Pulmonary epithelial integrity in children: relationship to ambient ozone exposure and swimming pool attendance.

Airway irritants such as ozone are known to impair lung function and induce airway inflammation. Clara cell protein (CC16) is a small anti-inflammatory protein secreted by the nonciliated bronchiolar Clara cells. CC16 in serum has been proposed as a noninvasive and sensitive marker of lung epithelial injury. In this study, we used lung function and serum CC16 concentration to examine the pulmonary responses to ambient [O.sub.3] exposure and swimming pool attendance. The measurements were made on 57 children 10-11 years of age before and after outdoor exercise for 2 hr. Individual [O.sub.3] exposure was estimated as the total exposure dose between 0700 hr until the second blood sample was obtained (mean [O.sub.3] concentration/[m.sup.3] x hours). The maximal 1-hr value was 118 [micro]g/[m.sup.3] (59 ppb), and the individual exposure dose ranged between 352 and 914 [micro]g/[m.sup.3]hr. These [O.sub.3] levels did not cause any significant changes in mean serum CC16 concentrations before or after outdoor exercise, nor was any decrease in lung function detected. However, children who regularly visited chlorinated indoor swimming pools had significantly lower CC16 levels in serum than did nonswimming children both before and after exercise (respectively, 57 [+ or -] 2.4 and 53 [+ or -] 1.7 [micro]g/L vs. 8.2 [+ or -] 2.8 and 8.0 [+ or -] 2.6 [micro]g/L; p < 0.002). These results indicate that repeated exposure to chlorination by-products in the air of indoor swimming pools has adverse effects on the Clara cell function in children. A possible relation between such damage to Clara cells and pulmonary morbidity (e.g., asthma) should be further investigated. Key words: airway irritants, children, Clara cell protein (CC16), nitrogen trichloride, ozone, swimming pool. Environ Health Perspect 112:1768-1771 (2004). doi:10.1289/ehp.7027 available via http://dx.doi.org/[Online 13 September 2004]

**********

Ozone is an important component of air pollution. Ground-level [O.sub.3] in urban air is formed in a photochemical reaction between oxygen and nitrogen dioxide from fossil fuel emissions under the influence of sunlight and volatile hydrocarbons. Therefore, the [O.sub.3] levels tend to be high in areas with high intensity of ultraviolet radiation and high emissions of N[O.sub.2] from car traffic or industries using fossil fuels (American Thoracic Society 1996; de Marco et al. 2002).

Epidemiologic and controlled human studies as well as animal experiments on exposure to [O.sub.3] have reported airway inflammation and/or a decrease in lung function at ambient concentrations [reviewed by Balmes (1993); Krishna et al. 1995]. Human experimental exposure to [O.sub.3] has demonstrated a spectrum of acute airway responses. Among these are decrements in forced vital capacity (FVC) and forced expiratory volume in 1 sec ([FEV.sub.1]), increased airway resistance (Blomberg et al. 1999; Seal 1993), altered airway permeability, and antioxidant defenses (Blomberg et al. 1999; Mudway et al. 2001), as well as a neutrophilic airway inflammation (Schelegle et al. 1991; Stenfors et al. 2002). Most of these studies were short-term exposures on healthy adults. Acute lung function changes in children have been shown in field exposures. Associations between ambient [O.sub.3] levels and reductions in FVC and [FEV.sub.1] in children in summer camps and big cities have been shown in several studies (American Thoracic Society 1996; Kopp et al. 2000). Recently, short-term effects of [O.sub.3] were observed in children as an increased frequency of emergency visits for asthma (Fauroux et al. 2000)

Repeated exposure to other environmental and occupational gases (e.g., sulfur dioxide and chlorine) also increase the risk of airway irritation and asthmalike symptoms (Olin et al. 2002). Several studies have shown that competitive swimmers have an increased prevalence of airway inflammation, bronchial hyperresponsiveness, and asthma (Helenius and Haahtela 2000; Potts 1996). This was attributed to inhalation of chlorine gas and its derivatives formed by chlorination of ammonia derived from organic matter in swimming pool water [e.g., nitrogen trichloride, trichloramine, or chlorine azide ([NCl.sub.3])].

In recent years there has been a growing interest in noninvasive indicators as a means to detect early effects of air irritants (Bernard et al. 1992; Broeckaert et al. 1999). Several reports describe different lung-specific secretory proteins, which may be used to detect changes in the number of and/or integrity of epithelial secretory cells (Hermans and Bernard 1999). One of these, Clara cell protein (CC16), is a small, 16-kDa protein produced and secreted by the nonciliated bronchiolar Clara cells and detectable in serum. CC16 has antioxidant properties, and the levels in serum increase when lung epithelium permeability is adversely affected by air pollutants or other lung toxicants (Broeckaert et al. 1999; Hermans and Bernard 1996). On the other hand, reduced levels of CC16 in lung lavage fluid are described in several lung disorders (e.g., chronic bronchitis) and in smokers. This may be caused by a decrease in the production of CC16 depending on a decreased number of Clara cells (Hermans and Bernard 1999). In this study we validated CC16 and the lung surfactant proteins A, B, and D in blood as biomarkers of adverse pulmonary effects. The advantage of studying lung proteins in serum instead of in bronchoalveolar lavage, which has been commonly used to study inflammatory effects in the airways, is obvious. Blood samples are more easily obtained than is lung lavage. Besides, lung lavage is not a suitable method in studies on children.

The general aim of the present program, which was part of a European Union project (HELIOS), was to examine lung function and possible changes in the serum levels of CC16 in relation to ambient [O.sub.3] exposure in Italy, France, Belgium, and Sweden (Bernard et al. 2003). The effects of exposure to other environmental factors (e.g., chlorine and its by-products) in swimming pools were also examined. The present study was conducted in Umea, a town in northern Sweden with low to moderate [O.sub.3] levels, and was divided into four substudies. The first part was conducted on healthy adults in winter, when [O.sub.3] levels were low, and the second part in the summer when [O.sub.3] levels were known to be higher than in other seasons. Similar studies were then repeated in children. In the present report we describe the results from the summer study on children.

Materials and Methods

The winter study on children, also the basis for the summer study, was conducted in November 2001; 139 children, 63 girls and 76 boys, from four primary schools were recruited. Children with a history of asthma or kidney disease were not included. The recruitment of school children was done according to the same protocol as in the study by Bernard et al. (2003) in Brussels. Lung function measurements were performed, and peripheral blood samples were obtained for analysis of CC16 in serum. In the present, summer, study, 57 of the children from the winter study were included, 56 Caucasian and one Chinese. The study was conducted in May 2002. The local ethics committee at Umea University approved the study protocol. A written informed consent was obtained from the parents.

The selection of the 57 children (33 boys and 24 girls) from the larger study in November 2001 was based on the results from the lung function tests and a questionnaire answered by the parents. Subjects who reported pollen allergy or childhood asthma and/or who had an FVC or [FEV.sub.1] < 80% of the predicted value were not included, nor were children whose blood samples had difficulties or whose questionnaires were missing. The age (mean [+ or -] SD) of the participating children was 10.8 [+ or -] 0.4 years. Lung function testing and blood sampling (in November 2001) were repeated twice, before and after light exercise outdoors for 2 hr (range, 1.5-3.0 hr). The parents completed a questionnaire on, for example, current food intake, passive smoking, and airway illness since the larger, winter study. The participating children answered questions on outdoor activities and swimming pool attendance. Nearly 40% of the children were regular indoor pool visitors (i.e., they had visited an indoor pool for at least 1 hr/month during 6 months or longer). Sodium hypochlorite (1% chlorine) was used to disinfect the pool water. Based on swimming pool attendance according to the questionnaire, the children were divided into two subgroups, 34 non-pool visitors and 23 pool visitors.

We determined lung function parameters (e.g., FVC and [FEV.sub.1]) using a portable spirometer connected to a computerized data program (KoKo Spirometer and KoKo DigiDoser; Pulmonary Data Service Instrumentation, Inc., Louisville, KY, USA). The instruments were calibrated in the morning and after every 10th measurement. Changeable filter mouthpieces were purchased from Intramedic AB (Balsta, Sweden). One trained lung physiologist tested the lung function in all children. The tests were carried out in the standing position. The best reproducible flow/volume curves were used in the analysis. The computer program calculated the predicted normal values as a function of sex, age, height, and weight according to Polgar and Promadhat (1971).

Blood samples were obtained from the antecubital vein after local anesthesia with a cream or plaster (EMLA, AstraZeneca Ltd., Sodertalje, Sweden) immediately before (S1; four missing samples) and after (S2; three missing samples) the outdoor session. Two CC16 values were available for 20 pool visitors and 31 non-pool visitors. Blood (7.5 mL) was drawn in Sarstedt Monovette tubes for serum (Serum Z/9 mL, Sarstedt, Landskrona, Sweden). Each sample was allowed to clot for 1-2 hr at room temperature. After centrifugation at 3,000 rpm (within 2 hr after sampling), the serum was transferred to cryotubes and frozen at -80 [degrees]C. These samples were then sent to the Industrial Toxicology Unit at the Catholic University of Louvain in Brussels for analysis. CC16 was determined by a latex immunoassay using rabbit anti-CC16 antibody (Dakopatts, Glostrup, Denmark) and also CC 16 purified according to the standard in the laboratory (Bernard et al. 1992; Carbonelle et al. 2002). The assay has been validated by comparison with a monoclonal-antibody-based enzyme-linked immunosorbent assay (ELISA) (Hermans et al. 1998). All samples were run in duplicate at two different dilutions. The between- and within-run coefficients of variation ranged from 5 to 10%.

Outdoor [O.sub.3] was monitored continuously at the university campus where the children spent time outdoors, using a Dasibi ultra-violet photometry ozone analyzer (model 1108; Dasibi Environmental Corporation, Glendale, CA, USA). [O.sub.3] exposure was estimated as the total exposure of [O.sub.3] between 0700 hr and the time the second blood sample was taken, between 1300 and 1600 hr (mean [O.sub.3] concentration/[m.sup.3] x number of hours). Because the children spent part of that time indoors (mean, 4 hr) and because it is known from other studies that indoor concentrations of [O.sub.3] are lower than those outdoors (American Thoracic Society 1996), each individual's exposure dose was estimated by assuming an exposure level of 50% of the mean outdoor [O.sub.3] concentration during time spent indoors. This assessment was confirmed by measurement with passive diffusion samplers in the examination room. The filters were purchased from and analyzed at IVL Swedish Environmental Research Institute, Ltd. (Gothenburg, Sweden). The mean indoor [O.sub.3] level during the study period was 40 [micro]g/[m.sup.3] (20 ppb).

The statistical program SPSS 11 (SPSS Inc., Chicago, IL, USA) was used for statistical analyses. Differences in [FEV.sub.1] and CC16 before and after exercise and differences between groups were assessed with Student's t-test, paired and unpaired. Pearson correlation tests were used for the analyses of correlations. A p-value < 0.05 was considered statistically significant.

Results

The mean daytime outdoor [O.sub.3] concentration in the days studied ranged from 77 to 116 [micro]g/[m.sup.3], and the maximal 1-hr value was 118 [micro]g/[m.sup.3]. The estimated individual exposure dose varied from 352 to 914 [micro]g/[m.sup.3]hr.

[FEV.sub.1] was significantly higher after outdoor exercise than before in both children who had regularly attended chlorinated swimming pools and children not swimming (Table 1). These differences remained also if the percentages of the predicted [FEV.sub.1] ([FEV.sub.1]% predicted) were compared. The mean measured [FEV.sub.1] values varied between 91.2 and 93.0% of the predicted ones. There were no significant differences between pool visitors and non-pool visitors, when comparing [FEV.sub.1]% predicted either before (p = 0.43) or after exercise (p = 0.45, Student's t-test), nor was there any significant difference in body mass index (BMI) between the two groups of children.

The mean [+ or -] SD serum concentrations of CC16 in non-pool visitors were 8.2 [+ or -] 2.8 [micro]g/L before exercise and 8.0 [+ or -] 2.6 [micro]g/L after exercise. The corresponding values in pool visitors were 5.7 [+ or -] 2.4 and 5.3 [+ or -] 1.7 [micro]g/L (Table 2; range, 2.2-16.1 [micro]g/L). The BMI was 18.5 [+ or -] 2.9 kg/[m.sup.2]. Only one pool visitor and three nonvisitors were exposed to passive smoke. There were no significant correlations between the serum CC16 levels and parental smoking or BMI. No significant differences were found between pre- and postexposure levels of serum CC16, nor did the time spent outdoors (mean, 6 hr) during the 2 days preceding the test day have any influence on the CC16 levels. However, the average CC16 levels in pool visitors both before (S1) and after (S2) exercise were lower than in non-pool visitors (p < 0.01) (Table 2). Twenty-two children regularly visited an indoor swimming pool for 1-35 hr/month (median, 4 hr/month). The children had been visiting indoor swimming pools regularly between 6 months to 10 years (median, 3 years). Only two children had been swimming since they were babies. No statistically significant relationship was found for attending a swimming pool during the last days before the test, probably because only seven children had attended indoor swimming pools the last 2 days before the test. In our study, we did not find any correlation between parental smoking and effects on the airways of the children or CC16 levels, possibly because only one pool visitor and three non-pool visitors were exposed to passive smoke.

The correlations between [O.sub.3] exposure and CC16 levels before or after exercise outdoors were not statistically significant in the group as a whole. However, when CC16 after exercise (S2) was considered, there was a tendency toward a correlation in non-pool visitors after exercise (p < 0.06) (Table 3, Figure 1).

[FIGURE 1 OMITTED]

Discussion

In this study, moderate [O.sub.3] levels between 77 and 116 [micro]g/[m.sup.3] did not have any adverse effect on the lung function parameter [FEV.sub.1] after 2 hr of outdoor exercise. In fact, the [FEV.sub.1] was slightly increased at the second measurement. This could be an effect of better test performance after exercise than before. The ambient [O.sub.3] levels in our study are also lower than those reported to affect lung function parameters at ambient [O.sub.3] concentrations (Kinney et al. 1996; Nickmilder et al. 2003).

The serum CC16 levels found in this study did not correlate with BMI, a result that has been shown in other studies as well (Hermans et al. 1998). They were of the same magnitude as those recently reported in children of the same age in Belgium (Bernard et al. 2003; Carbonelle et al. 2002). In those studies, the serum levels of CC16 in children did not change significantly during swimming exercise. In the present study, we have compared the serum CC16 levels in pool visitors and in a control group not exposed to chlorination by-products and found significantly lower levels of serum CC16 in pool visitors suggesting adverse effects on Clara cells.

There were no significant differences between the levels of CC16 before and after outdoor exercise. Neither were there any statistically significant relationships between CC16 levels in serum and ambient [O.sub.3] exposure, although a marginally significant tendency was found among nonswimmers (Figure 1). The lack of statistical significance may be due to the limited number of subjects and/or the [O.sub.3] levels' not being high enough to cause a response. In the present study, the [O.sub.3] concentration was approximately one-fourth of the [O.sub.3] concentration that recently was found to increase the serum CC16 levels in adult subjects (n = 22) exposed for 2 hr in an exposure chamber to 400 [micro]g/[m.sup.3] [O.sub.3] (Blomberg et al. 2003). Another possible explanation for the lack of a clear relationship between serum CC16 and the [O.sub.3] dose in the present study is that the time period between the measurements was not long enough to cause a measurable change in CC16 levels. There may also be an interference with diurnal variation not corrected for in the present study. Such a diurnal variation was indicated in a recent study on adults (n = 19) (Helleday et al. 2003). The reason why Bernard et al. (1997) did not find any significant variations in serum CC16 between 0900-1000 hr and 1600-1750 hr in seven healthy adults may be the low number of subjects studied.

Lower CC16 levels among subjects regularly attending chlorinated swimming pools are in accordance with the findings by our Belgian partners in the HELIOS project (Bernard et al. 2003; Carbonelle et al. 2002). These authors found that the concentrations of CC16 in trained swimmers were negatively correlated with their cumulated pool attendance. Thus, swimmers seem to have a somewhat decreased pool of CC16 in the Clara cells in the lungs. The CC16 concentration in serum reflects both the epithelial permeability and the integrity of Clara cells (Hermans and Bernard 1999). Therefore, it is conceivable that a repeated exposure to disinfecting byproducts formed by hypochlorite and organic matter (e.g., urea and sweat) in pools may decrease the CC16 secretion because of Clara cell dysfunction or damage. Thus, a possible increase in the intravascular leakage of CC16 caused by, for example, [O.sub.3] exposure could be masked by a decrease in the production of CC16 in swimmers (Bernard et al. 2003; Carbonelle et al. 2002). That this could be the case also in our study is indicated by the tendency toward a correlation between short-term [O.sub.3] exposure and the serum CC16 levels in non-pool visitors, but not in pool visitors, after exercise.

The levels of chlorination by-products were not measured in this study, but evidently they were high enough to affect the lung epithelium in children regularly visiting indoor pools. Because sodium hypochlorite (1% chlorine) was used as a sanitizer of the pool water, increased levels of [NCl.sub.3] were likely to be present in the pool air. A limited number of measurements of [NCl.sub.3] in indoor air at the swimming pool most frequently used by the swimming children in our study had been performed in 1995. Levels were similar to those reported in the same year from France by Hery et al. (1995), who identified [NCl.sub.3] as the main component of chlorination by-products present in the air of indoor swimming pool areas. Hery et al. (1995) also reported that symptoms of irritation in the eyes and throat were correlated with the air levels of [NCl.sub.3]. Bernard et al. (2003) reported that [NCl.sub.3] in public pools typically are in the range of 0.1-1 mg/[m.sup.3] in air sampled 1.5 m above the water surface, that is, values similar to those reported by Hery et al. (1995) and the few Swedish measurements (Eriksson and Jacobsson, unpublished data).

Conclusions

Our results indicate that repeated exposure to chlorination by-products in the air of indoor swimming pools has an adverse effect on the Clara cell function in children, such that the anti-inflammatory role of CC16 in the lung could be diminished. A possible role of such influence on Clara cell function in inducing pulmonary morbidity (e.g., asthma) should be further studied. The lung function parameter [FEV.sub.1] was not adversely affected by outdoor exercise at a moderate [O.sub.3] concentration in either pool visitors or in non-pool visitors. A possible effect of ambient [O.sub.3] on serum CC16 levels (in nonswimming children) needs further investigation.

Research nurse M. Backman, Paediatrics, has contributed with excellent work in the practical part of this study. A. Hagenbjork-Gustafsson, the National Institute for Working Life, Umea, performed the ozone measurements.

Financial support has been given by the European Commission (HELIOS project, QLK4-1308), the Swedish Environmental Protection Agency, and Forskningsradet for Miljo, Areella Naringar och Samhalle.

The authors declare they have no competing financial interests.

Received 12 February 2004; accepted 13 September 2004.

REFERENCES

American Thoracic Society, Committee of the Environmental and Occupational Health Assembly. 1996. State of the art. Health effects of outdoor air pollution. Section 2. Ozone. Am J Crit Care Med 153:15-27.

Balmes JR. 1993. The role of ozone exposure in the epidemiology of asthma. Environ Health Perspect 101(suppl 4):219-224.

Bernard A, Carbonelle S, Michel O, Higuet S, de Burbure C, Buchet J-P, et al. 2003. Lung hyper-permeability and asthma prevalence in schoolchildren: unexpected associations with the attendance at indoor chlorinated swimming pools. Occup Environ Med 60:385-394.

Bernard A, Hermans C, Van Houte G. 1997. Transient increase of Clara cell protein (CC16) after exposure to smoke. Occup Environ Med 54:63-65.

Bernard A, Marchandise FX, Depelchin S, Lauwerys R, Sibille Y. 1992. Clara cell protein in serum and bronchoalveolar lavage. Eur Resp J 5:1231-1238.

Blomberg A, Mudway IS, Nordenhall C, Hedenstrom H, Kelly FJ, Frew AJ, et al. 1999. Ozone-induced lung function decrements do not correlate with early airway inflammatory or antioxidant responses. Eur Respir J 13:1418-1428.

Blomberg A, Mudway I, Svensson M, Hagenbjork-Gustafsson A, Thomasson L, Helleday R, et al. 2003. Clara cell protein as a biomarker for ozone-induced lung injury in humans. Eur Respir J 22:883-808.

Broeckaert F, Arsalane K, Hermans C, Bergamaschi E, Brustolin A, Mutti A, et al. 1999. Lung epithelial damage at low concentrations of ambient ozone. Lancet 353:900-901.

Carbonelle S, Francaux M, Dumont X, de Burbure C, Morel G, Bernard A. 2002. Changes in serum pneumoproteins caused by short-term exposure to nitrogen trichloride in indoor chlorinated swimming pools. Biomarkers 7:464-478.

de Marco R, Poli A, Ferrari M, Accordini S, Giammanaco G, Bugiani M, et al. 2002. The impact of climate and traffic-related N[O.sub.2] on prevalence of asthma and allergic rhinitis in Italy. Clin Exp Allergy 32:1405-1412.

Fauroux B, Sampil M, Quenel P, Lemoullec Y. 2000. Ozone a trigger for hospital pediatric asthma emergency room visits. Pediatr Pulmonol 30(1):41-46.

Helenius I, Haahtela T. 2000. Allergy and asthma in elite summer sports athletes. J Allergy Clin Immunol 106(3):444-452.

Helleday R, Segerstedt B, Forsberg B, Nordberg B, Svensson M, Bernard A, et al. 2003. Diurnal variation in serum levels of Clara cell protein (CC16) in humans [Abstract]. Eur Resp J 22(suppl 45):75s.

Hermans C, Bernard A. 1996. Clara cell protein: characteristics and potential applications as marker of lung toxicity. Biomarkers 1:3-8.

Hermans C, Bernard A. 1999. State of the art. Lung epithelium-specific proteins. Characteristics and potential applications as markers. Am J Respir Crit Care Med 159:646-678.

Hermans C, Osman A, Nyberg B G, Peterson C, Bernard A. 1998. Determinants of Clara cell protein (CC16) concentration in serum: a reassessment with two different methods. Clin Chim Acta 272:101-110.

Hery M, Hecht G, Gerber JM, Gendre JC, Hubert B. 1995. Exposure to chloramines in the atmosphere of indoor swimming pools. Ann Occup Hyg 39:427-439.

Kinney PL, Thurston GD, Raizenne M. 1996. The effects of ambient ozone on lung function in children: a reanalysis of six summer camp studies. Environ Health Perspect 104:170-174.

Kopp MV, Bohnet W, Frischer T, Ulmer C, Studnicka M, Ihorst G, et al. 2000. Effects of ambient ozone on lung function in children ovar a two-year summer period. Eur Respir J 16:893-900.

Krishna MT, Mudway I, Kelly FJ, Frew AJ, Holgate ST. 1995. Ozone, airways and allergic airway disease. Clin Exp Allegy 25:1150-1158.

Mudway IS, Stenfors N, Blomberg A, Helleday R, Dunster C, Marklund SL, et al. 2001. Differences in basal airway antioxidant concentrations are not predictive of individual responsiveness to ozone: a comparison of healthy and mild asthmatic subjects. Free Radic Biol Med 31(8):962-974.

Nickmilder M, Carbonelle S, de Burbure, Bernard A. 2003. Relationship between ambient ozone and exhaled nitric oxide in children [Letter]. JAMA 290:2546-2547.

Olin A-C, Granung G, Hagberg S, Adriansson M, Brisman J, Dalander 0, et al. 2002. Respiratory health among bleachery workers exposed to ozone and chlorine dioxide. Scand J Work Environ Health 28(2):117-123.

Polgar G, Promadhat V. 1971. Pulmonary Function Testing in Children: Techniques and Standards. Philadelphia:W.B. Saunders.

Potts J. 1996. Factors associated with respiratory problems in swimmers. Sports Med 21(4):256-261.

Schelegle ES, Siefkin AD, McDonald RJ. 1991. Time course of ozone-induced neutrophilia in normal humans. Am Rev Respir Dis 143(6):1353-1350.

Seal E Jr, McDonnell WF, House DE, Salaam SA, Dewitt PJ, Butler SO, et al. 1993. The pulmonary response of white and black adults to six concentrations of ozone. Am Rev Respir Dis 142(4):804-810.

Stenfors N, Pourazar J, Blomberg A, Krishna MT, Mudway I, Helleday R, et al. 2002. Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects. Respir Med 96(5):352-358.

Birgitta Json Lagerkvist, (1) Alfred Bernard, (2) Anders Blomberg, (3) Erik Bergstrom, (4) Bertil Forsberg, (1) Karin Holmstrom, (5) Kjell Karp, (5) Nils-Goran Lundstrom, (1) Bo Segerstedt, (7) Mona Svensson, (1) and Gunnar Nordberg (7)

(1) Environmental and Occupational Medicine, Department of Public Health and Clinical Medicine, Umea University, Umea, Sweden; (2) Unit of Industrial Toxicology, Catholic University of Louvain, Brussels, Belgium; (3) Respiratory Medicine and Allergy, Department of Public Health and Clinical Medicine, (4) Paediatrics, Department of Clinical Sciences, and (5) Clinical Physiology, Department of Surgical and Perioperative Sciences, Umea University, Umea, Sweden

Address correspondence to B.J. Lagerkvist, Environmental Medicine, Department of Public Health and Clinical Medicine, Umea University, S-901 87 Umea, Sweden. Telephone: [phone]. Fax: [phone]. E-mail: [e-mail]

Table 1. [FEV.sub.1] (L/sec) and [FEV.sub.1]% predicted before
(S1) and after (S2) outdoor exercise in children who do and do
not regularly visit pools (mean [+ or -] SD).

Category                              S1                   S2

All (n = 57)
  [FEV.sub.1]                2.19 [+ or -] 0.31   2.22 [+ or -] 0.32
  [FEV.sub.1]% predicted     91.3 [+ or -] 7.2    92.7 [+ or -] 7.6
Non-pool visitors (n = 34)
  [FEV.sub.1]                2.25 [+ or -] 0.32   2.29 [+ or -] 0.33
  [FEV.sub.1]% predicted     91.2 [+ or -] 5.6    92.6 [+ or -] 6.3
Pool visitors (n = 23)
  [FEV.sub.1]                2.09 [+ or -] 0.27   2.13 [+ or -] 0.28
  [FEV.sub.1]% predicted     91.5 [+ or -] 9.1    92.9 [+ or -] 9.5

                                                       P-Value
Category                           Diff S2-S1       (paired Nest)

All (n = 57)
  [FEV.sub.1]                0.033 [+ or -] 0.061     < 0.001
  [FEV.sub.1]% predicted       1.4 [+ or -] 2.5       < 0.001
Non-pool visitors (n = 34)
  [FEV.sub.1]                0.035 [+ or -] 0.063       0.003
  [FEV.sub.1]% predicted       1.4 [+ or -] 2.5         0.002
Pool visitors (n = 23)
  [FEV.sub.1]                0.031 [+ or -] 0.060       0.021
  [FEV.sub.1]% predicted       1.3 [+ or -] 25          0.018

Diff, difference.

Table 2. CC16 levels ([micro]g/L) in plasma of children who
do and do not regularly visit pools, before (S1) and after
(S2) outdoor exercise (mean [+ or -] SD).

Category                                         S1

All (n = 31)                              7.2 [+ or -] 2.9
Non-pool visitors (n = 31)                8.2 [+ or -] 2.8
Pool visitors (n = 20)                    5.7 [+ or -] 2.4
t-Test pool visitors versus nonvisitors      p < 0.002

Category                                         S2

All (n = 31)                              7.0 [+ or -] 2.7
Non-pool visitors (n = 31)                8.0 [+ or -] 2.6
Pool visitors (n = 20)                    5.3 [+ or -] 1.7
t-Test pool visitors versus nonvisitors       p < 0.001

Category                                  Paired t-test

All (n = 31)                                 p = 0.31
Non-pool visitors (n = 31)                   p = 0.68
Pool visitors (n = 20)                       p = 0.14
t-Test pool visitors versus nonvisitors

Table 3. Correlation between individual [O.sub.3] exposure
doses and serum CC16 concentrations in children after
exercise (S2).

Category                     Correlation (S2)   p-Value

All (n = 54)                       0.17          < 0.21
Non-pool visitors (n = 33)         0.34          < 0.06
Pool visitors (n = 21)            -0.08          < 0.74