Since the first studies on indoor air quality (IAQ), some common indoor air pollutants such as carbon monoxide, particulate matter, volatile organic compounds (VOC) and formaldehyde have been related to acute and chronic respiratory diseases,1 contributing to indoor air pollution burden of disease estimated at 2.7%.2 In modern society, children spend most of their time indoors (up to 85%),3 at home or in children day care centers (DCC), and they are particularly vulnerable to the effects of IAQ.4

Different studies have shown that IAQ and ventilation are associated with respiratory symptoms, allergy and infections.5–7 Furthermore, it was estimated that about one million deaths per year in children under 5, could be attributed to acute respiratory infections due to IAQ.8

Scientific and political authorities have proposed limits to the levels of air pollutants through recommendations and legislation. In 2006, Portuguese legislation was passed (Decreto-Lei 89/2006) setting reference levels for some indoor pollutants.9

The design and construction of buildings have focused on mechanical resistance, durability and comfort. Economic concerns related to energy efficiency in buildings have resulted in a decrease in ventilation flow rates and practices that have led to a deterioration in IAQ. It has been suggested that these might have been partly responsible for the rise in respiratory diseases in recent decades.10–12

Despite the large number of publications about the effects of indoor air pollution in schools, there is not much evidence about the association between IAQ in DCCs and health. This was addressed in some European reviews that systematized what has been done and what research is still required.13–15

Studies focused on IAQ in DCCs are important because of their specificity and potential impact on children's health. Furthermore, as there are no clear thresholds that define the level of exposure to indoor air pollutants that affects children's health at DCCs, studies of those environments could provide additional relevant information. Day care centres studies may help identify feasible interventions that could subsequently be implemented by policy makers. It has been shown that some interventions might be effective in modulating IAQ in schools16–18 and influencing respiratory symptoms.19

Because IAQ patterns and effects vary widely between different countries,8 local and national studies are important where interventions on population health are being considered. Each population is uniquely linked to its genetic background, specificities of the available infrastructures or construction typology, cultural behavior and climatic factors. This point has been made by Canha et al.20 who showed that there was a connection between ventilation strategies and lower ventilation rates and higher CO2 levels in Portuguese elementary school classrooms, when compared with Finnish classrooms in a rural setting. Other Portuguese studies targeting schoolchildren and adolescents have reported associations between IAQ and respiratory symptoms and wheezing.21,22

Outdoor and indoor environments should be considered simultaneously when studying the IAQ effects on human health. Different cross-sectional23–25 and longitudinal26–28 studies have shown the effects of both indoor and outdoor air pollutants on children's respiratory health. Background pollutant concentrations measured at central monitoring stations, traffic density/distance to major roads29,30 or dispersion models25 have been used29,31 to study outdoor air pollutants. Exposure to indoor environments, such as schools32,33 or houses, where children spend a large amount of time, has only more recently been taken into consideration.34,35

Some Portuguese studies have also reported IAQ deficiencies both in Lisbon36,37 and Oporto.22 Pegas et al.37 reported that ventilation problems are present throughout the year and different pollutant profiles can be found in different seasons of the year.

When studying children's illnesses it might also be important to consider their effects on parents, as diseases could enhance both child and parental stress. There is clear evidence that stress induces cortisol release and this hormone exerts an inhibitory action over the immune system.38,39 On the other hand, research has shown that the mothers of children with respiratory diseases, such as asthma, have higher levels of parenting stress, compared to those with healthy children,40–42 and children whose parents experience higher levels of stress show higher vulnerability to respiratory disease.43–45 Therefore, the parent–child relationship could influence the child's hypothalamus–pituitary–adrenal cortex (HPA) axis activation and cortisol release.46 In fact, cortisol may be one of the biological links between parent–child stress and respiratory disease in childhood. For these reasons ENVIRH also included an ancillary study to assess parental symptoms (depression/anxiety), psychosocial context (parental stress, marital satisfaction, life events) and a measure of the child's HPA axis activation (24-h cortisol production).

In view of the lack of studies about IAQ in Portuguese DCCs and the public health concern about this topic, we decided to conduct the ENVIRH study (Environment and Health in Children Day Care Centres). The main goals of the ENVIRH study were: (1) to characterize the respiratory health of the children attending DCCs; (2) to describe building characteristics and performance, including ventilation in DCCs; (3) to study the IAQ in DCCs, choosing those pollutants that might be more relevant to children's respiratory diseases, according to the literature: particulate matter, volatile organic compounds, formaldehyde, carbon dioxide, fungi and bacteria; (4) to assess how parental psychological and relational dimensions, together with child cortisol release (stress biomarker), may change according to child respiratory disease; (5) to study the relationship between the previous factors; and finally (6) to suggest recommendations for the authorities, the scientific community and the general public. This paper discusses the rationale of the ENVIRH Study and describes the methodology used by our team to carry it out.

Our study was focused on DCCs. These are institutions that provide childcare while parents are at work. Besides basic care, these centers offer children a wide range of activities for their entertainment, including some educational activities.

DCCs provide care for infants, toddlers and pre-school children (up to 5–6 years old). We chose to study DCCs from the two main Portuguese cities (Lisbon and Oporto) mainly for reasons of feasibility. These are busy cities with different air pollution sources which influence outdoor air quality. DCCs are located in different places: some are in apartments, others in isolated villas (with or without gardens, terraces or other forms of outdoor spaces), some in modern buildings and others in very old parts of town. In our country, children DCCs vary from fully State-run to joint or solely private ventures. We approached the Instituto da Segurança Social (the DCC national regulatory authority) in order to obtain a list of potential DCCs to be included in the study. For this study, only the private social security institutions (IPSS) included in the list were considered, for practical reasons, as they broadly provided all the age ranges that we aimed to study and their participation was expected to be easier. Those DCCs (IPSS) belonged either to Confederação Nacional de Instituições de Solidariedade or to União das Misericórdias Portuguesas, both of which supported our study.

The project was designed to be carried out in two different phases. During Phase I we collected data on building characteristics and some preliminary IAQ indicators, and used a small screening health questionnaire. In Phase II a new health questionnaire was distributed to the same children who attended the selected DCCs, and a more detailed assessment of IAQ, building characteristics and ventilation was carried out. The project flowchart can be found in Figure 1. Task description will be divided in the two phases.

Figure 1.

Study protocol.

(0.39MB).

Data collection was carried out successfully. Data analysis is still ongoing and almost completed. We have already presented preliminary results at scientific meetings and promoted public discussions about the study.60 Formal recommendations to national authorities and public education materials will be made available in written documents. The number of evaluations performed in each task is summarized in Table 2.

During Phase I we collected data on 45 DCCs (25 in Lisbon and 20 in Oporto). All participating and non-participating DCCs were included in the same management model (IPSS). As DCC selection procedure was stratified according to the district, the proportion of social strata/residential areas was expected to be similar in both participating and non-participating DCCs. Despite this, the non-participating DCCs could not be visited and no data was available to allow us to compare building characteristics with those of participating DCCs.

The age of the buildings was very diverse with a median age of 50 years (P25–P75: 20–134) – the oldest built in 1579 and the most recent in 2008.

In Phase I, CO2 was used to select the DCCs to be included in Phase II as it allowed sampling of DCCs in two different clusters insofar as ventilation was concerned. The first cluster included 4 DCCs in Lisbon and 5 in Oporto. Average CO2 levels, indoor temperature and relative humidity were 2137±368ppm, 18.7±2.2°C and 57.4±5.0%, equivalent to the cluster with worst ventilation conditions. The second cluster included 7 DCCs in Lisbon and 4 in Oporto. Average CO2 levels, indoor temperature and relative humidity were 1233±170ppm, 20.1±1.8°C and 45.6±5.9%, equivalent to the cluster with better ventilation. CO2 (p<0.001), indoor temperature (p=0.025) and relative humidity (p<0.001) differed significantly between the two clusters.

Phase I questionnaire allowed us to collect some data on parental education, medical history and smoking at home. It was also possible to estimate the prevalence of wheezing in the previous 12 months, atopic dermatitis, reported asthma and food allergy. As for the response rate, 3186 out of 5161 questionnaires were returned (61.7%). About half of the sample (50.5%) were boys and the average age was 3.1±1.5 years.

During Phase II we had one DCC drop out during Visit 1 and two others during Visit 2. Indoor air quality was evaluated in 19 DCCs during Visit 1 and in 17 DCCs during Visit 2. Measurements included thermal comfort parameters, CO, CO2, PM10, TVOC, formaldehyde, bacteria, molds and dust mites. Information regarding indoor and outdoor values was collected.

Objective ventilation measures were gathered during Visit 2 from 16 DCCs (7 in Lisbon and 9 in Oporto). The ventilation data are being used to create modeling software that may be useful in future studies.

In relation to the questionnaires used in Phase II, we had different return rates. Of the 2287 ISAAC questionnaires handed out, 1221 were returned (53.4%). The average age of the sample was 3.5±1.5 years, and 52.4% were boys.

Data concerning non-respondents were available only for the DCCs included in Phase II, as many institutions considered these data confidential. With regard to the ISAAC questionnaire, no statistical differences were found for gender (p=0.700) and age (children under 3 years versus children over 3 years: p=1.00) between respondents and non-respondents. When the Phase I and Phase II samples were compared in terms of gender and age, the only significant difference between Phase I and the ISAAC questionnaire was that children in Phase II were slightly older.

As for the questionnaires used to evaluate recent respiratory symptoms (follow-up questionnaires), 755 out of 1221 were answered during Visit 1 (53.3% of boys with an average age of 3.5±1.5 years) and 667 out of 1411 during Visit 2 (54% of boys with an average age of <3.2±1.6 years).

During Visit 2, nitrates/nitrites and pH were measured in 100 exhaled breath condensate samples collected in Lisbon DCCs, and 94 samples from children of different DCCs were collected and tested for respiratory viruses, out of the target population of 1411 children. Finally, parental stress questionnaires were collected during both visits.

ENVIRH is the first Portuguese study addressing the effect of IAQ on children less than six years old, in a multidisciplinary perspective through objective measurements of IAQ, ventilation, building characteristics and health outcomes. This paper reviews its rationale and outlines methods that might be implemented by other investigators conducting studies on the association between these variables.

The major strengths of the study are the inclusion of a sample of carefully selected DCCs, the longitudinal approach, the detailed characterization of each DCC and the complex statistical analyses that took into consideration the correlation structure of the observations for each classroom/DCC.

Our study has evaluated a considerable number of DCCs and children. Air pollutant patterns could vary according to the environment studied.61

It might have been interesting to include in our DCC sample institutions from urban and rural settings; however – like other Portuguese authors – we decided to study only urban areas, namely Lisbon20,36,37 and Oporto.62 This allowed us to study a larger number of DCCs and limited the environmental confounding that we would have found if we had also included DCCs from rural settings.

There were some concerns about the representativity of our sample in Phase II as we were unable to study a large sample of DCCs. To address this issue we designed the study in two main phases. During Phase I we assessed a larger sample, and during Phase II a more detailed evaluation was conducted.

The participation rate was lower than in other epidemiological studies which also used the ISAAC questionnaire in this age group.7,63 The different types of assessments (through questionnaire and biological fluids) and the requests for ongoing follow-up may have influenced the decision to participate in a study that could have taken up a substantial amount of time. The need for active consent64 might also have influenced the final number of participants as we only handed out questionnaires to those parents who gave written consent. Despite this limitation, the male–female proportion and the average age were similar across the different questionnaires.

CO2 levels were used to divide the Phase I DCCs into two different clusters in terms of ventilation, which were then included in Phase II. This was possible because CO2 is an established marker of building ventilation65 and together with occupancy could be used to develop effective models to predict room ventilation.66 Phase II was designed to include two evaluations at different times with a dual purpose: ensuring seasonal variety and enabling a longitudinal evaluation. The first and comprehensive evaluation of children's health took place during the Spring of 2011 using the ISAAC questionnaire. It set the respiratory and environmental baseline characteristics for the participating children. The first follow-up questionnaire was distributed shortly after, during the Spring, and allowed us to gather information about recent health events. This questionnaire was then repeated during the Winter, 9 months after the first evaluation.

IAQ parameters were also collected from the DCCs during Spring and Winter in order to study the association between IAQ parameters and respiratory events. Studies have shown that there are seasonal variations in indoor pollutant levels.37,61 Pegas et al. showed in a group of Lisbon elementary schools that bioaerossols and CO2 levels were above the recommended limits in different seasons and some volatile organic compounds concentrations were higher during the Winter.37

As different patterns of indoor air pollutants can be found throughout the year, seasonal studies are relevant in order to identify specific problems for the time of the year in order to plan effective actions.

The cross-sectional part of our study, performed in each phase, will allow us to study the association between DCC IAQ and respiratory symptoms in young children. For this purpose we took into consideration different variables reported in the literature as confounders. ISAAC questionnaire was used to collect information about those covariates related not only to the children, but also to the home environment.

ENVIRH also included a longitudinal part to study causal associations, as we planned to follow-up with the participating children. However, the decrease in the participation rate across the study will limit this objective.

Different indoor air pollutants have been associated with respiratory diseases or symptoms.15 ENVIRH was designed to study IAQ as thoroughly as possible. Adding PM2.5 would be interesting because of their impact on lower airways.1 However, it has been shown that exposure to PM10 is associated with respiratory disease15 and airway inflammation biomarkers,22 making it acceptable for our purposes.

Carbon dioxide level was measured in both phases and visits because it is a surrogate of ventilation. Although a more specific quantification was needed to create ventilation model software, PFT technique was used only in a few DCCs and during Visit 2, because it is burdensome. PFT technique to determine air infiltration rates at homes and buildings, has been reported in several studies,67,68 while its applicability and effectiveness have been discussed elsewhere.69,70 Ventilation studies are important because this variable not only reflects IAQ but, when considering interventions to improve exposure to pollutant, also shows that improving ventilation will positively affect all airborne contaminants. We also looked at air pollutants outside the DCCs in order to differentiate indoor from outdoor sources. It might be interesting in future studies to determine the contribution of the outdoor to IAQ by analyzing traffic patterns in the surrounding areas, as it may be another intervention target.

The ISAAC questionnaire has been an important validated tool in different epidemiological studies all over the world. In order to get a more objective picture of the airways, we measured some inflammatory biomarkers in the EBC; however, due to funding limitations we only collected 100 samples. EBC study could be very useful as it is a non-invasive and easy-to-collect technique and allows the measurement of different markers. Different studies showed that nitrogen products and pH in EBC were associated with asthma and FEV1 deterioration.71 As far as pH is concerned, an association was found between benzene, NO2 and PM10 levels and EBC acidity.22

Another constraint we faced was the low number of children from whom virus swabs were collected, despite the efforts of our team who held several public discussions with parents and DCC staff. This may be because of parents’ concern about the technique and not appreciating the importance of this aspect of the study. Although we should note that after the Influenza H1N1 alert in 2011 the study compliance increased dramatically.

Stress evaluation is an emerging issue in childhood respiratory disease studies. It has been shown that chronic stress in the family may lead to children's respiratory infections or diseases.72 Parental stress levels were assessed by validated instruments. Children's stress levels were inferred from cortisol levels measured in 24h urine samples to monitor for circadian rhythm alterations. This approach is validated, although results need to be interpreted carefully when assessing children <3 years of age, as urinary cortisol levels could rise during DCC attendance in this age group.73 This was not relevant to our study as urine was collected only in children over 4 years of age.

Our main concern in ENVIRH was to contribute to healthier environments for children. This issue is regarded as a priority by the World Health Organization and the European Commission.74,75 Improving children's health is such a major concern, not only for its short-term effects, but also because it will lead to a physically, psychologically and socially healthier individual. With that in mind, some particularly important outcomes for our study were: contributing to public education and making recommendations to national authorities; and creating modeling software for the analysis of ventilation patterns. Public education should be the mainstream of all interventions involving the population, contributing to a healthy and conscious lifestyle and to greater awareness of diseases and associated risks.

DCC staff representatives, parents’ associations and health authorities were invited to participate in the study seminars. The next objective will be to contribute to recommendations, as it has been shown that when there are formal IAQ programs it is more likely that institutions will implement the necessary changes and strategies to improve IAQ.76

In addition, legislation relating to DCCs may be a good starting point to improve respiratory health in children, regulating cost-effective changes for environments where children spend a significant amount of time. Modeling software for ventilation studies may also prove valuable toward these ends, from two perspectives: research and policy implementation. For research in other settings, models may prove as valuable, reliable and more affordable instruments to calculate IAQ parameters than direct measurements, even if they retain some limitations.77,78 They may also be useful for building design and ventilation strategies in some cases78,79 and could be used to predict the health impact and costs of specific situations and interventions, bringing together engineers, physicians and policy makers.

Studies on IAQ and children's respiratory health need to be carried out as there are few publications on this subject. Most data come from areas of the world with different building and environmental characteristics (building stock, construction materials, cleaning and ventilation habits in DCC and climate). For this reason, national studies designed to address specific regional needs, should be conducted.

We recommend addressing these issues in multidisciplinary teams, creating knowledge that could be readily transmitted to the community, thus helping to improve the health of populations in the short and long terms.

The activities of project ENVIRH were funded by research grants from national funds attributed by Fundação para a Ciência e Tecnologia (Project references – PTDC/SAU-ESA/100275/2008 and PEST-OE/MAT/UI0006/2011).

The authors have no conflicts of interest to declare.