Airborne bacteria levels in indoor urban environments: The influence of season and prevalence of sick building syndrome (SBS)

Abstract

This study investigated levels of airborne bacteria in 90 indoor sampling sites with no known complaints such as kindergartens, schools, homes and dorms over three seasons. Airborne bacteria samples were collected before and during the heating season in Çanakkale, Turkey. Also, a detailed occupant questionnaire was completed by one of the occupants (n = 90) in each sampling site. Total bacteria counts (TBCs) were found to be higher in the sampling sites where coal was used as a heating source and in kindergartens where the occupancy rate was highest (p < 0.05). Meteorological parameters were associated with outdoor TBC levels (p < 0.001). Associations were found (p = 0.05) for indoor TBC and other building factors (last floor covering/wall painting/renovation times, heating/cooking fuels and the age of the building). Indoor TBC levels together with the building-related factors (environmental tobacco smoke exposure, occupancy rate, distance to traffic, heating fuel, amount of carpet covering, and last wall painting time) and personal factors (gender and daily time spent indoors) were found to be associated with at least one group of the sick building syndrome (SBS) symptoms (p < 0.05). In particular, general SBS symptoms (headache, fatigue and dizziness) were observed more often in female participants (p < 0.05). In conclusion, there is a need to conduct more studies in larger populations to find the links between air pollutants and SBS symptoms.

Keywords

Airborne bacteria Indoor Outdoor Heating Sick building syndrome Çanakkale

Introduction

Airborne bacteria pose risks when the concentration is high enough to induce adverse health effects,¹ such as infections, allergies, toxic illness and asthma.^2,3 Dose–response relationships and safe values for airborne bacteria for indoor environments have not been determined yet. Levels of airborne bacteria in indoor environments are related to microbial composition of outdoor air and the sources or activities done within the indoor environments.⁴ When substantial indoor or outdoor airborne bacteria sources do not exist, the most significant factor that influences indoor bacteria levels is human presence.^5–8

The scope of sick building syndrome (SBS) was first defined by the World Health Organization (WHO) based on earlier studies conducted in northern Europe and the USA indicating unspecified, combined, adverse health effects due to poor indoor air quality (IAQ). The most distinctive feature of SBS is a decrease in symptoms or their severity after leaving the ‘sick building’. The main symptoms of SBS, as defined by WHO, are: (i) eye, nose and throat irritation; (ii) sensation of dry mucous membranes and skin; (iii) erythema; (iv) mental fatigue; (v) headaches, high frequency of airway infections, and cough; (vi) hoarseness, wheezing, itching and unspecific hypersensitivity and (vii) nausea, dizziness.⁹ Buildings are classified as ‘sick’ according to the complaint rate of the total occupants (e.g. 20%).¹⁰ Studies have attempted to find the correlation between SBS and its potential causes. Passarelli¹⁰ and Burge¹¹ pointed out the link between SBS and inadequate levels of maintenance. It was found that exposure to bioaerosols is associated with SBS and other hypersensitivity reactions.^12–14 Moreover, the prevalence of SBS was found to be associated with reduced employee well-being, absentee rates and psychosocial factors.^15,16 However, there are few studies that investigate the association between airborne bacteria levels and the prevalence of SBS symptoms, e.g.^17,18

Heating activity can affect IAQ in two ways: (i) heating activity is mainly based on a combustion process, and several factors such as heater type, heating fuel and the heating conditions can result in both indoor and outdoor air pollution and (ii) the main aim of the heating activity, saving the heat by trapping relatively warmer air inside the building and keeping the relatively colder outdoor air outside the building, may not allow for the introduction of fresh air into the indoor environment. No studies have been conducted in Turkey to compare the influence of the type of heating fuel on IAQ.

Since the perception and public awareness of IAQ in Turkey is quite recent, studies conducted on this issue are limited so far. The aims of this study were (i) to investigate the variations of airborne bacteria levels in four types of indoor environments with no known complaints (schools, kindergartens, dorms and homes) before and during the heating season; (ii) to compare outdoor total bacteria count (TBC) levels with indoor TBC levels, (iii) to find the links between several environmental factors and TBC levels; (iv) to estimate the SBS symptoms among the occupants and (v) to find the associations between SBS symptoms and environmental factors. The results of this study will help to increase public awareness of the importance of providing better IAQ in developing countries such as Turkey.

Materials and methods

Study design

Airborne bacteria samples were collected from different types of indoor environments (n = 90) and their adjacent outside environments in Çanakkale. Sampling sites were located in the Barbaros, Esenler, Ismetpasa and Kepez districts. Esenler represented a semi-urban site, while others were designated as urban. Locations of the sampling sites are given in Figure 1. Throughout the study, three sampling campaigns were conducted at all sampling sites: the first one was in October, 2012 before heating season (autumn, n = 90); the second was from December, 2012 to January, 2013 during the early stages of the heating season (winter, n = 90) and the third was in March, 2013, the middle of the heating season (spring, n = 90). In general, the heating period starts in November and ends just before April in Canakkale.

Figure 1.

Sampling locations (A: Ismetpasa; B: Kepez; C: Barbaros; D: Esenler).

Çanakkale is a small city, located on the west coast of Turkey. Its climate has a transient character between the Mediterranean and Black Sea regions: snowy days are limited, whereas rainy days are more frequent. The main rainy season in Çanakkale is spring, especially the month of May. The most prominent characteristics of the meteorology of Çanakkale are relatively high wind speed (wind speed ≥ 7 m s⁻¹ at 50 m above the sea level) and the annual number of windy days (average: 180 days year⁻¹). The mean outdoor temperature values in Çanakkale were 19.3, 7.9 and 10.3℃ during October, December–January and March, respectively. Exact values of the meteorological parameters at the time of the samplings were gathered from Turkish Meteorological Institute.

Parallel indoor and outdoor samplings were conducted for each sampling site. The indoor sampling sites were categorized into four groups: (i) homes (n = 33), (ii) kindergartens (n = 10), (iii) dormitories (n = 37) and (iv) schools (n = 5 for university classes and n = 5 for third-grade classes). All indoor environments were selected randomly. Any known dampness in the sampling sites and essential complaints by the occupants were not recorded prior to the study. Our primary aim was to choose sampling sites that were representative of the extensive socioeconomic conditions of the city. The majority of the sampling sites were in different buildings, with the exceptions being mostly some of the dorms and classes. Since it is possible that outdoor air composition may show fluctuations, parallel duplicate outdoor samples were collected for each sampling site. The average of the duplicate measurements was considered as outdoor TBC level of the sampling site concerned (n = 90 for each season).

A detailed questionnaire (see Appendix) was distributed to an occupant of each sampling point (n = 90). First the sampled environments were selected randomly and then participants (age > 18) for the questionnaire were chosen. One participant per sampling site was chosen. As much as possible, participants were chosen from among the occupants who spend most of his or her time in this environment and most probably available throughout the three sampling campaigns. Questionnaires were given to the teacher in charge of the class for kindergarten and school sampling sites. Maximum two or three people were available at the home sampling sites at the time of the first sampling campaign. Among them, one or rarely two people mentioned that they spent most of their time in the home. Finally, one person available at home for the three sampling campaigns volunteered and was selected. There were one or two people at the dorms at the time of first sampling campaign (two to three person rooms). Among them, there was only one person who spent most of his or her time and would be available for three sampling campaigns, and answered the questionnaire. Since none of the potential participants were informed regarding the objective of the questionnaire before the first sampling campaign, the participant population did not include only those with SBS symptoms, which could have minimised the selection bias. The questionnaire was applied to the same participants three different times. The sampling dates were not mentioned to the participants in advance. After the airborne bacteria samplings, we gave limited information about the questionnaire and administered it as an interview to the participants. Additionally, descriptive information regarding the indoor environment of the sampling sites was gathered.

Airborne bacteria sampling and enumeration

Cultivable airborne bacteria samples were collected by means of a single-stage Andersen sampler (SKC Inc.) including 400 holes with a flow rate of 1.698 m³h⁻¹.^19–22 The air flow was adjusted and controlled by a digital calibrator (Defender, SKC Inc.) before each sampling, average relative standard deviation (SD) of flow readings was <3%. As recommended by several authors referenced subsequently, maximum number of colony formation on a petri dish was assumed to be 200 colonies to prevent masking effect.^{3,7,19,20,23–25} Sampling time was 5 min. Samples were collected during daytime from 1.5 m height at the centre of the room for indoor air samples and at least 2 m away from the building for outdoor samples. No outdoor air sampling was conducted in rainy or snowy weather.

Cultivable airborne bacteria samples were impacted onto plate count agar to count the total bacteria colonies. All agar plates were ready to use and sterile (Salubris Co., Turkey). Bacteria samples were incubated at 37℃ for two days. After the incubation period, enumeration of bacteria colonies’ growth onto a plate was done under a light source by using a semiautomatic colony counter. TBC was calculated as colony-forming unit per drawn air in m³ (CFU m⁻³). Although some of the TBC values exceeded 200 colonies per plate (n_autumn = 4, n_winter = 14 and n_spring = 17), indicating potential sources to propagate bacteria may exist (i.e. mostly in kindergartens, dorms and homes, respectively), remaining TBC values were lower than 1400 CFU m⁻³. Limit of detection (LOD) value for TBC was 7 CFU m⁻³; however, colony formation was not observed on a few sample plates (n = 5).

To minimise the fluctuations in bacteria levels or to better see the actual airborne bacteria levels in the sampling sites, occupants were informed beforehand that they should not change their daily routine (e.g. do not clean and do not additionally ventilate your home/dorm/class, etc.).

Questionnaire

The questionnaire contained questions on (i) the factors that may affect TBC, (ii) the characteristics of the environment, (iii) socioeconomic conditions of the occupants, (iv) the prevalence of SBS and (v) the factors that may affect SBS. The questionnaire was completed by the same person (≥18 years old) in each sampling period. The participants were not informed about the TBC levels measured in their indoor environments throughout the entire study. The questionnaire (see Appendix) is similar in some respects to earlier studies,^11,26–29 but is primarily based on the revised version of Mentese et al.³⁰ The following topics–issues were addressed: (i) information regarding the environment (e.g. location, distance to traffic, age of the building, floor and wall covering, last renovation time, indoor pet/plant availability, stove type and heating/cooking fuel type); (ii) socioeconomic information (e.g. gender, age, education level, number of people living in the microenvironment, time spent indoors and smoking habits); (iii) general medical condition of the participant including the occurrence of SBS symptoms (general medical condition of the occupant, existence of any illness/illnesses related to SBS symptoms, if none; occurrence of SBS symptoms of fatigue, headache, coughing, eye irritation, sneezing, tickling in the throat, dizziness, dry skin and chest tightness) and (iv) concerns regarding the perceived indoor and ambient air quality. The answers to the first section of the questionnaire (i.e. indoor environmental aspects) were also cross checked by our visual inspection together with the participants at the time of surveying. If the visual inspection did not match with participants’ answers, the participants were asked if he or she was sure about the answer and then his or her final decision was accepted.

Statistical analyses

Statistical analyses were performed using either Statgraphics 3.0 software (free software) or Statistical Package for the Social Sciences 19.0. Airborne bacteria concentrations (TBC) were analysed as an independent variable; while sampling period (autumn, winter and spring), meteorological factors, individual SBS symptoms and sample type (indoor and outdoor) were analysed as dependent variables to relate the factors with airborne bacteria levels. TBC levels that were lower than LOD value were shifted with ½ LOD value for statistical analyses.

Distribution fitting of the TBC values was investigated with the Kolmogorov Smirnov test. Paired t-test and Wilcoxon signed rank test were applied to TBC pairs to test the statistically significant differences between the pairs in terms of their mean values and median values, respectively.

Several building-related environmental factors were taken into account statistically. Data from the measurements and environmental factors were analysed statistically by analysis of variance (ANOVA) tests, paired t-test and Spearman’s rank correlation. Bivariate regression analyses were applied to find the associations among the SBS symptoms, building/environmental-related factors and TBC levels (p < 0.2).

Logistic regression analyses were performed to find the associations between the observed SBS symptoms and the personal factors, building-related factors and TBC. SBS symptoms were categorized whether they were observed (yes = 1) or not (no = 0) among the healthy participants who had no related disease. Building related factors with p < 0.2, found from bivariate regression analyses, were included in the logistic regression models. All SBS symptoms in the survey, except chest tightness, were included in the regression models by classification of the SBS symptoms into three classes: (i) general symptoms (headache, fatigue and dizziness); (ii) dermal symptoms (dry skin) and (iii) mucosal symptoms (eye irritation, sneezing, tickly throat and coughing). Categorical variables in the data set were defined beforehand and coded logically according to their potential contribution to the factor or individual susceptibility from lower to higher (e.g. male: 0 and female: 1). Multiple logistic regression (enter mode) models were developed and odds ratios (ORs) and 95% confidence intervals (CIs) were used to evaluate risk factors associated with those three-symptom classes. Fitted regression models were presented in the table after removing the insignificant factors with p > 0.5 from the models. Finally, statistically significant models were obtained for general, mucosal, and dermal SBS symptoms for three seasons and only variables with p < 0.5 remained in the models.

Values with p < 0.05 were considered as statistically significant for all tests.

Results and discussion

Meteorological parameters

Meteorological parameters such as temperature, relative humidity (RH), wind speed and cloudiness at the time of sampling periods are given in Table 1. The potential relationships of these meteorological factors with measured indoor TBC levels were researched by Spearman’s rank correlation test (see Table 2). RH was found to be positively associated with indoor TBC levels during winter and spring, while wind speed was found to be negatively associated with TBC levels measured in autumn and winter periods. Also, associations between outdoor TBC levels and the meteorological factors were analysed by repeated-measure ANOVA (see Table 3). As can be seen from Table 3, outdoor TBC levels varied on a seasonal basis and there were statistically significant differences between outdoor TBC levels and the meteorological parameters of temperature, RH, wind speed and cloudiness (p < 0.001).

Table 1.

Meteorological parameters (M ± SD) measured during the study.

Period^a	Temperature (℃)	Wind speed (m s⁻¹)^b	Cloudiness^c	RH (%)
Autumn	23.6 ± 0.6	5.1 ± 1.6	2.3 ± 1.4	79.3 ± 7.9
Winter	7.8 ± 2.5	4.0 ± 1.9	4.0 ± 2.3	84.6 ± 7.3
Spring	11.7 ± 1.8	3.5 ± 2.5	3.6 ± 2.9	63.3 ± 9.4

Note: RH, relative humidity; M, mean; SD, standard deviation.

In each period, number of sampling days = 10.

Wind speed at 10 m above sea level (m s⁻¹).

Cloudiness ranges between 0 (no cloud) and 8 (completely covered by clouds).

Table 2.

Spearman’s rank correlations^a among the meteorological parameters and indoor TBC levels.

	Meteorological parameters
	Cloudiness	RH	Wind speed	Temperature
TBC^autumn	−0.17 (0.120)	−0.10 (0.340)	−0.26 (0.015)	−0.11 (0.330)
TBC^winter	0.01 (0.980)	0.48 (0.001)	−0.30 (0.005)	−0.11 (0.300)
TBC^spring	−0.07 (0.540)	0.29 (0.008)	−0.02 (0.870)	−0.14 (0.200)

Note: RH, relative humidity; TBC, total bacteria count.

p Values were shown in the brackets, and n = 90 for each season.

Table 3.

Results of repeated-measure ANOVA test among the outdoor TBC levels and the meteorological factors.^a

Factor	p Value
Season	0.001
Temperature	0.001
RH	0.001
Wind speed	0.001
Cloudiness	0.001

Note: ANOVA, analysis of variance; RH, relative humidity; TBC, total bacteria count.

n = 270.

Characteristics of the sampling sites

Descriptive information regarding the sampling sites was recorded during the measurements. Table 4 shows several environmental/building-related factors that may affect IAQ (i.e. heating/cooking fuels, distance to traffic and ventilation system) and potential airborne bacteria sources in indoor environments (i.e. presence of pets and plants, and densities of carpet and pressed wood products indoors). In brief, the environments shared the following characteristics: the most prevalent heating fuels were natural gas and coal; cooking fuel was butane–propane; traffic density was low or moderate; variable number of occupants indoors; low incidence of pets and pot plants indoors; variable carpet and pressed wood products; buildings mostly built >2 years ago; half of the sites were renovated within the last year and were generally naturally ventilated environments. Also, the frequency histogram of the number of occupants in the sampling sites is given in Figure 2.

Figure 2.

Frequency histogram for the number of people (n = 90).

Table 4.

Building factors of the sampling sites.^a

Parameter	Frequency (%)
Carpet covering
No carpet	20.2
Small area of carpet	40.4
Moderate area of carpet	31.5
Large area of carpet	7.9
Heating fuel
Natural gas	50.6
Coal	42.4
Butane–propane	7.2
Cooking fuel
Butane–propane	49.3
Natural gas	29.6
Electricity	21.1
Distance to traffic
Far from traffic	41.1
Moderate distance to traffic	53.3
Close to traffic	5.6
Pressed wood products
Small amount	21.3
Moderate	48.3
Large amount	30.4
Ventilation system
Natural	88.0
AC	1.2
Natural + AC	10.8
Age of the building
≤2	20.5
>2–10	41.0
>10	38.5
Floor of the building
Ground floor	15.7
First floor	28.9
Second floor	25.3
>Second floor	30.1
Last floor covering time (year)
≤2	27.3
>2–10	56.0
>10	16.7
Last wall painting time (year)
≤1	50.0
>1–2	20.3
>2	29.7
Last renovation time (year)
≤1	54.3
>1–2	16.7
>2	28.8
Presence of pets	19.6
Presence of plants	6.7

n = 90.

Statistically significant differences in TBC levels were found by the repeated ANOVA tests (see Table 5) for the number of people indoors, types of cooking and heating fuels, type and age of the building, and last floor covering, wall painting and renovation times (p < 0.05). In general, the lowest TBC levels were observed in sampling sites occupied by fewer people (i.e. ≤4 person), whereas the highest TBC levels were observed in densely occupied sites (e.g. kindergartens) throughout the heating periods (Figure 3). TBC levels increased and varied widely when the number of people indoors was above 4, regardless of the sampling period.

Figure 3.

TBC levels^a (CFU/m³) vs. number of people indoors for (a) Autumn, (b) Winter and (c) Spring periods. ^aBoxes show 25th–75th percentiles, lower and upper bars show the ranges, lines inside the boxes show the median values, symbols of ‘+’ inside the boxes represent the mean values and ‘□’ designate the outliers.

Table 5.

Results of repeated-measure ANOVA test among indoor TBC levels and the building-related factors.^a

Factor	p Value
Number of people indoors	0.001
Type of building (home/school/ kindergarten/dorm)	0.001
Age of the building	0.001
Last floor covering time	0.001
Last wall painting time	0.001
Last renovation time	0.001
Type of heating fuel	0.002
Type of cooking fuel	0.002
Amount of pressed woods	0.007
Amount of carpet	0.079
Type of ventilation (natural or Air Conditioning)	0.093

Note: ANOVA, analysis of variance; TBC, total bacteria count.

n = 270.

In terms of cooking fuel, TBC levels varied in the sampling sites with electrical systems, natural gas and butane–propane gas appliances, in ascending order. As can be seen in Figure 4, the highest TBC levels were found in the sampling sites where coal is used as a heating fuel, particularly in the winter period. In Çanakkale, coal used to be the main heating fuel. Natural gas has become the main heating fuel in most metropolitan areas of Turkey over the last decade. The residential and commercial natural gas subscription rate in Çanakkale has slightly exceeded 50% since 2008. Also, TBC levels were close to each other in measured sampling sites before the heating period, regardless of the type of the heating fuel used. A remarkable increase in TBC levels for the sampling sites using coal was especially noticeable throughout the heating season. Thus, we can conclude that heating activity, as well as the type of heating fuel, was associated with the TBC levels at the sampling sites.

Figure 4.

TBC levels (CFU/m³) vs. type of heating fuel for (a) Autumn, (b) Winter and (c) Spring periods.

Airborne bacteria levels at the sampling sites

TBC levels showed neither normal nor lognormal distributions for all seasons (better uniform). Both indoor (<7 to >1400 CFU m⁻³) and outdoor (<7 to 1054 CFU m⁻³) TBC levels varied over a large scale throughout the study (see Table 6).

Table 6.

TBC levels^a (CFU m⁻³) and I/O ratios.^a

Type of the building	Variable	TBC levels
Type of the building	Variable	Autumn	Winter	Spring
Kindergartens (n = 10)	I	745 (63–1400)	1400 (1337–1400)	1400 (882–1400)
	O	156 (105–284)	88 (28–473)	149 (21–574)
	I/O ratio	4.9 (0.5–13)	16 (1–50)	12 (2–66)
Homes (n = 33)	I	448 (<7–1400)	364 (<7–1253)	385 (28–1400)
	O	305 (74–602)	112 (<7–133)	130 (18–322)
	I/O ratio	1 (0.1–12)	4 (0.1–25)	3 (0.2–16)
Dorms (n = 37)	I	490 (49–1400)	630 (<7–1400)	357 (14–1400)
	O	151 (77–1054)	105 (<7–784)	203 (123–319)
	I/O ratio	2 (0.2–9)	4 (0.5–185)	1 (0.1–11)
Schools (n = 10)	I	357 (168–518)	67 (<7–266)	441 (189–1113)
	O	175 (147–305)	10 (<7–42)	119 (25–137)
	I/O ratio	1 (0.7–3)	3 (0.7–19)	6.5 (2–20)
All (n = 90)	I	448 (<7–1400)	434 (<7–1400)	420 (14–1400)
	O	165 (74–1054)	90 (<7–784)	140 (18–574)
	I/O ratio	2.3 (0.1–13)	4.3 (0.1–185)	2.9 (0.1–67)

Note: CFU, colony-forming unit; I/O, indoor to outdoor; TBC, total bacteria count.

Median (min–max) values.

Indoor TBC levels, observed during autumn, winter and spring periods varied on a large scale among the sampling site groups and the period. Before the heating season, the highest indoor TBC levels were seen in kindergartens within a broad range (10² to >1400 CFU m⁻³). As Fanger³¹ pointed out, IAQ in classrooms is generally poor due to inadequate ventilation, which resulted in elevated TBC levels especially during the heating season of this study. Since the occupancy rate of kindergartens (approximately 15 children in one class) is higher than other sampling groups, those elevated TBC levels occurring in kindergartens are most likely related to the elevated number of people indoors. Neither air-conditioning (AC) systems nor mechanical ventilation equipment was available in any of the kindergartens. In addition, TBC levels in kindergartens ranged in a broad scale with a median value of approximately 745 CFU m⁻³ during autumn, whereas TBC levels were found to be at the maximum level in the winter and spring due to lack of ventilation to save energy for heating purposes. Thus, occupancy rates together with inadequate fresh air intake can induce the promotion of bacteria propagation in such indoor environments (see also Figure 3). In another study, an increasing effect of occupancy in TBC levels was found in kindergartens.⁸ Qian et al.³² also observed a strong peak in bacterial genomes during occupancy. TBC levels in dorms and homes were similar (average: 500 CFU m⁻³), while the lowest TBC levels were seen in the schools (in university classes). The reason why indoor TBC levels in the schools were found to be the lowest among the sampling sites was that the classrooms were not occupied during the sampling. Note that airborne bacteria samplings in the classrooms of the schools were conducted right after the courses during the autumn and the spring campaigns, while airborne samples in all of the schools were collected before the courses during the winter campaign.

In the earlier phase of the heating season (winter), TBC levels observed at the kindergartens and dorms increased. In winter, TBC levels observed in the kindergartens showed no variation and were always found to be higher than 1400 CFU m⁻³, which was upper detection limit of airborne bacteria concentration for this study. The range of TBC levels in the dorms for the winter period was broader than those found before the heating season. Similar to autumn results, the lowest TBC levels in the winter period were seen in the schools. TBC levels measured in the kindergartens were again higher than 1400 CFU m⁻³, during the middle of the heating season (spring). TBC levels in the homes showed similar distribution for both winter and spring periods. TBC levels found in this study are similar to those found in other studies conducted in environments without complaints, including 300 CFU m⁻³ in offices,³³ 10–10³ CFU m⁻³ in high-rise buildings in Korea,³⁴ 401–743 CFU m⁻³ in Hong Kong markets,²⁵ 600–1800 CFU m⁻³ in Hong Kong shopping malls,³⁵ 10–10³ CFU m⁻³ in bars, cafeterias, classrooms and homes in Korea,³⁶ and 10²–10³ CFU m⁻³ in university rooms.³⁷

Since meteorological conditions have substantial effects on airborne bacteria levels,³⁸ investigating the effect of heating activity by comparing the TBC values before and during the heating season should be helpful to evaluate the sampling results accurately. The most remarkable change in TBC levels (i.e. to >1400 CFU m⁻³) during the heating season was observed in kindergartens (p < 0.05) (see Table 6). Since most kindergarten children in Turkey are under the age of 6, kindergartens are crucial buildings for child development, including the development of their immune systems. Children are prone to acquire some illnesses, while environmental conditions in the kindergartens can trigger some already existing diseases. Since concerns about the increasing worldwide prevalence of childhood asthma have been rising lately, sustaining ‘healthy’ IAQ in kindergartens is crucial.³⁹

Obviously ventilation decreases during the heating season, especially in naturally ventilated spaces. Ventilation frequency here reaches a maximum in July–August. On the other hand, the chosen sampling period before the heating season was October, which was neither too hot to ventilate indoors nor too cold to close windows like people usually do in January. Thus, observed TBC levels did not primarily originate from the ventilation frequency, since extreme differences in ventilation frequencies were not expected.

Indoor to outdoor (I/O) ratios of TBC values were also calculated. If the I/O ratio is >1.0, there is a difference between outdoor and indoor bacteria sources and the source exists in the indoor environment. I/O ratios of TBC levels at the sampling sites according to the building types (i.e. kindergarten, home, dorm and school) were calculated in each sampling period (see Table 6). As can be seen from the table, the median values of I/O ratios for the four building types were found to be ≥1.0 during the study. I/O ratios during the heating periods (winter and spring) were found to be higher than those observed before the heating season (autumn). Among the buildings monitored in this study, the highest I/O ratios were observed in kindergartens for each period.

Table 7 indicates that unlike indoor TBC levels, outdoor TBC levels showed statistically significant seasonal differences according to paired t-test and Wilcoxon signed rank (p ≤ 0.01). Table 6 also shows outdoor TBC levels measured at the sampling sites. Median values of TBC values measured indoors were found to be higher than outdoors at each sampling point, regardless of the season. As can also be seen from Table 7, indoor and outdoor TBC levels measured in three seasons were found to be significantly different according to the results of paired t-test and Wilcoxon signed rank test (p < 0.05). Thus, it can clearly be concluded that the major sources of indoor bacteria might be any source or activity done in those environments. Reponen et al.⁴⁰ specified that indoor bacteria levels were not correlated with outdoor bacteria values, as the most important source of bacteria was occupancy.^8,33,41 Indoor bacteria levels were found to be higher than outdoor levels in a study conducted in homes, offices, schools, shopping malls and restaurants in Hong Kong,⁵ in Athens and Chania,⁷ and in Ankara, Turkey.^8,19,20

Table 7.

Paired sample comparisons at 95% confidence level.

Variable pairs		Mean difference ± SD^a	z^b	p ^b
TBC_autumn-in	TBC_winter-in	−55.7 ± 505.6	−0.36^c	0.720
TBC_autumn-in	TBC_spring-in	−81.6 ± 555.9	−0.95^d	0.330
TBC_winter-in	TBC_spring-in	−25.9 ± 443.3	−0.88^d	0.380
TBC_autumn-out	TBC_winter-out	141.1 ± 267.0	−4.94^c	0.001
TBC_autumn-out	TBC_spring-out	90.7 ± 230.9	−2.55^c	0.010
TBC_winter-out	TBC_spring-out	−50.3 ± 139.7	−4.23^d	0.001
TBC_autumn-in	TBC_autumn-out	279.7 ± 391.4	−5.83^c	0.001
TBC_winter-in	TBC_winter-out	476.4 ± 457.6	−7.69^c	0.001
TBC_spring-in	TBC_spring-out	452.0 ± 476.4	−7.22^c	0.001

Note: SD, standard deviation; TBC, total bacteria count.

Paired t-test was applied.

Wilcoxon signed rank test was applied.

z Value based on possible positive ranks.

z Value based on possible negative ranks.

Personal factors and observation frequency of SBS symptoms

Table 8 shows the descriptive information regarding the occupants. In brief, half of the participants were younger than 25 (M ± SD = 35 ± 16); two thirds of the participants were female; approximately half of the participants were employees; two thirds of the participants were university students or had a college degree; and daily time spent in indoors was on average approximately 14 h. Since the introduction of the revised Law on Prevention and Control of Hazards of Tobacco Products in 2008,⁴² smoking is banned in all closed environments except homes and smoking habits in indoor spaces decreased drastically. For this reason, the question asked was ‘do you smoke in this indoor environment’? not ‘are you a smoker’? Thus, the calculated smoker ratio given in Table 8 shows the participants who smoked indoors (which can also be called the environmental tobacco smoke, ETS, exposure). More than half smoked <10 cigarettes daily, whereas the others smoked more than 10 cigarettes per day. A similar survey, conducted in 2011 in central and in Çan, a rural province, found the smoking rate to be higher than 50%.³⁰ Hence, we can conclude that the smoking prohibition law and increasing public health concerns related to smoking might have helped to lower this rate.

Table 8.

Characteristics of the occupants.^a

Parameter	Frequency (%)
Age
<25	48.3
25–40	17.2
>40	34.5
Gender
Female	70.5
Male	29.5
Employment status
Unemployed	53.4
Employed	30.7
Retired	15.9
Educational level
Primary school	15.1
High school	18.6
University	66.3
Time spent indoors (h)
0–10	29.9
>10–16	43.7
>16	26.4
Number of people who smoke indoors
0	72.7
1	13.6
2	0
>2	13.7
Total daily cigarette consumption indoors
0	72.7
2–10	15.9
>10	11.4

n = 90.

Weekly observation frequencies of SBS symptoms among the occupants with no known/diagnosed related disease/diseases (‘healthy participants’) are shown in Table 9. None of the participants were already diagnosed for asthma. Eighty percent of the participants had no existing disease, which may have symptoms similar to SBS (i.e. healthy participants). The recorded diseases of unhealthy participants (accounting for 20% of the participants) were mostly bronchitis, sinusitis and migraine. Among the healthy participants, fatigue, headache and coughing were the most frequently occurring SBS symptoms. When the prevalence of SBS symptoms was compared seasonally, assuming that autumn was a reference season, prevalence of all SBS symptoms increased in the spring with the exception of headache.

Table 9.

Prevalance (%) of Weekly SBS Symptoms among Healthy Participants

Symptom	Frequency (%)
Symptom	Autumn^a	Winter^b	Spring^c
Fatigue	44.3	33.8	43.1
Headache	30.0	27.9	23.6
Coughing	18.6	23.5	22.2
Eye irritation	20.0	14.7	23.6
Sneezing	18.6	19.1	22.2
Tickly throat	11.4	14.7	19.4
Dizziness	8.6	16.2	16.7
Dry skin	10.0	13.2	12.5

Note: for healthy subjects ^an = 70, ^b68, and ^c72 for the autumn, winter, and spring campaigns, respectively; and ^a18, ^b16, and ^c18 subjects with a disease that could cause at least one of the symptoms were excluded for the autumn, winter and spring campaigns, respectively.

Prevalence of the number of observed SBS symptoms among the healthy participants was also examined. Almost one third of the healthy participants did not suffer from any of the SBS symptoms (i.e. control group of the healthy participants), while the remaining healthy participants suffered from at least one of the SBS symptoms (i.e. case group of the healthy participants).

As mentioned in the ‘Statistical analyses’ section, SBS symptoms were categorized into three groups to apply multiple logistic regression models. For each subgroup of SBS symptoms, multiple logistic regression models were used for each of the three sampling periods as well. The factors (p < 0.2) used as input for the models were gathered from bivariate analyses, given in Table 10. Statistically significant models were obtained for general, mucosal and dermal SBS symptoms for three seasons (see Table 11) as mentioned in Statistical analyses section. Only factors that have ORs > 1 and p < 0.05 were taken into account as statistically significant. It is obvious that the associations are not the same for each season, indicating that there is a lot of effect from modification of environmental associations related to season.

Table 10.

Bivariate analysis results between the SBS symptoms and the building/occupant-related factors.^a

Symptom category	Autumn (n = 88)	Winter (n = 84)	Spring (n = 90)
General symptoms	TBC_autumn-out (0.02) Type of building (0.01) Distance to traffic (0.01) Age of the building (0.01) Last floor covering time (0.02) Amount of carpet (0.16) Amount of pressed woods (0.09) Number of people indoors (0.01) Gender (0.07) Age of the occupant (0.07) Education level of the occupant (0.04) Employment status (0.14)	Type of building (0.02) Floor of the building (0.13) Distance to traffic (0.00) Age of the building (0.01) Last floor covering time (0.01) Amount of pressed woods (0.01) Last wall painting time (0.14) Last renovation time (0.13) Type of heating fuel (0.08) Number of people indoors (0.06) Daily time spent indoors (0.04) Gender (0.01) Age of the occupant (0.08) Education level of the occupant (0.02)	TBC_spring-in (0.13) Distance to traffic (0.01) Age of the building (0.00) Type of ventilation (0.18) Last floor covering time (0.01) Amount of pressed woods (0.01) Last wall painting time (0.01) Last renovation time (0.01) Type of cooking fuel (0.07) Presence of plants (0.02) Number of people who smoke indoors (0.14) Daily cigarette consumption indoors (0.12) Gender (0.01) Age of the occupant (0.02) Employment status (0.02)
Dermal symptoms	TBC_autumn-in (0.04) Floor of the building (0.02) Age of the building (0.01) Type of cooking fuel (0.01) Amount of carpet (0.16) Type of ventilation (0.01) Number of people indoors (0.06) Daily time spent indoors (0.14)	TBC_winter-out (0.04) Type of building (0.01) Floor of the building (0.01) Age of the building (0.01) Type of ventilation (0.08) Last floor covering time (0.02) Last wall painting time (0.19) Amount of pressed woods (0.02) Number of people indoors (0.02) Age of the occupant (0.02) Gender (0.16) Daily time spent indoors (0.12) Employment status (0.10)	TBC_spring-out (0.02) Type of building (0.13) Age of the building (0.01) Type of ventilation (0.01) Type of cooking fuel (0.13) Presence of pets (0.17) Amount of pressed woods (0.10) Number of people indoors (0.01) Gender (0.13)
Mucosal symptoms	TBC_autumn-in (0.01) TBC_autumn-out (0.11) Type of building (0.01) Distance to traffic (0.20) Amount of pressed woods (0.06) Type of heating fuel (0.15) Number of people indoors (0.20) Number of people who smoke indoors (0.01) Daily cigarette consumption indoors (0.01) Education level of the occupant (0.18) Employment status (0.01)	Type of building (0.01) Distance to traffic (0.04) Type of ventilation (0.14) Amount of pressed woods (0.19) Last renovation time (0.05) Number of people who smoke indoors (0.07) Daily cigarette consumption indoors (0.04) Gender (0.13) Employment status (0.06)	TBC_spring-in (0.14) TBC_spring-out (0.01) Type of building (0.16) Distance to traffic (0.04) Age of the building (0.01) Floor of the building (0.05) Type of ventilation (0.15) Type of cooking fuel (0.07) Last floor covering time (0.01) Last wall painting time (0.02) Last renovation time (0.03) Number of people indoors (0.03) Age of the occupant (0.01)

Note: TBC, total bacteria count; SBS, sick building syndrome.

p Values are given in the brackets and personal factors are shown in italics.

Table 11.

Multiple Logistic Regression Models between the SBS Symptoms and the Building/Personal Related Factors^a

Symptom category	Autumn (n = 88)		Winter (n = 84)		Spring (n = 90)
Symptom category	Variable	ORs (95% CI)	Variable	ORs (95% CI)	Variable	ORs (95% CI)
General Symptoms	TBC_autumn-out^b Gender Age of the occupant Amount of carpet covering	1.24 (0.96–1.59)^^ 4.50 (1.24–16.29)^* 0.95 (0.92–0.99)* 2.33 (1.08–5.04)*	Floor of the building Age of the building Last floor covering time Daily time spent indoors Gender	0.51 (0.27–0.97)^* 0.90 (0.75–1.06) 1.28 (0.93–1.75)^*** 1.37 (1.13–1.67)^* 7.94 (0.94–66.80)^**	Daily time spent indoors Number of people who smoke indoors Age of the occupant TBC_spring-in^b	1.16 (1.04–1.31)^* 1.86 (0.74–4.65)^*** 0.95 (0.91–0.99)^* 1.15 (1.01–1.31)^*
Dermal Symptoms	Daily time spent indoors TBC_autumn-in^b Floor of the building	0.93 (0.78–1.12) 1.14 (0.90–1.44)^*** 1.32 (0.73–2.39)	Floor of the building Daily time spent indoors Number of people indoors Age of the occupant	0.79 (0.42–1.50) 0.91 (0.76–1.09) 2.08 (1.02–4.26)^* 0.97 (0.90–1.05)	Amount of pressed woods Number of people indoors Gender	0.29 (0.06–1.17)^* 1.62 (0.89–2.95)^* 6.20 (0.70–55.19)^***
Mucosal Symptoms	TBC_autumn-in^b Type of sampling site Distance to traffic Type of heating fuel Daily cigarette consumption indoors	0.83 (0.67–1.03)^** 0.59 (0.23–1.50) 7.70 (1.46–40.71)^* 2.57 (1.23–5.37)^* 1.12 (1.01–1.24)^*	Distance to traffic Amount of pressed woods Last renovation time Daily cigarette consumption indoors	3.09 (0.62–15.32)^* 9.20 (0.45–187.61)^* 0.62 (0.32–1.19)^*** 1.12 (1.02–1.23)^*	TBC_spring-in^b Floor of the building Last wall painting time Type of cooking fuel Number of people indoors TBC_spring-out^b	0.97 (0.94–1.00)^* 0.65 (0.35–1.21)^*** 2.66 (1.12–6.33)^* 2.84 (0.96–8.41)^ 1.97 (0.82–4.74)^* 1.86 (0.66–5.20)

ORs: Odds ratios; 95% Confidence Interval (CI); statistically significant factors (p < 0.05) with OR > 1.0 were underlined and with OR < 1.0 were shown in Italics.

All variables with p < 0.2, gathered from the bivariate analysis results for each season and each symptom category, were included in the logistic models and then variables with p < 0.5 remained in the final models. ^bTBC values were included in the model as “Concentration/100”; for other exposure variables ORs were expressed per one unit change.

p <0.05; ^**p<0.1; ^***p<0.2.

Gender was found to be associated with general symptoms observed in autumn (ORs = 4.50), indicating that female occupants suffered from the general symptoms of SBS at least 4.5 times more than male occupants. Amount of carpet covering was related with the general SBS symptoms observed in the autumn (ORs = 2.33). Daily time spent (h) by the occupants indoors was associated with general symptoms of SBS both in the winter (ORs = 1.37) and spring (ORs = 1.16). TBC levels measured in the spring were found to be associated with general SBS symptoms (ORs = 1.15). A 100 CFU m⁻³ increment in TBC levels increased the odds of general SBS symptoms by 15%. Besides, some of the factors with ORs < 1.0 and p < 0.05 were found to be ‘preventive’ for the occurrence of the general SBS symptoms, which are the age of the occupant and floor number of the sampling site. Thus, the probability of observation of general SBS symptoms among young occupants was less than for old ones. Also, as the floor of the site increases, occurrence of general SBS symptoms decreases. When the height of the site is close to ground level, it is possible to be exposed to pollutants coming from traffic sources or due to the resuspension of particles or bioaerosols from the soil.

The number of people indoors was found to be associated with dermal symptoms occurring in the winter (ORs = 2.08). One person increment of the occupants at the sampling sites resulted in 2.08 times greater odds of having dermal SBS symptoms. Since climatic conditions indoors are more comfortable for people than staying outdoors during the winter period, people mostly prefer to spend their time indoors.

Daily cigarette consumption indoors (i.e. ETS exposure) were found to be associated with the mucosal SBS symptoms occurring both in the autumn and winter (ORs = 1.12). Every single cigarette consumed resulted in a 12% increase in the odds of having mucosal SBS symptoms. Distance of the sampling site to traffic was found to be associated with exposure to mucosal SBS symptoms in the autumn with high odds (ORs = 7.70). Distance of sampling sites to traffic sources was classified as located ‘on the street’ and ‘on the/close to highway’. Living close to high traffic sources resulted in 7.7 times odds of having mucosal SBS symptoms. Proximity to traffic sources can result in exposure to other pollutants such as particles and organic or oxygenated compounds,^43,44 which can also induce SBS symptoms. Also, type of heating fuel of the sampling site (ORs = 2.57) was associated with mucosal SBS symptoms observed in the autumn. In terms of heating fuel types, natural gas was assumed to be the reference fuel, while butane–propane gas and coal were assumed to be the categories. When the other fuels were used as heating source, the probability of having mucosal SBS symptoms increased 2.57 times compared to locations using natural gas. Finally, it was found that as the time since last wall painting increases, the probability of occurrence of mucosal symptoms increased during the spring (ORs = 2.66). Since microbiological pollutants can grow on the walls,⁴⁵ environmental factors such as temperature and humidity might have increased the exposure to biological pollutants over time.

According to the ANOVA tests, some of the individual SBS symptoms showed dose–response relationships with measured TBC levels (p < 0.05). They are fatigue, headache and dry skin. TBC levels were found to be higher when the participant had these problems. Harrison et al.¹⁸ also found positive associations between symptom prevalence rates and levels of airborne viable bacteria in buildings with similar ventilation systems.

The occurrence of SBS symptoms according to gender was also researched by Spearman’s rank correlation test. Similar to outcomes gathered from the multiple logistic regression models, significant associations were found for the occurrence of the general SBS symptoms (Spearman’s R = 0.39, p = 0.002) observed in female participants during the winter and spring seasons. As stated, females had a higher prevalence of SBS than males,^46–53 probably due to calculated higher time spent indoors in this study (M ± SD; for females = 17.4 ± 5.8 h and males = 10.7 ± 4.4 h) and having higher sensitivity than men.

The prevalence of the SBS symptoms among different age groups was also studied. Since the sampling sites were selected randomly, a relatively high ratio of young participants (1/2 of the participants were <25, especially in dorms and schools) was included to the study. General and mucosal SBS symptoms (Spearman’s R = −0.33, p = 0.06) were related to the ages of the occupants throughout the study period. Norlén and Andersson⁵⁴ pointed out that the relationship between age and SBS seemed to be rather complex; in addition several studies^55,56 showed no relationship between age and SBS. Links between SBS symptoms and the age of the occupants were considered here in spite of some confounding factors. However, it should be noted that younger people may suffer less from any related disease for several reasons: a strong immune system response, having fewer symptoms of any unknown disease and acquiring a new disease less often. Also, elderly people are more susceptible to diseases. On the other hand, even though we asked the participants whether they had any related disease that can induce SBS-like symptoms, it might be possible that they already acquired a disease but they were unaware of it.

Conclusions

This study has shown that indoor TBC levels were found to be higher in the sampling sites where coal is used as a heating source during the heating season (p < 0.05).

Statistically significant differences were found between TBC levels measured indoors and outdoors (p < 0.001). Additionally, calculated I/O ratios were found to be >1.0, indicating bacteria proliferation due to indoor sources.

Outdoor TBC levels had associations with the meteorological factors of wind speed, RH, temperature and cloudiness (p < 0.001).

The most prominent difference in indoor TBC levels was seen between the kindergartens and the schools. This result can be explained by two relationships: (i) bacteria levels were strongly correlated with occupancy; this was verified by a dose–response relationship between the TBC levels and occupancy rate, and (ii) lower outdoor TBC levels indicate that indoor bacteria originates from indoor sources.

Indoor TBC levels showed statistically significant relationships among the building factors of occupancy rate, type and age of the building, last floor/wall painting/renovation times and type of heating/cooking fuels (p < 0.05).

80% of the participants had no existing disease/diseases, which may induce similar symptoms to SBS (i.e. healthy participants). Almost 30% of the healthy participants did not suffer from any SBS symptoms. Among the healthy participants, fatigue, headache and coughing were the most frequently occurring SBS – symptoms.

Multiple logistic models showed that season has an obvious effect on associations between the environmental conditions and occurring SBS symptoms.

Multiple logistic models also showed that indoor TBC levels, personal factors (gender and daily time spent indoors) and building-related factors (occupancy rate, ETS exposure, distance to traffic, type of heating fuel, amount of carpet covering and last wall painting time) were found to be associated with at least one group of the SBS symptoms.

The general SBS symptoms (headache, fatigue and dizziness) occurred significantly in female participants, probably due to more time spent indoors and higher sensitivity than men.

The study also has some limitations that should be taken into account, particularly when conducting further studies:

While estimating SBS symptoms among people, several points should be taken into consideration: participants should be in the age ranges from 18–25 to 50–60, depending on the average life expectancy and socioeconomic status of the country involved.

Finally, this study was a small study and associations between air pollutants (i.e. TBC and other pollutants) and SBS symptoms should be investigated in a larger study population, including more than one person at each site to minimise selection bias.

Footnotes

Authors’ contribution

Sibel Mentese was responsible for all steps of the study and the evaluation and publication of the study results. Deniz Tasdibi was responsible for the application of the questionnaires.

Acknowledgements

The authors thank Env. Eng. students Tuğba Böce, Melis B. Mutlu, Saime S. Özdemirpençe, Salih Y. Nişancı, Elif Palaz, Burak Çetin, Burak Selçuk and Sevil Karagöz for their help during the field studies; Dr. Margaret Hawthorne and Catherine Yigit for reading the manuscript; Dr. Coskun Bakar and Dr. Sibel Cevizci for the recommendation about the logistic regression, and the occupants of the sampling points for their patience.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This study was supported by Çanakkale Onsekiz Mart University Scientific Research Center. Project no: 2010/168.

Appendix – Questionnaire

References

Goyer N, Lavoie J, Lazure L, Marchand G, Allard R and Bhérer L. Bioaerosols in the workplace: evaluation, control and prevention guide: studies and research projects. Technical Guide T-24. Québec: Institut de recherché Robert-Sauvé en santé et en sécurité du travail du Québec, 2001, p. 94.

Ross

Curtis

Scheff

Hryhorczuk

Ramakrishnan

Wadden

Persky

. Association of asthma symptoms and severity with indoor bioaerosols. Allergy 2000; 55: 705–711.

Stetzenbach

Buttner

Cruz

. Detection and enumeration of airborne biocontaminants. Curr Opin Biotech 2004; 15: 170–174.

Lee

Grinshpun

Martuzevicius

Adhikari

Crawford

Luo

Reponen

. Relationship between indoor and outdoor bioaerosols collected with a button inhalable aerosol sampler in urban homes. Indoor Air 2006; 16: 37–47.

Lee

Guo

Chan

. Inter-comparison of air pollutant concentrations in different indoor environments in Hong Kong. Atmos Environ 2002; 36: 1929–1940.

Green

Scarpino

Gibbs

. Assessment and modeling of indoor fungal and bacterial bioaerosol concentrations. Aerobiologia 2003; 19: 159–169.

Kalogerakis

Paschali

Lekaditis

Pantidou

Eleftheriadis

Lazaridis

. Indoor air quality – bioaerosol measurements in domestic and office premises. J Aerosol Sci 2005; 36: 751–761.

Mentese

Arisoy

Rad

Güllü

. Bacteria and fungi levels in various indoor and outdoor environments in Ankara, Turkey. CLEAN-Soil Air Water 2009; 37(6): 487–493.

World Health Organisation (WHO). Indoor air pollutants: exposure and health effects. EURO Reports and Studies, No. 78. Copenhagen: WHO, 1983.

10.

Passarelli

. Sick building syndrome: an overview to raise awareness. J Build Appraisal 2009; 5: 55–56.

11.

Burge

. Sick building syndrome. Occup Environ Med 2004; 61: 185–190.

12.

American Conference of Governmental Industrial Hygienists (ACGIH). Bioaerosol: assessment and control, 1st ed. Cincinnati, OH: ACGIH, 1999.

13.

Carrer

Maroni

Alcini

Cavallo

. Allergens in indoor air: environmental assessment and health effects. Sci Total Environ 2001; 270: 33–42.

14.

Douwes

Thorne

Pearce

Heederik

. Bioaerosol health effects and exposure assessment: progress and prospects. Ann Occup Hyg 2003; 47: 187–200.

15.

Redman

Hamilton

Malloch

Kleymann

. Working here makes me sick! The consequences of sick building syndrome. Human Resource Manag J 2011; 21: 14–27.

16.

Runeson

Wahlstedt

Wieslander

Norback

. Personal and psychosocial factors and symptoms compatible with sick building syndrome in the Swedish workforce. Indoor Air 2006; 16: 445–453.

17.

Hsu

Tai

. Indoor pollution and sick building syndrome symptoms among workers in day-care centers. Arch Environ Health Int J 1997; 52: 200–207.

18.

Harrison

Pickering

Faragher

Austwick

Little

Lawton

. An investigation of the relationship between microbial and particulate indoor air pollution and the sick building syndrome. Respiratory Med 1992; 86: 225–235.

19.

Mentese

Rad

Arısoy

Güllü

. Seasonal and spatial variations of bioaerosols in indoor urban environments, Ankara, Turkey. Indoor Built Environ 2012; 21: 797–810.

20.

Mentese

Rad

Arısoy

Güllü

. Multiple comparisons of organic, microbial, and fine particulate pollutants in typical indoor environments: diurnal and seasonal variations. J Air Waste Manag Assoc 2012; 62: 1380–1393.

21.

Andersen

. New sampler for the collection, sizing, and enumeration of viable airborne particles. J Bacteriol 1958; 76: 471–478.

22.

National Institute for Occupational Safety and Health (NIOSH). Method 0800 – bioaerosol sampling (indoor air), culturable organisms: bacteria, fungi, thermophilic actinomycetes. NIOSH, 1998.

23.

Saldanha

Manno

Saleh

Ewaze

Scott

. The influence of sampling duration on recovery of culturable fungi using the Andersen N6 and RCS bioaerosol samplers. Indoor Air 2008; 18: 464–472.

24.

Macher

Chatigny

Burge

Air sampling instruments for evaluation of atmospheric contaminants. In: Cohen

Hering

(eds). American Conference of Governmental Industrial Hygienists, Cincinnati, OH: AIHA Inc, 1995, pp. 279–321.

25.

Guo

Lee

Chan

. Indoor air quality investigation at air-conditioned and non-air-conditioned markets in Hong Kong. Sci Total Environ 2004; 323: 87–98.

26.

Takigawa

Saijo

Morimoto

Nakyama

Shibata

Tanaka

Yoshimura

Chikara

Kishi

. A longitudinal study of aldehydes and volatile organic compounds associated with subjective symptoms related to sick building syndrome in new dwellings in Japan. Sci Total Environ 2012; 417–418: 61–67.

27.

AuYeung

Chow

Lam

. Sick building syndrome – a case study. Build Environ 1991; 26: 319–330.

28.

Godish

. Indoor environmental quality, Boca Raton: Lewis Publishers, 2001.

29.

Norbäck

Michel

Widström

. Indoor air quality and personal factors related to the sick building syndrome. Scandin J Work Environ Health 1990; 16: 121–128.

30.

Mentese S, Mutlu M and Sen B. Comparison of indoor and ambient air quality perceptions in Can and Central Provinces of Canakkale, Turkey. In: Proceedings of Healthy Buildings, Brisbane, Australia, July 8–12 2012.

31.

Fanger

. What is IAQ? Indoor Air 2006; 16: 328–334.

32.

Qian

Hospodsky

Yamamoto

Nazaroff

Peccia

. Size resolved emission rates of airborne bacteria and fungi in an occupied classroom. Indoor Air 2012; 22: 339–351.

33.

Pastuszka

Tha Paw

Lis

Wlazlo

Ulfig

. Bacterial and fungal aerosol in indoor environment in Upper Silesia, Poland. Atmos Environ 2000; 34: 3833–3842.

34.

Lee

. Characteristics of indoor and outdoor bioaerosols at Korean high-rise apartment buildings. Environ Res 2006; 101: 11–17.

35.

Lee

Chan

. Indoor air quality at nine shopping malls in Hong Kong. Sci Total Environ 2001; 273: 27–40.

36.

Seo

. Indoor and outdoor bioaerosol levels at recreation facilities, elementary schools, and homes. Chemosphere 2005; 61: 1570–1579.

37.

Stryjakowska-Sekulska

Piotraszewska-Pajak

Szyszka

Nowicki

Filipiak

. Microbiological quality of indoor air in university rooms. Pol J Environ Stud 2007; 16: 623–632.

38.

Jones

Harrison

. The effects of meteorological factors on atmospheric bioaerosol concentrations? A review. Sci Total Environ 2004; 326: 151–180.

39.

Mentese

Are the kindergartens healthy enough for the children? In: Thompson

(ed.). Kindergartens: programs, functions and outcomes, Hauppauge, NY: Nova Science Publishers, 2010, pp. 147–161.

40.

Reponen

Nevalainen

Raunemaa

. Bioaerosol and particle mass levels and ventilation in Finnish homes. Environ Int 1989; 15: 203–208.

41.

Bonetta

Mosso

Sampò

Carraro

. Assessment of microbiological indoor air quality in an Italian office building equipped with an HVAC system. Environ Monit Assess 2010; 161: 473–483.

42.

Ministry of Health of Turkey. Law on prevention and control of hazards of tobacco products. National Law. Law no: 5727. Turkey, 3 Jan 2008.

43.

Martuzevicius

Grinshpun

Lee

Biswas

Reponen

LeMasters

. Traffic-related PM_2.5 aerosol in residential houses located near major highways: indoor versus outdoor concentrations. Atmos Environ 2008; 42: 6575–6585.

44.

Baek

Kim

Perry

. Indoor air quality in homes, offices and restaurants in Korean urban areas-indoor/outdoor relationships. Atmos Environ 1997; 31: 529–544.

45.

Altenburgera

Kämpferb

Makristathisc

Lubitza

Bussea

. Classification of bacteria isolated from a medieval wall painting. J Biotechnol 1996; 47: 39–52.

46.

Hedge

Sterling

Collett

Sterling

Nie

. Indoor air quality and health in two office buildings with different ventilation systems. Environ Int 1989; 15: 115–128.

47.

Engvall

Norrby

Norback

. Sick building syndrome in relation to building dampness in multi-family residential buildings in Stockholm. Int Arch Occup Environ Health 2001; 74: 270–278.

48.

Jaakkola

JJK

Heinonen

Seppanen

. Sick building syndrome, sensation of dryness and thermal comfort in relation to room temperature in an office building: need for individual control of temperature. Environ Int 1989; 15: 163–168.

49.

Runeson-Broberg

Norback

. Sick building syndrome (SBS) and sick house syndrome (SHS) in relation to psychosocial stress at work in the Swedish workforce. Int Arch Occup Environ Health 2013; 86: 915–922.

50.

Skov

Valbjorn

Pedersen

. Influence of personal characteristics, job related factors and psychological factors on the sick building syndrome. Scand J Work Environ Health 1989; 15: 286–295.

51.

Stenberg

Wall

. Why do women report ‘sick building symptoms’ more often than men? Soc Sci Med 1995; 40: 491–502.

52.

Brasche

Bullinger

Morfeld

Gebhardt

Bischof

. Why do women suffer from sick building syndrome more often than men? – subjective higher sensitivity versus objective causes. Indoor Air 2001; 11: 217–222.

53.

Engvall

Hult

Corner

Lampa

Norback

Emenius

. A new multiple regression model to identify multi-family houses with a high prevalence of sick building symptoms “SBS”, within the healthy sustainable house study in Stockholm (3H). Int Arch Occup Environ Health 2010; 83: 85–94.

54.

Norlén U and Andersson K. The indoor climate in the Swedish housing stock. Document. Stockholm: Swedish Council for Building Research, 1993.

55.

Bjornsson

Janson

Norback

Boman

. Symptoms related to the sick building syndrome in a general population sample: associations with atopy, bronchial hyper-responsiveness and anxiety. Int J Tuberc Lung Dis 1998; 2: 1023–1028.

56.

Norback

Edling

. Environmental, occupational, and personal factors related to the prevalence of sick building syndrome in the general population. Br J Ind Med 1991; 48: 451–468.