Tobacco smoke exposure reduction strategies - do they work?

Abstract

Objective:

Many children experience tobacco smoke exposure (TSE) and parents may take preventive measures to reduce TSE. The study goal is to assess if these strategies are associated with lower cotinine values, an objective biological measure of TSE.

Methods:

Families admitted to Children’s Hospital Colorado from 2014–2018 who screened positive for TSE were invited to participate in a tobacco smoking cessation/reduction program. Caregivers were consented and asked about demographics, beliefs around smoking, and strategies to reduce TSE. Child urine samples were collected, tested for cotinine levels, and analyzed using geometric means. Bi-variable comparisons and multivariable linear regression were completed using SAS v9.4.

Results:

213 children (81.4%) are included in this analysis. The median ages of children and parents were 4 and 32 years. 57% of children were male, 36% were Hispanic, and 55% were white. 56% of parents had at least some college education and 69% had an annual income less than $50K. The median daily cigarettes smoked per day was 10. 88% reported using at least one type of protective measure to prevent TSE and 90% believed they protect other household members from TSE. None of the strategies had a significant relationship with lower cotinine levels on bi-variable or multivariable analyses.

Conclusions:

Parental strategies to decrease TSE did not result in lower cotinine levels. Many measures are not evidence-based and do not protect children. Parent’s clothing and homes may create a reservoir for nicotine. Education should focus on exposure elimination and cessation rather than protective measures.

Introduction:

Prevalence of tobacco smoke exposure (TSE) among nonsmokers has declined in the past twenty years.¹ Despite this success, children are still more heavily exposed to TSE than adults.¹ Children 3–11 years old and adolescents 12–19 years old experience the highest TSE prevalence at 40.6% and 33.8% respectively.¹ Higher TSE prevalence was also observed in those living below the poverty level (43.2%), persons living in rental housing (36.8%), and people with a grade 11 or less education (27.6%).¹ Estimates using biochemical measures demonstrate that over half of U.S. children are exposed to secondhand smoke (SHS)², and the main place young children are exposed is in the home.²

Children who experience TSE may be exposed to third hand smoke (THS) as well as SHS.⁹ SHS refers to the aerosols present during or directly after the act of smoking.³ Exposure to SHS is often via inhalation and has a typical time scale of several hours.³ Alternatively, THS refers to the residual tobacco smoke pollutants left after a cigarette has been extinguished.⁴ THS can deeply penetrate surfaces and dust, react with atmospheric antioxidants to create harmful byproducts, and be re-emitted into the gas phase and resuspended in the air.⁴ THS exposure pathways include dermal uptake and ingestion, and can occur up to months after the initial exposure.³ Elevated nicotine levels have been found in homes and cars of smokers when compared to non-smokers or those with a smoking ban in place.⁴ Young children who frequently put their hands or foreign objects in their mouth are particularly at risk for exposure when coming into contact with these contaminated environments.³

Exposure to tobacco smoke has a significant negative impact on child health.² TSE can worsen symptoms related to asthma exacerbation as well as respiratory infections such as bronchiolitis and pneumonia.² Additionally it is associated with a greater risk for sudden unexplained infant death (SUID).² In the event of hospitalization, TSE can also have negative effects on a child’s hospital course. TSE is associated with increased severity of illness for children admitted with bronchiolitis.⁵ Further, children who experience TSE and get influenza have a 70% increased length of stay and their risk of ICU admission is 4.7 times greater than for children with influenza who are not exposed.⁶ Finally, children with TSE who have asthma exacerbation are more likely to be intubated than those not exposed.⁷

Smoke free indoor air policies have been shown to effectively reduce TSE.⁸ Additionally, there are indications that having a smoke free home or vehicle can help a person quit smoking and can reduce adolescents’ risk of becoming smokers themselves.² Parents may take additional protective measures to help reduce TSE in their children, however there is little existing data on what types of protective measures they use and how well they work. Cotinine is the proximate metabolite of nicotine and is commonly used to measure TSE.⁹ Cotinine has an average half life of 16–24 hours and can be used as an indicator of exposure over the preceding 2–4 days.⁹ A cotinine cutoff level that indicates potential active smoking is 10 ng/ml.⁹ Levels that are greater than 10 ng/ml suggest light smoking from direct sources, and levels that are less than 10 ng/ml suggest heavy TSE.⁹ The objective of this study was to assess if the different exposure prevention strategies that parents report using truly result in lower biological measures of TSE. Additionally, the study examines a parent’s reported belief in the protective measures compared to any observed differences in urine cotinine.

Methods:

Study population: Patients and their parents were recruited as part of a randomized controlled trial of a parent smoking cessation intervention delivered at Children’s Hospital, Colorado from 2014 to 2018. Families were screened for eligibility using the question “Does anyone who lives in your home or who cares for your child use tobacco?”. Providers could also recommend potential families to the study team. Families who met the following inclusion criteria were invited to participate in the study: an admitted patient who was less than 17 years of age and had at least one custodial parent or guardian who smoked cigarettes daily. Families were excluded if the parent did not speak English or Spanish, there was unclear or non-permanent custody such as foster care, the child did not live with the smoking parent, or the child had been admitted to hospital care for more than five days at the time of approach. Informed consent was obtained of all participants and $25 was provided on a debit card in compensation for time spent completing study activities. The study was approved by the Colorado Multiple Institutional Review Board and was registered on www.ClinicalTrials.gov at NCT02281864.

Study activities: All enrolled families were given a survey that asked about demographics, smoking habits, home and car smoking rules or bans, use of e-cigarettes or vaping, beliefs about smoking, and strategies used to reduce TSE. Parents were given the option of specifically checking off one or more of the following strategies: use of a smoking jacket while smoking, use of a smoking hat while smoking, use of breath mints after smoking, use of air filters or purifiers, washing hands after smoking, as well as the option to list other strategies they use. Urine samples were collected from infants, young children, and developmentally disabled patients with cotton balls or a urine bag placed in the diaper; children who were able provided a urine sample in a container. Samples were collected within 120 hours (5 days) of Emergency Department or hospital presentation. Samples were stirred for 30 minutes for even distribution, separated into 10 mL aliquots, frozen at −80°C, and batched and sent on dry ice to the laboratory at the University of California San Francisco for urine cotinine testing.

Analysis of urinary tobacco biomarkers: Urinary cotinine (COT) was analyzed using liquid chromatography-tandem mass spectrometry. The limit of quantification (LOQ) for COT was 0.050 ng/ml. Laboratory blank and quality control samples were simultaneously processed and analyzed to assure the quality of the analytical results. Subjects with COT levels greater than 10 ng/ml were considered active smokers and excluded from the analysis.

Statistical Analysis: The study cohort was limited to families who completed the baseline survey and had urine collected and successfully analyzed. Characteristics of the pediatric participants were summarized with count (percent) for categorical data and median (IQR) for continuous data. Any subjects with cotinine levels below the level of quantitation (LOQ) were assigned a value of LOQ/2. Due to the highly skewed nature of the laboratory data, cotinine values were analyzed on the log transformed scale. Transformed means and their confidence intervals were exponentiated to obtain geometric means and confidence intervals. Separate linear regression models were used to compare each protective method with log transformed cotinine outcome. Multivairable linear regression models were used to adjust for having a home smoking ban, number of smokers in home, number of cigarettes smoked per day, time since presentation to urine collection (>=24 hrs, <24hrs), car rules concerning smoking, and use of e-cigarettes. All data was analyzed using SAS v9.4 and statistical tests were performed with a level of 0.05 significance.

Results:

8421 families were screened and 5510 were either excluded for not meeting eligibility or not approached due to parent availability. Of the 2911 families who were eligible after screening, 263 (13%) agreed to participate in the randomized controlled trial (Figure 1). Of these, 214 had both complete survey data and cotinine data. After excluding one teenager with cotinine levels >10 ng/mL (signifying an active smoker), 213 (81%) were included in these analyses. Overall, the median parent age was 32 years (IQR: 27, 36), the median child age was 4 years (IQR: 1,8), and 57% of children were male. The racial and ethnic composition of the patient cohort reflects the population of Colorado: 55% white, 11% black or African American, and 22% multiracial; with 36% reporting Hispanic or Latino ethnicity (Table 1). 69% of families reported an annual income less than $50,000 per year, 56% of parents claimed to have at least some college education, and 60% reported living in standalone housing. The median number of cigarettes parents smoked on an average was 10 (IQR: 5,15). 12% of families reported zero smokers in the home, 34% reported one smoker, and 53% reported two or more smokers (Table 2).

Figure 1:

Study flowchart

Table 1.

Demographics

Response		N (%)
Child’s Gender	Male	121 (57%)
	Female	92 (43%)
Child’s Ethnicity	Not Hispanic/Latino	136 (64%)
	Hispanic/Latino	76 (36%)
	Missing	1 (0%)
Child’s Race	White	117 (55%)
	Black or African American	24 (11%)
	Multiracial	47 (22%)
	Other	20 (9%)
	Missing	5 (2%)
Parent’s Marriage Status	Married or member of a couple	133 (62%)
	Single (never married)	49 (23%)
	Divorced, widowed, separated	29 (14%)
	Missing	2 (1%)
Parent’s Education Level	Some high school or less	32 (15%)
	Grade 12 or GED (high school graduate)	57 (27%)
	College 1 to 3 years (some college or technical school)	95 (45%)
	College 4 years or more (college graduate)	24 (11%)
	Missing	5 (2%)
Family’s annual income	Less than $20,000	67 (31%)
	$20,000 – $50,000	80 (38%)
	More than $50,000	46 (22%)
	Missing	20 (9%)
Type of housing family lives in	Standalone housing	128 (60%)
	Attached housing	81 (38%)
	Missing	4 (2%)

Table 2.

Simple and Multivariable Linear Regression for Urinary Cotinine

			Bivariable association		Multivariable association*
Response		N (%)	Geometric Mean (95% CI)	P value	Geometric Mean Ratio (95% CI)	P value
Protection method
Washing hands after smoking	Yes	168 (79%)	0.97 (0.76, 1.22)	0.85	1.00 (0.58,1.70)	0.99
	No	45 (21%)	1.01 (0.65, 1.57)		ref
Smoking jacket	Yes	85 (40%)	1.22 (0.90, 1.67)	0.08	1.35 (0.87,2.10)	0.18
	No	128 (60%)	0.84 (0.64, 1.10)		ref
Breath mints	Yes	63 (30%)	1.11 (0.77, 1.60)	0.41	1.27 (0.80,2.00)	0.31
	No	150 (70%)	0.92 (0.72, 1.19)		ref
Other	Yes	34 (71%)	0.86 (0.46, 1.64)	0.98	0.95 (0.30, 3.03)	0.93
	No	14 (29%)	0.88 (0.46, 1.69)		ref
Air filters/Purifiers	Yes	37 (17%)	1.75 (1.04, 2.95)	0.0098	1.82 (1.04,3.17)	0.0362
	No	176 (83%)	0.86 (0.69, 1.08)		ref
Smoking hat	Yes	13 (6%)	1.07 (0.44, 2.62)	0.82	1.09 (0.46,2.58)	0.84
	No	200 (94%)	0.98 (0.78, 1.20)		ref
Brush Teeth	Yes	8 (17%)	1.13 (0.70, 1.81)	0.63	1.93 (0.47,7.89)	0.35
	No	40 (83%)	0.83 (0.47, 1.46)		ref
Change Clothes	Yes	8 (17%)	0.66 (0.21, 2.09)	0.61	0.63 (0.16,2.52)	0.50
	No	40 (83%)	0.92 (0.53, 1.59)		ref
Number of smokers in home
1		116 (54%)	0.72 (0.55, 0.95)	0.002
2		77 (36%)	1.24 (0.89, 1.71)	0.11
≥3		20 (9%)	2.24 (1.01, 4.96)	ref
Number of protective methods (of 6) used per subject
0		26 (12%)
1		73 (34%)
2		71 (33%)
≥3		43 (20%)

^1.Without adjusting for multiple covariates, we have 80% power to detect standard effect sizes between 0.4 – 0.6 as significant which transforms to geometric mean ratios between 1.8 – 2.6.

^*adjusted for home smoking ban, number of smokers in home, number of daily cigarettes, time since presentation to urine collection, car smoking rules, use of e cigs

88% of families reported using at least one of the six most common protective methods (using a smoking jacket or hat, washing hands after smoking, using breath mints after smoking, using air filters or purifiers, brushing teeth after smoking, or changing clothes after smoking) and 64% of families stated that there was a home smoking ban (Table 2). 48 families selected “other” protective method of which 14 were classified into existing categories. The remaining 34 could not be grouped further and included tactics such as gum, candy, or air fresheners.

When asked about beliefs around smoke exposure, 77% said that there was not a safe level of exposure to secondhand smoke, 89% said that they were at least slightly worried about their child’s exposure to secondhand smoke, and 83% of families agreed that they protected all other household members from any harm related to their smoking. Of all the methods analyzed, none had a significant relationship with lower cotinine levels on both bivariable and multivariable models (Table 2). Use of air filters or purifiers were associated with a significantly higher level of cotinine (1.75 vs. 0.86; p<0.01). Children whose parents say there is a home smoking ban had lower geometric mean cotinine levels than those without (0.88 and 1.29 respectively), although this difference was not significant (p=0.09). To account for THS, families were considered to have a home smoking ban or vehicle smoking ban (car rules concerning smoking) only if no smoking was reported allowed in the home or car at all, regardless of children being present.

The multivariable models analyzed the following confounders: the number of smokers in the home, the number of cigarettes smoked per day, the presence of a home smoking ban, time since presentation to urine collection (>=24 hrs, <24 hrs), car rules concerning smoking, and use of e-cigarettes. On this analysis, air filters remained the only significant difference (p=0.0362).

Discussion:

In this cohort the majority of families were well educated in understanding that there is no safe level of TSE and most reported being worried about their child’s exposure. Additionally, most parents were utilizing at least some type of preventative measure. However, over 80% of families believed that the strategies they were using were successfully protecting other members of the household from TSE, including their children. Lower cotinine levels were not observed among children whose caregivers used any of the measures listed. Many of these measures are not evidence-based, and do not actually protect children from TSE sources such as a parent’s clothing, their home, or their car.³ The finding that air filters are associated with higher cotinine levels is interesting and contrasts with prior studies¹⁰; it is possible either that parents who use air filters are less careful about exposure, or that parents who know their exposure is high may choose to use the air filter. Future research is needed to understand how air filters may increase a child’s TSE, however the data analyzed in this study indicates that they are not sufficient in protecting children from tobacco smoke.

When potential confounders were taken into account, including the number of smokers in the home, the number of cigarettes smoked per day, the presence of a home or car smoking ban, use of e-cigarettes, and when the urine was collected, these results do not change. Children in homes where parents smoke fewer cigarettes per day and have fewer smokers in the home have significantly lower cotinine levels, however their levels may still be high enough to potentially cause harm.¹¹

This study provides clear evidence that there is need for education in how to reduce a child’s exposure from all forms of TSE, both from SHS and THS. It is vitally important to remove smoke from homes and cars, and restrict child’s visits to other homes or with people who do not practice these same precautions. Cessation and elimination remains the most reliable preventative strategy to reduce TSE in children.

Limitations:

Data was self-reported via survey by the parents and could not be corroborated, and so it is possible that parents claimed to have taken certain protective steps when they actually did not. There is an assumption that the TSE is coming from the parents themselves and not from an external source such as neighbors. Older children could be more susceptible to these nonparental sources, especially unknown exposure to e-cigarettes. Since the parents were participating in a randomized controlled trial, we were limited to children of parents who were willing and eligible to participate in an intervention (13% of families approached).The low participation rate and characteristics of this sample may have biased the results. Finally, the sample is from children who were admitted to a hospital in Colorado, USA. Therefore, results may not be generalizable outside of this population and cannot be accurately applied to all children.

Conclusions:

Parents are willing to utilize preventative measures and have demonstrated that they want to protect their children from TSE. Challenging the belief that these measures work and providing adequate education will be the cornerstone for any behavior change. Even low levels of TSE can cause exceptional harm in children¹¹ and parents should be supported in their concerns about exposure. Since cessation is the only way to completely protect children from tobacco smoke, this study highlights the critical need to provide evidence-based smoking cessation programs for parents in a variety of settings, as well as policies to prevent tobacco use among adolescents and young adults. Additional education for caregivers should focus on eliminating children’s exposure to tobacco smoke.

What’s New:

This research evaluates the types of protective measures parents use to lower their child’s tobacco smoke exposure and their effectiveness. Physicians can use this information to make evidenced based recommendations to parents about smoking cessation and elimination.