- Email: [email protected]
Decreasing trend in passive tobacco smoke exposure and association with asthma in U.S. children
Environmental Research 166 (2018) 35–41
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Decreasing trend in passive tobacco smoke exposure and association with asthma in U.S. children Xiao Zhanga, Natalie Johnsonb, Genny Carrillob, Xiaohui Xua, a b
T
⁎
Department of Epidemiology and Biostatistics, School of Public Health, Texas A&M University, United States Department of Environmental and Occupational Health, School of Public Health, Texas A&M University, United States
A R T I C LE I N FO
A B S T R A C T
Keywords: Passive smoke Cotinine Asthma Trend Children Prevalence
In this study, we assessed trends of serum cotinine levels over time among US children ages 3–11 years and compared the risk of asthma in groups exposed to passive tobacco smoke. We utilized National Health and Nutrition Examination Survey (NHANES) data collected from 2003 to 2014 (n = 8064). Serum cotinine level, household smoker status, asthma status, and sociodemographic information were extracted for multiple regression analyses. The adjusted biannual change in log (cotinine) in comparison to earlier NHANES survey cycles was − 0.196 (p < 0.001) overall, − 0.055 (p = 0.089) among children with household smoker(s), and − 0.129 (p < 0.001) among children without. The proportion of children living with household smokers decreased from 24.9% in the 2003–2004 cycle to 11.4% in the 2013–2014 cycle. The adjusted odds ratios (ORs) for asthma were 1.34 (95% confidence interval (CI): 1.00–1.80; 2nd tertile vs 1st tertile) and 1.69 (95%CI: 1.25–2.29; 3rd tertile vs1st tertile), respectively. Highly exposed asthmatic children, in the 3rd cotinine tertile (> 0.13 ng/mL), were primarily Non-Hispanic Black (61.0%) and whose family incomes were below poverty guidelines. Overall results reveal passive smoke exposure level among children ages 3–11 in the US decreased over the study period. Nevertheless, higher exposure to passive smoke is still associated with higher odds of childhood asthma. Targeted smoking cessation interventions in clinical practices are needed to reduce tobacco smoke exposure and related asthma risk in children, particularly in low-income and minority groups.
1. Introduction Passive tobacco smoke is defined as smoke emitted from burning cigarettes or exhaled smoke by the person smoking (Centers for disease control and prevention (CDC), 2015). Passive tobacco smoke contains thousands of chemical compounds, including the same chemicals inhaled by smokers (American Cancer Society, 2015). These chemical mixtures are reported to have detrimental health risks, and at least 70 chemical constituents are classified as carcinogens (U.S. Department of Health and Human Services, 2006). According to CDC, roughly 2,500,000 nonsmokers have died from health problems caused by passive tobacco smoke exposure since 1964, and the cost of lost productivity due to tobacco smoke exposure is estimated at ~$6.6 billion per year (Max et al., 2012). Compared to adults, children are more vulnerable to environmental passive tobacco smoke since most of the time they have no control over the exposure and also represent a biologically susceptible group due to higher breathing rates and developing immune systems (Silvestri et al., 2015). Children can be exposed to passive tobacco smoke at home if
they live with smokers or have smoking visitors. Even if smokers do not reside in their homes, children may still be exposed to environmental tobacco smoke in public places. In the U.S., governmental actions regarding the regulation of tobacco sales, marketing, and use have been implemented since 1960s. According to the Pro-Children Act of 1994, all federal funded children's services must be smoke-free (CDC, 2013b). As of December 31, 2010, 26 states had enacted comprehensive smokefree laws (Tynan et al., 2011). In 2008–2009, the American Academy of Pediatrics and the U.S. Public Health Services updated guidelines to recommend pediatricians and clinicians to offer advice and interventions to smoking youths and their smoking patients (American Academy of Pediatrics, 2009; Fiore et al., 2008). The World Health Organization (WHO) has developed six key tobacco control policies called MPOWER (WHO, 2009) to protect people from tobacco smoke exposure (WHO FCTC, 2003). Studies have reported declines in prevalence of passive tobacco smoke exposure among children from 1988 to 2010 (Kit et al., 2013), 1999–2000–2011–2012 (Homa et al., 2015; Jain, 2016). However, whether there is a continuously decreasing trend of children's passive tobacco smoke exposure levels recently is not readily known,
⁎ Correspondence to: Department of Epidemiology & Biostatistics, School of Public Health, Texas A&M University, MS 1266, 212 Adriance Lab Road, College Station, TX 77843-1266, USA. E-mail address: [email protected] (X. Xu).
https://doi.org/10.1016/j.envres.2018.05.022 Received 12 April 2018; Received in revised form 15 May 2018; Accepted 20 May 2018 0013-9351/ © 2018 Elsevier Inc. All rights reserved.
Environmental Research 166 (2018) 35–41
X. Zhang et al.
home?” If the answer was yes, then we classified the participant as having likely been exposed to the home-source tobacco smoke; if the answer was no we classified the participant as having not been exposed to home-source tobacco smoke, suggesting that he/she might have been primarily exposed to non-home source tobacco smoke or have not been exposed to tobacco smoke from any source at all. For 2013–2014 survey cycle, the question was removed, and therefore we selected another question asking for “the number of people who live here smoke tobacco” to identify children with household smokers. If the answer was 0, then we classified the participant as having no home-source tobacco smoke exposure.
specifically at the national level across multiple demographic groups. Moreover, considering the potential disparities in passive tobacco smoke levels among children with household smokers and those without, it is worthwhile to evaluate differences in the trends of exposure levels to support targeted interventions. Previous investigations have linked passive tobacco smoke with various diseases in children, among which asthma is significant because it is a prevalent disease in children representing the leading cause of school absences. Although there is a biological plausibility for passive tobacco smoke exposure and childhood asthma, a proportion of studies lack strong evidence or have not identified a positive association (Martín-Pujol et al., 2013; Vlaski et al., 2011). These inconsistent findings may be partially due to differences in exposure assessment methods across studies. The majority of previous studies used questionnaires, an indirect exposure assessment method, to collect information about passive tobacco smoke exposure status. While this approach is easy to obtain and not costly, exposure misclassification issues often result. By comparison, direct measurement, such as biomarkers of tobacco smoke exposure, are able to quantify exposure levels, especially when self-reported measures (e.g., hours per day exposed to passive smoke, proximity to smokers, room ventilation) are much harder to quantify. Serum cotinine, the major proximate metabolite of nicotine, is considered a valid marker of tobacco smoke exposure (Benowitz, 1996). Thus, using serum cotinine measurements to assess exposure levels may help to obtain more accurate effect estimates of passive smoke exposure on risk of children's asthma. Additionally, direct exposure assessment through biomarker measurement can easily discriminate population subgroups at the upper and lower exposure levels. In this study, we evaluated the difference in the trends of passive tobacco smoke exposure levels over time among US children ages 3–11 years using National Health and Nutrition and Examination Survey (NHANES) data from 2003 to 2014 (n = 8064). Serum cotinine levels were used as a direct exposure measurement to examine the association between passive tobacco smoke exposure levels by tertile and asthma, as well as to identify sociodemographic groups of asthmatic children with high passive tobacco smoke exposure levels.
2.3. Respiratory health assessment Determination of asthma status was based on the answer to “Ever been told you have asthma” and “Still have asthma” in the medical condition questionnaire. Children answered “Yes” to both questions were classified as having asthma. “Don’t know” and “Refused” were treated as missing data. 2.4. Covariates
2. Materials and methods
Data on age, gender, race-ethnicity, and survey cycle were extracted from the demographic data files. Family income/poverty ratio was used as a surrogate measure of socioeconomic status (SES), and lower values indicate higher poverty level. It is a ratio of family income to poverty guidelines, which is calculated by dividing family income by the poverty guidelines specific to the survey cycle. The poverty guidelines are issued every year in the Federal Register, which vary by family size and geographic location. Detailed information has been described elsewhere (CDC, 2013a). All these variables were considered as potential confounders in the relationships between serum cotinine level and adverse respiratory outcomes based on previous studies (Ding et al., 2015; Kanchongkittiphon et al., 2014) and adjusted for in the logistic regression models. Except for survey cycle, all of these variables were considered as confounders and added in the multiple linear regression models to examine the trends of passive tobacco smoke exposure in children over time.
2.1. Study population
2.5. Statistical Analysis
We utilized the data from six consecutive cycles of the 2003–2014 NHANES, a national cross-sectional survey assessing the health and nutritional status of the U.S. population. Informed consent was obtained from all survey participants in the NHANES. Study protocols were approved by the National Center for Health Statistics Research Ethics Review Board. Demographic data file, cotinine data file (laboratory data), household smoker questionnaire, and medical condition questionnaire were downloaded from each survey cycle and merged. This study was limited to children 3–11 years.
2.5.1. Descriptive analysis To account for the right-skewness of the serum cotinine data, natural log transformation was performed. Log (serum cotinine) in each sociodemographic group is presented using weighted geometric mean and 95% confidence interval (CI). We used the Wald test to compare the weighted geometric means in each of the categories of the covariates. We also divided serum cotinine levels into tertiles using data from all children aged 3–11 years, therefore the cutoff points are the same for children in different sociodemographic groups. Proportion of children who had asthma and proportion of asthmatic children who were in the 3rd cotinine tertile (highest exposure) are presented for each sociodemographic group. Comparisons of cotinine levels between the groups with and without household smokers for each survey cycle are presented in box plots. Percentage change of serum cotinine levels comparing the later survey cycle to 2003–2004 and proportions of children with household smokers for each survey cycle are presented in a line chart.
2.2. Passive tobacco smoke exposure assessment Serum cotinine level was collected from participants aged 3 years and older at each of the six survey cycles (2003–2014). Serum samples were obtained at the Mobile Examination Center (MEC). In the original laboratory data files, results below limit of detection (LOD) were replaced by LOD / 2 . The LOD was constant for all survey cycles. Detailed information regarding the measurement methods and detection limits are described elsewhere (CDC, 2010). The proportions of participants whose serum cotinine was below LOD was 13.68%, 17.62%, 16.96%, 23.84%, 25.89%, and 34.53% for each of the six survey cycles, respectively. For 2003–2012, determination of the presence of household smoker (s) was based on the answer to the question in the household smoker questionnaire open to all participants - “Does anyone smoke inside
2.5.2. Regression analysis (1) To evaluate the trend in passive tobacco smoke exposure in children over time, multiple linear regression models describing the association between log (serum cotinine) and survey cycle (continuous) were fitted to the data stratified by home source tobacco smoke exposure status (Yes/No). Categorical variables, including age (3–6, 36
Environmental Research 166 (2018) 35–41
X. Zhang et al.
7–9, 10–11; in years), gender (Male/Female), race-ethnicity (NonHispanic White, Non-Hispanic Black, Hispanic, Other race including multiracial), and family income/poverty ratio (0–1, > 1–2, > 2–3, > 3) were adjusted for in the regression models. Number of cigarettes smoked inside room was also added in the models when assessing the association among children with household smoker(s). (2) To examine the association between passive tobacco smoke exposure and asthma in children, multiple logistic regression analyses relating tertiles of serum cotinine levels to asthma (Yes/No) were performed. Categorical variables, including age (3–6, 7–9, 10–11; in years), gender (Male/Female), race-ethnicity (Non-Hispanic White, Non-Hispanic Black, Hispanic, Other race including multiracial), family income/poverty ratio (0–1, > 1–2, > 2–3, > 3), and survey cycle (2003–2004, 2005–2006, 2007–2008, 2009–2010, 2011–2012, 2013–2014) were adjusted for in the regression models. Odds of asthma were compared between children in the 3rd (high exposure) or 2nd (median exposure) tertile and those in the 1st (low exposure) tertile group.
Table 1 Summary statistics for serum cotinine levels (ng/mL) by selected characteristics, NHANES, 2003–2014. Cotinine (ng/mL) Na (%a,b)
Meanb (95%CI)c
Total
8064 (100.0)
0.076 (0.067, 0.087) 0.084)
4116 (52.0) 3948 (48.0)
0.075 (0.065, 0.086) 0.078 (0.067, 0.089)
0.5250
3253 (39.2) 2802 (35.0) 2009 (25.8)
0.090 (0.076, 0.106) 0.073 (0.063, 0.084) 0.063 (0.054, 0.072)
< 0.0001
2197 (54.4) 2154 (14.4) 2985 (23.1) 728 (8.1)
0.082 0.185 0.039 0.064
(0.067, (0.156, (0.035, (0.052,
0.100) 0.220) 0.043) 0.080)
< 0.0001
2824 2118 1022 1660
0.182 0.102 0.060 0.034
(0.150, (0.085, (0.049, (0.029,
0.220) 0.122) 0.074) 0.039)
< 0.0001
Gender Male Female Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other racee Income/poverty ratio 0–1 > 1–2 > 2–3 >3 Any smoker inside home No Yes Number of cigarettes smoked 0 1–10 11–20 21–30 31–39 > = 40 Survey cycle 2003–2004 2005–2006 2007–2008 2009–2010 2011–2012 2013–2014 Asthmaf Yes No
Statistical analyses were conducted using Stata (version 14.0, Stata Corp, College Station, Texas, USA) and R (R version 3.3.3 (2017-0306)). Sampling weight was 1/6 × examination sample weight and implemented into each analysis. Sampling design variables were included to account for the complex and multistage survey design. P-values < 0.05 were considered significant. 3. Results A total of 8064 children ages 3–11 years were analyzed in our study. As shown in Table 1, serum cotinine level was significantly associated with age, race-ethnicity, family income/poverty ratio, whether there is any household smoker, number of cigarettes smoked inside home, survey cycle, and asthma status. Without adjustment, children of younger age, lower family income/poverty ratio, greater number of cigarettes smoked inside home, Non-Hispanic Black ethnicity, and who had asthma were more likely to have higher geometric means of serum cotinine level. The lower and upper tertile cutoff points of serum cotinine levels among children ages 3–11 years were 0.02 ng/mL and 0.13 ng/mL, respectively. More than 50% of asthmatic children who are Non-Hispanic Black or whose family incomes were below poverty guideline (family income/poverty ratio < 1) were in the 3rd tertile of serum cotinine levels (Table 2). As presented in Table 3, a decreasing trend of serum cotinine level was observed among children without household smokers. The biannual change in log (serum cotinine) was − 0.129 (p < 0.001), corresponding to a ratio of 0.88 in serum cotinine levels when comparing later survey cycle to the two years’ earlier cycle, after adjusting for age, sex, race-ethnicity, and family income/poverty ratio. The adjusted biannual change in log (serum cotinine) among children with household smokers was − 0.055. This difference was not statistically significant (p = 0.089). Irrespective of the household smoker (s) status, the biannual change in log (serum cotinine) was − 0.196 (p < 0.001) among all children. Fig. 1 shows the differences in the serum cotinine levels between those with household smoker(s) and those without. The cotinine level among children with household smokers was remarkably and consistently higher than that among children without. Fig. 2 shows the continuously decreasing trend of cotinine level with slight fluctuations among children for both groups. Fig. 3 indicates overall the proportion of children with household smokers decreasing over survey cycles. As shown in Table 3, among children with household smokers, older children were more likely to have lower serum cotinine levels compared to younger children after adjusting for other covariates, with differences in log (serum cotinine) equal to − 0.233 (p = 0.005) for 7–9 year olds and − 0.384 (p < 0.001) for 10–11 year olds, both compared
p-valued
Characteristics
(25.7) (23.7) (15.6) (30.7)
6746 (83.7) 1269 (15.7) inside home 6746 (83.7) 660 (7.5) 227 (3.3) 67 (1.0) 10 (0.1) 72 (1.2)
0.045 (0.042, 0.049) 1.188 (1.026, 1.375)
< 0.0001
0.045 0.975 1.918 2.990 3.191 2.511
(0.042, (0.841, (1.581, (2.357, (1.873, (1.876,
0.049) 1.130) 2.326) 3.795) 5.438) 3.362)
< 0.0001
1265 1300 1344 1359 1348 1448
0.144 0.079 0.098 0.061 0.058 0.049
(0.094, (0.064, (0.070, (0.048, (0.046, (0.037,
0.222) 0.097) 0.139) 0.076) 0.073) 0.065)
(17.1) (16.7) (16.2) (16.3) (16.6) (17.0)
863 (9.5) 7181 (90.5)
0.109 (0.089, 0.133) 0.073 (0.064, 0.084)
0.0003
< 0.0001
a
Numbers may not add up to total because of missing values. Weighted percentage (%), and geometric mean. c Confidence interval (CI). d P is calculated based on Wald test on ln(cotinine) scale. e “Other race” group includes multi-racial. f Answered “Yes” to “Ever been told you have asthma” and “Still have asthma”in the medical condition questionnaire. b
to 3–6 year olds. Compared to Non-Hispanic Whites, Hispanics tended to have lower serum cotinine levels. We also found that higher family income/poverty ratio was significantly associated with lower serum cotinine level. We didn’t find a statistically significant association between gender and serum cotinine after adjustment. Similar associations between serum cotinine level and sociodemographic characteristics were found among children without household smokers, except for the relationship between race-ethnicity and cotinine. Compared to NonHispanic Whites, geometric mean of serum cotinine in Non-Hispanic Black children was 0.391 higher (p < 0.001). Table 4 shows the results for the associations between serum cotinine levels and asthma status from the logistic regression models. The odds ratio (OR) for asthma were 1.34 (95% CI: 1.00–1.80) in the 2nd tertile and 1.69 (95%CI: 1.25–2.29) in the 3rd tertile with a reference to the 1st tertile of serum cotinine levels, respectively, after adjusting for age, sex, race-ethnicity, family income/poverty ratio, and survey cycle.
37
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Table 2 Prevalence of asthma and proportion of asthmatic children in the upper tertile of cotinine levels (ng/mL), by sociodemographic characteristics, NHANES, 2003–2014. Characteristics
Gender Male Female Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other raced Income/poverty ratio 0–1 > 1–2 > 2–3 >3 a b c d
% children who have asthma (95%CI)a,b
Table 3 Biannual change a in log (serum cotinine) and association between log (serum cotinine) and selected characteristics among children aged 3–11 years, with and without household smokers, NHANES, 2003–2014. Regression coefficientb
% asthmatic children with high exposurea,c
11.0 (9.6, 12.2) 8.1 (7.1, 9.2)
41.1 42.7
8.3 (7.3, 9.6) 9.8 (8.5, 11.2) 10.9 (9.5, 12.5)
46.5 40.4 37.9
8.1
40.9
15.8
61.0
8.6 10.1
24.9 33.9
11.4 (9.8, 13.3) 9.3 (7.5 11.4) 9.3 8.3
62.6 46.5 34.2 18.2
Weighted percentage. Missing CI due to single sampling unit in the stratum. In the upper tertile of cotinine levels. “Other race” group includes multi-racial.
Characteristics
With household smoker(s) in home
Without household smoker(s) in home
Biannual change Gender Female Male Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other racec Number of cigarettes smoked Income/poverty ratio 0–1 > 1–2 > 2–3 >3
− 0.055
− 0.129*
Reference 0.028
Reference 0.010
Reference − 0.233* − 0.384*
Reference − 0.198* − 0.253*
Reference
Reference
− 0.115
0.391*
− 0.841* 0.056 0.030*
− 0.765* − 0.096 –
Reference − 0.208* − 0.529* − 0.307*
Reference − 0.452* − 0.890* − 1.334*
a
Comparing later cycle to earlier cycle. Regression coefficients in the multiple linear regression model, with log (serum cotinine) as dependent variable, cycle (indicating biannual change), age (3–6, 7–9, 10–11; in years), sex (female, male), race-ethnicity (Non-Hispanic white, Non-Hispanic black, Hispanic, Other race), and income/poverty ratio (0–1, > 1–2, > 2–3, > 3) as independent variables. Number of cigarettes smoked inside home was only added in the model among children with smoker (s) inside home. The unadjusted biannual change for children with household smokers is − 0.113 (p-value = .022), and − 0.122 (p-value < .001) for children without household smokers. c “Other race” group includes multi-racial. * p-value < 0.05.
4. Discussion
b
In this study utilizing a representative U.S. population sample, we found the trend of passive smoke exposure levels among children is overall decreasing from 2003 to 2014. The decreasing trend of exposure reflected by lower serum cotinine levels among children without household smokers was more pronounced than among those with household smokers. Additionally, more than 50% of asthmatic children of Non-Hispanic Black ethnicity or family income below poverty guidelines were in the highest tertile of passive tobacco exposure. Importantly, children with highest exposure in the population were more likely to have asthma. The main finding of this study demonstrating a significantly decreasing trend of serum cotinine level in children without household smokers is consistent with a recent study assessing passive tobacco exposure among children and nonsmoker adolescents (Jain, 2016), indicating the effectiveness of smoke free policies in public places and other governmental actions to control tobacco smoke in the U.S. Results from another study conducted earlier (Pirkle et al., 2006) showed a substantial decline of roughly 70% in serum cotinine concentrations in non-smokers during the period 1988–2002. In a similar cross-sectional study in 2013, which also used NHANES data, found a decrease in environmental tobacco smoke exposure among asthmatic children and adolescents from 1988 to 2010 (Kit et al., 2013). Together these findings suggested a continuously declining trend of tobacco smoke exposure level in public places thanks to the implementation of worldwide tobacco control policies and nationwide tobacco free laws and regulations. The increased public awareness regarding the harmful impact of tobacco smoke and significant declines in cigarette smoking rates (CDC, 2009) as a result of public health promotion education may also contribute to the observed decreasing trend. Additionally, the proportion of children with household smokers decreased from 24.9% in the 2003–2004 cycle to 11.4% in the 2013–2014 cycle, which is consistent with the overall decreasing trend of geometric mean of cotinine levels over time. Notably, the cotinine levels for children with household smokers were remarkably higher than those of children without household smokers across survey cycles. This may be due to the relatively small activity space in home and close distance between the
smoking parents/family members and children. In a previous study (Jain, 2016), an adjusted biannual increase of 1.05 ng/mL serum cotinine levels for children with household smokers was observed over the period of 1999–2012, whereas in our study an adjusted biannual decrease of 0.88 ng/mL was observed, though this decreasing trend was not statistically significant. According to findings from Jain et al. and our analyses, the geometric mean of serum cotinine levels among children ages 3–11 years with household smokers was 0.83 ng/mL in 1999–2000, 1.46 ng/mL in 2007–2008, and steadily decreasing to 0.75 ng/mL in 2013–2014. Whether this decreasing trend remains constant requires further observation. Under no public supervision or mandatory laws to constrain parents/family members’ behaviors at home, it might be harder to control home-source exposure levels. In the clinical settings, pediatricians are prompted to discuss health impacts of tobacco smoke exposure to children as well as their parents (American Academy of Pediatrics, 2009), yet the effectiveness of message delivery and behavioral changes should be considered. Future interventions including effective smoking cessation guidance for smoking parents/ family members are still needed, especially for those living with younger children who may not have any control over the passive smoke exposure. A positive association between cotinine tertiles and asthma was found in this study, which agrees with findings from prior studies conducted elsewhere (Hassanzad et al., 2015; Tsai et al., 2010). Passive tobacco smoke facilitates sensitization to perennial indoor allergens, and chemicals contained in tobacco smoke such as nicotine increase the 38
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Fig. 1. Difference of cotinine levels between the groups with and without household smokers among US children aged 3–11 years, NAHNES 2003–2014.
other places, and spend less time with parents at home. For children without household smoker(s), older children are more likely to protect themselves when exposed to tobacco smoke in public places. In other words, older children seem to have more control over passive tobacco smoke as compared to younger children. An alternative explanation could be the improved ability of metabolism as children age; older children may have higher rate for excretion of cotinine (Dempsey et al., 2000, 2013). Additionally, we found that socioeconomic class was also strongly related to children's serum cotinine levels: lower SES (i.e. lower family income/poverty ratio, higher poverty level) was related to higher serum cotinine, which is consistent with previous studies (Collaco et al., 2014; Kelly et al., 2013). Poor families were more likely to live in poorly managed communities where children may be exposed to higher level of tobacco smoke due to poorer ventilation system or more crowded room (Kelly et al., 2013) and fewer smoke-free spaces in the neighborhood. It is possible that mothers with low SES are less aware and less likely to take immediate actions when finding children exposed in the high tobacco smoke environment. We also found that among children who had asthma, over 50% of Non-Hispanic Black children or children whose family income was below federal poverty guidelines were in the highest tertile of serum cotinine levels. Though it is difficult to tell whether their asthma developed after the high level of
hyper-responsiveness of the airway and slow lung growth rates of children (Wang et al., 2015). A recent systematic review (Feleszko et al., 2014) of 28 studies found that tobacco smoke increased total immunoglobulin E, an indicator of allergic sensitization, and this effect was most pronounced in preschool children (OR=1.20, 95%CI: 1.05–1.38). This immunoregulatory mechanism explains, at least in part, the increased risk of asthma among children with passive tobacco smoke exposure. In our study, we combined information regarding history of passive tobacco smoke exposure from both questionnaire and laboratory data, which improved the accuracy of exposure assessment. We also considered using spirometry examination data to assess the lung function of the children; however, spirometry data are available for limited cycles, and only children older than six years old were eligible for this examination at the time of survey, therefore we used medical condition questionnaire for outcome assessment. This study also indicates that younger children are more likely to have higher serum cotinine levels compared to their older counterparts. This association was significant for both children with and without household smoker(s). For children with household smoker(s), this can be explained by the fact that parents may spend more time with younger children. Regardless of the parents’ smoking status, when children get older they may hang out with friends at school, in parks or
Fig. 2. Percentage change of geometric mean of cotinine levels, comparing the later survey cycles to the 2003–2004 cycle among children ages 3–11 years with and without household smoker(s). 39
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Fig. 3. Proportion of children with household smoker(s), by survey cycle, NHANES 2003–2014.
Table 4 Association between serum cotinine levels and asthma 3–11 years, NHANES, 2003–2014.
a
Exposure level
Crude odds ratio (OR)
95%CI
Adjusted ORb
95%CI
1st tertile 2nd tertile
1.00 1.35
1.00 1.34
3rd tertile
1.77
– [1.02, 1.79] [1.38, 2.25]
– [1.00, 1.80] [1.25, 2.29]
1.69
misclassification of presence of household smoker(s) may affect our evaluation of the trend in passive tobacco smoke exposure from different sources. Despite the misclassification on the source of tobacco smoke exposure, the overall trend of serum cotinine level among children aged 3–11 was decreasing during the study period, which is good news.
among children aged
5. Conclusions An overall decreasing trend of passive tobacco smoke exposure level among children was observed in this study. A consistent higher serum cotinine levels among children with household smokers and a positive association between high passive tobacco smoke exposure and asthma found in this study suggests effective smoking cessation interventions in homes are still needed to reduce adverse respiratory outcomes in children, especially those in the high exposure groups.
a Answered “Yes” to “Ever been told you have asthma” and “Still have asthma” in the medical condition questionnaire. b Adjusted for age, sex, race-ethnicity, family income/poverty ratio, and survey cycle.
tobacco smoke exposure due to the nature of study, these asthmatic children were at high risk of asthma exacerbation and other tobacco smoke related adverse health outcomes in either case. Overall, strengths of our study include that it was a nationwide population-based study with large sample size, therefore provided great precision for the coefficient estimates. Importantly, serum cotinine level was used as the direct exposure assessment method, preventing recall bias, which is often the issue in the cross-sectional studies using selfreport questionnaires to collect exposure information. In addition, the stratified analysis on the association in children with and without smoker(s) respectively enabled us to detect the possible different trends of passive tobacco smoke exposure among children exposed from different sources. Limitations of our study should also be noted. First, serum cotinine levels we observed reflect the tobacco exposure level up to one week prior to the examination. We assumed that the household smoker's smoking status or habit did not change over time, and the current serum cotinine level corresponded to the average daily passive tobacco smoke exposure level, which may or may not be true. Second, there may be residual confounding that we were not able to adjust for in the model describing the association between passive tobacco smoke exposure and asthma, including presence of other allergies at home. Third, we assumed that serum cotinine level among children with household smoker(s) reflected the tobacco smoke exposure from home source, and that among children without household smoker(s) reflected the passive tobacco smoke exposure from non-home source. However, there is no easy way to differentiate the exposure sources; children who had smoking parent(s) may have also been exposed to tobacco smoke in the public places, whereas children without smoking parents reported in the questionnaire may in fact have smoking parents. This potential
Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Conflict of interest The authors declare no conflict of interest. References American Academy of Pediatrics, 2009. Tobacco use: a pediatric disease. Policy Statement. http://dx.doi.org/10.1542/peds.2009-2114. American Cancer Society, 2015. Health risks of secondhand smoke [WWW Document]. URL 〈https://www.cancer.org/cancer/cancer-causes/tobacco-and-cancer/ secondhand-smoke.html〉 (Accessed 21 April 2017). Benowitz, N.L., 1996. Cotinine as a biomarker of environmental tobacco smoke exposure. Epidemiol. Rev. http://dx.doi.org/10.1093/oxfordjournals.epirev.a017925. Centers for Disease Control and Prevention, 2009. Cigarette smoking among adults and trends in smoking cessation - United States, 2008, MMWR. Morbidity and mortality weekly report. mm5844a2 [pii]. Centers for Disease Control and Prevention, 2013a. NHANES 2011–2012: Demographic variables and sample weights data documentation, codebook, and frequencies [WWW Document]. URL 〈https://wwwn.cdc.gov/Nchs/Nhanes/2011-2012/DEMO_ G.htm〉 (Accessed 7 May 2018). Centers for Disease Control and Prevention, 2013b. Smoking and tobacco use [WWW Document]. Centers Dis. Control Prev. URL 〈https://www.cdc.gov/tobacco/data_ statistics/by_topic/policy/legislation/index.htm〉 (Accessed 23 April 2017). Centers for Disease Control and Prevention, 2015. Smoking and tobacco use; fact sheet; secondhand smoke [WWW Document]. Centers Dis. Control Prev. URL 〈https:// www.cdc.gov/tobacco/data_statistics/fact_sheets/secondhand_smoke/general_facts/ 〉 (Accessed 21 April 2017).
40
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Lundon, N., Waclawczyk, A., Richter, J.C., Keall, M., Chapman, P.H., Baker, M.G., 2013. Exposure to harmful housing conditions is common in children admitted to wellington hospital (Article). N. Z. Med. J. 126, 108–127. Kit, B.K., Simon, A.E., Brody, D.J., Akinbami, L.J., 2013. US prevalence and trends in tobacco smoke exposure among children and adolescents with asthma. Pediatrics 131, 407–414. http://dx.doi.org/10.1542/peds.2012-2328. Martín-Pujol, A., Fernández, E., Schiaffino, A., Moncada, A., Ariza, C., Blanch, C., Martínez-Sánchez, J.M., 2013. Tobacco smoking, exposure to second-hand smoke, and asthma and wheezing in schoolchildren: a cross-sectional study. Acta Paediatr. Int. J. Paediatr. 102, 305–309. http://dx.doi.org/10.1111/apa.12232. Max, W., Sung, H.-Y., Shi, Y., 2012. Deaths from secondhand smoke exposure in the United States: economic Implications. Am. J. Public Health 102, 2173–2180. http:// dx.doi.org/10.2105/AJPH.2012.300805. Pirkle, J.L., Bernert, J.T., Caudill, S.P., Sosnoff, C.S., Pechacek, T.F., 2006. Trends in the exposure of nonsmokers in the U.S. population to secondhand smoke: 1988–2002. Environ. Health Perspect. 114, 853–858. http://dx.doi.org/10.1289/ehp.8850. Silvestri, M., Franchi, S., Pistorio, A., Petecchia, L., Rusconi, F., 2015. Smoke exposure, wheezing, and asthma development: a systematic review and meta-analysis in unselected birth cohorts. Pediatr. Pulmonol. 50, 353–362. http://dx.doi.org/10.1002/ ppul.23037. Tsai, C.-H., Huang, J.-H., Hwang, B.-F., Lee, Y.L., 2010. Household environmental tobacco smoke and risks of asthma, wheeze and bronchitic symptoms among children in Taiwan. Respir. Res. 11, 11. http://dx.doi.org/10.1186/1465-9921-11-11. Tynan, M., Babb, S., MacNeil, A., Griffin, M., 2011. (Morb. Mortal. Wkly. Rep). State smoke-free laws for worksites, restaurants, and bars — United States, 2000–2010. CDC (Accessed 23 April 2017). https://www.cdc.gov/mmwr/preview/mmwrhtml/ mm6015a2.htm. U.S. Department of Health and Human Services, 2006. The health consequences of involuntary exposure to tobacco smoke: a report of the surgeon general, U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Vlaski, E., Stavric, K., Seckova, L., Kimovska, M., Isjanovska, R., 2011. Do household tobacco smoking habits influence asthma, rhinitis and eczema among 13-14 year-old adolescents? Allergol. Immunopathol. 39, 39–44. http://dx.doi.org/10.1016/j.aller. 2010.03.006. Wang, Z., May, S.M., Charoenlap, S., Pyle, R., Ott, N.L., Mohammed, K., Joshi, A.Y., 2015. Effects of secondhand smoke exposure on asthma morbidity and health care utilization in children: a systematic review and meta-analysis. Ann. Allergy, Asthma Immunol. 115 (396–401), e2. http://dx.doi.org/10.1016/j.anai.2015.08.005. WHO, 2009. WHO Report On The Global Tobacco Epidemic, 2009: Implementing smokefree environments. WHO %6, %& 〈http://dx.doi.org/10.1596/978-0-8213-7859-5〉. WHO FCTC, 2003. WHO Framework Convention on Tobacco Control. WHO.
Centers for disease control and prevention, 2010. NHANES 2007-2008: Cotinine - Serum & Total NNAL - Urine Data Documentation, Codebook, and Frequencies [WWW Document]. URL 〈https://wwwn.cdc.gov/Nchs/Nhanes/2007-2008/COTNAL_E. htm〉. Collaco, J.M., Aherrera, A.D., Ryan, T., McGrath-Morrow, S.A., 2014. Secondhand smoke exposure in preterm infants with bronchopulmonary dysplasia. Pediatr. Pulmonol. 49, 173–178. http://dx.doi.org/10.1002/ppul.22819. Dempsey, D., Jacob, P., Benowitz, N.L., 2000. Nicotine metabolism and elimination kinetics in newborns. Clin. Pharmacol. Ther. 67, 458–465. http://dx.doi.org/10.1067/ mcp.2000.106129. Dempsey, D.A., Sambol, N.C., Jacob 3rd, P., Hoffmann, E., Tyndale, R.F., Fuentes-Afflick, E., Benowitz, N.L., 2013. CYP2A6 genotype but not age determines cotinine half-life in infants and children. Clin. Pharmacol. Ther. 94, 400–406. http://dx.doi.org/10. 1038/clpt.2013.114. Ding, G., Ji, R., Bao, Y., 2015. Risk and protective factors for the development of childhood asthma. Paediatr. Respir. Rev. 16, 133–139. http://dx.doi.org/10.1016/j.prrv. 2014.07.004. Feleszko, W., Ruszczyński, M., Jaworska, J., Strzelak, A., Zalewski, B.M., Kulus, M., 2014. 15 Environmental tobacco smoke exposure and risk of allergic sensitisation in children: a systematic review and meta-analysis. Arch. Dis. Child. 985–992. http://dx. doi.org/10.1136/archdischild-2013-305444. Fiore, M.C., Jaén, C.R., Baker, T.B., Bailey, W.C., Benowitz, N.L., Curry, S.J., Dorfman, S.F., Froelicher, E.S., Goldstein, M.G., Healton, C.G., Henderson, P.N., Heyman, R.B., Koh, H.K., Kottke, T.E., Lando, H.A., Mecklenburg, R.E., Mermelstein, R.J., Mullen, P.D., Orleans, C.T., Robinson, L., Stitzer, M.L., Tommasello, A.C., Villejo, L., Wewers, M.E., 2008. A clinical practice guideline for treating tobacco use and dependence: 2008 update. Am. J. Prev. Med. 35, 158–176. http://dx.doi.org/10.1016/j.amepre. 2008.04.009. Hassanzad, M., Khalilzadeh, S., Nobari, S.E., Bloursaz, M., Sharifi, H., Mohajerani, S.A., Nejad, S.T., Velayati, A.A., 2015. Cotinine level is associated with asthma severity in passive smoker children. Iran. J. Allergy, Asthma Immunol. 14, 67–73. Homa, D.M., Neff, L.J., King, B.A., Caraballo, R.S., Bunnell, R.E., Babb, S.D., Garrett, B.E., Sosnoff, C.S., Wang, L., 2015. Vital signs: disparities in nonsmokers' exposure to secondhand smoke — United States, 1999–2012. Morb. Mortal. Wkly. Rep. 64, 103–108 (mm6404a7 )(pii). Jain, R.B., 2016. Trends in exposure to second hand smoke at home among children and nonsmoker adolescents. Sci. Total Environ. 542, 144–152. http://dx.doi.org/10. 1016/j.scitotenv.2015.10.076. Kanchongkittiphon, W., Gaffin, J.M., Phipatanakul, W., 2014. Review article The indoor asthma environment and inner-city childhood asthma. JAsian Pac. J. Allergy Immunol. 32, 103–110. Kelly, A., Denning-Kemp, G., Geiringer, K., Abdulhamid, A., Albabtain, A., Beard, M., Brimble, J., Campbell, A., Feng, S., Haminudin, M., Hunter, J., Kunac, N., Lin, P.,
41
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Decreasing trend in passive tobacco smoke exposure and association with asthma in U.S. children Xiao Zhanga, Natalie Johnsonb, Genny Carrillob, Xiaohui Xua, a b
T
⁎
Department of Epidemiology and Biostatistics, School of Public Health, Texas A&M University, United States Department of Environmental and Occupational Health, School of Public Health, Texas A&M University, United States
A R T I C LE I N FO
A B S T R A C T
Keywords: Passive smoke Cotinine Asthma Trend Children Prevalence
In this study, we assessed trends of serum cotinine levels over time among US children ages 3–11 years and compared the risk of asthma in groups exposed to passive tobacco smoke. We utilized National Health and Nutrition Examination Survey (NHANES) data collected from 2003 to 2014 (n = 8064). Serum cotinine level, household smoker status, asthma status, and sociodemographic information were extracted for multiple regression analyses. The adjusted biannual change in log (cotinine) in comparison to earlier NHANES survey cycles was − 0.196 (p < 0.001) overall, − 0.055 (p = 0.089) among children with household smoker(s), and − 0.129 (p < 0.001) among children without. The proportion of children living with household smokers decreased from 24.9% in the 2003–2004 cycle to 11.4% in the 2013–2014 cycle. The adjusted odds ratios (ORs) for asthma were 1.34 (95% confidence interval (CI): 1.00–1.80; 2nd tertile vs 1st tertile) and 1.69 (95%CI: 1.25–2.29; 3rd tertile vs1st tertile), respectively. Highly exposed asthmatic children, in the 3rd cotinine tertile (> 0.13 ng/mL), were primarily Non-Hispanic Black (61.0%) and whose family incomes were below poverty guidelines. Overall results reveal passive smoke exposure level among children ages 3–11 in the US decreased over the study period. Nevertheless, higher exposure to passive smoke is still associated with higher odds of childhood asthma. Targeted smoking cessation interventions in clinical practices are needed to reduce tobacco smoke exposure and related asthma risk in children, particularly in low-income and minority groups.
1. Introduction Passive tobacco smoke is defined as smoke emitted from burning cigarettes or exhaled smoke by the person smoking (Centers for disease control and prevention (CDC), 2015). Passive tobacco smoke contains thousands of chemical compounds, including the same chemicals inhaled by smokers (American Cancer Society, 2015). These chemical mixtures are reported to have detrimental health risks, and at least 70 chemical constituents are classified as carcinogens (U.S. Department of Health and Human Services, 2006). According to CDC, roughly 2,500,000 nonsmokers have died from health problems caused by passive tobacco smoke exposure since 1964, and the cost of lost productivity due to tobacco smoke exposure is estimated at ~$6.6 billion per year (Max et al., 2012). Compared to adults, children are more vulnerable to environmental passive tobacco smoke since most of the time they have no control over the exposure and also represent a biologically susceptible group due to higher breathing rates and developing immune systems (Silvestri et al., 2015). Children can be exposed to passive tobacco smoke at home if
they live with smokers or have smoking visitors. Even if smokers do not reside in their homes, children may still be exposed to environmental tobacco smoke in public places. In the U.S., governmental actions regarding the regulation of tobacco sales, marketing, and use have been implemented since 1960s. According to the Pro-Children Act of 1994, all federal funded children's services must be smoke-free (CDC, 2013b). As of December 31, 2010, 26 states had enacted comprehensive smokefree laws (Tynan et al., 2011). In 2008–2009, the American Academy of Pediatrics and the U.S. Public Health Services updated guidelines to recommend pediatricians and clinicians to offer advice and interventions to smoking youths and their smoking patients (American Academy of Pediatrics, 2009; Fiore et al., 2008). The World Health Organization (WHO) has developed six key tobacco control policies called MPOWER (WHO, 2009) to protect people from tobacco smoke exposure (WHO FCTC, 2003). Studies have reported declines in prevalence of passive tobacco smoke exposure among children from 1988 to 2010 (Kit et al., 2013), 1999–2000–2011–2012 (Homa et al., 2015; Jain, 2016). However, whether there is a continuously decreasing trend of children's passive tobacco smoke exposure levels recently is not readily known,
⁎ Correspondence to: Department of Epidemiology & Biostatistics, School of Public Health, Texas A&M University, MS 1266, 212 Adriance Lab Road, College Station, TX 77843-1266, USA. E-mail address: [email protected] (X. Xu).
https://doi.org/10.1016/j.envres.2018.05.022 Received 12 April 2018; Received in revised form 15 May 2018; Accepted 20 May 2018 0013-9351/ © 2018 Elsevier Inc. All rights reserved.
Environmental Research 166 (2018) 35–41
X. Zhang et al.
home?” If the answer was yes, then we classified the participant as having likely been exposed to the home-source tobacco smoke; if the answer was no we classified the participant as having not been exposed to home-source tobacco smoke, suggesting that he/she might have been primarily exposed to non-home source tobacco smoke or have not been exposed to tobacco smoke from any source at all. For 2013–2014 survey cycle, the question was removed, and therefore we selected another question asking for “the number of people who live here smoke tobacco” to identify children with household smokers. If the answer was 0, then we classified the participant as having no home-source tobacco smoke exposure.
specifically at the national level across multiple demographic groups. Moreover, considering the potential disparities in passive tobacco smoke levels among children with household smokers and those without, it is worthwhile to evaluate differences in the trends of exposure levels to support targeted interventions. Previous investigations have linked passive tobacco smoke with various diseases in children, among which asthma is significant because it is a prevalent disease in children representing the leading cause of school absences. Although there is a biological plausibility for passive tobacco smoke exposure and childhood asthma, a proportion of studies lack strong evidence or have not identified a positive association (Martín-Pujol et al., 2013; Vlaski et al., 2011). These inconsistent findings may be partially due to differences in exposure assessment methods across studies. The majority of previous studies used questionnaires, an indirect exposure assessment method, to collect information about passive tobacco smoke exposure status. While this approach is easy to obtain and not costly, exposure misclassification issues often result. By comparison, direct measurement, such as biomarkers of tobacco smoke exposure, are able to quantify exposure levels, especially when self-reported measures (e.g., hours per day exposed to passive smoke, proximity to smokers, room ventilation) are much harder to quantify. Serum cotinine, the major proximate metabolite of nicotine, is considered a valid marker of tobacco smoke exposure (Benowitz, 1996). Thus, using serum cotinine measurements to assess exposure levels may help to obtain more accurate effect estimates of passive smoke exposure on risk of children's asthma. Additionally, direct exposure assessment through biomarker measurement can easily discriminate population subgroups at the upper and lower exposure levels. In this study, we evaluated the difference in the trends of passive tobacco smoke exposure levels over time among US children ages 3–11 years using National Health and Nutrition and Examination Survey (NHANES) data from 2003 to 2014 (n = 8064). Serum cotinine levels were used as a direct exposure measurement to examine the association between passive tobacco smoke exposure levels by tertile and asthma, as well as to identify sociodemographic groups of asthmatic children with high passive tobacco smoke exposure levels.
2.3. Respiratory health assessment Determination of asthma status was based on the answer to “Ever been told you have asthma” and “Still have asthma” in the medical condition questionnaire. Children answered “Yes” to both questions were classified as having asthma. “Don’t know” and “Refused” were treated as missing data. 2.4. Covariates
2. Materials and methods
Data on age, gender, race-ethnicity, and survey cycle were extracted from the demographic data files. Family income/poverty ratio was used as a surrogate measure of socioeconomic status (SES), and lower values indicate higher poverty level. It is a ratio of family income to poverty guidelines, which is calculated by dividing family income by the poverty guidelines specific to the survey cycle. The poverty guidelines are issued every year in the Federal Register, which vary by family size and geographic location. Detailed information has been described elsewhere (CDC, 2013a). All these variables were considered as potential confounders in the relationships between serum cotinine level and adverse respiratory outcomes based on previous studies (Ding et al., 2015; Kanchongkittiphon et al., 2014) and adjusted for in the logistic regression models. Except for survey cycle, all of these variables were considered as confounders and added in the multiple linear regression models to examine the trends of passive tobacco smoke exposure in children over time.
2.1. Study population
2.5. Statistical Analysis
We utilized the data from six consecutive cycles of the 2003–2014 NHANES, a national cross-sectional survey assessing the health and nutritional status of the U.S. population. Informed consent was obtained from all survey participants in the NHANES. Study protocols were approved by the National Center for Health Statistics Research Ethics Review Board. Demographic data file, cotinine data file (laboratory data), household smoker questionnaire, and medical condition questionnaire were downloaded from each survey cycle and merged. This study was limited to children 3–11 years.
2.5.1. Descriptive analysis To account for the right-skewness of the serum cotinine data, natural log transformation was performed. Log (serum cotinine) in each sociodemographic group is presented using weighted geometric mean and 95% confidence interval (CI). We used the Wald test to compare the weighted geometric means in each of the categories of the covariates. We also divided serum cotinine levels into tertiles using data from all children aged 3–11 years, therefore the cutoff points are the same for children in different sociodemographic groups. Proportion of children who had asthma and proportion of asthmatic children who were in the 3rd cotinine tertile (highest exposure) are presented for each sociodemographic group. Comparisons of cotinine levels between the groups with and without household smokers for each survey cycle are presented in box plots. Percentage change of serum cotinine levels comparing the later survey cycle to 2003–2004 and proportions of children with household smokers for each survey cycle are presented in a line chart.
2.2. Passive tobacco smoke exposure assessment Serum cotinine level was collected from participants aged 3 years and older at each of the six survey cycles (2003–2014). Serum samples were obtained at the Mobile Examination Center (MEC). In the original laboratory data files, results below limit of detection (LOD) were replaced by LOD / 2 . The LOD was constant for all survey cycles. Detailed information regarding the measurement methods and detection limits are described elsewhere (CDC, 2010). The proportions of participants whose serum cotinine was below LOD was 13.68%, 17.62%, 16.96%, 23.84%, 25.89%, and 34.53% for each of the six survey cycles, respectively. For 2003–2012, determination of the presence of household smoker (s) was based on the answer to the question in the household smoker questionnaire open to all participants - “Does anyone smoke inside
2.5.2. Regression analysis (1) To evaluate the trend in passive tobacco smoke exposure in children over time, multiple linear regression models describing the association between log (serum cotinine) and survey cycle (continuous) were fitted to the data stratified by home source tobacco smoke exposure status (Yes/No). Categorical variables, including age (3–6, 36
Environmental Research 166 (2018) 35–41
X. Zhang et al.
7–9, 10–11; in years), gender (Male/Female), race-ethnicity (NonHispanic White, Non-Hispanic Black, Hispanic, Other race including multiracial), and family income/poverty ratio (0–1, > 1–2, > 2–3, > 3) were adjusted for in the regression models. Number of cigarettes smoked inside room was also added in the models when assessing the association among children with household smoker(s). (2) To examine the association between passive tobacco smoke exposure and asthma in children, multiple logistic regression analyses relating tertiles of serum cotinine levels to asthma (Yes/No) were performed. Categorical variables, including age (3–6, 7–9, 10–11; in years), gender (Male/Female), race-ethnicity (Non-Hispanic White, Non-Hispanic Black, Hispanic, Other race including multiracial), family income/poverty ratio (0–1, > 1–2, > 2–3, > 3), and survey cycle (2003–2004, 2005–2006, 2007–2008, 2009–2010, 2011–2012, 2013–2014) were adjusted for in the regression models. Odds of asthma were compared between children in the 3rd (high exposure) or 2nd (median exposure) tertile and those in the 1st (low exposure) tertile group.
Table 1 Summary statistics for serum cotinine levels (ng/mL) by selected characteristics, NHANES, 2003–2014. Cotinine (ng/mL) Na (%a,b)
Meanb (95%CI)c
Total
8064 (100.0)
0.076 (0.067, 0.087) 0.084)
4116 (52.0) 3948 (48.0)
0.075 (0.065, 0.086) 0.078 (0.067, 0.089)
0.5250
3253 (39.2) 2802 (35.0) 2009 (25.8)
0.090 (0.076, 0.106) 0.073 (0.063, 0.084) 0.063 (0.054, 0.072)
< 0.0001
2197 (54.4) 2154 (14.4) 2985 (23.1) 728 (8.1)
0.082 0.185 0.039 0.064
(0.067, (0.156, (0.035, (0.052,
0.100) 0.220) 0.043) 0.080)
< 0.0001
2824 2118 1022 1660
0.182 0.102 0.060 0.034
(0.150, (0.085, (0.049, (0.029,
0.220) 0.122) 0.074) 0.039)
< 0.0001
Gender Male Female Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other racee Income/poverty ratio 0–1 > 1–2 > 2–3 >3 Any smoker inside home No Yes Number of cigarettes smoked 0 1–10 11–20 21–30 31–39 > = 40 Survey cycle 2003–2004 2005–2006 2007–2008 2009–2010 2011–2012 2013–2014 Asthmaf Yes No
Statistical analyses were conducted using Stata (version 14.0, Stata Corp, College Station, Texas, USA) and R (R version 3.3.3 (2017-0306)). Sampling weight was 1/6 × examination sample weight and implemented into each analysis. Sampling design variables were included to account for the complex and multistage survey design. P-values < 0.05 were considered significant. 3. Results A total of 8064 children ages 3–11 years were analyzed in our study. As shown in Table 1, serum cotinine level was significantly associated with age, race-ethnicity, family income/poverty ratio, whether there is any household smoker, number of cigarettes smoked inside home, survey cycle, and asthma status. Without adjustment, children of younger age, lower family income/poverty ratio, greater number of cigarettes smoked inside home, Non-Hispanic Black ethnicity, and who had asthma were more likely to have higher geometric means of serum cotinine level. The lower and upper tertile cutoff points of serum cotinine levels among children ages 3–11 years were 0.02 ng/mL and 0.13 ng/mL, respectively. More than 50% of asthmatic children who are Non-Hispanic Black or whose family incomes were below poverty guideline (family income/poverty ratio < 1) were in the 3rd tertile of serum cotinine levels (Table 2). As presented in Table 3, a decreasing trend of serum cotinine level was observed among children without household smokers. The biannual change in log (serum cotinine) was − 0.129 (p < 0.001), corresponding to a ratio of 0.88 in serum cotinine levels when comparing later survey cycle to the two years’ earlier cycle, after adjusting for age, sex, race-ethnicity, and family income/poverty ratio. The adjusted biannual change in log (serum cotinine) among children with household smokers was − 0.055. This difference was not statistically significant (p = 0.089). Irrespective of the household smoker (s) status, the biannual change in log (serum cotinine) was − 0.196 (p < 0.001) among all children. Fig. 1 shows the differences in the serum cotinine levels between those with household smoker(s) and those without. The cotinine level among children with household smokers was remarkably and consistently higher than that among children without. Fig. 2 shows the continuously decreasing trend of cotinine level with slight fluctuations among children for both groups. Fig. 3 indicates overall the proportion of children with household smokers decreasing over survey cycles. As shown in Table 3, among children with household smokers, older children were more likely to have lower serum cotinine levels compared to younger children after adjusting for other covariates, with differences in log (serum cotinine) equal to − 0.233 (p = 0.005) for 7–9 year olds and − 0.384 (p < 0.001) for 10–11 year olds, both compared
p-valued
Characteristics
(25.7) (23.7) (15.6) (30.7)
6746 (83.7) 1269 (15.7) inside home 6746 (83.7) 660 (7.5) 227 (3.3) 67 (1.0) 10 (0.1) 72 (1.2)
0.045 (0.042, 0.049) 1.188 (1.026, 1.375)
< 0.0001
0.045 0.975 1.918 2.990 3.191 2.511
(0.042, (0.841, (1.581, (2.357, (1.873, (1.876,
0.049) 1.130) 2.326) 3.795) 5.438) 3.362)
< 0.0001
1265 1300 1344 1359 1348 1448
0.144 0.079 0.098 0.061 0.058 0.049
(0.094, (0.064, (0.070, (0.048, (0.046, (0.037,
0.222) 0.097) 0.139) 0.076) 0.073) 0.065)
(17.1) (16.7) (16.2) (16.3) (16.6) (17.0)
863 (9.5) 7181 (90.5)
0.109 (0.089, 0.133) 0.073 (0.064, 0.084)
0.0003
< 0.0001
a
Numbers may not add up to total because of missing values. Weighted percentage (%), and geometric mean. c Confidence interval (CI). d P is calculated based on Wald test on ln(cotinine) scale. e “Other race” group includes multi-racial. f Answered “Yes” to “Ever been told you have asthma” and “Still have asthma”in the medical condition questionnaire. b
to 3–6 year olds. Compared to Non-Hispanic Whites, Hispanics tended to have lower serum cotinine levels. We also found that higher family income/poverty ratio was significantly associated with lower serum cotinine level. We didn’t find a statistically significant association between gender and serum cotinine after adjustment. Similar associations between serum cotinine level and sociodemographic characteristics were found among children without household smokers, except for the relationship between race-ethnicity and cotinine. Compared to NonHispanic Whites, geometric mean of serum cotinine in Non-Hispanic Black children was 0.391 higher (p < 0.001). Table 4 shows the results for the associations between serum cotinine levels and asthma status from the logistic regression models. The odds ratio (OR) for asthma were 1.34 (95% CI: 1.00–1.80) in the 2nd tertile and 1.69 (95%CI: 1.25–2.29) in the 3rd tertile with a reference to the 1st tertile of serum cotinine levels, respectively, after adjusting for age, sex, race-ethnicity, family income/poverty ratio, and survey cycle.
37
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Table 2 Prevalence of asthma and proportion of asthmatic children in the upper tertile of cotinine levels (ng/mL), by sociodemographic characteristics, NHANES, 2003–2014. Characteristics
Gender Male Female Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other raced Income/poverty ratio 0–1 > 1–2 > 2–3 >3 a b c d
% children who have asthma (95%CI)a,b
Table 3 Biannual change a in log (serum cotinine) and association between log (serum cotinine) and selected characteristics among children aged 3–11 years, with and without household smokers, NHANES, 2003–2014. Regression coefficientb
% asthmatic children with high exposurea,c
11.0 (9.6, 12.2) 8.1 (7.1, 9.2)
41.1 42.7
8.3 (7.3, 9.6) 9.8 (8.5, 11.2) 10.9 (9.5, 12.5)
46.5 40.4 37.9
8.1
40.9
15.8
61.0
8.6 10.1
24.9 33.9
11.4 (9.8, 13.3) 9.3 (7.5 11.4) 9.3 8.3
62.6 46.5 34.2 18.2
Weighted percentage. Missing CI due to single sampling unit in the stratum. In the upper tertile of cotinine levels. “Other race” group includes multi-racial.
Characteristics
With household smoker(s) in home
Without household smoker(s) in home
Biannual change Gender Female Male Age (years) 3–6 7–9 10–11 Race-ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Other racec Number of cigarettes smoked Income/poverty ratio 0–1 > 1–2 > 2–3 >3
− 0.055
− 0.129*
Reference 0.028
Reference 0.010
Reference − 0.233* − 0.384*
Reference − 0.198* − 0.253*
Reference
Reference
− 0.115
0.391*
− 0.841* 0.056 0.030*
− 0.765* − 0.096 –
Reference − 0.208* − 0.529* − 0.307*
Reference − 0.452* − 0.890* − 1.334*
a
Comparing later cycle to earlier cycle. Regression coefficients in the multiple linear regression model, with log (serum cotinine) as dependent variable, cycle (indicating biannual change), age (3–6, 7–9, 10–11; in years), sex (female, male), race-ethnicity (Non-Hispanic white, Non-Hispanic black, Hispanic, Other race), and income/poverty ratio (0–1, > 1–2, > 2–3, > 3) as independent variables. Number of cigarettes smoked inside home was only added in the model among children with smoker (s) inside home. The unadjusted biannual change for children with household smokers is − 0.113 (p-value = .022), and − 0.122 (p-value < .001) for children without household smokers. c “Other race” group includes multi-racial. * p-value < 0.05.
4. Discussion
b
In this study utilizing a representative U.S. population sample, we found the trend of passive smoke exposure levels among children is overall decreasing from 2003 to 2014. The decreasing trend of exposure reflected by lower serum cotinine levels among children without household smokers was more pronounced than among those with household smokers. Additionally, more than 50% of asthmatic children of Non-Hispanic Black ethnicity or family income below poverty guidelines were in the highest tertile of passive tobacco exposure. Importantly, children with highest exposure in the population were more likely to have asthma. The main finding of this study demonstrating a significantly decreasing trend of serum cotinine level in children without household smokers is consistent with a recent study assessing passive tobacco exposure among children and nonsmoker adolescents (Jain, 2016), indicating the effectiveness of smoke free policies in public places and other governmental actions to control tobacco smoke in the U.S. Results from another study conducted earlier (Pirkle et al., 2006) showed a substantial decline of roughly 70% in serum cotinine concentrations in non-smokers during the period 1988–2002. In a similar cross-sectional study in 2013, which also used NHANES data, found a decrease in environmental tobacco smoke exposure among asthmatic children and adolescents from 1988 to 2010 (Kit et al., 2013). Together these findings suggested a continuously declining trend of tobacco smoke exposure level in public places thanks to the implementation of worldwide tobacco control policies and nationwide tobacco free laws and regulations. The increased public awareness regarding the harmful impact of tobacco smoke and significant declines in cigarette smoking rates (CDC, 2009) as a result of public health promotion education may also contribute to the observed decreasing trend. Additionally, the proportion of children with household smokers decreased from 24.9% in the 2003–2004 cycle to 11.4% in the 2013–2014 cycle, which is consistent with the overall decreasing trend of geometric mean of cotinine levels over time. Notably, the cotinine levels for children with household smokers were remarkably higher than those of children without household smokers across survey cycles. This may be due to the relatively small activity space in home and close distance between the
smoking parents/family members and children. In a previous study (Jain, 2016), an adjusted biannual increase of 1.05 ng/mL serum cotinine levels for children with household smokers was observed over the period of 1999–2012, whereas in our study an adjusted biannual decrease of 0.88 ng/mL was observed, though this decreasing trend was not statistically significant. According to findings from Jain et al. and our analyses, the geometric mean of serum cotinine levels among children ages 3–11 years with household smokers was 0.83 ng/mL in 1999–2000, 1.46 ng/mL in 2007–2008, and steadily decreasing to 0.75 ng/mL in 2013–2014. Whether this decreasing trend remains constant requires further observation. Under no public supervision or mandatory laws to constrain parents/family members’ behaviors at home, it might be harder to control home-source exposure levels. In the clinical settings, pediatricians are prompted to discuss health impacts of tobacco smoke exposure to children as well as their parents (American Academy of Pediatrics, 2009), yet the effectiveness of message delivery and behavioral changes should be considered. Future interventions including effective smoking cessation guidance for smoking parents/ family members are still needed, especially for those living with younger children who may not have any control over the passive smoke exposure. A positive association between cotinine tertiles and asthma was found in this study, which agrees with findings from prior studies conducted elsewhere (Hassanzad et al., 2015; Tsai et al., 2010). Passive tobacco smoke facilitates sensitization to perennial indoor allergens, and chemicals contained in tobacco smoke such as nicotine increase the 38
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Fig. 1. Difference of cotinine levels between the groups with and without household smokers among US children aged 3–11 years, NAHNES 2003–2014.
other places, and spend less time with parents at home. For children without household smoker(s), older children are more likely to protect themselves when exposed to tobacco smoke in public places. In other words, older children seem to have more control over passive tobacco smoke as compared to younger children. An alternative explanation could be the improved ability of metabolism as children age; older children may have higher rate for excretion of cotinine (Dempsey et al., 2000, 2013). Additionally, we found that socioeconomic class was also strongly related to children's serum cotinine levels: lower SES (i.e. lower family income/poverty ratio, higher poverty level) was related to higher serum cotinine, which is consistent with previous studies (Collaco et al., 2014; Kelly et al., 2013). Poor families were more likely to live in poorly managed communities where children may be exposed to higher level of tobacco smoke due to poorer ventilation system or more crowded room (Kelly et al., 2013) and fewer smoke-free spaces in the neighborhood. It is possible that mothers with low SES are less aware and less likely to take immediate actions when finding children exposed in the high tobacco smoke environment. We also found that among children who had asthma, over 50% of Non-Hispanic Black children or children whose family income was below federal poverty guidelines were in the highest tertile of serum cotinine levels. Though it is difficult to tell whether their asthma developed after the high level of
hyper-responsiveness of the airway and slow lung growth rates of children (Wang et al., 2015). A recent systematic review (Feleszko et al., 2014) of 28 studies found that tobacco smoke increased total immunoglobulin E, an indicator of allergic sensitization, and this effect was most pronounced in preschool children (OR=1.20, 95%CI: 1.05–1.38). This immunoregulatory mechanism explains, at least in part, the increased risk of asthma among children with passive tobacco smoke exposure. In our study, we combined information regarding history of passive tobacco smoke exposure from both questionnaire and laboratory data, which improved the accuracy of exposure assessment. We also considered using spirometry examination data to assess the lung function of the children; however, spirometry data are available for limited cycles, and only children older than six years old were eligible for this examination at the time of survey, therefore we used medical condition questionnaire for outcome assessment. This study also indicates that younger children are more likely to have higher serum cotinine levels compared to their older counterparts. This association was significant for both children with and without household smoker(s). For children with household smoker(s), this can be explained by the fact that parents may spend more time with younger children. Regardless of the parents’ smoking status, when children get older they may hang out with friends at school, in parks or
Fig. 2. Percentage change of geometric mean of cotinine levels, comparing the later survey cycles to the 2003–2004 cycle among children ages 3–11 years with and without household smoker(s). 39
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Fig. 3. Proportion of children with household smoker(s), by survey cycle, NHANES 2003–2014.
Table 4 Association between serum cotinine levels and asthma 3–11 years, NHANES, 2003–2014.
a
Exposure level
Crude odds ratio (OR)
95%CI
Adjusted ORb
95%CI
1st tertile 2nd tertile
1.00 1.35
1.00 1.34
3rd tertile
1.77
– [1.02, 1.79] [1.38, 2.25]
– [1.00, 1.80] [1.25, 2.29]
1.69
misclassification of presence of household smoker(s) may affect our evaluation of the trend in passive tobacco smoke exposure from different sources. Despite the misclassification on the source of tobacco smoke exposure, the overall trend of serum cotinine level among children aged 3–11 was decreasing during the study period, which is good news.
among children aged
5. Conclusions An overall decreasing trend of passive tobacco smoke exposure level among children was observed in this study. A consistent higher serum cotinine levels among children with household smokers and a positive association between high passive tobacco smoke exposure and asthma found in this study suggests effective smoking cessation interventions in homes are still needed to reduce adverse respiratory outcomes in children, especially those in the high exposure groups.
a Answered “Yes” to “Ever been told you have asthma” and “Still have asthma” in the medical condition questionnaire. b Adjusted for age, sex, race-ethnicity, family income/poverty ratio, and survey cycle.
tobacco smoke exposure due to the nature of study, these asthmatic children were at high risk of asthma exacerbation and other tobacco smoke related adverse health outcomes in either case. Overall, strengths of our study include that it was a nationwide population-based study with large sample size, therefore provided great precision for the coefficient estimates. Importantly, serum cotinine level was used as the direct exposure assessment method, preventing recall bias, which is often the issue in the cross-sectional studies using selfreport questionnaires to collect exposure information. In addition, the stratified analysis on the association in children with and without smoker(s) respectively enabled us to detect the possible different trends of passive tobacco smoke exposure among children exposed from different sources. Limitations of our study should also be noted. First, serum cotinine levels we observed reflect the tobacco exposure level up to one week prior to the examination. We assumed that the household smoker's smoking status or habit did not change over time, and the current serum cotinine level corresponded to the average daily passive tobacco smoke exposure level, which may or may not be true. Second, there may be residual confounding that we were not able to adjust for in the model describing the association between passive tobacco smoke exposure and asthma, including presence of other allergies at home. Third, we assumed that serum cotinine level among children with household smoker(s) reflected the tobacco smoke exposure from home source, and that among children without household smoker(s) reflected the passive tobacco smoke exposure from non-home source. However, there is no easy way to differentiate the exposure sources; children who had smoking parent(s) may have also been exposed to tobacco smoke in the public places, whereas children without smoking parents reported in the questionnaire may in fact have smoking parents. This potential
Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Conflict of interest The authors declare no conflict of interest. References American Academy of Pediatrics, 2009. Tobacco use: a pediatric disease. Policy Statement. http://dx.doi.org/10.1542/peds.2009-2114. American Cancer Society, 2015. Health risks of secondhand smoke [WWW Document]. URL 〈https://www.cancer.org/cancer/cancer-causes/tobacco-and-cancer/ secondhand-smoke.html〉 (Accessed 21 April 2017). Benowitz, N.L., 1996. Cotinine as a biomarker of environmental tobacco smoke exposure. Epidemiol. Rev. http://dx.doi.org/10.1093/oxfordjournals.epirev.a017925. Centers for Disease Control and Prevention, 2009. Cigarette smoking among adults and trends in smoking cessation - United States, 2008, MMWR. Morbidity and mortality weekly report. mm5844a2 [pii]. Centers for Disease Control and Prevention, 2013a. NHANES 2011–2012: Demographic variables and sample weights data documentation, codebook, and frequencies [WWW Document]. URL 〈https://wwwn.cdc.gov/Nchs/Nhanes/2011-2012/DEMO_ G.htm〉 (Accessed 7 May 2018). Centers for Disease Control and Prevention, 2013b. Smoking and tobacco use [WWW Document]. Centers Dis. Control Prev. URL 〈https://www.cdc.gov/tobacco/data_ statistics/by_topic/policy/legislation/index.htm〉 (Accessed 23 April 2017). Centers for Disease Control and Prevention, 2015. Smoking and tobacco use; fact sheet; secondhand smoke [WWW Document]. Centers Dis. Control Prev. URL 〈https:// www.cdc.gov/tobacco/data_statistics/fact_sheets/secondhand_smoke/general_facts/ 〉 (Accessed 21 April 2017).
40
Environmental Research 166 (2018) 35–41
X. Zhang et al.
Lundon, N., Waclawczyk, A., Richter, J.C., Keall, M., Chapman, P.H., Baker, M.G., 2013. Exposure to harmful housing conditions is common in children admitted to wellington hospital (Article). N. Z. Med. J. 126, 108–127. Kit, B.K., Simon, A.E., Brody, D.J., Akinbami, L.J., 2013. US prevalence and trends in tobacco smoke exposure among children and adolescents with asthma. Pediatrics 131, 407–414. http://dx.doi.org/10.1542/peds.2012-2328. Martín-Pujol, A., Fernández, E., Schiaffino, A., Moncada, A., Ariza, C., Blanch, C., Martínez-Sánchez, J.M., 2013. Tobacco smoking, exposure to second-hand smoke, and asthma and wheezing in schoolchildren: a cross-sectional study. Acta Paediatr. Int. J. Paediatr. 102, 305–309. http://dx.doi.org/10.1111/apa.12232. Max, W., Sung, H.-Y., Shi, Y., 2012. Deaths from secondhand smoke exposure in the United States: economic Implications. Am. J. Public Health 102, 2173–2180. http:// dx.doi.org/10.2105/AJPH.2012.300805. Pirkle, J.L., Bernert, J.T., Caudill, S.P., Sosnoff, C.S., Pechacek, T.F., 2006. Trends in the exposure of nonsmokers in the U.S. population to secondhand smoke: 1988–2002. Environ. Health Perspect. 114, 853–858. http://dx.doi.org/10.1289/ehp.8850. Silvestri, M., Franchi, S., Pistorio, A., Petecchia, L., Rusconi, F., 2015. Smoke exposure, wheezing, and asthma development: a systematic review and meta-analysis in unselected birth cohorts. Pediatr. Pulmonol. 50, 353–362. http://dx.doi.org/10.1002/ ppul.23037. Tsai, C.-H., Huang, J.-H., Hwang, B.-F., Lee, Y.L., 2010. Household environmental tobacco smoke and risks of asthma, wheeze and bronchitic symptoms among children in Taiwan. Respir. Res. 11, 11. http://dx.doi.org/10.1186/1465-9921-11-11. Tynan, M., Babb, S., MacNeil, A., Griffin, M., 2011. (Morb. Mortal. Wkly. Rep). State smoke-free laws for worksites, restaurants, and bars — United States, 2000–2010. CDC (Accessed 23 April 2017). https://www.cdc.gov/mmwr/preview/mmwrhtml/ mm6015a2.htm. U.S. Department of Health and Human Services, 2006. The health consequences of involuntary exposure to tobacco smoke: a report of the surgeon general, U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Vlaski, E., Stavric, K., Seckova, L., Kimovska, M., Isjanovska, R., 2011. Do household tobacco smoking habits influence asthma, rhinitis and eczema among 13-14 year-old adolescents? Allergol. Immunopathol. 39, 39–44. http://dx.doi.org/10.1016/j.aller. 2010.03.006. Wang, Z., May, S.M., Charoenlap, S., Pyle, R., Ott, N.L., Mohammed, K., Joshi, A.Y., 2015. Effects of secondhand smoke exposure on asthma morbidity and health care utilization in children: a systematic review and meta-analysis. Ann. Allergy, Asthma Immunol. 115 (396–401), e2. http://dx.doi.org/10.1016/j.anai.2015.08.005. WHO, 2009. WHO Report On The Global Tobacco Epidemic, 2009: Implementing smokefree environments. WHO %6, %& 〈http://dx.doi.org/10.1596/978-0-8213-7859-5〉. WHO FCTC, 2003. WHO Framework Convention on Tobacco Control. WHO.
Centers for disease control and prevention, 2010. NHANES 2007-2008: Cotinine - Serum & Total NNAL - Urine Data Documentation, Codebook, and Frequencies [WWW Document]. URL 〈https://wwwn.cdc.gov/Nchs/Nhanes/2007-2008/COTNAL_E. htm〉. Collaco, J.M., Aherrera, A.D., Ryan, T., McGrath-Morrow, S.A., 2014. Secondhand smoke exposure in preterm infants with bronchopulmonary dysplasia. Pediatr. Pulmonol. 49, 173–178. http://dx.doi.org/10.1002/ppul.22819. Dempsey, D., Jacob, P., Benowitz, N.L., 2000. Nicotine metabolism and elimination kinetics in newborns. Clin. Pharmacol. Ther. 67, 458–465. http://dx.doi.org/10.1067/ mcp.2000.106129. Dempsey, D.A., Sambol, N.C., Jacob 3rd, P., Hoffmann, E., Tyndale, R.F., Fuentes-Afflick, E., Benowitz, N.L., 2013. CYP2A6 genotype but not age determines cotinine half-life in infants and children. Clin. Pharmacol. Ther. 94, 400–406. http://dx.doi.org/10. 1038/clpt.2013.114. Ding, G., Ji, R., Bao, Y., 2015. Risk and protective factors for the development of childhood asthma. Paediatr. Respir. Rev. 16, 133–139. http://dx.doi.org/10.1016/j.prrv. 2014.07.004. Feleszko, W., Ruszczyński, M., Jaworska, J., Strzelak, A., Zalewski, B.M., Kulus, M., 2014. 15 Environmental tobacco smoke exposure and risk of allergic sensitisation in children: a systematic review and meta-analysis. Arch. Dis. Child. 985–992. http://dx. doi.org/10.1136/archdischild-2013-305444. Fiore, M.C., Jaén, C.R., Baker, T.B., Bailey, W.C., Benowitz, N.L., Curry, S.J., Dorfman, S.F., Froelicher, E.S., Goldstein, M.G., Healton, C.G., Henderson, P.N., Heyman, R.B., Koh, H.K., Kottke, T.E., Lando, H.A., Mecklenburg, R.E., Mermelstein, R.J., Mullen, P.D., Orleans, C.T., Robinson, L., Stitzer, M.L., Tommasello, A.C., Villejo, L., Wewers, M.E., 2008. A clinical practice guideline for treating tobacco use and dependence: 2008 update. Am. J. Prev. Med. 35, 158–176. http://dx.doi.org/10.1016/j.amepre. 2008.04.009. Hassanzad, M., Khalilzadeh, S., Nobari, S.E., Bloursaz, M., Sharifi, H., Mohajerani, S.A., Nejad, S.T., Velayati, A.A., 2015. Cotinine level is associated with asthma severity in passive smoker children. Iran. J. Allergy, Asthma Immunol. 14, 67–73. Homa, D.M., Neff, L.J., King, B.A., Caraballo, R.S., Bunnell, R.E., Babb, S.D., Garrett, B.E., Sosnoff, C.S., Wang, L., 2015. Vital signs: disparities in nonsmokers' exposure to secondhand smoke — United States, 1999–2012. Morb. Mortal. Wkly. Rep. 64, 103–108 (mm6404a7 )(pii). Jain, R.B., 2016. Trends in exposure to second hand smoke at home among children and nonsmoker adolescents. Sci. Total Environ. 542, 144–152. http://dx.doi.org/10. 1016/j.scitotenv.2015.10.076. Kanchongkittiphon, W., Gaffin, J.M., Phipatanakul, W., 2014. Review article The indoor asthma environment and inner-city childhood asthma. JAsian Pac. J. Allergy Immunol. 32, 103–110. Kelly, A., Denning-Kemp, G., Geiringer, K., Abdulhamid, A., Albabtain, A., Beard, M., Brimble, J., Campbell, A., Feng, S., Haminudin, M., Hunter, J., Kunac, N., Lin, P.,
41