

The context of a study influences the reporting of symptoms. Int Arch Occup Environ Health 2003; 76: 621-624.



No consistent risk factor pattern for symptoms related to the sick building syndrome: a prospective population based study. Int Arch Occup Environ Health 2006.



The sick building syndrome - a chicken and egg situation? Int Arch Occup Environ Health 2006.

This thesis was done at the Department of Occupational Medicine, Copenhagen University Hospital in Glostrup during my employment as a staff specialist (2002-2005). First, I want to thank my three supervisors. I want to thank Sigurd Mikkelsen who made me interested in the subject and made it possible for me to carry out this study. He has taught me so much and given me an indescribable support throughout the study. Thank you very much for your encouragement and patience. I am also grateful to Henrik Kolstad and Palle Ørbæk for their thorough remarks and inspiring discussions. I also wish to thank Peder Skov for his collaboration in the validation study, his good discussions and his friendship. To the secretaries at our department, especially Gitte Grubbe Baunsgård, I would like to say a big thank you for your assistance in co-ordinating the questionnaire survey. And to my colleagues and fellow members of the Center for Indoor environment and Stress research, thank you for your interest, support and discussions. Last but not least, I would like to thank all the participants. Charlotte Brauer, Copenhagen, July 2005

The study has received financial support from the Danish Working Environment Authority, The Danish Working Environment Council and the Danish Medical Research Council.

5

CONTENTS

BACKGROUND __________________________________________________ 7 Introduction _________________________________________________________7 History of health aspects related to the indoor environment __________________8 Definition of the sick building syndrome __________________________________9 Previous research on SBS ______________________________________________9 Comparison of SBS with other conditions with non-specific symptoms _________9 Factors influencing the SBS symptoms __________________________________10

PRESENT STUDY AND AIMS _____________________________________ 11 VALIDATION STUDY ____________________________________________ 12 Study population ____________________________________________________12 The Glostrup Questionnaire ___________________________________________12 Statistical analyses ___________________________________________________12 Results _____________________________________________________________14 Discussion __________________________________________________________14

THE PROSPECTIVE STUDY ______________________________________ 16 Study population ____________________________________________________16 Questionnaire _______________________________________________________16 Statistical analyses ___________________________________________________18 Results _____________________________________________________________21 Discussion __________________________________________________________22

OVERALL CONCLUSIONS AND IMPLICATIONS____________________ 26 ENGLISH SUMMARY____________________________________________ 27 DANSK RESUMÉ________________________________________________ 29 REFERENCES __________________________________________________ 31 APPENDICES___________________________________________________ 39 Tables and manuscripts

6

BACKGROUND

me. The interviews began with an open-ended question about symptoms that the employee considered related to the indoor environment and were followed by systematic questions about the character, frequency and workrelatedness of the symptoms. In the open-ended questions almost 25% of the interviewed persons mentioned well-defined diseases caused by microorganisms or symptoms without a biologically plausible explanation in the indoor environment, for example herpes zoster, candida infection, a taste of metal, irregular menstrual cycles, dental problems, ear ache, sinusitis, pneumonia, and cystitis. When I compared interview information with the employees’ responses in questionnaires, I found only a moderate agreement concerning the responses as to whether the symptoms were experienced especially at work. I experienced that statements about this work-relatedness often seemed vague, when they were explored in detail. Instead of referring to the presence of symptoms when staying in the specific building, a common argument for perceiving the symptoms as work-related was that many of their colleagues had similar symptoms. Some of the employees described their symptoms in almost identical words or sentences. However, many of the symptoms that the employees perceived as similar turned out to be rather different when specified further. In two of the workplaces another argument for perceiving a symptom as work-related was that mould growth had been discovered in the workplace. Many of the employees told that they had not attributed the symptoms to the indoor environment until they were informed of the mould, but now they were certain that their symptoms were caused by the mould. Thus the in-depth interviews indicated that employees may ascribe their symptoms to the indoor environment in the workplace, if their colleagues report apparently similar symptoms or if they have become aware of a potential indoor environment problem. It seemed to include not only non-specific symptoms, but also clinically defined diseases that had other well-known causes. Hence the interplay between perceived health and indoor environment may be very complex. The experience from these interviews made me interested in studying whether awareness affects the reporting of symptoms and whether

Introduction The Department of Occupational Medicine, Copenhagen University Hospital, Glostrup, was established in 1989. The uptake area is the Copenhagen County with a population of approximately 600,000 persons. During the years quite a few of the patients referred to us by general practitioners had symptoms attributed the indoor environment. We were also regularly contacted by workplaces because of suspected indoor climate problems and symptoms among the employees. In many of these workplaces complaints about the indoor environment had persisted for a long time, even years. Often the occupational health services or the Labour Inspection Authority had been involved and frequently surveys about symptom prevalence already have been conducted. These surveys were conducted with a number of different questionnaires. The questionnaires varied considerably with respect to the symptoms queried about, the time frames of perceived symptoms and how to evaluate their workor building-relatedness. Furthermore, they had no well-established reference values with which to compare the survey results of complaints in a specific building. This resulted in different expert assessments of whether a specific building was “sick” or not and consequently in different decisions on the action to be taken to “cure” the building. On this background we established a working group in 1995 in order to develop a standardised Danish questionnaire on symptoms and perceived indoor environment. The group consisted of specialists in occupational medicine and public health medicine as well as representatives from the occupational health services and the Danish Working Environment Authority. I became involved in this process in 1996 when the questionnaire was to be pilot tested and validated. I conducted the validation study and a subsequent study that included collection of normative data [22]. The questionnaire is now called "the Glostrup Questionnaire". It can be downloaded in Danish on the website: www.cis.suite.dk. During the validation study, approximately 175 employees at four different workplaces were interviewed by two of my colleagues and

7

construction materials now tended to remain in buildings longer and at higher concentrations. A working group was assembled by the World Health Organization (WHO) in 1979 to evaluate health aspects related to indoor air quality [126]. Apart from the cancer risk related to asbestos and radon, the group described health effects in terms of irritation of the eye and respiratory tract due to release of formaldehyde from building material as well as eye irritation, coughing and possibly nausea due to environmental tobacco smoke. In 1982 a WHO working group once more discussed the impact of indoor air pollutants on health [127]. The group reported that an increasing number of case stories especially from the Scandinavian countries and the United States reported similar non-specific symptoms that they attributed to indoor climate problems. The group distinguished between temporarily sick buildings and permanently sick buildings. The former category comprised newly constructed or newly remodelled buildings, where the symptoms would decrease in time and mostly disappear. The latter category comprised buildings, where normally no obvious cause was evident and where the symptoms would persist for years and sometimes would be resistant to extensive remedial action. The symptoms described were: eye, nose and throat irritation; sensation of dry mucous membranes and skin; erythema; mental fatigue; headaches, high frequency of airway infections and cough; hoarseness, wheezing, itching and unspecific hypersensitivity; nausea and dizziness. The WHO group found it reasonable to assume that a true environmental health problem had emerged and introduced the term sick building syndrome. In 1994, a dramatic change in suspected adverse health effects of the indoor environment was seen when a cluster of the serious disease "acute pulmonary haemorrhage" was reported among infants in Cleveland, Ohio. Initially, water-damage with growth of a toxin-producing mould in the homes of the infants was suspected to have caused these outbreaks of disease. However, a subsequent thorough review of the case studies found inadequate evidence supporting a causal relationship between these cases of illness and the mould [3]. Since the Cleveland cluster of disease, mould in the indoor environment has received extensive media coverage and is often mentioned in sensational headlines. In Denmark, buildings have been evacuated and torn down due to

perceived exposures in the indoor environment is associated with symptoms with a biologically plausible explanation in the indoor environment as well as with symptoms without such an explanation. History of health aspects related to the indoor environment The health impact of the indoor environment has attracted much interest of hygienists, public health scientists and engineers for centuries. In the 19th century the importance of living in good sanitary conditions to prevent the spread of infectious diseases was well known, although the direct cause of the infections was unknown. The German hygienist Max von Pettenkofer, who was closely involved in investigating cholera epidemics, claimed that cleanliness, a nutritious diet and fresh air were important to improve public health. In the 1850s he proposed that the indoor air was unclean due to volatile organic substances from the occupants' breath and skin and suggested a limit value of carbon dioxide in the indoor air as a measure of an adequately ventilated room [86]. In 1881 Elias Heyman, the first professor of Hygiene in Sweden, wrote the book "The indoor air of our homes" in which he discussed health effects of indoor air in rooms that were insufficiently ventilated [51]. In his experience unclean air, draught and dry air could cause nausea, headache, rheumatism, pain in the nerves, common cold, catarrh or irritation in the lower respiratory system and a sensation of dryness in the throat. He recommended that a dwelling was kept dry, clean, well ventilated, and not too crowded. In the 20th century concern about allergic diseases related to indoor allergens has appeared. The role of fungi and mould growth in asthma and hay fever has been considered since the 1930s and the house dust mite was suspected as being a producer of house dust allergens in 1964 [68, 96]. Since then the risk of allergic sensitisation of house dust mite and mould spores in dwellings has been extensively studied [44, 48, 62]. In the 1970s much attention was drawn to asbestos and radon in buildings due to the cancer risk of the occupants. The energy crisis in 1973-1974 resulted in energy-conserving measures being taken with insulation of buildings and reduction in ventilation. This in turn led to concern about an increasing risk of exposure to indoor air pollutants, because emissions from the

8

79, 85, 103]. Some studies have examined the indoor environment in dwellings, but most studies have concentrated on non-industrial workplaces such as offices, schools, day-care centres and hospitals. In the 1980s focus was on ventilation, particulates, formaldehyde, temperature, relative humidity, number of persons occupying the room, man made mineral fibres, and wall-to-wall carpeting. In the 1990s attention was drawn to volatile organic compounds, visual display units, and damp and mouldy buildings. Cross-sectional studies examining the SBS have found numerous associations between reported symptoms and self-reported exposures in the indoor environment. Mucous membrane symptoms have been associated with perception of dry air, too little air movement, noise, static electricity, and dust [79, 80, 83]. General symptoms have been related to reported odours, humidity, temperature, too little air movement, static electricity, and noise [11, 79, 83]. Cross-sectional studies with objective measurements of exposures have often found statistically significant associations, but there is little consistency in the findings. Ventilation seems to be associated with perceived air quality and SBS symptoms [9, 57, 58, 113, 123]. Signs of dampness appear to increase the risk of pulmonary symptoms and SBS symptoms [18, 19, 61]. However, the evidence that indoor mould growth plays a role as a causative agent for health effects in terms of SBS symptoms is weak [18, 19, 47, 61, 91]. Studies about health effects due to airborne particles in non-industrial buildings are inconclusive [98]. The only longitudinal epidemiological study of which I am aware found only a few statistically significant associations between SBS symptoms and an extensive panel of objectively measured exposure variables [26], notably mould in seat dust, but not in air samples or floor dust. Mechanical ventilation, floor dust and other “established” indoor environmental factors had no significant association with SBS symptoms. Thus, despite extensive research no objectively measured factors in the indoor environment have consistently been associated with SBS symptoms.

mould and a few lawsuits have been brought to court because of chronic diseases ascribed to mould in the workplace. In the United States, mould litigation is frequently occurring, and insurance companies have paid millions of dollars in claims to occupants of mould damaged buildings [55]. Thus, despite a considerable improvement in the building and housing standard in the industrial countries during the past centuries, the indoor environment is still considered a potential health hazard and almost a more serious health hazard than a hundred years ago. Definition of the sick building syndrome The WHO group did not state a precise definition of the sick building syndrome (SBS) and even today there is no general consensus on a definition. There is no agreement on how the temporal relationship between the building and the symptoms is defined and on how often the occupants have to experience the symptoms. There is also disagreement about whether the term refers to a population or an individual. Some researchers define the SBS as a population in a specific "sick" building and require an excess of symptoms in the building [39, 85]. Others make a case definition of SBS on an individual basis, and some define a constellation of core symptoms as diagnostic of SBS [13, 38, 52, 71, 73, 74, 110, 115]. However, there seems to be agreement that the SBS typically includes mucous membrane symptoms, general symptoms such as fatigue and headache, and possibly skin symptoms. Some authors also include lower respiratory symptoms. In the following this group of symptoms will be referred to as SBS symptoms. Previous research on SBS Extensive research has tried to identify factors in the indoor environment, which could be the cause(s) of SBS symptoms. Several large scale studies have been performed, especially in the United States and Europe but also in Asia [1, 4, 9, 17, 20, 25, 26, 34, 36, 39, 52, 58, 65, 70, 72, 78-80, 83, 85, 87, 103, 109, 112, 120, 121, 128]. Several investigations have been performed in so-called "sick" buildings in which a high proportion of workers have experienced symptoms, which they attributed to the indoor environment already before the study took place [11, 33, 53, 69, 81]. However, a high prevalence of symptoms has also been found in buildings without pre-existing assumptions that these symptoms were building related [39, 77,

Comparison of SBS with other conditions with non-specific symptoms The symptoms associated with SBS are nonspecific symptoms that are quite common in the

9

Symptoms for which there is no biologically plausible explanation in the indoor environment has also been reported in buildings with reported problems in the indoor environment, for example bad nerves, constipation, low back pain or other musculoskeletal symptoms [11, 87, 88]. Thus the SBS has many similarities with the other environmental illnesses: The symptoms are related to multiple organs and no specific cause can be identified. Nevertheless, the symptoms are attributed to environmental factors that are tolerated by the majority of people.

general population and are seldom accompanied by pathological abnormalities. They can accompany other illnesses and can occur without connection to the environment [17, 37, 56, 63, 64]. Not only the indoor environment, but also other environmental exposures with low-level concentrations of chemicals or emissions have been associated with the occurrence of non-specific symptoms. Some of the conditions have been described as clinical entities or syndromes such as Gulf war syndrome, waste disposal site syndrome, multiple chemical sensitivity, and electrical hypersensitivity [27, 50, 102, 106, 124]. As research has failed to establish a link between the suspected environmental cause and the symptoms, a working group convened by WHO recommended in 1996 to use the term idiopathic environmental intolerance (IEI) instead of multiple chemical sensitivity [2]. Today this term usually also includes the other abovementioned conditions [50, 107]. In spite of the fact that the exposures are rather different, many of the symptoms are quite similar to the symptoms related to the SBS. Several theories have been formulated for the mechanisms behind IEI [60, 107]. According to the toxicogenic theory, chemical sensitivity to very low levels of almost any environmental agent accounts for the multi-system symptoms, while the psychogenic theory suggests that it is a functional somatic syndrome characterised by an overvalued idea attributing symptoms to environmental agents [108]. Provocation challenge studies of IEI have demonstrated that symptoms could only be reliably associated with chemical exposures at levels above the olfactory or irritant threshold, suggesting that perception and cognition mediate symptoms [29]. Likewise a provocation test of persons with alleged electrical hypersensitivity did not show relationships between symptoms and the actual presence of electromagnetic fields [6]. Many of the symptoms in the SBS can theoretically be explained by some chemical or physical exposures in the environment e.g. headache by noise, eye irritation by dust. Provocation challenge studies of SBS symptoms have shown that exposure to low concentrations of volatile organic compounds, office dust or office equipment is associated with reported mucous membrane symptoms and headache [12, 49, 76, 89, 122]. However, only indications of objective health effects have been found, and some of these indications were in biologically unexplainable directions [49].

Factors influencing the SBS symptoms Psychosocial work characteristics have been associated with the SBS symptoms in several studies [79, 80, 84, 104, 120, 128]. These psychosocial factors included work satisfaction in general, dissatisfaction with superiors or colleagues, quantity of work, a high level of stress, and climate of co-operation. Only a few authors have examined the impact of personality traits on SBS symptoms [11, 14, 94, 95]. A low sense of coherence was associated with a higher prevalence of SBS symptoms, but the personality trait negative affectivity was not. Female sex, low age and atopy have also been associated with an increased prevalence of symptoms [16, 80, 95, 104, 109]. Studies on psychosomatic symptoms have shown that persons with a general tendency to exaggerate or overrate their work stress may report more symptoms [41]. A tendency to report physical and emotional stress symptoms has been associated with chronic heart disease [45]. Among 264 patients presenting with supposed environmental illness, symptom reporting was associated with psychiatric and somatic conditions rather than an environmental exposure in the majority of the cases [21]. Toxic chemicals were regarded as the most probable cause in only five of these cases. The relationship between symptom reporting and SBS symptoms has not been investigated. Several studies have shown that environmental worry and awareness of a potential environmental hazard may influence the reporting of non-specific symptoms [30, 66, 75, 93, 106]. Assuming that SBS symptoms are related to the indoor environment in a building, the symptoms presumably either are present only when a person stays in that specific building or get worse in that building. Hence research on the SBS should attempt to get a

10

The main aims of this PhD thesis were to examine:

description of the temporal relationship between the occurrence of symptoms and staying in the building, for example the workplace. However, if worry and awareness influence the reporting of symptoms, is it then possible to get a valid description of this relationship? In only a few published studies, the validity of an SBS questionnaire is described [7, 24, 35, 73, 125]. Most of these studies have either compared questionnaire information with a physician's diagnosis or assessed the test-retest agreement. One study took account of "awareness” and found that the prevalence of work-related SBS symptoms increased when the employees were informed that a study of the indoor environment was going on [73]. This suggests that awareness and beliefs may influence self-reports on the buildingrelatedness of SBS symptoms.

1.

2.

3. 4.

PRESENT STUDY AND AIMS The thesis includes a 1-year prospective study of a random sample of the Danish adult population as well as a validation study. The results of the prospective study are reported in the three manuscripts, which are included in this thesis (appendix II to IV). Parts of the validation study are mentioned in one of the manuscripts. Initially, I will describe the validation study in further details. After this, the prospective study is described.

Whether symptoms are more likely to be reported as work-related, if attention is focused on the indoor environment at work (Appendix II). How the perceived indoor environment is associated with the prevalence, incidence and persistence of the non-specific symptoms that are traditionally included in the sick building syndrome (SBS symptoms) (Appendix III). How a general tendency to report symptoms may influence the reporting of SBS symptoms (Appendix III). Whether perceived exposures in the indoor environment is associated with symptoms with a biologically plausible explanation in the indoor environment as well as with symptoms without such an explanation (Appendix IV).

Validation of the questions about SBS symptoms and perceived indoor environment was another purpose of the study (Appendix III and thesis).

11

VALIDATION STUDY Prior to the prospective study, the questions from the Glostrup Questionnaire about symptoms related to the SBS and the indoor environment were validated.

headache, and difficulty in concentrating; range 0-3). Perceived exposures in the indoor environment (17 items) were assessed with the following question: "Have you been exposed to any of the following factors in your work environment during the past four weeks?" with the same four response options as for the questions on symptoms. Exposures were regarded as relevant if they were present several times a week or daily. Based on previous research the questions were grouped into a thermal index (draught, too low temperature, too high temperature, temperature variations and draught along the floor), a stuffy air index (stuffy air and unpleasant odour), a noise index (noise in the room, noise from other rooms and noise from outside), and a light index (illumination problems and reflective surfaces) [28]. The remaining 5 perceived exposures (dry air, static electricity, environmental tobacco smoke, cramped for space and poor cleaning) were not initially grouped, but were tested for dimensionality to assess whether these could be grouped in an index. In addition the questionnaire included question about age, sex and allergic diseases. Atopy was defined as reporting either allergy to pollen, furry animals or house dust mite or a history of asthma, hay fever or childhood eczema.

Study population The validation took place in eight workplaces that had contacted our department in the years 1998-2000 because of suspected indoor climate problems and symptoms among the employees. All the workplaces were located in Greater Copenhagen. In all, 947 employees participated. They had a mean age of 44 years (range 18-68) and were predominantly women (73%). The characteristics of the eight workplaces and employees are shown in table 1. The Glostrup Questionnaire The “Glostrup Questionnaire” comprised questions about the following 13 symptoms: eye irritation, nose irritation, nasal congestion, irritation of the throat, hoarseness, cough, wheezing in the chest, breathlessness, dry skin on the hands or arms, flushing face, fatigue, headache, and difficulty in concentrating. The symptoms were chosen as relevant for SBS on the basis of the existing literature. The participants were asked to consider symptoms experienced during the past four weeks, and each question had four response options: "no", "yes, sometimes", "yes, several times a week", and "yes, daily". Data were dichotomised so that a symptom occurring several times a week or daily was a positive answer. In addition, the participants had to specify for each symptom, if that particular symptom was more pronounced "on days off" or "on working days". Alternative response options were "no difference days off/working days" and "I don't know". Figure 1 shows an example of a question about SBS symptoms. If a symptom was more pronounced on working days, it was considered to be workrelated. We constructed four symptom indices by the sum of the dichotomised variables, irrespective of the symptoms being reported as work-related or not. The symptom indices were: mucous membrane index (eye irritation, nose irritation, nasal congestion, irritation of the throat, and hoarseness; range 0-5); pulmonary index (cough, wheezing in the chest, and breathlessness; range 0-3); skin index (dry skin on the hands or arms, and flushing face; range 0-2); and general symptom index (fatigue,

Statistical analyses The constructed indices of symptoms and perceived indoor environment were examined with respect to item bias, unidimensionality, and internal consistency. These analyses were carried out in the validation sample (N=947) and were cross-validated in the baseline study sample from the prospective study (N=2164), see below. The test-retest reliability was assessed in two of the eight workplaces from the validation sample. The tests for unidimensionality and item bias were conducted with the software program DIGRAM, which can be downloaded on the web site http://www.biostat.ku.dk/~skm/skm/. All other statistical analyses were done with SAS® System version 8.2. Item bias Item bias analyses test whether each item in an index functions in the same way, irrespective of

12

Figure 1. An example of a question about SBS symptoms in the questionnaire During the past 4 weeks have you been bothered by stuffy or runny nose? no....................................................................... □ yes, sometimes .................................................. □ yes, several times a week.................................. □ yes, daily............................................................ □ If yes, Are the symptoms most pronounced? on days off ......................................................... □ on working days................................................. □ no difference days off / working days ................ □ I don’t know........................................................ □

and general symptoms). Initially, we crosstabulated the symptom indices two by two checking the standardised residuals. Large residuals at the edges and small residuals at the diagonal suggested different dimensions, corresponding to more than expected scoring low on one index and high on the other index. Subsequently, we used the Martin-Löf test to test the hypothesis that the two indices were one dimension [119]. The Martin-Löf test is a confirmatory test that is suitable for dichotomous items in contrast to factor analysis. The same procedure was followed for the questions on perceived exposures in the indoor environment.

subgroup investigated. Because age, sex, and atopy have been shown to influence the reporting of SBS, we examined item bias in relation to these subgroups. Variables like sex and age are called exogenous variables, as they are not included in the items or index. We used a method based on three-way contingency tables of item score by the exogenous variables of interest stratified by the index score [15, 43]. Item bias analyses were done on the four symptom indices: mucous membrane index, pulmonary index, skin index, and general symptom index. Respondents who had one or more items in the index unanswered were excluded from the analysis. The partial gamma was calculated for each three-way table. The gamma coefficient ranges from -1 to +1, and a gamma outside the interval -0.31 to +0.31 can be interpreted as moderate to large item bias [15]. Two sided p-values0 on the symptom index. At follow-up, the outcomes of interest were the incidence and the persistence of mucous membrane symptoms and general symptoms. Incidence of symptoms was defined as a score>0 on the symptom index at follow-up among participants with a score=0 on the respective score index at baseline. Persistence of symptoms was defined as having a score>0 at follow-up as well as at baseline. Analyses were done in a three-stage process. Firstly, we examined the effect of each of the indoor environment indices, adjusting for sex, age, atopy and symptom reporting tendency. These potential confounders were kept in all subsequent models whether significant or not. Secondly, we in addition adjusted for other personal factors and factors in the psychosocial work environment. Potential personal or psychosocial confounders comprised marital status, smoking, negative affectivity, type A behaviour, a tendency to worry about health, self-efficacy, support from family and friends, job demands, job control, job strain, work support, and effort-reward imbalance. We used backward elimination to choose which of these

Statistical analyses Power Power calculations were done before the study. It was computed that if an outcome occurred with a prevalence of 5%, it would require a sample size of approximately 1160 persons to detect an odds ratio (OR) of 2 [alpha=0.05 and beta=0.90]. With an expected participation rate of 70% and assuming that 80% of these participants were in employment at baseline, the sample size of 4000 persons was considered sufficient.

18

are referred to as dummy symptoms in the following. None of the dummy symptoms were included in the WHO description of the SBS symptoms. The dummy symptoms were the same as the symptoms used to describe symptom reporting tendency (table 7), but now they were used as an outcome and not as a potential confounder. The variable for symptom reporting tendency was defined as at least four of the 15 symptoms that were reported to occur very often or often. However, in order to have a sufficient number of incident cases with dummy symptoms it was necessary to re-define a positive outcome. Responses of "never" or "seldom" were assigned a value of 0, while "sometimes", "often" and "very often" were assigned a value of 1, 2 and 3, respectively. The scores were summed to obtain an overall score for "dummy" symptoms ranging from 0 to 45. At baseline the prevalence of dummy symptoms was defined as having a score of at least 7, which corresponds to reporting 7 or more specific symptoms at least "sometimes". The incidence was defined as a score of at least 7 at follow-up among persons with a score less than 7 at baseline. This cut-off point corresponded to the 75th percentile and was chosen because it paralleled the dichotomization of the mucous membrane symptoms and general symptoms. It was fixed before any analyses were done regarding relationships between exposure and outcome. Baseline data were used to examine the cross-sectional association between the perceived indoor environment and the prevalence of mucous membrane symptoms, general symptoms, and dummy symptoms. Longitudinal data were used to examine the association between indoor environment at baseline and the incidence of SBS symptoms and dummy symptoms at 1-year follow-up as well as the reverse association: whether symptoms at baseline predicted that a person began to report new exposures in the indoor environment at follow-up (figure 3). New reports of exposures in the indoor environment was defined as having at least a score of one on the index at follow-up among persons who had a score of zero on the respective index at baseline. Analyses were done in the same three stage process as described above adjusting for sex, age, atopy, other personal factors, psychosocial work characteristics and indoor environment indices. The only difference was that analyses were performed without symptom reporting

covariates to keep in the model and chose 0.10 as the significance level of the Wald chi-square for keeping a variable in the model. Finally, we additionally adjusted each indoor environment index for the other indoor environment indices, testing whether the effect for individual environment factors could be explained by the other indoor environment factors. Again we used backward elimination at a 0.10 significance level to decide which of the other indoor environment indices to keep in the final model together with the potential confounders chosen at stage two. To test for a dose-response effect we used a test for trend analysing the models with the indoor environment indices as continuous variables in the models. An association was regarded as significant if the top level of the index had a 95% CI not including 1.0 or if the p-value was below 0.05 in the linear tests for trend. In analyses of follow-up data, we used information from the baseline questionnaire on the indoor environment and other risk factors as predictors of developing new symptoms or having persistent symptoms after one year. In order to keep the indoor environment as constant as possible, analyses at follow-up were restricted to the participants who were still employed in the same company and who still lived in the same dwelling as they did at baseline. The models were tested for goodness of fit with the Hosmer and Lemeshow method [54]. The influence of a symptom reporting tendency In the models described above, symptom reporting tendency was forced in all models as a potential confounder. In order to test whether this was an over-adjustment, we repeated the analyses without symptom reporting tendency in the models. Perceived indoor environment, SBS symptoms and “dummy” symptoms Logistic regression was used to examine the association between perceived indoor environment at work and SBS symptoms as well as other symptoms that cannot plausibly be linked to the indoor environment. The SBS symptoms were grouped into mucous membrane symptoms and general symptoms, and the outcomes of interest were the prevalence and incidence of these symptom groups as described above. The other symptoms that were considered not to be SBS symptoms

19

Baseline

Follow-up

SBS symptoms: mucous membrane symptoms and general symptoms

SBS symptoms: mucous membrane symptoms and general symptoms

  ‘ Perceived indoor environment

Perceived indoor environment

‘   Other symptoms: "dummy" symptoms

Other symptoms: "dummy" symptoms

Figure 3. Models on the predictive relationships between the perceived indoor environment and symptoms. "1" indicates cross-sectional analysis at baseline, "2" indicates longitudinal analysis with normal causal direction between exposure and outcome, and "3" indicates the reverse causation in longitudinal analysis.

20

Perceived indoor environment and prevalence, incidence and persistence of SBS symptoms (appendix III) Mucous membrane symptoms and general symptoms were common, as 28% of the population at baseline reported to have mucous membrane symptoms regularly and 23% reported to have general symptoms. The most common symptoms were nasal congestion (16%) followed by fatigue (15%), headache (12%), nose irritation (11%) and eye irritation (10%). At follow-up the incidence of mucous membrane symptoms and general symptoms was 15% and 9%, respectively, and approximately 50% recovered from symptoms during the one-year follow-up period. No difference was found between manual and nonmanual workers in prevalence or one-year incidence of mucous membrane symptoms or general symptoms. In the indoor environment at baseline the most common reported exposure was noise (46%) followed by reports about cramped space or dust (38%) and perception of dry air (30%). The majority of participants remained at the same level of indoor environment complaints (63% to 94%), and only few changed more than one score on an index (1% to 7%). Table 6 shows the changes in perceived indoor environment between baseline and follow-up.

tendency as a covariate in the models. In the analyses of the reverse association, we adjusted only for sex, age, and atopy because the small number of cases did not permit further adjustments in the models. To test for trend we also analysed the models with the indoor environment indices as continuous variables in the models. An association was regarded as significant if the top level of the index had a 95% CI not including 1.0 or if the p-value was below 0.05 in the linear tests for trend. Analyses at follow-up were restricted to the 1402 participants who worked and lived in the same place as they did at baseline. All data processing were done with the SAS® System version 8.2 [97]. Results Focus on the indoor environment at work and work-related symptoms (appendix II) A slight difference in the accompanying letter replacing the word "work" with "home" in two sentences influenced the responses considerably. The prevalence estimates of symptoms were similar in group W and group H, but differences were found as regards the reporting of work-relatedness and homerelatedness. If the information letter focused slightly more on the indoor environment at the workplace than in the home, the participants were more likely to report that their symptoms were work-related (significant unadjusted odds ratios (OR) between 1.8 and 5.5). Adjusting the results for several potential confounders mainly led to higher estimates of OR. Likewise, the participants were more likely to report that their symptoms were more pronounced at home, if focus was on the environment at home (significant unadjusted OR between 5.7 and 20.6). As regards the psychosocial work characteristics, persons who reported to have high job demands and low support at work also tended to report work-related general symptoms (significant adjusted OR ranging between 1.9 and 4.0). Work-related mucous membrane symptoms were associated only with high job demands (significant adjusted OR between 1.9 and 2.6).

We found no consistent risk factor pattern in the cross-sectional and the longitudinal analyses for the associations between perceived indoor environment factors at work and symptoms. Whereas mucous membrane symptoms in the cross-sectional analysis were significantly associated with self-reported high temperature and dry air, the prospective analyses showed that onset of mucous membrane symptoms was associated with the sensation of draught, dry air, and noise. Persistent mucous membrane symptoms were associated only with perception of stuffy air. General symptoms were associated with self-reported stuffy air and dry air in the cross-sectional analysis, while draught was the only predictor of onset of general symptoms. We found no predictors in the indoor environment for the persistence of general symptoms. The influence of a symptom reporting tendency (appendix III and thesis) Table 7 shows the symptom check list used to record a symptom reporting tendency, which was defined as having at least four of the 15

21

to temperature conditions, stuffy air, noise, and light conditions. Persons with general symptoms began to report exposures to draught, temperature conditions, stuffy air, noise, and light conditions while persons with dummy symptoms reported draught, temperature conditions, and stuffy air.

symptoms that cannot plausibly be linked to the indoor environment. Symptom reporting tendency was associated with the prevalence of mucous membrane symptoms as well as general symptoms with odds ratios of approximately 2.3 and 7.2, respectively (table 8). In addition, symptom reporting tendency was similarly associated with the incidence and persistence of mucous membrane symptoms. Elevated odds ratios were also found regarding the incidence and persistence of general symptoms, but the associations were of borderline significance (table 8). When the analyses were repeated without symptom reporting tendency as a covariate in the models, the estimates of the effects of the indoor environment factors changed only slightly (table 9 and table 10).

Discussion In my study there was no risk factor in the perceived indoor environment that was consistently and plausibly associated with the prevalence, incidence and persistence of mucous membrane or general symptoms. Some of the perceived indoor environment factors predicted not only mucous membrane or general symptoms, but also symptoms without a biologically plausible explanation in the indoor environment (dummy symptoms). However, SBS symptoms and dummy symptoms also predicted that the participants began to report exposures in the indoor environment at followup. Even a small increase in focus on the workplace influenced the responses as to whether a symptom was reported as workrelated or not. A general tendency to report symptoms was an independent risk factor of mucous membrane symptoms and general symptoms, but not a confounder in this study.

Perceived indoor environment, SBS symptoms and dummy symptoms (appendix IV and thesis) In the cross-sectional analysis, the perceived indoor environment was associated with symptoms that are traditionally included in the SBS as well as with dummy symptoms. Mucous membrane symptoms were associated with dry air, dust, temperature and stuffy air (table 9, model without symptom reporting tendency). General symptoms were associated with stuffy air and light (table 10, model without symptom reporting tendency). Dummy symptoms were associated with stuffy air index, dry air index, noise index and patches of dampness (appendix IV). In the longitudinal analysis only few of the perceived indoor environment factors predicted the development of any of the symptom groups. Draught, dry air and noise were associated with incident mucous membrane symptoms, and draught was associated with incident general symptoms (table 9, table 10 and appendix IV). Incident dummy symptoms were associated with dry air. Although many were not statistically significant, the patterns of association for the dummy symptoms resembled those of the mucous membrane symptoms and general symptoms. In the reverse longitudinal analysis, we found that persons with symptoms at baseline were more likely to report exposures in the indoor environment than persons without symptoms were. Both the SBS symptoms and the dummy symptoms predicted several new reported exposures in the indoor environment (appendix IV). Persons with mucous membrane symptoms at baseline began to report exposures

Comparison with other findings Except for a few intervention studies and one longitudinal epidemiological study, previous studies on the SBS are cross-sectional. In these cross-sectional studies several self-reported exposures, but no objective measurements have consistently been associated with mucous membrane symptoms and general symptoms. The only findings from previous studies that could be reproduced in the present study were that mucous membrane symptoms were associated with the perception of dry air and noise, and that general symptoms were associated with perceived stuffy air. There is, however, no obvious biological pathway between noise and mucous membrane symptoms. In a recent longitudinal epidemiological study mould in seat dust, selfreported lack of office cleanliness and selfreported number of people working in the office were associated with SBS symptoms [26]. Perceived exposure to dust or dirt was not a risk factor in my study. The participants were asked a question about the number of people working

22

a temporal relationship measuring the exposure before the outcome, on the assumption that the exposure was relatively stable during the oneyear follow-up period. Longitudinal studies are more informative in etiological analyses. However, cross-sectional studies may support a causal relationship, if the studied symptoms or disease develop shortly after the exposure. SBS symptoms are believed to be acute reactions to factors in the indoor environment rather than chronic symptoms. Hence cross-sectional studies could be suitable to assess acute reactions to the indoor environment, if it was possible to measure the outcome objectively and thus reduce reporting bias. I chose a longitudinal design because of the possibility to assess whether the reported exposures preceded the symptom. A limitation is the interval of one year between baseline and follow-up. This choice was made to control for seasonal variations. I have no information about changes in the indoor environment during the one-year followup period, but I excluded participants who had changed job in order to keep exposure conditions as constant as possible. However, fluctuations in the exposure are likely to occur. Thus, misclassification of the exposure may have occurred, which would reduce the associations towards the null hypothesis of no association. More follow-up rounds with a closer interval than a year would have made it possible to assess more precisely how the symptoms emerged and disappeared in relation to exposure. This approach has been used in the only other longitudinal epidemiological study [26]. In that study, environmental measurements and questionnaire survey were done every six weeks during a year and included 98 participants. My power calculations before the study indicated that approximately 1160 participants were needed in a study with one follow-up round. It is difficult to assess, however, if a design with more repeated measurements would have been more costeffective. In addition, repeated measurements may influence the reporting of symptoms. For example, if the participants are more enthusiastic in the beginning of a study, a decrease in symptom prevalence will be observed during the follow-up period, which may affect the observed associations [26]. A limitation in the study is that I have only self-reports on the exposure. However, it was of course neither practically nor financially possible to make objective measurements in

in the office or workroom, but the validity of this question was doubtful. Hence I did not include this variable in the models. This variable, however, was not statistically significant in univariate analyses. A few other studies on SBS symptoms have focused on the association between exposures in the indoor environment and other symptoms, which cannot plausibly be related to the indoor environment. Wallace et al. found that pain in the back, hands or shoulder/neck was associated with glare, noise and odour of cosmetics [120]. Platt et al. found an increased frequency of backache and constipation among adults living in houses with visible damp or mould [88]. Pirhonen et al. found that self-reported dampness or mould problems in homes were associated not only with many different SBS symptoms, but also with backache and recurrent stomach ache [87]. These biologically unlikely results seem to be in good agreement with the results of my study. There exists no other study on the SBS that has examined the reverse association assessing how the presence of symptoms influences future reporting of exposure. However, in occupational stress research the reverse causation hypothesis has been used examining the impact of mental health on future reporting of work characteristics [31, 32, 41, 118]. De Lange et al found that health affected perceived work characteristics both positively and negatively, and the study by van Hoof et al. suggested that health and work characteristics may act reciprocally, similar to the findings in my study. One other study on SBS has also taken account of "awareness" by informing only some of the participants that a study about the indoor environment was going on [73]. They found that the prevalence of work-related SBS symptoms was three times higher among workers aware of the study relative to those blinded. Another study examined an outbreak of SBS in a school with mould growth and also found that the perception of symptoms increased after the problem was publicised [46]. This is in good accordance with my results. Thus altogether, studies that have examined the same type of problems seem to support my results. Strenghts and limitations of the study The strengths of the present study are the large population and the prospective design. The prospective design made it possible to evaluate

23

more than two thousand workplaces and homes. Although self-reports seem to reflect objectively measurable factors in the indoor environment to some degree [1, 23, 82, 90, 92, 114, 117], a risk of reporting bias exists resulting in bias towards more significant positive associations with symptoms. However, in the analyses of follow-up data the exposure and outcome is measured on different points in time. Assuming that the participants are unable to remember what they answered a year before, this will reduce but not eliminate reporting bias.

I had expected that the same factor in the indoor environment would induce symptoms as well as maintain symptoms, yielding a consistent risk factor pattern in the prevalence, incidence and persistence of SBS symptoms. However, as SBS symptoms are supposed to be acute recurrent symptoms, the symptoms may disappear and reappear during the follow-up period, which may reduce the estimates of incidence and persistence. Nevertheless, a consistent risk factor pattern would have supported a causal relationship, assuming that persons who are sensitive to develop SBS symptoms still exists in the population, that persons do not become immune to SBS symptoms once they have got it, and that the exposure still persists at follow-up. In the present study, the most consistent finding was that dry air was associated with the prevalence and incidence of mucous membrane symptoms. The prevalence could be explained by problems in distinguishing the sensation of dry air with irritation of the eye, but not the incidence. If there is a true biological relation between indoor environmental factors and SBS symptoms, one would expect these variables to be stronger associated with SBS symptoms than with other symptoms for which there is no plausible explanation. Thus, the lack of specificity of associations found in my study may speak against a true causal relation. Furthermore, a number of associations between the reported indoor exposures and SBS symptoms were also biologically implausible. The association between mucous membrane symptoms and noise, for example, was not easily understood. Pirhonen et al. found an increased prevalence of backache and stomach ache in damp buildings [87]. However, when they analysed data on a sample consisting solely of persons not complaining of backache or stomach ache, the strength of the associations between the indoor environment and SBS symptoms were significantly reduced. This indicates that some associations between SBS symptoms and perceived indoor environment factors could be due to over-reporting. In my study symptom reporting tendency was strongly related with both mucous membrane symptoms and general symptoms in the cross-sectional analysis. This may reflect persons with disorders in several organs, but may also be due to over-reporting. A possible explanation for the reverse association, showing that persons with symptoms were likely to begin to report

As the cohort is a random sample of the general population it is possible that the contrasts in exposure are insufficient to detect true associations, because only few presumably work in buildings with severe indoor climate problems. The prevalence of perceived exposures in the present study population corresponded to the prevalences I have found in 41 non-problem buildings [22]. The distribution of reported exposures in these 41 non-problem buildings showed a large or complete overlap with the similar distribution in the 8 problem buildings in my validation study [unpublished data]. Thus, it may be assumed that also the participants from the general population are exposed to a wide variation in physical or chemical factors in the indoor environment. In fact, a large variation in perceived indoor environment factors was found in the present study. The possibility of selection bias must be considered. The participants who were lost to follow-up did not differ from the cohort in baseline characteristics on perceived indoor environment and symptoms. Thus I consider the drop out to be less important. A response rate of 68% at baseline and 80% at follow-up can be considered satisfactory for a population study, if there is no selection bias. Possible explanations for the findings In the longitudinal analyses I chose to use the symptoms as an outcome without considering whether the participants reported them as more pronounced at work or not. This choice was made because the baseline study had shown that information about work-relatedness may be seriously biased. Nevertheless, this might lead to underestimation of possible effects, because the work-related symptoms are diluted with symptoms that are not temporarily related to the workplace.

24

respect to work-relatedness of symptoms is likely to occur when surveys are conducted in buildings with suspected indoor environment problems, because attention has already been drawn to the problems, and a survey will only further increase the focus on the problems. In particular the risk of bias exists when the exposure and outcome are measured simultaneously, and especially if information on exposures as well as outcomes relies exclusively on the participants' subjective evaluation. Thus, a study conducted in a problem building may not be suitable for detecting causal effects of indoor environment factors, even if a non-problem building is included as a control, which has been the design of some other studies [11, 53]. Studies examining problem buildings may, however, be useful to generate important hypotheses and provide information from extreme exposure situations that are rarely encountered in systematic studies. In addition, cases initiated because of concern about health effects among the occupants of a building may also be useful for qualitative studies, including personal and contextual factors in the models as mediators or effect modifiers in order to explain the causal pathway from environmental exposure to outcome.

exposures in the indoor environment, is that symptomatic persons may be more susceptible in that specific organ to irritants or other problems in the indoor environment [67]. However several of the findings in my study are not easily understood, if such a mechanism is the explanation of “reversed causation”. It is, for example, difficult to explain why persons with symptoms such as stomach ache or heart palpitation should begin to complain more about draught or temperature conditions. It seems reasonable that persons with irritated eyes could be more bothered by glare and reflections, but it is not conceivable that persons with irritation in the nose or throat should complain about the light. Repeating the analyses with eye irritation and nose/throat irritation separately did not alter the associations [data not shown]. Thus it does not seem likely that individual susceptibility solely explains the reverse associations. An alternative explanation is that people who have many health complaints look for potential causes for example stress at work or problems in the indoor environment. The result is an increased awareness on external factors, which may produce a spurious association between e.g. the indoor environment and health. I chose to study SBS symptoms in a random sample of the general population instead of studying occupants of a “sick building”. Based on my experience from interviewing employees in problem buildings, I presumed that selfreports could be biased in cases initiated by concern about health effects among the occupants of a building. Subsequently my study has shown that even little awareness may introduce bias in self-reported health. Other researchers have proposed that the sick building syndrome could be due to or modified by stressrelated factors at work or a group behaviour as a "contagious" condition [105, 116]. In my study the participants completed the questionnaire at home out of context with the workplace, so no group behaviour ought to be present. In the term "sick building syndrome" the cause of the syndrome is incorporated in the definition of the syndrome. This easily leads to arguing in a circle: a "sick" building may be defined as a building where many occupants have similar symptoms and, conversely, the SBS are symptoms occurring in a sick building. Does this mean that an excess of symptoms in a building is required to call it SBS, and that SBS symptoms can be investigated only in a sick building? My results suggest that bias with

25

bias towards a tendency to over-report. According to my study symptoms such as stomach ache, heart palpitations and muscle tension were also associated with perceived factors in the indoor environment. This suggests that these symptoms just as well as mucous membrane symptoms and general symptoms could have been included in the SBS, if the occupants of a building were of the conception that these symptoms were related to the indoor climate. This does not exclude that there may be associations between specific indoor environment factors and health. However, it suggests that the SBS symptoms do not constitute a well-defined syndrome.

OVERALL CONCLUSIONS AND IMPLICATIONS In my study three important criteria of causal associations between the indoor environment and symptoms could not be met: temporality, biological plausibility, and consistency. The results suggest that many people experience mucous membrane symptoms and general symptoms regularly. As the reporting of workrelatedness is severely biased, many symptoms may be misclassified as "building related symptoms" and the term should be avoided. The validation study showed that the symptoms included in the SBS represent different dimensions and thus should not be grouped to form an SBS index or an SBS-diagnosis. The reverse associations found in my study indicate

26

ENGLISH SUMMARY This PhD thesis was done during my employment at the Department of Occupational Medicine at the Copenhagen University Hospital in Glostrup.

focused slightly more on the indoor environment at home. In two sentences of the letter the word "work" was replaced with "home". I found that persons who had been focused on the workplace were more likely to report their symptoms as more pronounced at work. In addition, persons who received the other letter with focus on the home were more likely to report their symptoms as more pronounced at home. The findings suggested that reports on work-relatedness may be severely biased. Some statistically significant associations were found between reported exposures in the indoor environment and the prevalence of mucous membrane symptoms and general symptom. However, not all associations were biologically plausible, and no consistent pattern was found between the associations found as regards the prevalence, incidence and persistence of these symptoms. I also examined the association between reported exposures in the indoor environment and a group of other symptoms that were biologically implausible to associate with the indoor environment (dummy symptoms). Several of the indoor environment factors were associated also with the dummy symptoms, thus indicating reporting bias. Furthermore I investigated the reverse order relationship: that symptoms may predict future reports of exposures in the indoor environment. I found that persons having mucous membrane symptoms, general symptoms or dummy symptoms were more likely to begin to report exposures in the indoor environment than persons without symptoms. Finally, a tendency to report many symptoms in general was related with mucous membrane symptoms and general symptoms at baseline. The findings suggest a bias towards a tendency to over-report. The validation study took place prior to the follow-up study in 8 workplaces with suspected indoor environment problems (N=947). It was cross-validated in the baseline sample from the prospective study. It showed a high degree of reliability of the questionnaire. However, item bias was found related to sex, age and atopy as regards the construction of two symptom indices, a skin index and a pulmonary index. Hence these indices were not used in the subsequent analyses of the follow-up study. Testing for unidimensionality showed that mucous membrane symptoms, skin symptoms,

During the last 20 years several cross-sectional studies have examined relationships between indoor environment factors and symptoms such as irritation of the mucous membranes, irritation of the skin, fatigue and headache. If the symptoms are attributed to the indoor environment, this constellation of symptoms has been described as the sick building syndrome (SBS). However, the cause(s) of the SBS symptoms are unknown and little is known about the temporal relationship between exposures in the indoor environment and the development of symptoms. In addition, it has been suggested that awareness of a potential environmental hazard and a personal tendency to overrate exposures or symptoms can influence how people report SBS symptoms and thus affect the association between the perceived indoor environment and SBS symptoms. The main aims of the present study were to examine the temporal relationship between reported exposures and SBS symptoms and to examine whether awareness of the indoor environment and a tendency to overrate may affect the association between the perceived indoor environment and SBS symptoms. Another aim was to validate the questionnaire used for the study. The study was designed as a follow-up study with two questionnaire surveys conducted with an interval of one year in a random sample of the general population aged 18-59 years. The study population consisted of 2164 participants at baseline, of whom 1402 were also eligible for inclusion at the one-year follow-up. The outcomes of interest were the prevalence, incidence and persistence of mucous membrane symptoms (an index consisting of eye irritation, nose irritation, nasal congestion, throat irritation and hoarseness) and general symptoms (an index consisting of fatigue, headache and difficulty in concentrating). At baseline a slight difference in the introduction letter was made so that some participants received a letter with a bit more attention on the indoor environment at work, and other participants received at letter that

27

pulmonary symptoms and general symptoms belong to different dimensions and thus should not be combined in a common index. In conclusion, the results indicates a rather confusing web of pathways between the symptoms and perceived exposures, where it is difficult to determine what existed first: the outcome or the exposure. In addition, the reported indoor environment exposures were associated with symptoms that cannot plausibly be explained by the indoor environment. The findings do not exclude that there may be associations between specific indoor environment factors and health. However, it suggests that there is a risk of reporting bias when assessing non-specific symptoms and that the SBS symptoms do not constitute a welldefined syndrome.

28

ene gruppe blev der lagt lidt mere vægt på indeklimaet på arbejdspladsen end i hjemmet. I den anden gruppe var det omvendt. Det drejede sig om at ordene "på arbejde", der blev skiftet ud med "i hjemmet" i to sætninger i brevet. Det viste sig, at personer, som havde fået brevet med mere vægt på arbejdspladsen, var mere tilbøjelige til at angive, at deres symptomer var værst, når de var på arbejde. Desuden var de personer, der havde fået det andet brev, mere tilbøjelige til at angive at symptomerne var værst i hjemmet. Disse resultater tyder på, at der kan være store fejlkilder forbundet med de svar man får på spørgsmål, der vedrører om et symptom er værst på arbejdspladsen. I den første spørgeskemaundersøgelse blev der fundet flere statistisk signifikante sammenhænge mellem selv-rapporterede faktorer i indeklimaet og slimhindesymptomer samt almensymptomer. Der var imidlertid ikke noget konsistent mønster mellem disse fund og de longitudinelle analyser med hensyn til hvilke faktorer i indeklimaet, der var associeret med forekomsten af indeklimasymptomer eller udvikling eller vedligeholdelse af symptomer. Den vigtigste risikofaktor for at udvikle nye indeklimasymptomer var, at man havde mange andre symptomer, som ikke er biologisk plausible at relatere til indeklimaet. Der fandtes endvidere også flere statistisk signifikante sammenhænge mellem disse uplausible symptomer og rapporterede påvirkninger i indeklimaet. Dette kunne tyde på rapporterings bias. Der syntes at være en vekselvirkning mellem klager over indeklimaet og symptomer, idet personer som havde symptomer ved undersøgelsens begyndelse, men som på dette tidspunkt ikke klagede over indeklimaet, begyndte at opleve problemer i indeklimaet i løbet af den 1-årige opfølgningsperiode. Undersøgelsen af validiteten blev udført på 8 arbejdspladser, hvor man mistænkte at der var indeklimaproblemer (N=947) samt i studiepopulationen fra den første spørgeskemaundersøgelse. Der blev fundet en høj grad af reproducerbarhed. To indexer omfattende hhv. hudsymptomer og nedre luftvejssymptomer viste sig at være forbundet med item-bias i forhold til køn, alder og atopi. Derfor blev disse to indexer ikke anvendt i analyserne af opfølgningsundersøgelsen. Analyser af dimensionalitet viste, at slimhindeindexet, hud-indexet, luftvejs-indexet og almensymptom-indexet tilhørte forskellige

DANSK RESUMÉ Ph.d. afhandlingen er udarbejdet under min ansættelse på Arbejdsmedicinsk Klinik, Amtssygehuset i Glostrup. I de seneste 20 år har adskillige tværsnitsstudier undersøgt mulige sammenhænge mellem forhold i indeklimaet og symptomer som irritation af slimhinder og hud, træthed og hovedpine. Hvis man mener, at symptomerne skyldes indeklimaet, kalder man ofte denne sammensætning af symptomer for "indeklimasyndrom". Årsagen eller årsagerne til indeklimasyndrom er imidlertid ukendt, og man kender kun lidt den tidsmæssige sammenhæng mellem udsættelse for kemiske eller fysiske forhold i indeklimaet og udviklingen af symptomer. Desuden er der noget, der tyder på, at forhold som opmærksomhed omkring årsager til indeklimaproblemer og en tendens til at overrapportere kan påvirke angivelsen af, om man har disse symptomer. Formålet med nærværende undersøgelse har været at undersøge den tidsmæssige sammenhæng mellem en persons oplevelse af indeklimaet og de uspecifikke symptomer, som er blevet beskrevet ved indeklimasyndromet, samt at undersøge hvorvidt opmærksomhed omkring indeklimaet og en overrapporteringstendens er associeret med symptomerne. Et andet formål var at validere det spørgeskema, som er blevet anvendt i undersøgelsen. Undersøgelsen er en opfølgningsundersøgelse baseret på spørgeskemaer fra en tilfældig stikprøve af danske statsborgere i alderen 18-59 år. Spørgeskemaerne er udfyldt to gange med et års mellemrum. I alt 2164 personer deltog ved første spørgeskemaundersøgelse. Af disse var der 1402 personer, der også besvarede det andet spørgeskema og som kunne indgå i opfølgningsundersøgelsen, idet de stadig var ansat på samme arbejdsplads og boede samme sted. De helbredsmål, der havde interesse for undersøgelsen, var slimhindesymptomer (et index bestående af øjenirritation, næseirritation, stoppet/løbende næse, irritation i svælget og hæshed) samt almensymptomer (et index bestående af hovedpine, træthed og koncentrationsbesvær). Ved den første spørgeskemaundersøgelse blev populationen delt op i to grupper, som blev præsenteret for formålet med undersøgelsen på en lidt forskellig måde. I ledsagebrevet til den

29

dimensioner og derfor ikke bør samles i et fælles index. Sammenfattende viste resultaterne et komplekst netværk af sammenhænge mellem symptomer og oplevede påvirkninger i indeklimaet, hvor det er vanskeligt at adskille hvad der kom først: eksponering eller symptomer. Desuden, var de rapporterede indeklimapåvirkninger associeret med symptomer som ikke er biologisk plausible at kæde sammen med indeklimaet. Disse fund udelukker ikke, at specifikke forhold i indeklimaet kan påvirke helbredet, men viser at der kan være bias forbundet med om man angiver disse symptomer og at symptomerne i indeklimasyndromet ikke udgør et veldefineret syndrom.

30

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

related illness (BRI) and Sick Building Syndrome (SBS) surveys. Levy, F and Maroni, M. 47-64. 1991. Oslo, Norway, NATO/CCMS.

Indoor air quality and work environment study: EPA headquarters building, Vol III: Relating employee responses to the follow-up questionnaire with environmental measurements of indoor air quality. 1991. U.S.Environmental Protection Agency, North Carolina. Report of Multiple Chemical Sensitivities (MCS) Workshop: International Programme on Chemical Safety (IPCS)/ German Workshop on Multiple Chemical Sensitivities. Berlin, Germany, 21-23 February 1996. Int Arch. Occup Environ Health.(1997): 224-6 From the Centers for Disease Control and Prevention. Update: pulmonary hemorrhage/hemosiderosis among infants--Cleveland, Ohio, 1993-1996. JAMA (2000); 283:1951-3 Abbritti G, Muzi G, Accattoli MP, Fiordi T, dell'Omo M, Colangeli C, Gabrielli AR, Fabbri T, D'Alessandro A (1992) High prevalence of sick building syndrome in a new air-conditioned building in Italy. Arch. Environ. Health 47:16-22 Andersen JH, Thomsen JF, Overgaard E, Lassen CF, Brandt LP, Vilstrup I, Kryger AI, Mikkelsen S (2003) Computer use and carpal tunnel syndrome: a 1-year follow-up study. JAMA 289:2963-9 Andersson B, Berg M, Arnetz BB, Melin L, Langlet I, Liden S (1996) A cognitive-behavioral treatment of patients suffering from "electric hypersensitivity". Subjective effects and reactions in a double-blind provocation study. J Occup Environ Med. 38:752-8 Andersson K (1998) Epidemiological approach to indoor air problems. Indoor Air Suppl 4:32-9 Andersson, K and Stridh, G. The use of standardized questionnaires in building-

31

9.

Apte MG, Fisk WJ, Daisey JM (2000) Associations between indoor CO2 concentrations and sick building syndrome symptoms in U.S. office buildings: an analysis of the 1994-1996 BASE study data. Indoor Air 10:246-57

10.

Avlund K, Kreiner S, Schultz-Larsen K (1993) Construct validation and the Rasch model: functional ability of healthy elderly people. Scand. J. Soc. Med. 21:233-46

11.

Bachmann MO, Myers JE (1995) Influences on sick building syndrome symptoms in three buildings. Soc. Sci. Med. 40:245-51

12.

Bako-Biro Z, Wargocki P, Weschler CJ, Fanger PO (2004) Effects of pollution from personal computers on perceived air quality, SBS symptoms and productivity in offices. Indoor Air 14:178-87

13.

Bauer RM, Greve KW, Besch EL, Schramke CJ, Crouch J, Hicks A, Ware MR, Lyles WB (1992) The role of psychological factors in the report of building-related symptoms in sick building syndrome. J. Consult Clin. Psychol. 60:213-9

14.

Berglund B, Gunnarsson AG (2000) Relationships between occupant personality and the sick building syndrome explored. Indoor Air 10:15269

15.

Bjorner JB, Kreiner S, Ware JE, Damsgaard MT, Bech P (1998) Differential item functioning in the Danish translation of the SF-36. J Clin. Epidemiol. 51:1189-202

16.

Bjornsson E, Janson C, Norback D, Boman G (1998) Symptoms related to the sick building syndrome in a general population sample: associations with

Service Center (The Danish Working Environment Council).

atopy, bronchial hyper-responsiveness and anxiety. Int. J. Tuberc. Lung Dis. 2:1023-8

23.

Broder, I., Pilger, C., and Corey, P. Building Related Discomfort Is Associated with Perceived Rather than Measured Levels of Indoor Environmental Variables. Indoor Air '90 1, 221-226. 1990.

17.

Blom, P, Levy, F, and Skåret, E. Omfanget av inneklimaproblemer i Norge. 97. 1992. Norges byggforskningsinstitut.

18.

Bornehag CG, Blomquist G, Gyntelberg F, Jarvholm B, Malmberg P, Nordvall L, Nielsen A, Pershagen G, Sundell J (2001) Dampness in buildings and health. Nordic interdisciplinary review of the scientific evidence on associations between exposure to "dampness" in buildings and health effects (NORDDAMP). Indoor Air 11:72-86

24.

Burge PS, Robertson AS, Hedge A (1991) Comparison of self-administered questionnaire with physician diagnosis in the diagnosis of the sick building syndrome. Indoor Air 1:422-7

25.

Burge S, Hedge A, Wilson S, Bass JH, Robertson A (1987) Sick building syndrome: a study of 4373 office workers. Ann. Occup. Hyg. 31:493-504

Bornehag CG, Sundell J, Bonini S, Custovic A, Malmberg P, Skerfving S, Sigsgaard T, Verhoeff A (2004) Dampness in buildings as a risk factor for health effects, EUROEXPO: a multidisciplinary review of the literature (1998-2000) on dampness and mite exposure in buildings and health effects. Indoor Air 14:243-57

26.

Chao HJ, Schwartz J, Milton DK, Burge HA (2003) The work environment and workers' health in four large office buildings. Environ. Health Perspect. 111:1242-8

27.

Chester AC, Levine PH (1994) Concurrent sick building syndrome and chronic fatigue syndrome: epidemic neuromyasthenia revisited. Clin. Infect. Dis 18 Suppl 1:S43-S48

28.

Christiansen, M. L., Franck, C., and Petersen, R. Indeklimasymptomer og oplevelser af indeklimaet på hospitaler og kontorer. 1-108. 2001. Arbejdsmiljørådets Service Center

29.

Dalton P, Hummel T (2000) Chemosensory function and response in idiopathic environmental intolerance. Occup. Med. 15:539-56

30.

Dalton P, Wysocki CJ, Brody MJ, Lawley HJ (1997) The influence of cognitive bias on the perceived odor, irritation and health symptoms from chemical exposure. Int Arch Occup Environ Health 69:407-17

31.

de Jonge J, Dormann C, Janssen PPM, Dollard MF, Landeweerd JA, Nijhuis FJN (2001) Testing resiprocal relationships between job characteristics and psychological well-being: A cross-

19.

20.

21.

22.

Bornehag CG, Sundell J, Sigsgaard T (2004) Dampness in buildings and health (DBH): Report from an ongoing epidemiological investigation on the association between indoor environmental factors and health effects among children in Sweden. Indoor Air 14 Suppl 7:59-66 Bornschein S, Hausteiner C, Zilker T, Forstl H (2002) Psychiatric and somatic disorders and multiple chemical sensitivity (MCS) in 264 'environmental patients'. Psychol Med. 32:1387-94 Brauer, C. and Mikkelsen, S. Indoor climate, psychosocial work environment, and symptoms in Denmark - Normative data [in Danish]. [Indeklima, psykisk arbejdsmiljø og symptomer i Danmark Et normalmateriale til Glostrupskemaet]. ISBN 87-90879-63-5, 1-101. 2002. Copenhagen, Arbejdsmiljørådets

32

John Wiley and Sons, New York, pp 212-36

lagged structural equation model. J Occup Organizational Psychol 74:29-46 32.

de Lange AH, Taris TW, Kompier MA, Houtman IL, Bongers PM (2005) Different mechanisms to explain the reversed effects of mental health on work characteristics. Scand. J Work Environ Health 31:3-14

33.

Engelhart S, Burghardt H, Neumann R, Ewers U, Exner M, Kramer MH (1999) Sick building syndrome in an office building formerly used by a pharmaceutical company: a case study. Indoor Air 9:139-43

34.

Engvall K, Norrby C, Norback D (2001) Sick building syndrome in relation to building dampness in multi-family residential buildings in Stockholm. Int Arch. Occup Environ Health 74:270-8

35.

36.

Engvall K, Norrby C, Sandstedt E (2004) The Stockholm Indoor Environment Questionnaire: a sociologically based tool for the assessment of indoor environment and health in dwellings. Indoor Air 14:24-33 Erdmann CA, Apte MG (2004) Mucous membrane and lower respiratory building related symptoms in relation to indoor carbon dioxide concentrations in the 100-building BASE dataset. Indoor Air 14 Suppl 8:127-34

37.

Eriksen HR, Ursin H, Ursin G, Svendsrod R (1998) Prevalence of subjective health complaints in the Nordic European countries in 1993. Eur. J Public Health 8:294-8

38.

Eriksson N, Hoog JH, Stenberg B, Sundell J (1996) Psychosocial factors and the "Sick Building-Syndrome". A case-referent study. Indoor Air 6:101-10

39.

Finnegan MJ, Pickering CA, Burge PS (1984) The sick building syndrome: prevalence studies. Br. Med. J. (Clin. Res. Ed). 289:1573-5

40.

Fleiss JL (1981) The measurement of interrrater agreement. In: Fleiss JL (ed)

33

41.

Frese M (1985) Stress at work and psychosomatic complaints: a causal interpretation. J Appl Psychol. 70:31428

42.

Gandek B, Ware JE, Aaronson NK, Apolone G, Bjorner JB, Brazier JE, Bullinger M, Kaasa S, Leplege A, Prieto L, Sullivan M (1998) Cross-validation of item selection and scoring for the SF12 Health Survey in nine countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 51:11718

43.

Groenvold M, Bjorner JB, Klee MC, Kreiner S (1995) Test for item bias in a quality of life questionnaire. J Clin. Epidemiol. 48:805-16

44.

Hallas TE, Korsgaard J (1983) Annual fluctuations of mites and fungi in Danish house-dust: an example. Allergol. Immunopathol. (Madr. ) 11:195-200

45.

Hallman T, Burell G, Setterlind S, Oden A, Lisspers J (2001) Psychosocial risk factors for coronary heart disease, their importance compared with other risk factors and gender differences in sensitivity. J Cardiovasc. Risk 8:39-49

46.

Handal G, Leiner MA, Cabrera M, Straus DC (2004) Children symptoms before and after knowing about an indoor fungal contamination. Indoor Air 14:87-91

47.

Hardin BD, Kelman BJ, Saxon A (2003) Adverse human health effects associated with molds in the indoor environment. J Occup Environ Med. 45:470-8

48.

Harving H, Korsgaard J, Dahl R (1993) House-dust mites and associated environmental conditions in Danish homes. Allergy 48:106-9

49.

Hauschildt P, Molhave L, Kjaergaard SK (1999) Reactions of healthy persons and persons suffering from allergic

syndrome. An experimental epidemiological study. Indoor 1:111-21

rhinitis when exposed to office dust. Scand. J. Work Environ. Health 25:4429 50.

Henningsen P, Priebe S (2003) New environmental illnesses: what are their characteristics? Psychother. Psychosom. 72:231-4

51.

Heyman E (1881) The indoor air of our homes [Om luften i våra bostäder]. Stockholm

52.

Hodgson MJ, Frohliger J, Permar E, Tidwell C, Traven ND, Olenchock SA, Karpf M (1991) Symptoms and microenvironmental measures in nonproblem buildings. J. Occup. Med. 33:527-33

53.

Hodgson MJ, Morey P, Leung WY, Morrow L, Miller D, Jarvis BB, Robbins H, Halsey JF, Storey E (1998) Building-associated pulmonary disease from exposure to Stachybotrys chartarum and Aspergillus versicolor. J. Occup. Environ. Med. 40:241-9

54.

Hosmer DW, Lemeshow S (2000) Applied logistic regression. John Wiley & Sons, Inc.,

55.

Hutchinson, C. T., Powell, H. R., Hardin, B. D., Robbins, C. A., Saxon, A., and Kelman, B. J. The growing hazard of mold litigation. 1-67. 2003. Waschington, D.C., U.S.Chamber Institute for Legal Reform and the Center for Legal Policy at the Manhattan Institute. Civil Justice Report 8.

and Air

59.

Karasek R, Theorell T (1990) Healthy work: stress, productivity, and the reconstruction of working life. Basic Books, New York

60.

Kipen HM, Fiedler N (2002) The role of environmental factors in medically unexplained symptoms and related syndromes: conference summary and recommendations. Environ. Health Perspect. 110 Suppl 4:591-5

61.

Kolstad HA, Brauer C, Iversen M, Sigsgaard T, Mikkelsen S (2002) Do indoor molds in nonindustrial environments threaten workers' health? A review of the epidemiologic evidence. Epidemiol. Rev. 24:203-17

62.

Korsgaard J (1983) House-dust mites and absolute indoor humidity. Allergy 38:85-92

63.

Lebowitz MD, Cassell EJ, McCarroll J (1972) Health and the urban environment. XI. The incidence and burden of minor illness in a healthy population: methods, symptoms, and incidence. Am Rev. Respir. Dis 106:824-34

64.

Lebowitz MD, Cassell EJ, McCarroll J (1972) Health and the urban environment. XII. The incidence and burden of minor illness in a healthy population: duration, severity, and burden. Am Rev. Respir. Dis 106:83541

56.

Ihlebaek C, Eriksen HR, Ursin H (2002) Prevalence of subjective health complaints (SHC) in Norway. Scand J Public Health 30:20-9

65.

Li CS, Hsu CW, Tai ML (1997) Indoor pollution and sick building syndrome symptoms among workers in day-care centers. Arch. Environ. Health 52:200-7

57.

Jaakkola JJ, Miettinen P (1995) Type of ventilation system in office buildings and sick building syndrome. Am. J. Epidemiol. 141:755-65

66.

58.

Jaakkola JJK, Heinonen OP, Seppanen O (1991) Mechanical ventilation in office buildings and the sick building

Lipscomb JA, Satin KP, Neutra RR (1992) Reported symptom prevalence rates from comparison populations in community-based environmental studies. Arch. Environ. Health 47:263-9

67.

Lundin L (1999) Allergic and nonallergic students' perception of the same

34

high school environment. Indoor Air 9:92-102 68.

69.

70.

71.

72.

73.

74.

75.

Malling HJ, Agrell B, Croner S, Dreborg S, Foucard T, Kjellman M, Koivikko A, Roth A, Weeke B (1985) Diagnosis and immunotherapy of mould allergy. I. Screening for mould allergy. Allergy 40:108-14 McDonald JC, Armstrong B, Benard J, Cherry NM, Farant JP (1993) Sick building syndrome in a Canadian office complex. Arch. Environ. Health 48:298304 Mendell MJ, Naco GM, Wilcox TG, Sieber WK (2003) Environmental risk factors and work-related lower respiratory symptoms in 80 office buildings: an exploratory analysis of NIOSH data. Am J Ind Med. 43:630-41 Meyer HW, Larsen FO, Jacobi HH, Poulsen LK, Clementsen P, Gravesen S, Gyntelberg F, Norn S (1998) Sick building syndrome: association of symptoms with serum IgE specific to fungi. Inflamm. Res. 47 (Suppl 1):S9-10 Meyer HW, Wurtz H, Suadicani P, Valbjorn O, Sigsgaard T, Gyntelberg F (2004) Molds in floor dust and buildingrelated symptoms in adolescent school children. Indoor Air 14:65-72 Mikatavage MA, Rose VE, Funkhouser E, Oestenstad RK, Dillon K, Reynolds KD (1995) Beyond air quality--factors that affect prevalence estimates of sick building syndrome. Am Ind Hyg Assoc J 56:1141-6 Mizoue T, Reijula K, Andersson K (2001) Environmental tobacco smoke exposure and overtime work as risk factors for sick building syndrome in Japan. Am J Epidemiol. 154:803-8 Moffatt S, Mulloli TP, Bhopal R, Foy C, Phillimore P (2000) An exploration of awareness bias in two environmental epidemiology studies. Epidemiology 11:199-208

35

76.

Molhave L, Bach B, Pedersen OF (1986) Human reactions to low cencentrations of volatile organic compounds. Environment International 12:167-75

77.

Muzi G, Abbritti G, Accattoli MP, dell'Omo M (1998) Prevalence of irritative symptoms in a nonproblem airconditioned office building. Int. Arch. Occup Environ Health 71:372-8

78.

Muzi G, dell'Omo M, Abbritti G, Accattoli P, Fiore MC, Gabrielli AR (1998) Objective assessment of ocular and respiratory alterations in employees in a sick building. Am. J. Ind. Med. 34:79-88

79.

Nelson NA, Kaufman JD, Burt J, Karr C (1995) Health symptoms and the work environment in four nonproblem United States office buildings. Scand. J. Work Environ. Health 21:51-9

80.

Norback D, Edling C (1991) Environmental, occupational, and personal factors related to the prevalence of sick building syndrome in the general population. Br. J. Ind. Med. 48:451-62

81.

Norback D, Michel I, Widstrom J (1990) Indoor air quality and personal factors related to the sick building syndrome. Scand. J. Work. Environ. Health 16:121-8

82.

Nordstrom K, Norback D, Akselsson R (1994) Effect of air humidification on the sick building syndrome and perceived indoor air quality in hospitals: a four month longitudinal study. Occup. Environ. Med. 51:683-8

83.

Nordstrom K, Norback D, Akselsson R (1995) Influence of indoor air quality and personal factors on the sick building syndrome (SBS) in Swedish geriatric hospitals. Occup. Environ. Med. 52:170-6

84.

Ooi PL, Goh KT (1997) Sick building syndrome: an emerging stress-related disorder? Int J Epidemiol 26:1243-9

85.

risk factors in Singapore. Environ Med 55:188-93

Occup

86.

Pettenkofer M (1858) Über den Luftwechsel in Wohngebäuden. Literarisch-Artistische Anstalt, München

87.

Pirhonen I, Nevalainen A, Husman T, Pekkanen J (1996) Home dampness, moulds and their influence on respiratory infections and symptoms in adults in Finland. Eur. Respir. J 9:261822

88.

Platt SD, Martin CJ, Hunt SM, Lewis CW (1989) Damp housing, mould growth, and symptomatic health state. BMJ 298:1673-8

89.

Podlekareva D, Pan Z, Kjaergaard S, Molhave L (2002) Irritation of the human eye mucous membrane caused by airborne pollutants. Int. Arch. Occup. Environ. Health 75:359-64

90.

Reinikainen LM, Jaakkola JJ (2001) Effects of temperature and humidification in the office environment. Arch. Environ. Health 56:365-8

91.

92.

93.

94.

personality, measured by the Karolinska Scales of Personality (KSP), on symptoms among subjects in suspected sick buildings. Indoor Air 14:394-404

Ooi PL, Goh KT, Phoon MH, Foo SC, Yap HM (1998) Epidemiology of sick building syndrome and its associated

Robbins CA, Swenson LJ, Nealley ML, Gots RE, Kelman BJ (2000) Health effects of mycotoxins in indoor air: a critical review. Appl Occup Environ Hyg 15:773-84 Robertson AS, McInnes M, Glass D, Dalton G, Burge PS (1989) Building sickness, are symptoms related to the office lighting? Ann. Occup. Hyg 33:47-59 Roht LH, Vernon SW, Weir FW, Pier SM, Sullivan P, Reed LJ (1985) Community exposure to hazardous waste disposal sites: assessing reporting bias. Am J Epidemiol 122:418-33 Runeson R, Norback D, Klinteberg B, Edling C (2004) The influence of

36

95.

Runeson R, Norback D, Stattin H (2003) Symptoms and sense of coherence - a follow-up study of personnel from workplace buildings with indoor air problems. Int. Arch. Occup. Environ. Health 76:29-38

96.

Salvaggio J, Aukrust L (1981) Postgraduate course presentations. Mold-induced asthma. J Allergy Clin. Immunol. 68:327-46

97.

SAS Institute Inc. [Version 8.2]. 1996. Cary, North Carolina, USA.

98.

Schneider T, Sundell J, Bischof W, Bohgard M, Cherrie JW, Clausen PA, Dreborg S, Kildeso J, Kaergaard SK, Lovik M, Pasanen P, Skyberg K (2003) 'EUROPART'. Airborne particles in the indoor environment. A European interdisciplinary review of scientific evidence on associations between exposure to particles in buildings and health effects. Indoor. Air 13:38-48

99.

Schultz-Larsen K, Avlund K, Kreiner S (1992) Functional ability of community dwelling elderly. Criterion-related validity of a new measure of functional ability. J. Clin. Epidemiol. 45:1315-26

100.

Setterlind S, Larsson G (1995) The stress profile: A psychosocial approach to measuring stress. Stress Med 11:8592

101.

Siegrist J (1996) Adverse health effects of high-effort/low-reward conditions. J Occup Health Psychol. 1:27-41

102.

Silny J (1999) Electrical hypersensitivity in humans--fact or fiction? Zentralbl. Hyg Umweltmed. 202:219-33

103.

Skov P, Valbjorn O (1987) The "sick" building syndrome in the office environment: The Danish Town Hall Study. Environment International 13:339-49

104.

with the Prevalence of Occupant Symptoms. Indoor Air 3:82-93

Skov P, Valbjorn O, Pedersen BV (1989) Influence of personal characteristics, job-related factors and psychosocial factors on the sick

113.

Sundell J, Lindvall T, Stenberg B (1994) Associations between type of ventilation and air flow rates in office buildings and the risk of SBS-symptoms among occupants. Environment International 20:239-51

114.

Tang SK (1997) Performance of noise indices in air-conditioned landscaped office buildings. J Acoust. Soc. Am 102:1657-63

building syndrome. Danish Indoor Climate Study Group. Scand. J. Work. Environ. Health 15:286-95 105.

Spurgeon A (2002) Models of unexplained symptoms associated with occupational and environmental exposures. Environ Health Perspect 110 Suppl 4:601-5

106.

Spurgeon A, Gompertz D, Harrington JM (1996) Modifiers of non-specific symptoms in occupational and environmental syndromes. Occup Environ Med 53:361-6

115.

Thorn A (1998) Building-related health problems: reflections on different symptom prevalence among pupils and teachers. Int. J. Circumpolar. Health 57:249-56

107.

Staudenmayer H (2001) Idiopathic environmental intolerances (IEI): myth and reality. Toxicol. Lett. 120:333-42

116.

Thorn A (2000) Emergence and preservation of a chronically sick building. J Epidemiol Community Health 54:552-6

108.

Staudenmayer H, Binkley KE, Leznoff A, Phillips S (2003) Idiopathic environmental intolerance: Part 1: A causation analysis applying Bradford Hill's criteria to the toxicogenic theory. Toxicol. Rev. 22:235-46

117.

Toftum J (2004) Air movement--good or bad? Indoor Air 14 Suppl 7:40-5

118.

van Hooff ML, Geurts SA, Taris TW, Kompier MA, Dikkers JS, Houtman IL, van den Heuvel FM (2005) Disentangling the causal relationships between work-home interference and employee health. Scand. J Work Environ Health 31:15-29

119.

Verhelst N (2001) Testing the unidimensionality assumptoin of the Rasch model. MPR-online 6:231-71

120.

Wallace LA, Nelson CJ, Highsmith R, Dunterman G (1993) Association of personel and workplace characteristics with health, comfort and odor: A survey of 3948 office workers in three buildings. Indoor Air 3:193-205

121.

Wan GH, Li CS (1999) Dampness and airway inflammation and systemic symptoms in office building workers. Arch. Environ. Health 54:58-63

122.

Wargocki P, Lagercrantz L, Witterseh T, Sundell J, Wyon DP, Fanger PO (2002) Subjective perceptions, symptom

109.

Stenberg B, Eriksson N, Hoog J, Sundell J, Wall S (1994) The Sick Building Syndrome (SBS) in office workers. A case- referent study of personal, psychosocial and buildingrelated risk indicators. Int. J. Epidemiol. 23:1190-7

110.

Stenberg B, Wall S (1995) Why do women report 'sick building symptoms' more often than men? Soc. Sci. Med. 40:491-502

111.

Streiner DL, Norman GR (1995) Health measurement scales. A practical guide to their development and use. Second edition. Oxford University Press,

112.

Sundell J, Andersson B, Andersson K, Lindvall T (1993) Volatile Organic Compounds in Ventilating Air in Buildings at Different Sampling Points in the Buildings and their Relationship

37

German-speaking area. Zentralbl. Hyg. Umweltmed. 202:435-70

intensity and performance: a comparison of two independent studies, both changing similarly the pollution load in an office. Indoor Air 12:74-80 123.

Wargocki P, Sundell J, Bischof W, Brundrett G, Fanger PO, Gyntelberg F, Hanssen SO, Harrison P, Pickering A, Seppanen O, Wouters P (2002) Ventilation and health in non-industrial indoor environments: report from a European multidisciplinary scientific consensus meeting (EUROVEN). Indoor. Air 12:113-28

124.

Wessely S, Nimnuan C, Sharpe M (1999) Functional somatic syndromes: one or many? Lancet 354:936-9

125.

Wiesmuller GA, Steup A, Ranft U (1999) Inner room-questionnaire for the

38

126.

World Health Organization. Health aspects related to indoor air quality report on a WHO working group. 21, 137. 1979. World Health Organzation, Regional Office for Europe. EURO Reports and Studies 21.

127.

World Health Organization. Indoor air pollutants: exposure and health effects Report on a WHO meeting. 78, 1-48. 1983. WHO Regional Office for Europe. EURO Reports and Studies 78.

128.

Zweers T, Preller L, Brunekreef B, Boleij JSM (1992) Health and indoor Climate Complaints of 7043 office workers in 61 buildings in the Netherlands. Indoor Air 2:127-36

Table 1. Characteristics of the participants in the validation study, N=947. Suspected indoor environment problem

N

Participation rate (%)

Women (%)

Age mean (range)

School No. 1

Mould growth in the roof construction

69

97

68

49 (26-63)

Hospital ward

Dust after renovation of the building

40

91

100

42 (24-68)

Social and health

Mould growth in the outside walls

26

>85*

100

45 (26-63)

School No. 2

Dust and mineral fibres from ceilings

52

95

67

45 (25-63)

Nursing home

Water damage

204

93

96

42 (18-66)

Technical University

Dust after renovation of the building

64

100

59

39 (22-65)

62

>85*

68

46 (24-63)

430

94

63

44 (18-65)

Type of workplace

care college

and ventilation problems School No. 3

Water damage and ventilation problems

Town hall

Dust after renovation of the building and ventilation problems

* The exact number of employees was not reported, only an approximate number

Table 2. Illustration of item bias with respect to atopy of the item “breathlessness” in the pulmonary symptom index. If there was no item bias the percentage of positive answers to the item "breathlessness" at the index scores 1 and 2 would be the same for non-atopics and atopics. However, here atopics are more likely to have a positive answer to the item "breathlessness" (shown in the hatched fields). At a score=0 and a score=3, both non-atopics and atopics, by definition, have 0% and 100%, respectively. Breathlessness n*

No (%)

Yes (%)

Index score 0 Non-atopic Atopic

483 250

100 100

0 0

Index score 1 Non-atopic Atopic

72 52

94 81

6 19

Index score 2 Non-atopic Atopic

7 18

57 44

43 56

Index score 3 Non-atopic Atopic

5 11

0 0

100 100

2 χ =6.0 d.f.=2 p=0.044

gamma=0.58 p=0.032 (two-sided)

*N=898, because respondents with any item unanswered in the index are excluded.

Table 3. Test-retest reliability and internal consistency of symptom indices and indoor environment indices Test-retest reliability, N=110

Internal consistency. Cronbach coefficient alpha

Full agreement (%)

kappa

(95% CI)

Validation sample, N=947

Baseline study sample, N=2164

Symptom indices Mucous membrane index Pulmonary index Skin index General symptom index

76 89 84 94

0.78 0.65 0.69 0.93

(0.70 - 0.86) (0.46 - 0.84) (0.56 - 0.83) (0.88 - 0.98)

0.71 0.61 0.49 0.66

0.64 0.53 0.41 0.57

Indoor environment indices Draught index Temperature index Stuffy air index Environmental tobacco smoke Dry air index Noise index Light index Space/dust index

80 77 73 85 88 76 87 77

0.64 0.58 0.63 0.79 0.77 0.67 0.63 0.74

(0.49 - 0.78) (0.43 - 0.73) (0.51 - 0.74) (0.63 - 0.94)* (0.65 - 0.89) (0.55 - 0.79) (0.45 - 0.81) (0.62 - 0.86)

0.64 0.58 0.60 0.44 0.59 0.58 0.45

0.72 0.58 0.61 0.50 0.68 0.57 0.41

* Simple kappa. All other kappa values are the weighted kappa.

Table 4. Baseline characteristics according to status at one-year follow-up. P-values computed with the Chi Square test unless otherwise noted. Non-respondents, N=424

Respondents, N=1740

n

(%)

(%)

P-value

Women Marital status, living as married Age, years 18-29 30-39 40-49 50-59 Employment self-employed non-manuals skilled manual worker unskilled manual worker

207 283

(48.8) (73.3)

907 1307

(52.1) (80.7)

0.22 0.001

107 113 105 99

(25.2) (26.7) (24.8) (23.4)

290 484 508 458

(16.7) (27.8) (29.2) (26.3)

0.001

46 232 55 71

(11.4) (57.4) (13.6) (17.6)

129 1107 224 231

(7.6) (65.5) (13.3) (13.7)

0.005

Smoking, daily Atopy

156 164

(37.1) (38.9)

539 658

(31.1) (37.9)

0.02 0.72

Symptoms More than one mucous membrane symptom*

122

(28.8)

485

(27.9)

0.72

104

(24.5)

396

(22.8)

0.45

49

(11.6)

150

(8.7)

0.06

Perceived indoor environment Draught index Temperature index Stuffy air index Environmental tobacco smoke Dry air index Noise index Light index Space/dust indext Patches of damp/mildew

107 108 92 113 104 193 54 147 26

(26.8) (27.3) (23.2) (28.5) (26.2) (48.6) (13.6) (37.1) (6.7)

464 467 387 421 527 757 293 650 108

(27.5) (27.7) (23.0) (25.1) (31.2) (44.9) (17.4) (38.5) (6.5)

0.77 0.86 0.94 0.17 0.05 0.18 0.07 0.60 0.91

Psychosocial work characteristics High job demands Low job control Poor support, co-workers and supervisors Effort-reward imbalance Poor social climate

156 77 141 61 44

(38.3) (19.1) (34.8) (15.1) (10.9)

668 245 601 311 244

(39.4) (14.4) (35.4) (18.3) (14.4)

0.68 0.02 0.83 0.13 0.07

53 87 52 52

(12.6) (20.7) (12.4) (12.3)

158 307 167 204

(9.1) (17.7) (9.7) (11.8)

0.03 0.16 0.10 0.76

102

(24.5)

338

(19.6)

0.02

More than one general symptom





Symptom reporting tendency

n

§

Personality traits Negative affectivity Type A behaviour Self-efficacy Worry about health Poor family support SF12 Physical Health Summary Scale (mean) Mental Health Summary Scale (mean)

50.8 50.3

51.8 50.2

* Mucous membrane symptoms: eye irritation, nose irritation, nasal congestion, throat irritation, and hoarseness. † General symptoms: fatigue, headache, and concentration difficulty ‡ At least four symptoms among 15 "non-SBS" symptoms § A score of one or more on the index ¶ T-test



0.02 ¶ 0.87

Table 5. Questions about psychosocial work characteristics and personality traits. Characteristic measured

Questions

Answers

Job demands

How demanding do you find your work altogether?

Extremely demanding, very demanding, rather demanding, somewhat demanding, not very demanding, very little demanding

Job control

How much influence do you have in planning and carrying out your work?

Very much, rather much, moderate, not so much, rather little, very little

Job strain* Support at work

If you have problems, is it possible to obtain the Always, often, sometimes, seldom, never necessary support from colleagues or superiors?

Effort-reward imbalance

Do you think that your job performance is sufficiently appreciated?

To a very high degree, to a high degree, to some degree, to a minor degree, to a low degree, to a very low degree

Social climate at work

How is the atmosphere and psychological climate at your workplace?

Very good, rather good, quite good, not so good, rather poor, very poor

Negative affectivity

Do you as a person tend to be worried, Not at all, very little, a little, somewhat, quite a bit, rather much, very much nervous, or somewhat pessimistic?

Type A behaviour

Do you as a person tend to be competitive, Not at all, very little, a little, somewhat, quite a bit, rather much, very much jealous, ambitious, and somewhat impatient

Self-efficacy

Are you a person who is usually able to solve difficult problems, manage unexpected situations and reach your goals?

Worry about health

Do you as a person tend to be worried about Not at all, very little, a little, somewhat, quite a bit, rather much, very much your health?

Social support

If you have problems, is it possible to obtain the Always, often, sometimes, seldom, never necessary support from family or friends?

Not at all, very little, a little, somewhat, quite a bit, rather much, very much

Bold indicates response alternatives with a hypothesized high risk of non-specific symptoms compared to non-bold response alternatives. * Job strain defined as the combination of high job demands and low job control

Table 6. Changes in perceived indoor environment between baseline and follow-up. Number and (%). Better

Unchanged

Worse

A change>1

A change=1

A change=1

A change>1

Draught

63

(4.7)

159

(12.0)

945

(71.1)

114

(8.6)

48

(3.6)

Temperature

49

(3.7)

160

(12.1)

955

(72.0)

121

(9.1)

41

(3.1)

Stuffy air

41

(3.1)

137

(10.4)

1010

(76.3)

107

(8.1)

28

(2.1)

155

(11.8)

1063

(80.7)

99

(7.5)

Environmental tobacco smoke Dry air

28

(2.1)

187

(14.1)

990

(74.6)

111

(8.4)

11

(0.8)

Noise

95

(7.2)

154

(11.6)

828

(62.6)

176

(13.3)

70

(5.3)

Light

27

(2.0)

101

(7.6)

1086

(82.0)

85

(6.4)

25

(1.9)

Space/dust

37

(2.8)

176

(13.3)

943

(71.0)

155

(11.7)

17

(1.3)

50

(3.8)

1225

(93.7)

32

(2.5)

Patches of damp/mildew

-

-

-

-

Table 7. The symptom check list used to record “symptom reporting tendency”. Results from baseline questionnaire. N=2164. Symptoms marked in bold are omitted from the total score of "symptom reporting tendency", because they are often included in the SBS. Prevalence very often/often sometimes (%) (%) 6.6 Stomach ache or problem with the stomach 4.1 1.0 20.1 1.6 Chest pain 4.6 0.7 9.9 2.8 Heart palpitations 4.6 0.8 8.8 2.7 Shortness of breath 4.6 0.8 8.5 3.0 Vertigo 4.5 0.8 11.9 24.5 Muscle tension 3.4 1.3 31.8 10.8 Sweating 4.0 1.1 21.1 4.3 Powerlessness or helplessness 4.4 0.9 12.2 5.2 Sadness or depression 4.1 1.0 20.6 7.3 Restlessness 4.1 1.0 20.9 7.1 Nervousness 4.1 1.0 20.4 12.1 Fatigue and weakness 3.8 1.1 27.7 9.6 Sleeping problems 4.1 1.1 19.8 4.8 Tendency to cry 4.3 0.9 14.0 10.4 Unable to relax 3.9 1.1 22.9 5.8 Concentration problems 4.1 1.0 21.2 4.1 Difficulty in making decisions 4.2 0.9 18.1 11.8 Forgetfulness 3.7 1.1 28.7 4.3 Difficulty in thinking clearly 4.2 0.9 16.4 Symptoms experienced during the past four weeks. The response options were: very often, often, sometimes, seldom, never. * The score ranges from 1 to 5, where 1=very often and 5=never. Symptom

Mean score*

SD

Table 8. Associations between "symptom reporting tendency" and the prevalence, incidence and persistence of mucous membrane symptoms and general symptoms. Odds ratios and 95% confidence interval.

tendency

reporting

Symptom

Mucous membrane symptoms (N=607)

General symptoms (N=500)

Prevalence

Incidence

Persistence

Prevalence

Incidence

Persistence

Min.*

2.3 (1.6 to 3.2)

2.5 (1.3 to 4.8)

2.2 (1.2 to 4.3)

7.2 (5.0 to 10.6)

2.2 (0.9 to 5.2)

1.4 (0.7 to 2.7)

Max.

2.3 (1.6 to 3.3)

2.7 (1.4 to 5.2)

2.4 (1.3 to 4.8)

7.3 (5.0 to 10.8)

2.4 (0.9 to 5.7)

1.7 (0.9 to 3.2)

*Minimum and maximum odds ratios from the models with each of the nine indoor environment indices.

Table 9. Associations between mucous membrane symptoms and perceived indoor environment. Models with and without "symptom reporting tendency". Adjusted odds ratios (OR) and 95% confidence interval (95%CI). Mucous membrane symptoms Prevalence, N=607

Incidence, N=152

Persistence, N=183

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR* (95% CI)

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR* (95% CI)

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR* (95% CI)

0

1

1

1

1

1

1

1

1.65

(1.2 to 2.2)

1.71

(1.3 to 2.3)

1.02

(0.6 to 1.8)

1.03

(0.6 to 1.8)

0.93

(0.5 to 1.6)

1.05

(0.6 to 1.8)

2

1.18

(0.8 to 1.8)

1.18

(0.8 to 1.8)

1.14

(0.5 to 2.2)

1.22

(0.6 to 2.4)

1.70

(0.8 to 3.7)

1.85

(0.9 to 4.0)

3

1.13

(0.7 to 1.8)

1.24

(0.8 to 2.0)

2.51

(1.2 to 5.1)

2.57

(1.2 to 5.2)

1.13

(0.5 to 2.8)

1.26

(0.5 to 3.1)

0

1

1

1.35

(1.0 to 1.8)

1.34

(1.0 to 1.8)

1.18

(0.7 to 1.9)

1.19

(0.7 to 1.9)

1.16

(0.7 to 2.0)

1.20

(0.7 to 2.1)

2

1.67

(1.1 to 2.5)

1.68

(1.1 to 2.5)

0.98

(0.5 to 2.0)

1.01

(0.5 to 2.0)

1.16

(0.6 to 2.3)

1.30

(0.7 to 2.5)

0

1

1

1.27

(0.9 to 1.7)

1.28

(0.9 to 1.7)

1.19

(0.7 to 2.0)

1.23

(0.7 to 2.1)

1.97

(1.2 to 3.4)

2.12

(1.3 to 3.6)

2

1.30

(0.9 to 1.9)

1.42

(1.0 to 2.1)

1.82

(0.9 to 3.7)

1.95

(1.0 to 3.9)

2.05

(1.0 to 4.2)

2.46

(1.2 to 5.0)

Environmental tobacco smoke

0

1

1

1.05

Dry air index

0

1

1

2.20

(1.7 to 2.8)

2.23

(1.7 to 2.9)

1.43

(0.9 to 2.3)

1.49

(0.9 to 2.3)

1.18

(0.7 to 1.9)

1.21

(0.7 to 2.0)

2

3.55

(2.5 to 5.1)

3.51

(2.5 to 5.0)

1.85

(0.9 to 3.6)

1.76

(0.9 to 3.4)

1.16

(0.6 to 2.3)

1.18

(0.6 to 2.3)

0

1

1

1.35

(1.0 to 1.8)

1.37

(1.1 to 1.8)

1.51

(0.9 to 2.4)

1.49

(0.9 to 2.4)

0.87

(0.5 to 1.5)

0.92

(0.5 to 1.6)

2

1.22

(0.9 to 1.7)

1.22

(0.9 to 1.7)

1.60

(0.9 to 2.7)

1.62

(0.9 to 2.7)

0.66

(0.3 to 1.2)

0.70

(0.4 to 1.3)

3

0.99

(0.7 to 1.5)

1.05

(0.7 to 1.5)

2.37

(1.2 to 4.4)

2.53

(1.3 to 4.7)

1.37

(0.6 to 3.2)

1.34

(0.6 to 3.1)

0

1

1

1.13

(0.8 to 1.6)

1.18

(0.9 to 1.6)

1.76

(1.0 to 2.9)

1.83

(1.1 to 3.0)

0.95

(0.5 to 1.8)

1.03

(0.6 to 1.9)

2

1.22

(0.8 to 1.9)

1.27

(0.8 to 2.0)

0.95

(0.3 to 2.3)

0.97

(0.4 to 2.3)

0.88

(0.4 to 2.0)

0.98

(0.4 to 2.2)

0

1

1

1.21

(0.9 to 1.5)

1.23

(1.0 to 1.6)

1.22

(0.8 to 1.9)

1.26

(0.8 to 1.9)

1.14

(0.7 to 1.9)

1.13

(0.7 to 1.9)

2

1.26

(0.9 to 1.8)

1.30

(0.9 to 1.9)

1.28

(0.6 to 2.5)

1.31

(0.7 to 2.5)

1.72

(0.9 to 3.5)

1.71

(0.8 to 3.5)

0

1

1

0.95

Risk factor Draught index

Temperature index

Stuffy air index

Noise index

Light index

Space/dust index

Patches of damp/mildew

Values

1

1

1

1

1 (0.8 to 1.3)

1.06

1

1

1.71

1.76

* OR: Odds ratios adjusted for sex, age, atopy, other personal factors and psychosocial factors according to the tables in appendix III.

1

1

1 (0.9 to 3.4)

1.48

(0.7 to 1.8)

1

1

1 (0.8 to 3.3)

1.13 1

1

1

1 (0.7 to 1.5)

1 (0.7 to 1.9)

1

1

1

1.17 1

1

1

1

1 (0.7 to 1.6)

1

1

1

1.04

1

1

1 (0.7 to 1.6)

1

1

0.99

1.02

1

1

1 (0.8 to 1.3)

1

(0.6 to 1.4)

1

1 (0.6 to 3.9)

1.69

(0.7 to 4.4)

Table 10. Associations between general symptoms and perceived indoor environment. Models with and without symptom reporting tendency. Adjusted odds ratios (OR) and 95% confidence interval (95% CI). General symptoms Prevalence, N=500

Incidence, N=103

Persistence, N=146

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR* (95% CI)

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR* (95% CI)

Model with symptom reporting tendency OR* (95% CI)

Model without symptom reporting tendency OR: (95% CI)

0 1

1 1.05

(0.7 to 1.5)

1 1.15

(0.8 to 1.6)

1 1.48

(0.8 to 2.8)

1 1.47

(0.8 to 2.7)

1 0.88

(0.4 to 1.8)

1 0.97

(0.5 to 1.9)

2

1.11

(0.7 to 1.7)

1.11

(0.7 to 1.7)

1.65

(0.7 to 3.7)

1.60

(0.7 to 3.6)

1.10

(0.5 to 2.4)

1.15

(0.5 to 2.5)

3

1.25

(0.8 to 2.0)

1.41

(0.9 to 2.2)

2.40

(1.0 to 5.7)

2.48

(1.0 to 5.9)

0.81

(0.3 to 2.1)

0.86

(0.3 to 2.2)

0

1

1

1.11

(0.8 to 1.5)

1.12

(0.8 to 1.5)

1.53

(0.8 to 2.7)

1.53

(0.8 to 2.7)

0.84

(0.5 to 1.5)

0.87

(0.5 to 1.6)

2

1.07

(0.7 to 1.6)

1.09

(0.7 to 1.6)

1.21

(0.5 to 2.7)

1.22

(0.5 to 2.7)

1.33

(0.6 to 3.0)

1.38

(0.6 to 3.1)

0

1

1

1.91

(1.4 to 2.6)

1.90

(1.4 to 2.5)

1.52

(0.8 to 2.8)

1.47

(0.8 to 2.7)

1.70

(0.9 to 3.2)

1.76

(0.9 to 3.3)

2

2.79

(1.9 to 4.1)

3.13

(2.2 to 4.5)

1.13

(0.4 to 2.7)

1.14

(0.4 to 2.8)

1.47

(0.7 to 3.1)

1.54

(0.7 to 3.2)

Environmental tobacco smoke

0

1

1

1.09

Dry air index

0

1

1

1.21

(0.9 to 1.6)

1.25

(0.9 to 1.6)

1.10

(0.6 to 1.9)

1.16

(0.7 to 2.0)

1.18

(0.7 to 2.1)

1.21

(0.7 to 2.2)

2

1.41

(0.9 to 2.1)

1.32

(0.9 to 1.9)

1.75

(0.8 to 3.6)

1.75

(0.8 to 3.6)

1.02

(0.5 to 2.1)

1.02

(0.5 to 2.2)

0

1

1

1.27

(0.9 to 1.7)

1.29

(1.0 to 1.7)

0.69

(0.4 to 1.2)

0.68

(0.4 to 1.2)

1.01

(0.5 to 1.9)

1.03

(0.5 to 2.0)

2

1.66

(1.2 to 2.3)

1.65

(1.2 to 2.3)

0.96

(0.5 to 1.9)

0.93

(0.5 to 1.8)

0.55

(0.3 to 1.1)

0.55

(0.3 to 1.1)

3

0.93

(0.6 to 1.4)

1.08

(0.7 to 1.6)

0.73

(0.3 to 1.7)

0.76

(0.3 to 1.7)

1.27

(0.5 to 3.2)

1.30

(0.5 to 3.3)

0

1

1

1.24

(0.9 to 1.7)

1.34

(1.0 to 1.9)

1.31

(0.7 to 2.4)

1.30

(0.7 to 2.4)

1.17

(0.6 to 2.3)

1.22

(0.6 to 2.4)

2

1.54

(0.9 to 2.5)

1.66

(1.0 to 2.7)

1.29

(0.4 to 3.4)

1.25

(0.4 to 3.3)

0.46

(0.2 to 1.1)

0.49

(0.2 to 1.2)

0

1

1

0.94

(0.7 to 1.2)

1.00

(0.8 to 1.3)

1.82

(1.1 to 3.0)

1.80

(1.1 to 3.0)

0.84

(0.5 to 1.5)

0.87

(0.5 to 1.5)

2

0.90

(0.6 to 1.4)

1.00

(0.7 to 1.5)

1.51

(0.7 to 3.2)

1.52

(0.7 to 3.2)

0.68

(0.3 to 1.4)

0.72

(0.3 to 1.5)

0

1

1

1.09

Risk factor Draught index

Temperature index

Stuffy air index

Noise index

Light index

Space/dust index

Patches of damp/mildew

Values

1

1

1

1

1 (0.8 to 1.4)

1.07

1

1

1.48

1.44

* OR: Odds ratios adjusted for sex, age, atopy, other personal factors and psychosocial factors according to the tables in appendix III

1

1

1 (0.6 to 3.0)

0.73

(0.4 to 1.4)

1

1

1 (0.6 to 3.1)

0.78 1

1

1

1 (0.8 to 1.9)

1 (0.4 to 1.4)

1

1

1

0.78 1

1

1

1

1 (0.4 to 1.1)

1

1

1

0.69

1

1

1 (0.4 to 1.1)

1

1

1.25

0.68

1

1

1 (0.8 to 1.4)

1

(0.7 to 1.7)

1

1 (0.3 to 1.9)

0.81

(0.3 to 2.0)