Adolescent Tobacco Smoke Exposure, Respiratory Symptoms, and Emergency Department Use

A secondary analysis of wave 2 data (October 3, 2014–October 30, 2015) from the Population Assessment of Tobacco and Health (PATH) Study, a longitudinal cohort study in which researchers measure tobacco use behavior and related health outcomes in US adolescents and adults,²¹ was conducted. A total of 7389 nonsmoking adolescents were included after excluding those who reported that they had asthma diagnoses (n = 2198) and were current smokers (n = 143). PATH Study procedures are described elsewhere.²¹ A university-based institutional review board considered the study to be not human subjects research, and exempt from review.

We used 3 exposure variables that included adolescent self-report of the following: (1) lives with a smoker who may not smoke inside the home and whether anyone who lives in the home now smokes cigarettes, cigars, cigarillos, or filtered cigars; (2) home TSE and whether smoking tobacco products that are burned are allowed inside the home by anyone who lives in the home, including visitors or workers; and (3) having ≥1 hour of TSE (how many hours they were around others who were smoking in the past 7 days, including time at home, in a car, at school, or outdoors).

TSE-related symptoms included adolescent self-report of the following: shortness of breath compared with peers, finding it harder to exercise compared with peers, chest sounding wheezy during or after exercise in the past 12 months, wheezing or whistling in the chest in the past 12 months, wheezing that disturbed sleep in the past 12 months, speech that is limited to only 1 to 2 words between breaths because of wheezing in the past 12 months, and dry cough at night that is not associated with a cold or chest infection in the past 12 months. Wheezing-related questions were only asked among adolescents who reported having experienced wheezing or whistling within their lifetime (n = 942). We assessed self-reported health indicators, for which responses were collapsed into 2 categories because of skewed distributions: overall health, physical health (poor, fair, or good versus very good or excellent), and frequency of missing school because of illness (never or rarely versus sometimes, often, or very often).

For ED and/or UC use, we used parent report variables because adolescent report was not publicly available. The variables included the following: (1) visited an ED and/or UC for a health problem in the past 12 months and (2) the number of ED and/or UC visits for a health problem in the past 12 months.

We selected the following covariates a priori: adolescent sex (male or female), age (12–14 years or 15–17 years), race (white, African American, or other), ethnicity (non-Hispanic or Hispanic), and parent education level (less than high school, high school or equivalent, some college, bachelor’s degree, or advanced degree). Adolescent age was dichotomized in the PATH Study for public use.

Analyses were performed by using R version 3.3.0.²² We used sampling weights provided in the PATH Study to compensate for sampling design factors (eg, oversampling), nonresponse adjustment factors, differential nonresponse rates, possible sampling frame deficiencies, and variable selection probabilities. We performed χ² tests to examine the relationship between TSE measures and covariates. We report raw sample sizes and weighted percentages to provide estimates that are generalizable to the US nonsmoking adolescent population without asthma diagnoses. We built a series of logistic regression models, adjusting for the covariates, to examine the association between TSE and TSE-related symptoms, other health indicators, and ED and/or UC use. We report adjusted odds ratios (aORs) and 95% confidence intervals (CIs). We built multiple Poisson regression models to assess the relationship between TSE and the number of ED and/or UC visits while controlling for the covariates. We present weighted least squares means, weighted SEs, and β coefficients. A sensitivity analysis was performed to assess whether removing adolescents who may not have been diagnosed and treated for asthma and/or reported asthma symptoms of wheezing or whistling in the chest during the past 12 months (n = 312) from analysis skewed the study’s results. Trends remained similar despite the smaller sample size and lower power. Analyses were 2 sided, with P < .05 indicating significance. Incomplete cases were excluded before each analysis.

Of the 7389 nonsmoking adolescents included after excluding those with asthma diagnoses (n = 2198) and current smokers (n = 143), 49.7% (n = 3651) were boys and 42.4% (n = 3181) were 12 to 14 years old. A total of 77.2% (n = 5078) were non-Hispanic, 71.1% (n = 4832) were white, 15.0% (n = 1054) were African American, and 13.9% (n = 1053) were of another race. Parent education level varied, with 12.4% obtaining less than a high school diploma (n = 1033), 21.7% completing high school or equivalent (n = 1595), 30.5% obtaining some college (n = 2098), 22.4% completing a bachelor’s degree (n = 1389), and 13.0% completing an advanced degree (n = 755). A total of 24.7% (n = 1868) of adolescents lived with a smoker, 17.7% (n = 1338) had home TSE, and 35.3% (n = 2554) had ≥1 hour of TSE.

Boys and girls had similar rates of living with a smoker (24.4% and 24.9%, respectively) and home TSE (17.7% and 17.7%, respectively). There was a significant difference between boys (31.9%) and girls (37.7%) for TSE ≥1 hour (P < .001; Table 1). Adolescent age differed by living with a smoker (P = .003) and TSE ≥1 hour (P < .001), with higher rates reported among 15- to 17-year-olds (26.2% and 37.3%, respectively) compared with 12- to 14-year-olds (23.1% and 33.2%, respectively). Race differed on the basis of living with a smoker (P < .001) and home TSE (P < .001), with the highest rates reported among African American adolescents (30.2% and 24.0%, respectively), followed by white adolescents (24.5% and 17.3%, respectively) and adolescents of other races (23.2% and 14.9%, respectively). Ethnicity differed on the basis of TSE (all P < .001), with non-Hispanic adolescents reporting higher rates of living with a smoker, home TSE, and TSE ≥1 hour (25.9%, 18.5%, and 36.9%, respectively) compared with Hispanic adolescents (20.5%, 14.7%, and 29.5%, respectively). Parent education level differed on the basis of TSE (all P < .001; Table 1). Participants with parents who completed high school or an equivalent had the highest rates of living with a smoker (35.9%), home TSE (26.3%), and TSE ≥1 hour (44.4%).

Adolescents who lived with a smoker, had home TSE, and had ≥1 hour of TSE were at increased risk of reporting shortness of breath, finding it harder to exercise compared with peers, wheezing during or after exercise, and dry cough at night than were unexposed participants (all P < .001; Table 2). Adolescents who lived with a smoker (P < .001) and had home TSE (P = .02) were at increased odds of reporting wheezing or whistling in the chest, and only adolescents with home TSE (P = .02) were at increased risk of reporting wheezing that disturbed sleep. Regarding other health indicators, adolescents who lived with a smoker, had home TSE, and had ≥1 hour of TSE were less likely to report very good or excellent overall health and physical health status (all P < .001; Table 2). Participants who lived with a smoker, had home TSE, and had ≥1 hour of TSE were ∼1.50 times more likely to report they sometimes, often, or very often missed school because of illness (all P < .001).

Participants who lived with a smoker and had ≥1 hour of TSE were significantly more likely to have had an ED and/or UC visit (both P < .001; Tables 3 and 4). Participants who lived with a smoker, had home TSE, and had ≥1 hour of TSE were at increased risk of having a higher number of ED and/or UC visits compared with unexposed participants. Specifically, participants with TSE had a mean of 1.62 to 1.65 visits compared with unexposed participants, who had a mean of 1.42 to 1.48 visits (Tables 3 and 4).

We found that more than one-third (35.3%) of the nonsmoking adolescent population without asthma diagnoses nationwide were around others who smoked for ≥1 hour in the past 7 days. Homa et al² found similar rates (33.8%) of biochemically validated TSE, as measured by serum cotinine with a mean half-life of ∼16 hours,²³ thus reflecting relatively short-term TSE. We found that one-quarter (24.7%) of adolescents lived with a smoker, and 17.7% lived in homes in which smoking is allowed inside the home by anyone. Similarly, Merianos et al¹⁴ found that one-quarter (24.1%) of parents nationwide reported that their children (aged 0–17 years) lived with a smoker, but only 5% had home TSE. Thus, we report exceptionally high rates of self-reported home TSE in adolescents without asthma diagnoses. One potential reason for the difference is that parents may underestimate their smoking prevalence,^24,25 but we used adolescent report of home TSE. It should be noted that adolescents who live with smokers who do not smoke inside the home are still potentially exposed to tobacco smoke and its related toxicants.²⁶ Despite an overall recent decline in TSE,² we highlight the need for continued TSE reduction initiatives, especially for adolescents, nationwide.

We report demographic differences based on TSE. Girls had higher rates of TSE ≥1 hour. No sex differences were reported based on living with a smoker or home TSE, which differed from previous research that revealed that boys had slightly higher rates of living with a smoker and home TSE.¹⁴ We found that older adolescents had relatively higher rates of living with a smoker and TSE ≥1 hour. Non-Hispanic adolescents had high rates of all TSE types, and African American adolescents had high rates of living with a smoker and home TSE, which is comparable to previous work.² African American and white adolescents had similar rates of TSE ≥1 hour. Parent education level varied; the lower the grade level, with the exception of less than a high school diploma, the higher the TSE prevalence, which parallels existing literature.^2,27 

One of our objectives was to examine the association between TSE and potentially related symptoms and other health indicators. As posited, we found that adolescents without asthma diagnoses who had any TSE were at increased odds of having TSE-related symptoms, including shortness of breath and dry cough at night. Adolescents who lived with a smoker had higher risk of reporting wheezing or whistling in the chest, and adolescents with home TSE had increased risk of wheezing or whistling in the chest and wheezing that disturbed sleep. Wheeze-related symptoms are frequently reported as TSE-attributable symptoms.⁵ One potential explanation for the variations in symptoms by TSE measures is that those who live with a smoker and have home TSE are regularly exposed compared with those who may have temporary TSE for ≥1 hour in the past 7 days. For example, adolescents who live with someone who smokes tobacco inside the home may experience exposure at night, thus leading to wheezing that disturbed sleep. Future researchers should capture the frequency and amount of TSE and use biochemically validated measures to provide better insight into the differences found in this study.

TSE negatively impacts adolescents’ overall health and physical health.^1,10 We found that TSE translated into a lower likelihood of having very good or excellent overall health and physical health status, as hypothesized. We found that regardless of TSE measure, adolescents with TSE reported that it was harder to exercise compared with peers and wheezing during or after exercise. Previous work reveals that TSE negatively impacts exercise capacity in adults,²⁸ which was reported in our study. TSE may cause overall decreased maximum oxygen uptake because TSE reduces oxygen delivery to the myocardium and potentially reduces the myocardium’s ability to use the delivered oxygen effectively.⁵ This physiologic response caused by TSE inhibits exercise. Additionally, participants who lived with a smoker, had home TSE, and were around TSE for ≥1 hour were at increased risk of frequently missing school because of being sick. This expanded on previous findings that living with a smoker contributed to increased school absenteeism.⁸ TSE contributes to increased morbidity and related consequences in adolescents without asthma diagnoses.

Another study objective was to examine the relationship between TSE and ED and/or UC use among adolescents who are often high users of these settings. The literature reveals mixed findings among youth; thus, this study was conducted to further assess these relationships by using 3 distinct TSE measures. As hypothesized, we found that adolescents who were exposed by living with a smoker or had ≥1 hour of TSE were at an elevated risk of ED and/or UC use compared with adolescents who did not live with a smoker or had no TSE in the past 7 days. These findings align with previous work by Merianos et al,¹⁴ which found that children who lived with a smoker were at an increased risk, and children with TSE levels (as measured by serum cotinine) that were indicative of active smoking (≥3 ng/mL) were 3.5 times more likely to select an ED as the place they most often go for health care.¹³ 

Interestingly, we found that all TSE types increased the risk of having a higher number of ED and/or UC visits. These venues should be considered high-priority settings for TSE prevention intervention because of the increased interaction with adolescents who are exposed. EDs are typically the main sources for primary care and preventable health care services for Americans who have limited access to primary health care settings.²⁹ Previous TSE reduction initiatives reveal that health care providers and parents who smoke and bring their children to ED and/or UC settings are highly receptive to these feasible efforts.^30,31 In addition, delivering interventions in these settings may increase caregiver quit attempts and successful quits and ultimately decrease child TSE.^32,–34 This includes the implementation of electronic health record–embedded clinical decision support systems that can be easily incorporated into busy ED and/or UC providers’ workloads.^34,35 However, despite these practical options, prevention initiatives in these settings, including the first step of TSE screening, are infrequent.^36,–38 Efforts supported by evidence-based policies and procedures, including the implementation of systems to prevent, identify, and treat tobacco dependence, may help to increase ED and/or UC professionals’ engagement with adolescents and their families about eliminating TSE to protect their overall health.^39,40 

Study strengths include examining the relationship between several TSE measures and ED and/or UC use by using a sample that is generalizable to the US nonsmoking adolescent population without asthma diagnoses and controlling for sociodemographics that may influence this relationship. Despite the study strengths, the limitations should be noted. We were limited to public-use variables provided in the PATH Study, including relying on adolescent report of TSE, which may have been underreported,^41,42 rather than using biomarkers. We relied on parent report of ED and/or UC use because adolescent report was unavailable. Although authors of other studies have used population-based survey data that rely on parent report of a child’s health care use,^13,14 future researchers should examine the validity of parent report by comparing responses to medical records. Residual confounding may have biased results, although we adjusted for well-known confounders. We adjusted for parent education level as a proxy for socioeconomic status because the PATH Study does not provide publicly available information on income level or health insurance. Although it is possible that some adolescents who were not diagnosed yet were included, the proportion of adolescents with asthma who were excluded (23%) was higher than in other national data, which indicate that ∼18% have had lifetime asthma.⁴³ Finally, inferences are not causal or temporal because we analyzed 1 wave of data. Because the PATH Study continues to collect data, it would be of interest to examine the longitudinal relationship between TSE and related symptoms and ED and/or UC use among nonsmoking adolescents.

With this study, we add to the literature by examining the potential impact TSE has on TSE-related symptoms and ED and/or UC use among solely nonsmoking adolescents without asthma diagnoses and revealing that a considerable proportion are exposed by living with a smoker, at home, and for ≥1 hour. We found that different TSE types uniquely increased the risk of TSE-related symptoms, other health indicators, and ED and/or UC use. However, all TSE types increased the risk of having a higher number of visits to an ED and/or UC, and this has not been previously reported. Because adolescents are high users of EDs and/or UCs for primary care reasons, these venues are high-volume settings that should be used to offer interventions to adolescents with TSE and their families. They should be educated on the importance of TSE elimination to increase their overall health and wellness, especially because they may not be receiving preventive care elsewhere. There are feasible options available for EDs and UCs to implement such efforts that would not interfere with clinical flow. Targeting initiatives to adolescents with potential TSE-related symptoms and illnesses may assist in reducing TSE and related morbidity. Sustained TSE reduction research is needed to better understand TSE-related symptoms and ED and/or UC use. Physicians at EDs and UCs should consider referring adolescents with asthma symptoms (eg, recurrent wheeze, shortness of breath, or difficulty exercising) for a formal evaluation to ensure that undiagnosed cases are properly treated. Longitudinal research with objective measures would provide better insight into the influence that TSE has on related symptoms, overall health, and ED and/or UC use over time.

aOR

adjusted odds ratio

CI

confidence interval

ED

emergency department

PATH

Population Assessment of Tobacco and Health

TSE

tobacco smoke exposure

UC

urgent care