INTRODUCTION
In order to detect themagnitude of passive smoking, parental questionnaires have beenused extensively. Epidemiological studies on respiratory effect ofpassive smoking in children suggested urinary cotinine excretionfor determining the intensity of exposure to tobacco smoke (1-4).In some other studies, carbon monoxide in expired air has beenreported to be an indirect measurement for the quantity of passivesmoking and CO poisoning (5, 6). Since measurement of cotinineexcretion is expensive and require more sophisticated laboratorywork-up, measurement of carbon monoxide in expired air may be analternative to estimate magnitude of environmental tobacco smokeespecially for nonaffluent countries and for epidemiologicalstudies. However, CO is produced in vivo in many tissues ofthe body by an enzyme called Heme oxygenase-1 (7). Heme oxygenaseis present in the pulmonary vascular endothelium (8) and alveolarmacrophages (9) and is upregulated by the oxidative stress (9) andinflammatory cytokines (10). On the basis of the fact that there isan inflammatory process in the pathogenesis of asthma, we examinedwhether asthmatic children produced more CO than do healthychildren, besides investigating the relationship between CO inexpired air and the intensity of passive smoking determined by aparental questionnaire in healthy and asthmaticchildren.
MATERIAL ANDMETHODS
This study ofcross-sectional design was performed in 235 healthy children in daycare centers, nursery schools and primary schools in summer monthsthat there was no air pollution and no need to heat with stove. All54 asthmatic children followed up at Pediatric outpatient clinicsof Dicle University Hospital included in the study had mildpersistent to moderate asthma (before admission; > 2 attack perweek, night symptoms > 2 a month or > 1 a week, FEV1 > 60%of predicted). Thirty of them were being treated with inhaledcorticosteroids (budesonid 200-400 µg/day,twice a day) at least for the last 4 weeks prior to the COmeasurements, but the rest 24 asthmatics were admitted to thehospital with acute asthma attack and CO measurements wereperformed after 2 weeks of attack while patients were receivingonly ß2 agonists. They did notreceive inhaled steroids during CO measurements, but after 1-2months, 9 of them could not be controlled with beta-agonists and wehad to treat them with corticosteroids. Thus 15 patients innon-steroid group had less severe asthma symptoms and 9 of them hadsimilar asthma severity compared to the steroid-received group. Allof the asthmatic children were included in the study provided thatthey were at symptom-free intervals at least for the 2 weeks duringtheir routine follow-up and they did not have any symptoms ofinfection. Knowledge about parental smoking habits at home wasobtained from parents using a questionnaire and an informed consentwas taken. In addition to questions about parental smoking habitsand how many cigarettes they smoked in house daily, some questionswere added to the questionnaire to clarify whether children haddoctor-diagnosed asthma, wheezing attacks or recent upperrespiratory infections. According to answers, children with thepast and current history of asthma, wheezing or recent upperrespiratory infections were excluded from the healthy group. Wehave taken the smoking habits of parents for the last 6 months intoaccount. All children were examined by the same physician,including lung auscultation. Carbon monoxide in expired air wasmeasured by using MicroCO Meter (Micro Medical, England). Time ofmeasurement was 0800 and 0900 hours a.m. inhealthy children and in the asthmatic group with no steroidtherapy, and 2-3 hours after receiving inhaled steroids in thesteroid receiving asthmatic group. The devide is based onelectrochemical fuel cell, which works through the reaction ofcarbon monoxide with an electrolyte at one electrode, and oxygen atthe other. This reaction generates an electrical currentproportional to CO concentration (5). Children were asked to expirethrough the device and the best value was recorded after threeattempts. Measurements of CO in 23 healthy and 5 younger asthmaticchildren were done by lack of adequate co-operation and thosechildren were excluded from the study. Asthma diagnosis wasestablished by using asthma criteria of American Thoracic Society(11). None of the asthmatics and healthy children was activesmoker.
Statisticalanalysis: results are expressed as mean ± Standarddeviation (SD). Due to skewed distribution of CO valuesnon-parametric Kruskal-Wallis one-way ANOVA and Mann-Whitney Utests were performed for comparison of data belonging to differentgroups. Spearman correlation analysis was performed to assess therelationship between the number of cigarettes smoked by parents perday and CO concentrations of children. Chi-square test was used totest differences between parental smoking habits of healthy andasthmatic children. P value less than 0.05 was accepted assignificant.
RESULTS
The mean age of healthysubjects was 4.4 ± 2.3 years (3-10 yrs) and asthmaticchildren was 4.5 ± 1.7 years. Male to female ratio was 1.4:1for healthy children and 1.8:1 for asthmatics. Parental smokinghabits of healthy and asthmatic children and their CO levels wereshown in table I. Parental smoking habits of asthmatic childrenwere similar to the healthy subjects (37.0% vs 33.6%non-smokers and 63.0% vs 66.4% smokers). When parentalsmoking habits were not taken into consideration, the COconcentration (mean ± SD) of asthmatics (1.32 ± 1.50ppm) was higher than those of healthy children (0.86 ± 1.35ppm, p = 0.028).
| Table I Influence of contact with smokers on carbon monoxidein expired air of healthy and asthmatic children (mean ±SD) | ||||||
| Healthy children (n = 235) | Asthmatics (n = 54) | Significance | ||||
| *n(%) | **CO(ppm) | *n(%) | **CO(ppm) | *P | **P | |
| Amount smoked in the home (cigarettes per day) | ||||||
| Zero | 79(33.6) | 0.37± 0.53 | 20(37.0) | 1.05± 1.55 | 0.01 | |
| 1-5 | 22(9.4) | 0.41± 0.33 | 11(20.4) | 0.77± 0.59 | NS | 0.03 |
| 6-10 | 63(26.8) | 0.87± 1.05 | 13(24.1) | 1.68± 1.75 | 0.03 | |
| 11-20 | 64(27.2) | 1.49± 1.60 | 10(18.5) | 2.02± 2.25 | NS | |
| >20 | 7(3.0) | 1.75± 2.31 | -- | -- | -- | |
| Parental smoking in the home | ||||||
| None | 79(33.6) | 0.37± 0.53 | 20(37.0) | 1.05± 1.55 | 0.01 | |
| Fatheralone | 24(10.3) | 0.94± 1.14 | 9(16.7) | 1.04± 0.73 | NS | NS |
| Motheralone | 37(15.7) | 1.10± 1.19 | 11(20.4) | 1.25± 0.96 | NS | |
| Bothparents | 95(40.4) | 1.18± 1.45 | 14(25.9) | 1.93± 2.42 | NS | |
| Contact with smokers other than parents | ||||||
| Yes | 128(54.4) | 0.92± 1.47 | 23(42.6) | 1.42± 1.52 | NS | NS |
| No | 107(45.6) | 0.79± 1.14 | 31(57.4) | 1.25± 1.40 | NS | |
NS: notsignificant, CO: carbon monoxide, *P: difference between parentalsmoking habits, according to Chi-square test; **P: differencebetween healthy and asthmatic children at each row according toMann-Whitney U test. | ||||||
In comparison ofasthmatic and healthy children of non-smoker parents, higher COconcentrations were found in asthmatic children than in healthysubjects (1.05 ± 1.55 ppm vs 0.37 ± 0.53 ppm,p = 0.01). Asthmatic children whose parents smoke 1-5 and 6-10cigarettes per day had also higher CO concentrations compared tohealthy children with similar parental smoking habits (p = 0.03, p= 0.03). There was no difference in CO levels between asthmatic andhealthy children whose parents smoke 11 or over cigarettes per dayat home (p > 0.05) (table I).
Children, whose neitherparents were smoker, had lowest exhaled CO concentrations. The meanCO levels of children whose mother or father smoke alone or bothparents smoke were similar (p > 0.05) (table I).
There was no differencebetween CO concentrations of boys and girls with Mann-Whitney Utest (p > 0.05). Age was not a predictor for exhaled COconcentration with Spearman's correlation analysis (p >0.05).
Significant relationshipbetween the number of smoking cigarettes in the house per day andexhaled CO concentrations were found in healthy (r = 0.35, p =0.003) and in asthmatic children (r = 0.44, p = 0.01) withSpearman's correlation analysis (table II).
| Table II Carbon monoxide levels according to gender and agein healthy children | |||
| n(%) | CO (ppm)(mean ± SD) | p | |
| Gender | |||
| Boys | 124(52.8) | 0.84± 1.47 | NS* |
| Girls | 111(47.2) | 0.87± 1.26 | |
| Age ofchild (Years) | |||
| 3-4 | 74(31.5) | 0.89 ±0.77 | |
| 5-6 | 62(26.3) | 0.75 ±0.71 | NS** |
| 7-8 | 57(24.3) | 0.86 ±1.96 | |
| 9-10 | 42(17.9) | 1.00 ±2.40 | |
| Total | 235(100) | 0.86± 1.35 | |
NS: notsignificant, *according to Mann-Whithey U test, **according toKruskal-Wallis test. | |||
Asthmatic children whodid not receive corticosteroids (n = 24) had higher COconcentrations (1.78 ± 1.53 ppm) than healthy children (0.86± 1.77 ppm, p = 0.022), and steroid-treated asthmaticpatients (0.96 ± 0.95 ppm, p = 0.02), whereas asthmaticchildren who received inhaled corticosteroids (n = 30) had similarCO concentrations in comparison with healthy children (p > 0.05)(table III).
| Table III Exhaled carbon monoxide concentrations (mean± SD) of asthmatic children who has taken or not inhaledsteroids, according to parental smoking habits | |||||
| Steroid group | Non-steroid group | Significance | |||
| Parentalsmoking habits | n | CO(ppm) | n | CO(ppm) | p |
| Number ofdaily smoked cigarettes | |||||
| Non-smoker | 12 | 0.58± 0.65 | 8 | 1.75± 1.45 | 0.024 |
| 1-5 | 6 | 0.57± 0.77 | 5 | 1.02± 1.05 | NS |
| 6-10 | 5 | 1.41± 1.15 | 8 | 1.85± 1.51 | NS |
| 11-20 | 7 | 1.64± 1.95 | 3 | 2.92± 2.05 | NS |
| >20 | -- | -- | |||
| Total | 30 | 0.96± 0.95 | 24 | 1.78± 1.53 | 0.02 |
In order to eliminate theeffect of passive smoking on the CO concentrations of asthmatics,we compared asthmatic children of non-smoker parents either treatedor not treated with inhaled corticosteroids (budesonid,200-400 µg per day). Thus, we foundsignificantly higher CO levels in children who did not receivesteroids (1.75 ± 1.45 ppm) than those who did (0.58 ±0.65 ppm) (p = 0.024) (table III).
DISCUSSION
In this study, wecompared data about passive smoking obtained from parentalquestionnaire and CO in expired air measured by a portable device.We found significant relationship between concentrations of COmeasured in expired air and intensity of exposure to tobacco smoke(ETS) assessed by parental questionnaire. Children having parentalsmoking are often exposed to higher levels of ETS. In the study ofIrvine, et al (1) many children of 501 families were exposed tohigh levels of environmental tobacco smoke and their cotininelevels were heavily dependent upon to the parental smoking. Sinceparents of asthmatic children usually learn the detrimental effectsof cigarette smoking on their children's health, they reduce theamount of indoor smoking near their children (12). This may be animportant factor that prevented the excessive levels of exhaled COin asthmatic children.
The effect of ETS asmeasured by the number of cigarettes smoked by parents, are likelyto be different between cold climate countries and other culturesin which exposure is effectively reduced because the lack of tightsealing in homes increases ventilation rates. Questionnaires aregenerally used to measure the history of exposure to ETS but totalexposure can be difficult to estimate from questionnaire becauseparents may change their smoking habits after the development ofsymptoms in their child (13). Our region is in South easternAnatolia with a hot climate that only in four months a year peopleneed indoor heating and there is no industrial factories. Wecarried out this study in spring and summer months and we cansuggest that the effect of air pollution on children was minimal.Passive smoking is responsible for respiratory morbidity inchildren (14, 15). Some studies have failed to show a causal effectof passive smoking on incidence of asthma and demonstratedincreased morbidity among asthmatic children (16). In most of thesestudies passive smoking exposure of the population was ascertainedby questionnaires (17, 18). In one study a correlation was foundbetween exposure to ETS and the child's carboxy hemoglobin (COHb)determined by direct measurement of capillary blood COHb levels(6). In the study of Rylander, et al (19) exposure to ETS wasestimated from urinary cotinine measurement in children withwheezing bronchitis and control subjects. They found that ETS is animportant risk factor for wheezing bronchitis and a single urinarycotinine measurement offer no major advantages to questionnairedata for assessment of long- term exposure to ETS (19).
Our results indicatedhigher exhaled CO concentrations in asthmatic children who did notreceive inhaled corticosteroids in comparison with healthy subjectsand asthmatic children who received inhaled steroids. These highlevels of exhaled CO concentration may reflect inflammation of thelung. In asthmatic inflammation many cytokines are involved,including interleukin-1, interleukin-6 and tumour necrosis factorwhich can upregulate heme oxygenase-1 activity in human tissues(10). The normal CO concentration in expired air of asthmaticchildren who received corticosteroids suggest that inhaledcorticosteroids downregulate heme oxygenase-1 activities throughreduction of inflammatory cytokines. In the study of Zayasu, et al(20) increased exhaled CO concentrations were found in adultasthmatic patients not receiving corticosteroids compared tocontrol subjects. Our study have shown similar results inchildren.
In conclusion, ourresults indicate that measurement of CO in expired air may be areliable index for exposure to tobacco smoke in healthy children aswell as a useful non-invasive marker of airway inflammation inasthmatics, since exhaled CO were found to be increased inasthmatic children and decreased with inhaled steroids.
ACKNOWLEDGEMENT
This paper was presentedin part (Allergy supl. 51, vol 54, p: 28) at the Joint meeting ofEuropean Respiratory Society Paediatric Assembly and EuropeanSociety of Pediatric Allergy and Clinical Immunology in Berlin,Germany in 26-29 May 1999.
