Abstract
Introduction
Parents/guardians can effectively reduce tobacco smoking and secondhand smoke exposure among youth by adopting and enforcing rules against indoor tobacco smoking (ie, home smoke-free policies). We investigate home smoke-free policies from childhood to adolescence in the United States and across rural, suburban, and urban households.
Aims and Methods
We analyzed 2019–2020 National Survey of Children’s Health data from n = 5,955 parents of youth aged 0–17, living at home with a tobacco smoker in the United States (U.S). Geographical categories were: rural, suburban, and urban. Home smoke-free policy reflected prohibiting tobacco smoking inside the home. Weighted logistic regressions examined the (1) association between youth age and home smoke-free policies, (2) interaction between geographic category and youth age, and (3) differing associations between youth age and home smoke-free policies by geography. Models controlled for youth race, ethnicity, sex, parental education, household annual income, and home structure.
Results
Approximately 13.2% of U.S. households with a smoker did not have a home smoke-free policy. Stratified analyses found one-year increase in youth age was associated with lower odds of having a home smoke-free policy in rural (aOR:0.91; 95%CI: 0.87–0.95) and urban (aOR: 0.96; 95%CI: 0.92–1.00; p = .039), but not suburban (aOR:1.00; 95%CI: 0.95–1.05) households, controlling for covariates.
Conclusion
Odds of having a smoke-free home in the U.S. declined significantly in rural (9% per year) and urban (4%) but not suburban (0%) households. We quantify declines in home smoke-free policies as children age and identify geographic disparities for this environmental determinant of health.
Implications
Health promotion efforts targeting secondhand smoke prevention is needed, particularly for parents of older youth. Furthermore, there is a clear geographic bias in secondhand smoke exposure among all youth particularly older youth. Tailored interventions are needed to address geographic disparities in secondhand smoke exposure among rural and urban youth.
Introduction
Secondhand smoke exposure during childhood and adolescence is linked to respiratory illness and infection1,2 and is a strong predictor of smoking initiation.3,4 State- and local-level regulations have effectively reduced adolescent secondhand smoke exposure in public locations (eg, restaurants, parks)5 and in vehicles.6 Unfortunately, approximately 38% of youth aged 3–11 and 28% of youth aged 12–17 reported exposure to secondhand smoke at home, in the United States (US).7 Rules against smoking combustible tobacco indoors, termed home smoke-free policies, are an effective method for reducing adolescent secondhand smoke exposure at home.8,9 As such, research is needed to identify possible gaps in home smoke-free policies so as to inform interventions aimed at reducing youth secondhand smoke exposure.
Geographic disparities in secondhand smoke and home smoke-free policies are present in the U.S. and disproportionately impact individuals in rural areas.10–14 One study found that rural households were 28% less likely to prohibit cigarette smoking in their homes.10 Another study found that rural youth were 26% more likely to be exposed to secondhand smoke at home.7 Health promotion interventions, including communication campaigns, have had success in changing perceptions of indoor smoking (ie, norms) and adoption of home smoke-free policies (ie, behaviors) across diverse populations15 and in various geographic settings.16 Though similar intervention methodologies can be utilized to address geographic disparities in secondhand smoke exposure for youth in rural areas, an extensive understanding of the role of geographic region on home-smoke free policies is an essential element in developing a tailored intervention.
Age is one factor to consider when exploring home smoke-free policies and possible geographic disparities. Parents are less likely to report having a home smoke-free policy as their child ages, according to nationally representative data of parents from 2016 to 2018.17 Similarly, high school students are more likely to report exposure to secondhand smoke, relative to middle school students.7 One reason for eliminating home smoke-free policies may be that parents reduce their monitoring and protective activities as their child ages.18 However, the role of youth age on home smoke-free policies has not been thoroughly studied.
Generally, rural homes are less likely to have (or enforce) home smoke-free policies.10–12 Differing social norms about indoor smoking have been found to be a driving factor for whether home smoke-free policies exist, with indoor tobacco smoking more normalized in rural settings.19 Thus, it is plausible that greater normalization of indoor smoking in rural homes10 results in greater exposure to secondhand smoke exposure at home for rural youth, relative to non-rural youth.7,14 To the best of our knowledge, the interaction of geographic category and youth age has not been thoroughly studied, though some have examined state-20 and local-level differences in secondhand smoke exposure.10
Study Aims and Hypotheses
First, we determine the association between child age and odds of being covered by a home smoke-free policy among a nationally representative sample of parents in the United States in 2019 and 2020. Second, we compare the association between child age and home smoke-free policy across urban, suburban, and rural areas. We hypothesize that the association between youth age and no home smoke-free policy will be greatest in rural and suburban households, relative to urban households. Findings from this study may inform future health promotion campaigns, particularly interventions targeting rural populations and/or parents of adolescents.
Methods
Study Sample and Population
We analyzed data from the National Survey of Children’s Health (NSCH) collected from June 2019 to January 2020 by the U.S. Census Bureau.21 The NSCH produces a nationally representative sample of parents of youth from 0 to 17 years of age in the United States. The NSCH instrument informs participants that survey completion is voluntary. Written consent was provided digitally or written at the completion of a questionnaire designed to determine if a volunteer meets inclusion criteria for participation.21
The NSCH (2019–2020) collected data from 72,210 parents or guardians of children aged 0-17. For this study, only 49,774 participants had their geography data released by NSCH and, thus, eligible for consideration to be included in the analytic sample. Living with a tobacco smoker was the first eligibility criterion of this study. Furthermore, 7,752 participants reported living with a tobacco smoker and had their geographic data released by NSCH.
Of the 7,752 eligible participants, 17.9% (n = 1,104) had no data on home smoke-free policies and an additional 224 had missing data on one or more study variables. This resulted in an analytic sample of 6,424 parents/guardians with complete data on all study variables. We elected to conduct a complete case analysis (CCA) over other methodologies for handling missing data such as multiple imputation (MI). Use of MI is best suited for situations of data missing at random, characterized by small amounts of missing data across several variables within a data structure.22 Conversely, CCA is considered a less biased option, relative to MI, when data are missing not at random and skewed to a single variable.22
An analytical consideration informed our approach to analyzing the 42,022 participants who reported not living with a tobacco smoker. These observations were retained to maintain fidelity with, and validity of, the survey weights of the NSCH dataset.21 We structured our multinomial logistic regression, described later, to maintain fidelity with, and validity of, the survey weights of the NSCH dataset.21 However, these observations are not included in the results reported and discussed in this paper because the computation would be interpreted as the differing risk of living with a tobacco smoker by age. As youth are not capable of deciding home smoke-free policies and addressing adult cessation is beyond the scope of this paper, we do not interpret the corresponding results.
Study Measures
Home Smoke-Free Policies
The primary outcome variable was home smoke-free policies within households with a tobacco smoker. First, participants were asked “Does anyone living in your household use cigarettes, cigars, or pipe tobacco?” Those who said “no” were excluded from this study. Participants who reported that anyone living in their household smokes cigarettes, cigars, or pipe tobacco were then asked: “Does anyone smoke insider your home?” Those who said “no” were considered to have a home smoke-free policy and those who said “yes” were considered to not have a home smoke-free policy.
Youth Age
The primary independent variable of this study is youth age, which ranged from 0 to 17 years. We categorized age into three categories: 0–5 years (young children; referent group), 6–11 years (children), and 12–17 years (adolescents). Youth age was analyzed as a continuous variable in preliminary analyses, described and presented below.
Geographic Category
Participants were grouped into the three mutually exclusive geographic categories—rural, suburban, and urban—following US Census Bureau definitions,21 with each physical address confirmed by the US Census Bureau at the point of data collection. First, participants were classified as being in a metropolitan statistical area (MSA), described as a “county or counties associated with at least one urbanized area of at least 50,000 population, plus adjacent counties having a high degree of social and economic integration with the core as measured through commuting ties.” Those who lived outside of an MSA were considered rural. Participants who did reside in an MSA were classified as suburban if they did not live within a principal city (ie, main core city in a metropolitan area). Participants who did reside in an MSA were classified as urban if they lived within a metropolitan principal city.
Covariates
Given the nature of these data, covariates related to the parent, youth, and household were included in our analytic model, per previous literature.17 Race and ethnicity were assessed and analyzed independently. First, race of youth was categorized as non-Hispanic White (referent), Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian and Other Pacific Islander, and two or more races. Next, a binary indicator of Hispanic ethnicity was included, derived from the question: “Is this child of Hispanic, Latino, or Spanish origin?” Sex assigned at birth was categorized as male (referent) and female. We also controlled for parental education, which was categorized into “high school graduate or less” (referent) and “more than high school.” Next, we controlled for annual household income, as a percentage of the Federal Poverty Level (FPL). Percentage of FPL was categorized into four groups: less than 100% of the FPL (referent), 100–199% of the FPL, 200–399% of the FPL, and 400% or more of the FPL. Additionally, we assessed for family structure, which was categorized as “two parents, currently married” (referent); “two parents, not currently married”; and any other structure, which included “single parent (mother or father)”, grandparent’s house, and “other family type.” Additionally, we controlled for maternal age at child birth, coded continuously from 18 to 45 (or older). Survey year was included to account for its random effect.
Statistical Analysis
Survey weights were applied for all analyses. To maintain the integrity of the population estimates, our analyses (ie, multinomial logistic regression) were conducted with all eligible participants in our dataset (n = 48,025). However, as our hypotheses only pertain to participants who reported living with a smoker, our reported results and corresponding interpretations are specific to an analytic sample of 6,424 parents/guardians who reported living with a smoker.
Prior to testing study hypotheses, we conducted four weighted, multinomial logistic regression models to assess the association between youth age as a continuous variable and odds of home smoke-free policies in the full sample, stratified by geographic category.
To test study hypotheses, first we conducted a weighted, multinomial logistic regression to determine the association between youth age-category and odds of being covered by a home smoke-free policy among the full sample. Households with young children (0–5 years of age) were the referent category, compared to children (6–11 years of age) and adolescents (12–17 years of age). Next, we tested the interaction between the two categorical variables of youth age and geographic category, among the full sample. For these multinomial logistic regression models, living in a household with a home smoke-free policy was the referent outcome, and was compared to living in a household without a home smoke-free policy.
We tested our hypothesis that the relationship between child age and home smoke-free policy would be modified by geographic category. We conducted three iterations of the weighted, multinomial logistic regression, stratified by urban, suburban, and rural households, Comparisons of effect size and 95% confidence intervals across all four estimates (ie, full sample, rural, suburban, urban) were used to discuss statistical and clinical differences in being protected by a home smoke-free policy across geographic category.
We modeled the interaction between child age category and the continuous measure of child age (0–17) using weighted, multinomial logistic regression. This interaction term modeled and compared the odds of having a home smoke-free policy for every one-year increase in youth age (continuously measured) for each category of age (young children, children, adolescents). This analysis was conducted to test for systematic differences in the effect of increasing youth age on odds of having a home smoke-free policy, within each of our age categories. We found no difference in the association between the continuous measure of child age and home smoke-free policies between young child (referent), children (p = .825), or adolescents (p = .658), indicating no bias in our coding of the child age categories.
All analyses controlled for sex, race, ethnicity, poverty level, household structure, parental education, parent/guardian age, number of youth in the household and survey year (2019, 2020). All analyses were conducted using STATA 14.2 (College Station, TX).
Results
Descriptive Statistics
Overall, 19.0% of the sample lived in rural areas; 51.0% lived in suburban areas; and 30.0% lived in urban areas. Youth age did not differ by geographic category. Approximately 13.2% of households with a tobacco smoker in the home did not have a home smoke-free policy. By geographic category, 17.6% of rural households did not have a home smoke-free policy, compared to 11.5% of suburban and 13.4% of urban households. Descriptive statistics of the study sample, stratified by geographic category, are available in Table 1. Descriptive statistics, stratified by home smoke-free policy, are available in Table 2.
Table 1.
Descriptive Statistics of Households with a Smoker in the United States, Stratified by Geographic Category (NSCH 2019/2020, n = 6,424; N = 8,557,701)
| Full sample | Rurala (n = 1,631) |
Suburbanb (n = 3,103) |
Urbanc (n = 1,819) |
p-value | |
|---|---|---|---|---|---|
| Percentage of Sample (95% CI) | 100% | 19.1% (17.5–20.8) |
50.9% (48.5–53.2) |
30.1% (27.8–32.4) |
|
| Age Category(of child) | |||||
| 0–5 years old | 29.0% (26.8–31.3) | 29.1% (24.6–33.9) | 29.8% (26.7–33.2) | 27.6% (23.6–31.9) | p = .932 |
| 6–11 years old | 35.6% (33.3–37.9) | 35.4% (31.1–39.9) | 35.1% (32.1–38.3) | 36.4% (31.7–41.5) | |
| 12–17 years old | 35.4% (33.3–37.6) | 35.6% (31.5–39.8) | 35.0% (32.1–38.0) | 35.9% (31.5–40.7) | |
| Sex(of child) | |||||
| Male | 50.8% (48.4–53.2) | 52.4% (47.8–57.1) | 49.7% (46.4–52.9) | 51.6% (46.7–56.4) | p = .627 |
| Female | 49.2% (46.8–51.6) | 47.6% (42.9–52.3) | 50.3% (47.0–53.6) | 48.4% (43.6–53.3) | |
| Racial Identity (of child) | |||||
| White | 69.1% (66.6–71.4) | 78.8% (74.5–82.6) | 74.4% (71.2–77.4) | 53.9% (48.9–58.8) | p < .001 |
| African American/Black | 14.7% (12.9–16.6) | 11.9% (09.0–15.5) | 09.2% (07.6–11.1) | 25.7% (21.4–30.6) | |
| Asian/Asian American | 2.9% (02.4–03.7) | 0.5% (0.22–1.1) | 3.2% (2.3–4.5) | 4.2% (2.9–5.8) | |
| Multiple Races | 10.8% (09.1–12.8) | 7.4% (04.9–10.9) | 10.7% (8.5–13.3) | 13.3% (9.7–18.1) | |
| Otherd | 02.4% (01.6–03.7) | 1.4% (0.89–2.3) | 2.6% (1.3–4.9) | 2.9% (1.7–4.8) | |
| Hispanic Ethnicity(of child) | |||||
| No | 79.9% (77.4–82.3) | 89.0% (83.9–92.7) | 80.6% (76.9–83.8) | 73.1% (67.7–77.8) | p < .001 |
| Yes | 20.1% (17.7–22.6) | 11.0% (07.3–16.1) | 19.4% (16.2–23.1) | 26.9% (22.2–32.3) | |
| Federal Poverty Line (FPL) e | |||||
| Below the FPL | 27.6% (25.3–29.9) | 32.5% (28.4–37.0) | 21.4% (18.6–24.5) | 34.8% (30.1–39.9) | p < .001 |
| FPL to 199% of FPL | 26.9% (24.9–29.2) | 30.5% (26.6–34.7) | 26.7% (23.7–29.9) | 25.2% (21.4–29.5) | |
| 200% to 399% of FPL | 29.7% (27.6–31.9) | 28.6% (24.4–33.2) | 31.8% (28.7–34.9) | 27.0% (22.9–31.5) | |
| 400% or more of FPL | 15.7% (14.4–17.1) | 08.4% (06.0–11.6) | 20.1% (18.1–22.3) | 12.9% (10.8–15.5) | |
| Home Structure | |||||
| Two Parent Household | 78.2% (76.0–80.2) | 73.2% (68.3–77.2) | 78.9% (75.9–81.6) | 80.1% (75.7–83.9) | p = .065 |
| Not Two Parent Household | 21.8% (19.8–23.9) | 26.8% (22.6–31.6) | 21.1% (18.4–24.1) | 19.9% (16.2–24.3) | |
| Education (of Parent) | |||||
| High School or Less | 42.5% (40.1–44.9) | 54.6% (49.9–59.1) | 38.7% (35.3–42.2) | 41.3% (36.5–46.2) | p < .001 |
| More than High School | 57.5% (55.1–59.9) | 45.4% (40.9–50.0) | 61.3% (57.9–64.7) | 58.7% (53.8–63.5) | |
| Age of Mother at Child’s Birth | |||||
| Mean (SD) | 27.6 (6.2) | 26.9 (7.0) | 27.9 (5.9) | 27.3 (6.2) | p = .011 |
| Year | |||||
| 2019 | 49.1% (46.7–51.5) | 49.1% (44.5–53.7) | 50.1% (46.9–53.4) | 47.2% (42.3–52.2) | p = .569 |
| 2020 | 50.9% (48.6–53.3) | 50.9% (46.3–55.5) | 49.9% (46.6–53.1) | 52.8% (47.8–57.7) | |
Bold reflects statistically significant (p < .05) difference.
Participants were classified as rural if they lived outside of a Metropolitan Statistical Area (MSA).
Participants were classified as suburban if they lived within an MSA but not live within a Metropolitan Principal City.
Participants were classified as urban if they lived within an MSA and lived within a Metropolitan Principal City.
Any other racial identity, including Native Hawaiian and Other Pacific Islanders
Corresponds to federal poverty line for each participant, by state, in 2019/2020
Table 2.
Descriptive Statistics of Households with a Smoker in the United States, Stratified by Home Smoke-Free Policies (NSCH 2019/2020, n = 6,424; N = 8,557,701)
| Home smoke-free policya (n = 5,635) |
No home smoke-free policya (n = 789) |
p-value | |
|---|---|---|---|
| Percent of Sample | 86.9% (85.3–88.4) | 13.1% (11.6–14.7) | |
| Geographic Region | |||
| Rural | 82.6% (79.1–85.5) | 17.4% (14.5–20.9) | p = .013 |
| Suburban | 88.6% (86.3–90.6) | 11.4% (9.4–13.7) | |
| Urban | 86.7% (83.3–89.5) | 13.3% (10.5–16.6) | |
| Age Category(of child) | |||
| 0–5 years old | 88.7% (84.9–91.7) | 11.3% (8.4–15.1) | p = .210 |
| 6–11 years old | 87.2% (84.5–89.6) | 12.8% (10.4–15.5) | |
| 12–17 years old | 85.1% (82.8–87.2) | 14.9% (12.8–17.2) | |
| Sex(of child) | |||
| Female | 87.6% (85.1–89.7) | 12.4% (10.3–14.9) | p = .407 |
| Male | 86.3% (84.0–88.2) | 13.7% (11.8–16.0) | |
| Racial Identity (of child) | |||
| White | 88.3% (86.5–89.9) | 11.7% (10.1–13.6) | p = .006 |
| African American/Black | 79.7% (74.8–83.9) | 20.3% (16.1–25.2) | |
| Asian/Asian American | 89.6% (77.8–95.5) | 10.4% (4.6–22.2) | |
| Multiple Races | 85.7% (78.5–90.8) | 14.3% (9.2–21.5) | |
| Otherb | 93.5% (81.0–98.0) | 6.% (2.1–19.1) | |
| Hispanic Ethnicity(of child) | |||
| No | 93.4% (89.6–95.9) | 6.6% (4.1–10.4) | p < .001 |
| Yes | 85.3% (83.5–86.9) | 14.7% (13.1–16.6) | |
| Federal Poverty Line (FPL) c | |||
| Below the FPL | 81.5% (77.7–84.8) | 18.5% (15.2–22.3) | p < .001 |
| FPL to 199% of FPL | 82.6% (78.7–85.9) | 17.4% (14.1–21.4) | |
| 200% to 399% of FPL | 91.0% (88.7–92.8) | 9.0% (7.2–11.3) | |
| 400% or more of FPL | 96.1% (94.7–97.1) | 3.9% (2.9–5.3) | |
| Family Structure | |||
| Two Parent Household | 87.9% (86.1–89.5) | 12.1% (10.5–13.9) | p = .012 |
| Not Two Parent Household | 83.3% (79.5–86.5) | 16.7% (13.5–20.5) | |
| Education (of Parent) | |||
| High School or Less | 82.8% (79.8–85.5) | 17.2% (14.5–20.2) | p < .001 |
| More than High School | 89.9% (88.2–91.5) | 10.1% (8.5–11.8) | |
| Age of Mother at Child’s Birth | |||
| Mean (SD) | 27.7 (6.3) | 28.9 (6.0) | p = .036 |
| Year | |||
| 2019 | 88.0% (85.7–89.9) | 12.0% (10.1–14.3) | p = .179 |
| 2020 | 85.9% (83.5–88.0) | 14.1% (12.0–16.5) | |
NOTE: p-value reflects chi-squared and t-test analyses.
Bold reflects statistically significant (p < .05) difference.
Self-reported rules on indoor tobacco smoking, among household anyone living in your household smokes combustible tobacco (cigarettes; cigars; pipe tobacco).
Any other racial identity, including Native Hawaiian and Other Pacific Islanders.
Corresponds to federal poverty line for each participant, by state, in 2019/2020.
Statistical Analyses
The initial analysis found that for every one-year increase in child age, odds of having a home smoke-free policy declined for the full sample (aOR: 0.96; 95% CI: 0.93–0.99) as well as for rural (aOR:0.91; 95% CI: 0.87–0.95) and urban (aOR: 0.96; 95% CI: 0.92–1.00; p = .039) youth. There was no statistical association between child age and home smoke-free policies among suburban youth (aOR: 1.00; 95% CI: 0.95–1.05). All analyses controlled for sex, race, ethnic identity, poverty level, household structure, parental education, parent/guardian age, number of child in the household and survey year (2019, 2020). These results established the requisite relationships between child age, home smoke-free policies, and geographic category. These results are available in Table 3.
Table 3.
Preliminary Analyses of Child Age, Home Smoke-Free Policies, and Geographic Category among Households with a Tobacco Smoking in the United States (NSCH 2019/2020)
| Full sample, n =6,424 | |
|---|---|
| Odds Ratio 95% Confidence Interval |
|
| Primary Association | |
| Model 1:Child Age (Continuous Measure) | |
| Child Age (0 to 17) | 0.96** (0.93–0.99) |
| Stratified By Geographic Category | |
| Model 2: Child Age by Geographic Category | |
| Rural Households: Child Age | 0.91** (0.93–0.99) |
| Suburban Households: Child Age | 1.00 (0.95–1.05) |
| Urban Households: Child Age | 0.94* (0.92–1.00) |
***p < .001; **p < .01; *p 0.05.
Bold reflects statistically significant (p < .05) difference.
Self-reported rule on indoor tobacco smoking, among household anyone living in your household smokes combustible tobacco (cigarettes; cigars; pipe tobacco).
Any other racial identity, including Native Hawaiian and Other Pacific Islanders
Corresponds to federal poverty line for each participant, by state, in 2019/2020.
As seen in Table 4, odds of having a smoke-free policy were 0.64 times lower (95% CI: 0.46–0.94) in homes with adolescents (12–17 years old) relative to young children (0–5 years old) among the full sample. There was no statistically significant difference in the odds of having a smoke-free policy in homes with children aged 6–11 (aOR: 0.79; 95% CI: 0.54–1.18).
Table 4:
Association Between Child Age Category and Home Smoke-Free Policy in Homes with a Tobacco Smoker, Stratified by Geographic Category (NSCH 2019/2020; n = 6,424; N = 8,718,9016)
| Full sample (n = 6,553) |
Rurala (n = 1,631) |
Suburbanb (n = 3,103) |
Urbanc (n = 1,819) |
|
|---|---|---|---|---|
| Odds Ratio 95% Confidence Interval |
Odds Ratio 95% Confidence Interval |
Odds Ratio 95% Confidence Interval |
Odds Ratio 95% Confidence Interval |
|
| Age Category (of child) | ||||
| 0–5 years old | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| 6–11 years old | 0.80 (0.54–1.18) | 0.49** (0.25–0.93) | 1.42 (0.79–2.54) | 0.51 (0.24–1.09) |
| 12–17 years old | 0.64* (0.44–0.93) | 0.29*** (0.16–0.56) | 1.04 (0.58–1.85) | 0.54 (0.28–1.05) |
| Sex(of child) | ||||
| Male | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| Female | 0.89 (0.69–1.14) | 0.73 (0.48–1.11) | 1.03 (0.67–1.58) | 0.84 (0.49–1.44) |
| Racial Identity (of child) | ||||
| White | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| African American/Black | 0.82 (0.57–1.17) | 0.82 (0.38–1.75) | 0.59 (0.33–1.06) | 0.98 (0.46–2.09) |
| Asian/Asian American | 1.53 (0.65–3.59) | 0.69 (0.12–4.12) | 1.12 (0.32–3.93) | 2.21 (0.65–7.47) |
| Multiple Races | 0.71 (0.44–1.14) | 0.43 (0.18–1.02) | 1.12 (0.47–2.68) | 0.56 (0.21–1.48) |
| Otherd | 1.42 (0.50–3.99) | 1.76 (0.49–6.26) | 1.36 (0.25–7.59) | 1.30 (0.16–10.5) |
| Hispanic Ethnicity(of child) | ||||
| Yes | 0.31*** (0.19–0.53) | 0.20** (0.07–0.59) | 0.17*** (0.70–0.41) | 0.58 (0.24–1.39) |
| Poverty Line e | ||||
| Below the FPL | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| FPL% to 199% of FPL | 1.04 (0.73–1.47) | 0.96 (0.52–1.76) | 0.92 (0.51–1.69) | 1.10 (0.55–2.21) |
| 200% to 399% of FPL | 2.14*** (1.45–3.14) | 1.97* (1.04–3.73) | 1.56 (0.84–2.89) | 3.64** (1.47–9.05) |
| 400% or more of FPL | 5.02*** (3.19–7.91) | 3.09* (1.15–8.29) | 3.83*** (1.90–7.70) | 9.86*** (3.84–25.3) |
| Home Structure | ||||
| Two Parent Household | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| Not Two Parent Household | 0.84 (0.61–1.15) | 0.81 (0.47–1.39) | 0.92 (0.53–1.59) | 0.74 (0.34–1.53) |
| Education (of Parent) | ||||
| High School or Less | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) | 1.00 (Ref) |
| More than High School | 1.47** (1.08–2.01) | 1.47 (0.92–2.35) | 1.99** (1.22–3.25) | 0.81 (0.42–1.57) |
| Age of Mother at Child’s Birth | ||||
| Mean (SD) | 0.99 (0.97–1.01) | 1.00 (0.97–1.03) | 0.99 (0.97–1.03) | 0.98 (0.94–1.03) |
***p < .001; **p < .01; *p < .05.
Bold reflects statistically significant (p < .05) difference.
Participants were classified as rural if they lived outside of a Metropolitan Statistical Area (MSA).
Particiapnts were classified as suburban if they lived within an MSA but not live within a Metroplitan Principal City.
Participants were classified as urban if they lived within an MSA and lived within a Metropolitan Principal City.
Any other racial identity, including Native Hawaiian and Other Pacific Islanders.
Corresponds to federal poverty line for each participant, by state, in 2019/2020.
Rural households with adolescents had 0.30 (95% CI: 0.16–0.56) lower odds of having a home smoke-free policy, relative to homes with youth children, controlling for covariates. Similarly, rural households with children had 0.49 (95% CI: 0.26–0.93) lower odds of having a home smoke-free policy, relative to homes with youth children, controlling for covariates. This reflects as a change in effect size of 53.1% for adolescents and 40.0% for children, from rural youth relative to the full sample.
Urban households with adolescents had 0.54 (95% CI: 0.28–1.05; p = .069) lower odds of having a home smoke-free policy, relative to homes with young children, controlling for covariates. Similarly, urban households with children had 0.51 (95% CI: 0.24–1.09; p = .084) lower odds of having a home smoke-free policy, relative to homes with young children, controlling for covariates. This reflects as a change in effect size of 15.6% for adolescents and 45.3% for children, from urban youth relative to the full sample.
Suburban households with adolescents had 1.04 (95% CI: 0.58–1.85; p = .899) greater odds of having a home smoke-free policy, relative to homes with young children, controlling for covariates. Similarly, suburban households with children had 1.42 (95% CI: 0.80–2.54; p = .231) greater odds of having a home smoke-free policy, relative to homes with young children, controlling for covariates. Although not statistically significantly, these changes would reflect a change in effect size of 62.5% for adolescents and 77.5% for children, from suburban youth relative to the full sample.
Discussion
Home smoke-free policies declined approximately 9% each year a child aged from birth through adolescence among rural households and 6% among urban households in the United States (Table 3). Gaps in home smoke-free policies were concentrated among adolescents (12–17 years of age) in rural households, and were more evenly distributed across developmental stages (ie, childhood, adolescence) in urban households (Table 4). Our findings are consistent with prior literature17 and expand on previously identified geographic disparities in adolescent secondhand smoke exposure.7 Specifically, this study identified potential areas for targeted intervention (ie, parents of adolescents, particularly in rural and urban areas). To the best of our knowledge, this is the first study to quantify the declining coverage of home smoke-free policies with youth age and to report geographic disparities in home smoke-free policies across age groups.
Health promotion effects, particularly communication campaigns, have had success in changing perceptions of indoor smoking (ie, norms) and adoption of home smoke-free policies (ie, behaviors) in nonrural settings16 and across diverse populations.15 Our findings indicate a need to build on the success of prior health promotion efforts by developing interventions tailored to parents/guardians as a method of increasing home smoke-free policies, particularly in rural areas. Furthermore, our study supports the need for similar efforts in urban communities, although youth age may not be as strong of a predictor for this geographic area. As such, there is a clear need for investment in educational campaigns and other methods of reaching parents/guardians of adolescents. As rural areas are less likely to have positive attitudes toward home smoke-free policies,10–14,23,24 formative research and community participatory methodologies are needed to develop, test, refine, and disseminate public health interventions aimed at reducing geographic disparities in secondhand smoke exposure.
One such approach may be community-based participatory research (CBPR), which has shown promise for recruiting minoritized populations for health promotion programs.25–27 Furthermore, incorporating CBPR with place-based efforts to reduce disparities in secondhand smoke exposure may be a viable pathway in which public health researchers can better involve the community to address health needs.28 Health promotion efforts, particularly communication campaigns, are only part of the solution for reducing adolescent secondhand smoke exposure in the home.
Two evidence-based methodologies for increasing adoption and enforcement of home smoke-free policies include (1) smoking cessation for adult smokers16,29–31 and (2) parent-focused clinical intervention during pediatric medical visits. The U.S. Surgeon General recommends clinical- and health system-based strategies for adult smoking cessation.32 However, declines in rural health infrastructure (eg, hospital closures) represents the primary impediment to accessing evidence-based smoking cessation services.16,29–31 Similarly, youth who live in areas that are insufficiently serviced by public and for-profit healthcare providers have fewer engagements with medical providers,33 which reduces the number of opportunities for parent-focused clinical intervention towards home smoke-free policies and secondhand smoke exposure. As such, addressing disparities in adult smoking prevalence and youth secondhand smoke exposure extends beyond the scope of public health programs and into healthcare policy, appropriations, and systems-level change.32 For these reasons, addressing geographic disparities in home smoke-free policies and secondhand smoke should be part of a comprehensive approach to tobacco control with appropriate investment from federal, state, and local government32 to support the healthcare infrastructure that delivers essential components of evidence-based tobacco prevention.32
This study has several implications for health disparities research. Our study expands on prior work by observing disparities in the age gradient-decline in home smoke-free policy among both rural and urban youth in the U.S. Hispanic households were significantly less likely to have a home smoke-free policy in the full sample as well as in rural and suburban households. Living in a household at 400% or more above the poverty level (ie, ~$103,000+ for a family of four)34 was the strongest correlate of having a home smoke-free policy in the full sample and in each geographic area. Living in a household with an annual income of 200%–399% of the FPL ($51,500–$102,999 for a family of four)34 was similarly correlated with having a home smoke-free policy. Households with Hispanic children were significantly less likely to have home smoke-free policies for the full sample as well as the rural and suburban samples. Findings related to income and ethnicity demonstrate that the geographic disparities may not be specific to rurality but to a broader area deprivation that reflects the socio-economic conditions of a geographic region.35 Future research is needed to explore the nuances in area deprivation and other socio-economic determinants of health for urban, suburban, and rural youth.
This study has limitations. First, the self-reported data from NSCH are subject to response bias. Second, the cross-sectional nature of the NSCH data prohibits causal inferences. Future longitudinal study is needed to examine home smoke-free policies as youth age increases. Third, we were unable to examine frequency of secondhand smoke exposure (ie, number of days per week, times per day) as well as the source of the tobacco smoking in the house (ie, from responding parent, from other guardian, from other members of the household). Fourth, the existence of a home smoke-free policy was not assessed among those who did not report living with a tobacco smoker. Specifically, the logic skip pattern of the home smoke-free policy question does not assess for the existence and enforcement of home smoke-free policies for guests, relatives, and other visitors, all of whom may be permitted by households to smoke tobacco indoors. Future research should increase sensitivity and specificity of home smoke-free policies.
Study findings expand our understanding of geographic disparities in home smoke-free policies and secondhand smoke exposure among youth. There is a clear need for public health interventions to reduce these disparities by reaching parents/guardians about home smoke-free policies, with evidence herein suggesting these interventions should target parents/guardians of adolescents. However, addressing home smoking behaviors must account for the complexities and realities of nicotine dependence among adult tobacco smokers. Thus, it is not enough to address psychosocial determinants of behavior (eg, knowledge, social norms) but also address the underlying nicotine dependence through evidence-based cessation services (eg, nicotine replacement therapy, counseling).32
Supplementary Material
Contributor Information
Dale S Mantey, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Onyinye Omega-Njemnobi, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Ethan T Hunt, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Kevin Lanza, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Benjamin Cristol, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Steven H Kelder, UT Health Science Center at Houston, UT Health, School of Public Health in Austin, 1616 Guadalupe, Suite 6.300, Austin, TX 78701, USA.
Funding
University of Texas Health Science Center at Houston School of Public Health Cancer Education and Career Development Program – National Cancer Institute/NIH Grant – National Cancer Institute/NIH Grant T32/CA057712. Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Declaration of Interests
None declared.
References
- 1. US Department of Health and Human Services. The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease; 2014. [Google Scholar]
- 2. US Department of Health and Human Services. A report of the surgeon general: How tobacco smoke causes disease: What it means to you. 2010. https://www.cdc.gov/tobacco/data_statistics/sgr/2010/consumer_booklet/pdfs/consumer.pdf. Accessed August 3, 2022.
- 3. Okoli CT, Kodet J.. A systematic review of secondhand tobacco smoke exposure and smoking behaviors: Smoking status, susceptibility, initiation, dependence, and cessation. Addict Behav. 2015;47:22–32. [DOI] [PubMed] [Google Scholar]
- 4. Parks MJ, Kingsbury JH, Boyle RG, Evered S.. Household implementation of smoke-free rules in homes and cars: a focus on adolescent smoking behavior and secondhand smoke exposure. Am J Health Behav. 2019;33(1):70–78. [DOI] [PubMed] [Google Scholar]
- 5. Soler R, Orenstein D, Honeycutt A, et al. Community-based interventions to decrease obesity and tobacco exposure and reduce health care costs: outcome estimates from Communities Putting Prevention to Work for 2010–2020. 2016. [DOI] [PMC free article] [PubMed]
- 6. Patel M, Thai CL, Meng Y-Y, et al. Smoke-free car legislation and student exposure to smoking. Pediatrics. 2018;141(Supplement_1):S40–S50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Mantey DS, Omega-Njemnobi O, Barroso CS.. Secondhand smoke exposure at home and/or in a vehicle: differences between urban and non-urban adolescents in the United States, from 2015 to 2018. Nicotine Tob Res. 2021;23(8):1327–1333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. US Department of Health and Human Services. The health consequences of smoking—50 years of progress: a report of the surgeon general. In: Paper presented at: US Department of Health and Human Services 2014.
- 9. US Department of Health and Human Services. Preventing Tobacco Use Among Youth and Young Adults: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease; 2012. [Google Scholar]
- 10. Stillman FA, Tanenbaum E, Wewers ME, et al. Variations in support for secondhand smoke restrictions across diverse rural regions of the United States. Prev Med. 2018;116:157–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Vander Weg MW, Cunningham CL, Howren MB, Cai X.. Tobacco use and exposure in rural areas: findings from the Behavioral Risk Factor Surveillance System. Addict Behav. 2011;36(3):231–236. [DOI] [PubMed] [Google Scholar]
- 12. Mills AL, White MM, Pierce JP, Messer K.. Home smoking bans among US households with children and smokers: opportunities for intervention. Am J Prev Med. 2011;41(6):559–565. [DOI] [PubMed] [Google Scholar]
- 13. Hahn EJ, Rayens MK, York N.. Readiness for smoke-free policy and overall strength of tobacco control in rural tobacco-growing communities. Health Promo Pract. 2013;14(2):238–246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Buettner-Schmidt K, Miller DR, Maack B.. Disparities in rural tobacco use, smoke-free policies, and tobacco taxes. Western J Nurs Res. 2019;41(8):1184–1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Parks MJ, Kegler MC, Kingsbury JH, Borowsky IW.. Reducing socioeconomic disparities in comprehensive smoke-free rules among households with children: a pilot intervention implemented through a National Cancer Program. Int J Environ Res Public Health. 2020;17(18):6787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Williamson M, Elbaum M, Pearson M, et al. Advancing Tobacco Prevention and Control in Rural America. 2019. [Google Scholar]
- 17. Jackson DB, Testa A.. The intersection between adverse childhood experiences and environmental tobacco smoke in US households with children. Nicotine Tob Res. 2021;23(4):732–740. [DOI] [PubMed] [Google Scholar]
- 18. Kerr M, Stattin H, Burk WJ.. A reinterpretation of parental monitoring in longitudinal perspective. J Res Adolesc. 2010;20(1):39–64. [Google Scholar]
- 19. Saito J, Shibanuma A, Yasuoka J, et al. Education and indoor smoking among parents who smoke: the mediating role of perceived social norms of smoking. BMC Public Health. 2018;18(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Tynan MA, Holmes CB, Promoff G, et al. State and local comprehensive smoke-free laws for worksites, restaurants, and bars—United States, 2015. MMWR-Morb Mortal Wkly Rep. 2016;65(24):623–626. [DOI] [PubMed] [Google Scholar]
- 21. Child and Adolescent Health Measurement Initiative. 2019 National Survey of Children’s Health: Guide to topics and Questions. www.childhealthdata.org. Accessed April 20, 2022
- 22. Hughes RA, Heron J, Sterne JA, Tilling K.. Accounting for missing data in statistical analyses: multiple imputation is not always the answer. Int J Epidemiol. 2019;48(4):1294–1304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. McMillen R, Wilson K, Tanski S, Klein JD, Winickoff JP.. Adult attitudes and practices regarding smoking restrictions and child tobacco smoke exposure: 2000 to 2015. Pediatrics. 2018;141(Supplement 1):S21–S29. [DOI] [PubMed] [Google Scholar]
- 24. McMillen RC, Winickoff JP, Gottlieb MA, et al. Public support for smoke-free section 8 public housing. West J Nurs Res. 2019; 41(8):1170–1183. [DOI] [PubMed] [Google Scholar]
- 25. Minkler M, Wallerstein N.. Community-Based Participatory Research for Health: From Process to Outcomes. John Wiley & Sons; 2011. [Google Scholar]
- 26. Israel BA, Schulz AJ, Parker EA, Becker AB.. Review of community-based research: assessing partnership approaches to improve public health. Annu Rev Public Health. 1998;19(1):173–202. [DOI] [PubMed] [Google Scholar]
- 27. Horowitz CR, Robinson M, Seifer S.. Community-based participatory research from the margin to the mainstream: are researchers prepared? Circulation. 2009;119(19):2633–2642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Petteway RJ, Sheikhattari P, Wagner F.. Toward an intergenerational model for tobacco-focused CBPR: integrating youth perspectives via photovoice. Health Promot Pract. 2019;20(1):67–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Carlson LE, Lounsberry JJ, Maciejewski O, et al. Telehealth-delivered group smoking cessation for rural and urban participants: Feasibility and cessation rates. Addict Behav. 2012;37(1):108–114. [DOI] [PubMed] [Google Scholar]
- 30. Rodriguez EM, Twarozek AM, Erwin DO, et al. Perspectives on smoking cessation in Northern Appalachia. J Commun Health. 2016;41(2):211–219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Coughlin LN, Bonar EE, Bohnert KM, et al. Changes in urban and rural cigarette smoking and Cannabis use from 2007 to 2017 in adults in the United States. Drug Alcohol Depen. 2019;205:107699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Centers for Disease Control and Prevention. Best Practices for Comprehensive Tobacco Control Programs—2014. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2014:162–169. [Google Scholar]
- 33. Frakt AB. The rural hospital problem. JAMA. 2019;321(23):2271–2272. [DOI] [PubMed] [Google Scholar]
- 34. Fins A. National Snapshot: Poverty among Women & Families, 2020. National Women’s Law Center. 2021. [Google Scholar]
- 35. Fairfield KM, Black AW, Ziller EC, et al. Area deprivation index and rurality in relation to lung cancer prevalence and mortality in a rural state. JNCI Cancer Spectrum. 2020;4(4):pkaa011. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.