Passive smoking and lower respiratory tract illnesses in children

PDF version

Ihab Hafez El-Sawy, Fatma M. Kamel Nasr, Ehsan Wafa E. Mowafy, Ola Atef M. Sharaki and Adel M. Abdel Bakey

ABSTRACT: The relationship between household tobacco smoke exposure and wheezing and nonwheezing lower respiratory tract illnesses in young children was investigated using both a questionnaire and the urinary cotinine/creatinine ratio to assess passive smoking. A case-control study was conducted on 60 children (≤5 years) with lower respiratory tract illnesses and 40 control children. The results showed that household passive smoking may be a predisposing and/or aggravating factor for lower respiratory tract illnesses in young children whether wheezing or nonwheezing, infective or noninfective. Parents' assessment of their own level of smoking may be inaccurate and objective biochemical measures of passive smoking are needed to identify its risks.

Le tabagisme passif et les maladies des voies respiratoires inférieures chez les enfants

RESUME: La relation entre l'exposition à la fumée de tabac dans le milieu familial et les maladies des voies respiratoires inférieures avec ou sans respiration sifflante chez les jeunes enfants a été étudiée au moyen d'un questionnaire et en utilisant le taux de créatine/cotinine urinaire pour évaluer le tabagisme passif. Une étude cas-témoins a été réalisée chez 60 enfants (≤ 5 ans) atteints de maladies des voies respiratoires inférieures et chez 40 enfants témoins. Les résultats ont montré que le tabagisme passif au foyer pouvait être un facteur prédisposant et/ou aggravant chez les jeunes enfants pour les maladies des voies respiratoires inférieures avec ou sans respiration sifflante, infectieuses ou non. L'évaluation faite par les parents de leur propre niveau de tabagisme peut être inexacte et des mesures biochimiques objectives du tabagisme passif sont nécessaires pour identifier les risques liés à celui-ci.

Introduction

Tobacco smoke is probably the most important source of air pollution in the home. Many studies have reported that children exposed to parental cigarette smoking may develop a chronic cough, phlegm and persistent wheezing (asthma), and may have higher risks of attacks of pneumonia and other respiratory illnesses [1-8]. Additional evidence suggests that parental smoke may adversely affect the rate of lung growth during childhood [9] and children whose parents smoke have also been found to differ in lung function [10,11].

Many researchers have investigated the relationship between passive smoking and wheezing illnesses [3,6,7,12,13] but few have examined the relationship with other types of lower respiratory tract illnesses (LRTIs) [14,15]. Also many studies have identified illness outcomes solely on the basis of parental recall, which may bias such reports [1,2,3,4,6,7,14]. The current study was conducted therefore to investigate the relationship of household tobacco smoke exposure to wheezing and nonwheezing LRTIs in young children using both a questionnaire and the urinary cotinine/creatinine ratio (CCR) to assess passive smoking.

Subjects and methods

A case-control study was conducted from September 1995 to October 1996 on 100 children (≤ 5 years) from low income families who were receiving health care at the Alexandria University children's hospital, Alexandria, Egypt. The study was approved by the Researches Committee of Alexandria Faculty of Medicine and informed consent to participate in the study was obtained from the caregivers of each child. The study population comprised 60 consecutive cases with LRTIs, whether wheezing or nonwheezing, and 40 children of matched age and sex with any presentation other than LRTIs and who also had no previous history of LRTIs. Children (cases and controls) who were reported by their caretakers to have been continuously exposed to household tobacco smoke since birth and those who were never exposed were included, while children who were intermittently exposed were excluded from the study population. Children with any hepatic or renal disorders were also excluded as the metabolism and excretion of cotinine could be affected by disturbed hepatic and renal functions.

Using a specially designed questionnaire, the accompanying parents/caregivers were interviewed in detail about smoking habits to identify smoking individuals at home (mother, father and/or others), duration of exposure, and the average daily cigarette consumption by all smokers at the place where the child was present (especially during the week before the interview). At the time of interview, the investigators completed a standardized form on the presence or absence of a list of symptoms and signs and made a diagnosis, which was confirmed by chest radiography. The severity of LRTIs was assessed by clinical and radiological findings, in addition to oxygen saturation using pulse oximetry [16].

Urinary cotinine/creatinine ratio

Urinary cotinine/creatinine ratio (CCR) was measured for all cases and control children to verify the reported smoking habits. A urine specimen was collected from each child at the time of interview, labelled with a code number and kept frozen at -20؛C until being analysed for urinary creatinine and cotinine without knowledge of the exposure state or the diagnosis of the child. Creatinine reacts in alkaline solution with picrate to form an orange complex which is measured colorimetrically [17]. The urinary cotinine was measured by an ¹²⁵I-radioimmunoassay procedure [18,19]. It is a liquid phase radioimmunoassay in which ¹²⁵I-labelled cotinine competes for a fixed time with cotinine in the patient's urine sample for antibody sites. Patient sample concentrations were read from a calibration curve and the results expressed in ng/ml.To adjust for the effect of variable dilution on the spot concentration of cotinine, urinary CCR was calculated, expressed as nanograms of cotinine per milligram of creatinine (ng/mg). Henderson et al. reported that a cutoff CCR of 30 ng/mg identifies children exposed to smoking at home with a high degree of sensitivity (80%) and specificity (100%) [20]. This cutoff was therefore used to categorize our patients into exposed and not exposed.

Statistical analysis

The chi-squared test (χ²), the t-test, the F-test, the least significant difference test (LSD), the Z-test and the correlation coefficient (r) were used as tests of significance. The odds ratio (OR) was used to estimate the risk of LRTIs associated with tobacco smoke exposure. The 95% confidence interval (95% CI) for the odds ratio was also used to conclude whether the association was statistically significant or not. The level of significance was considered at P < 0.05.

Results

The demographic features (age, sex and residency) which might affect the results were similar in children with LRTIs and those in the control group (Table 1). The different diagnoses of children with LRTIs and control children are also shown in Table 1.

Comparing the reported exposure variables, namely proportions having any smoker at home (reported exposure) and the mean daily cigarette consumption by all smokers, there was no significant difference between children with LRTIs and those in the control group (Table 2). However, the percentage of exposed children (measured exposure with CCR ≥ 30 ng/mg) was significantly higher in the LRTI group than the control group (85.0% versus 67.5% respectively, P = 0.03). It was also found that the exposed children (CCR ≥ 30 ng/mg) had a 2.7 times greater risk of developing LRTIs than unexposed children (CCR < 30 ng/mg). This risk was significant (OR = 2.7, 95% CI = 1.02-9.54). Moreover, the mean urinary CCR was higher in children with LRTIs than control children (138 ng/mg versus 46 ng/mg respectively); this difference was statistically highly significant (P = 0.00005). It was found that, in children with LRTIs, the percentage of exposed children (CCR ≥ 30 ng/mg) was significantly higher than the percentage of reported exposure (85.0% and 66.7% respectively, P = 0.02), while in the control group, no significant difference was found between measured and reported exposures (67.5% and 60.0% respectively, P > 0.05).

With regard to the reported degree of exposure (none, a little, some, a lot), the mean CCR level was significantly higher in children whose parents reported a greater degree of exposure whether in cases or controls (P < 0.05 for both). No smoking mothers were reported in the study population. According to the different smoking groups (none, only father, father and others), CCR levels were significantly higher in cases than controls (P r = 0.121, P > 0.05) or in control children (r = 0.050, P > 0.05).

Comparing the reported exposure variables of children with wheezing LRTIs, those with nonwheezing LRTIs and the controls, there was no significant difference in the proportion having any smoker at home or in the mean daily cigarette consumption by all smokers (P > 0.05 for both variables) (Table 3). The mean CCR values were 144 ng/mg in the wheezing group, 128 ng/mg in the nonwheezing group and 46 ng/mg in the control group. The least significant difference (LSD) test showed an absence of significant difference between wheezing and nonwheezing groups, while both had significantly higher values than the control group.

No significant difference was found between children with severe LRTIs, those with mild LRTIs and the controls as regards the percentage having any smoker at home (P > 0.05) (Table 4). On the other hand, the LSD test showed that the reported mean daily cigarette consumption by all smokers and the CCR level were significantly higher in the group with severe LRTIs (8.6 cigarette/day and 154 ng/mg respectively) compared that with mild LRTIs (3.8 cigarette/day and 88 ng/mg) and the control group (4.8 cigarette/day and 46 ng/mg). Comparing the same variables between the last two groups showed no significant difference.

The LSD test showed a significant difference in CCR between each of the diagnostic subgroups of LRTIs (being higher) in comparison with the control group (Table 5). On the other hand, no significant difference was found when each of the diagnostic subgroups was compared with the others.

There was no significant difference in the mean urinary CCR between children with LRTIs in the age group ≤ 2 years and those > 2 years (mean = 153 ± 131.1 ng/mg and 135 ± 112.6 ng/mg respectively, P > 0.05). There was also no significant difference in the mean urinary CCR between the same age groups of control children (mean = 51 ± 25.5 ng/mg and 45 ± 21.3 ng/mg respectively, P > 0.05). The mean age at first LRTI was lower in exposed children compared with unexposed ones (mean = 6.4 ± 5.3 months and 9.8 ± 9.3 months respectively, P < 0.05).

Discussion

Accurate biochemical measures of passive smoking are needed to identify its risks and to quantify the benefits of antismoking interventions. Cotinine, the major metabolite of nicotine found in saliva, blood and urine, is a stable and accurate indicator of cigarette smoke exposure [20-22]. Measurement of cotinine in biological fluids has been found to be the most reliable method of estimating individual levels of exposure to tobacco [18,20,21]. As the half-life of cotinine in the blood is about 19-40 hours, urinary cotinine levels might therefore reflect exposure to tobacco smoke during the preceding few days [21,23,24]. Also the probability of contamination during sampling is low because cotinine is specific to tobacco and exclusively a product of in vivo metabolism [21,23]. For these reasons, this study was undertaken using urinary CCR as an objective measure of passive smoking to investigate the relationship between household tobacco smoke exposure and wheezing and nonwheezing LRTIs in preschool children.

As to whether cotinine is found in the urine of all people or only in those known to have been exposed to cigarette smoke, Henderson et al. [20] studied children who were known to be either exposed or not exposed to cigarette smoke at home (verified by home air sampling for nicotine). Small amounts of cotinine were detected in some unexposed children's urine samples, but a urinary CCR of less than 30 ng/mg could be used as a threshold level to accurately classify children as not exposed. Considering this cutoff, 78% of our study population were exposed to tobacco smoke; this rate of household cigarette smoke exposure is disturbingly high. A higher measured exposure rate was found in the group with LRTIs but not in the control group. CCR levels in children with LRTIs whose caregivers reported absence of household tobacco exposure ranged from 0 to 100 ng/mg with a mean of 48 ± 19.7 ng/mg, which exceeded the cutoff level of home exposure. This inaccurate reporting of exposure may have been due to the parents becoming more sensitized to the issue of passive smoke exposure by the study questionnaire itself and perhaps wanting to minimize the reported exposure. Outdoor accidental exposure (which was not included in the questionnaire) may be another explanation for this discrepancy. Our results are consistent with other studies which have shown that cotinine is measurable, sometimes at high levels, in children with no reported exposure at home [19,25].

Measurement of urinary CCR showed that children with LRTIs (wheezing or nonwheezing) were definitely more exposed to cigarette smoke than control children. On the other hand, analysis of reported exposure variables (percentage having any smoker at home and daily cigarette consumption by all smokers) showed an absence of significant differences between those with LRTIs (wheezing or nonwheezing) and those in the control group. Moreover, there was no correlation between CCR values of children and the number of cigarettes smoked in their households. Therefore, depending only on the parents' subjective assessment of their own level of smoking to quantify their children's exposure to tobacco smoke is potentially inaccurate as factors, such as room ventilation, proximity to the smokers and the presence of other smokers, are likely to cause a wide variation in exposure. These factors, in addition to parents' recall (which may be biased), indicate the importance of using an objective measure of exposure to evaluate more critically a causal relationship between passive smoking and childhood ill-health.

In the present study, exposed children (CCR ≥ 30 ng/mg) were found to have a 2.7 times greater risk of developing LRTIs than unexposed children (CCR < 30 ng/mg). Also the significantly higher mean CCR of children with LRTIs persisted when each of the diagnostic subgroups of LRTI patients was compared with control children. However, there was no significant difference in CCR values when each of diagnostic subgroups was compared with the others. These findings prove that passive smoking is a risk factor for many types of LRTIs whether wheezing or nonwheezing, infective or noninfective. Our results are consistent with those of many investigators who have found that environmental tobacco smoke produces an increased risk of the development of acute lower respiratory tract irritation, asthma and acute lower respiratory tract infections in children exposed in the home [1-8,14,15]. Unlike our study, most of these studies were based only on parents' subjective assessment of their own level of smoking [1-4,7,14,15], while other studies using an objective measure of tobacco smoke exposure investigated mainly wheezy children with asthma and bronchiolitis [12,19,25,26].

Heavy exposure to tobacco smoke might be an aggravating factor for LRTIs in young children as indicated by the significantly higher level of CCR in patients with severe episodes of LRTIs compared with mild LRTI cases or control children. Our findings are consistent with those of Chen et al. who suggested that exposure to household cigarette smoke in early life increases the risk of severe respiratory illness [27]. In addition, Willers et al. [12] and Reese et al. [26] found that passive smoking might be an aggravating factor for childhood asthma and bronchiolitis respectively.

It should be noted that most North American and European literature [13,14] has linked LRTIs and passive smoking to maternal rather than paternal smoking, and the correlation has appeared to be most marked for the first two years of life and has shown a steady decline with increasing age. The authors have attributed these findings to the fact that mothers usually spend more time with their infants than do fathers, especially in early life. Also, the children in these studies attended day care centres when they reached the age of two years so they spent some time outdoors away from parental smoking. However, in the present study, maternal smoking was not reported in any family in the study population. This might be real or explained by cultural factors and the stigma regarding women smoking in the Egyptian community, which may have led to false answers. Therefore, in the present study the relation between household smokers and LRTIs was restricted to fathers or fathers and others. In addition, we demonstrated an absence of significant difference in the degree of exposure between children below or above two years of age. This could be explained by the fact that children in our study did not attend day care centres and spent most of the time indoors with their smoking fathers and/or others.

Regarding the age at which the first LRTI occurred, we found that exposed children had their first attack at a younger age than unexposed children. This is consistent with the findings of Wright et al., who reported that when the amount smoked was considered, children whose mothers were heavy smokers were younger at first LRTI and at both first wheezing and first nonwheezing LRTI than other children [14]. This relationship was also found by other investigators [28,29], who reported that younger children may be more susceptible to smoke exposure and its hazards because of their proximity to their parents and also because of the presence of a number of anatomical and physiological peculiarities of their airways or their immunological status.

Conclusion and recommendations

Household cigarette smoke exposure of preschool children may be a predisposing and/or aggravating factor for LRTIs whether wheezing or nonwheezing, infective or noninfective. A questionnaire for measurement of passive smoking may inadequately reflect the child's dose of environmental tobacco smoke and the use of an objective measure (e.g. CCR) is more accurate to evaluate critically a causal relationship between tobacco smoke exposure and childhood ill-health.

The findings of this study should prompt renewed efforts to discourage smoking in families, especially during the first five years of children's lives. It is extremely unlikely that we will ever be willing or able to regulate the smoking of adults in their own homes; therefore, we must employ strategies other than coercion to help parents reduce their smoking, both for their own health as well as for their children's well-being. It is suggested that paediatricians should increase their efforts to inform parents about the hazards of passive smoking to their children, especially the association between parental cigarette smoking and the increased risk of a child developing lower respiratory tract illnesses. This strategy may help parents stop smoking in the interest of their children's health.

References

1. El-Nawawy A et al. Effect of passive smoking on the frequency of respiratory illnesses and serum immunoglobulin-E (IgE) and interleukin-4 (IL-4) concentrations in exposed children. Journal of tropical pediatrics, 1996, 42:166-9.
2. Marbury MC, Maldonado G, Waller L. The indoor air and children's health study: methods and incidence rates. Epidemiology, 1996, 7:166-74.
3. Abramson MJ, Marks GB, Pattemore PK. Are non-allergenic environmental factors important in asthma? Medical journal of Australia, 1995, 163:542-5.
4. Volkmer RE et al. The prevalence of respiratory symptoms in South Australian preschool children. II. Factors associated with indoor air quality. Journal of paediatrics and child health, 1995, 31:116-20.
5. Environmental tobacco smoke. Health effects and prevention policies. Council on Scientific Affairs, American Medical Association. Archives of family medicine, 1994, 3:865-71.
6. Rylander E et al. Parental smoking and other risk factors for wheezing bronchitis in children. European journal of epidemiology, 1993, 9:517-26.
7. Duff AL et al. Risk factors for acute wheezing in infants and children: viruses, passive smoke, and IgE antibodies to inhalant allergens. Pediatrics, 1993, 92:535-40.
8. Holberg CJ et al. Child day care, smoking by caregivers and lower respiratory tract illness in the first three years of life. Pediatrics, 1993, 91:885-92.
9. Lebowitz MD, Sherrill D. Effects of passive smoking on lung growth in children. Pediatric pulmonology, 1992, 12:37-42.
10. Haby MM, Peat JK, Woolcock AJ. Effect of passive smoking, asthma and respiratory infection on lung function in Australian children. Pediatric pulmonology, 1994, 18:323-9.
11. Sherrill DL et al. Longitudinal effects of passive smoking on pulmonary function in New Zealand children. American review of respiratory disease, 1992, 145:1136-41.
12. Willers S, Svenonius E, Skarping G. Passive smoking and childhood asthma. Allergy, 1991, 46:330-4.
13. Weitzman M et al. Maternal smoking and childhood asthma. Pediatrics, 1990, 85:505-11.
14. Wright AL et al. Relationship of parental smoking to wheezing and nonwheezing lower respiratory tract illnesses in infancy. Group Health Medical Associate. Journal of pediatrics, 1991, 118:207-14.
15. Jin C, Rossignol AM. Effects of passive smoking on respiratory illness from birth to age eighteen months in Shanghai, People's Republic of China. Journal of pediatrics, 1993, 123:553-8.
16. Bishop J, Nolan T. Pulse oximetry in acute asthma. Archives of disease in childhood, 1991, 66:724-5.
17. Larsen K. Creatinine assay by a reaction-kinetic principle. Clinica chimica acta, 1972, 41:209-11.
18. Knight GJ et al. Exposure to environmental tobacco smoke measured by cotinine ¹²⁵I-radioimmunoassay. Clinical chemistry, 1989, 35:1036-9.
19. Ogborn CJ, Duggan AK, DeAngelis C. Urinary cotinine as a measure of passive smoke exposure in asthmatic children. Clinical pediatrics, 1994, 33:220-6.
20. Henderson FW et al. Home air nicotine levels and urinary cotinine excretion in preschool children. American review of respiratory disease, 1989, 140:197-201.
21. Wald NJ et al. Urinary cotinine as a marker of breathing other people's tobacco smoke. Lancet, 1984, 1:230-1.
22. Greenberg RA et al. Measuring the exposure of infants to tobacco smoke: nicotine and cotinine in urine and saliva. New England journal of medicine, 1984, 310:1075-8.
23. Benowitz NL et al. Cotinine deposition and effects. Clinical pharmacology and therapeutics, 1983, 34:604-11.
24. Jarvis MJ et al. Elimination of cotinine from body fluids: implication for non-invasive measurement of tobacco smoke exposure. American journal of public health, 1988, 78:696-8.
25. Elrich R et al. Childhood asthma and passive smoking. Urinary cotinine as a biomarker of exposure. American review of respiratory disease, 1992, 145:594-9.
26. Reese AC et al. Relationship between urinary cotinine level and diagnosis in children admitted to hospital. American review of respiratory disease, 1992, 146:66-70.
27. Chen Y, Li W, Yu S. Influence of passive smoking on admissions for respiratory illness in early childhood. British medical journal, 1986, 293:303-6.
28. Reid L. Influence of the pattern of structural growth of lung on susceptibility to specific infectious diseases in infants and children. Pediatric research, 1977, 11:210-5.
29. Martinez FD et al. Diminished lung function as a predisposing factor for wheezing respiratory illness in infants. New England journal of medicine, 1988, 319:1112-7.