Abstract
Background
Secondhand smoke exposure has been cross-sectionally associated with worse sleep health outcomes in children and shorter sleep duration in adolescents.
Objectives
We assessed longitudinal and cross-sectional associations between secondhand smoke (SHS) exposure and shorter sleep duration in children from the Fragile Families and Child Wellbeing Study, a longitudinal birth cohort. We additionally examined whether associations would persist after controlling for potential confounders.
Participants
Mothers (N=4898) were recruited using a stratified random sample of large United States cities and oversampling for nonmarital births.
Measurements
Mothers were asked about whether they smoked during pregnancy, whether their child spent time with someone who is smoking, and their child’s weekday sleep duration. Sociodemographic factors, asthma diagnosis, and bedtime routines were assessed as potential confounders. Data collected at ages 3, 5, and 9 years were analyzed using multivariable regression models (N=1912; 51.6% boys).
Results
SHS exposure at age 3 predicted 15.0 fewer minutes at age 5 (p=0.001) and 12.3 fewer minutes at age 9 (p=0.003). SHS exposure at age 9 was cross-sectionally associated with 14.4 fewer minutes of sleep duration at age 9 (p=0.002). Findings persisted after controlling for potential confounders.
Conclusion
These results provide associational support for the hypothesis that SHS exposure may have long-term consequences for childhood sleep duration. Future studies should investigate the relationship between SHS exposure and shorter sleep duration using objective measurements of serum cotinine and sleep actigraphy and by exploring potential mechanisms.
Keywords: secondhand smoke exposure, sleep, pediatrics, cohort study
Introduction
Sleep is an important facilitator of growth and health throughout youth. The American Academy of Sleep Medicine and National Sleep Foundation recommend that children obtain 10-13 hours of sleep as preschoolers (ages 3-5 years) and 9-12 hours as school-aged children (ages 6-13 years)1,2. Unfortunately, more than half of school-aged children in the United States are estimated to sleep fewer than 9 hours per night3. Insufficient sleep has been associated with a range of negative outcomes in children, including poor academic performance4, difficulty in mood regulation4, mental disorders5, attention disorders4, and obesity6.
Exposure to secondhand smoke (SHS) has also been associated with adverse health and well-being outcomes in children7,8, including wheezing9–11, mental health disorders, behavior disorders, and cognitive impairments12–14. Notably, SHS exposure has also been cross-sectionally associated with worse sleep health at different ages throughout youth15–22. One study of children from an underserved urban community found that SHS exposure was associated with longer sleep onset latency, sleep-disordered breathing, parasomnias, and daytime sleepiness in children ages 6-12 years with asthma, yet it did not find an association between SHS exposure and sleep duration18. Another study of urban children found that either smoking during pregnancy or smoking when the child was 3 years old was associated with greater risk of poor sleep, but exposure to both was not associated with increased risk9. These findings did not support the hypothesis of an additive effect of exposure to smoking being associated with greater risk of poor sleep. More recently, a study that looked at smoking both during pregnancy and postnatally found no effects of trouble sleeping when the child was 5 years old23.
While research on SHS exposure and sleep duration among school-aged children is limited18,21, prior research on adolescents has found a dose-response relationship. Greater frequency of SHS exposure was associated with more frequent restless sleep and decreased weekday and weekend sleep duration in adolescents 13-18 years old17. Another study found that SHS exposure at home was associated with greater odds of shorter sleep duration in adolescents 14-17 years old21. However, with several exceptions9,23,24, these studies on adolescents as well as most of the current literature predominantly relies on cross-sectional studies to examine SHS exposure and sleep. There have been few studies to date that investigate the longitudinal association of SHS exposure during childhood and sleep duration at a later age. Fewer still have examined both SHS exposure and sleep duration in underserved and urban communities9,18.
The relationship between SHS exposure and general sleep health may be partially explained by confounding mechanisms related to SHS exposure or sleep, such as asthma induced by SHS exposure or inconsistency around bedtimes or bedtime routines, even after adjustment for sociodemographic factors. For example, SHS exposure has been associated with an increased risk of asthma in children11,14, and children with asthma tend to have poorer sleep quality and greater daytime sleepiness than children without asthma25,26, even when controlling for SHS exposure27. Moreover, many children from families of low socioeconomic status have daily exposure to SHS28 but are also less likely to have regular bedtime routines29. Delayed or inconsistent bedtimes and bedtime routines may instead account for children’s shorter sleep duration.
This study aims to extend prior cross-sectional findings of SHS exposure and sleep duration in school-aged children18 and to also investigate longitudinal associations between SHS exposure and sleep duration at different childhood ages. To our knowledge, this is one of the first studies to evaluate such longitudinal associations during childhood. We hypothesized that SHS exposure would be longitudinally and cross-sectionally associated with shorter sleep duration in children at ages 5 and 9 years. We additionally examined whether these longitudinal and cross-sectional associations would persist after controlling for potential confounders, including sociodemographic factors, the child’s diagnosis of asthma, and whether the child had a regular bedtime and regular bedtime routine. Based on prior literature, we expected children whose mothers were not white29,30, whose family’s household income was below the poverty threshold30, and who had an asthma diagnosis25,26 to have shorter sleep duration. We also expected the presence of regular bedtimes and bedtime routines29 to be associated with longer sleep duration.
Methods
Participants
We analyzed observational data from the Fragile Families and Child Wellbeing Study (FFCWS), a longitudinal birth cohort from 20 large United States cities. The FFCWS is a sample of urban households, with an oversampling of nonmarital births31. Mothers (N=4898) were randomly sampled within each selected hospital and provided written consent upon recruitment31; mothers were excluded if she or the father were younger than 18 years old, she planned to place the child up for adoption, or the father or child were no longer alive at the time of the interview. Mothers were interviewed at the hospital at the time of their child’s birth and were subsequently interviewed when the child was 1, 3, 5, 9 and 15 years old. All procedures were approved by the Institutional Review Boards of the participating institutions.
In order to examine the same sample of children across ages, we only included in this study children who did not have missing data for any variables of interest at ages 3, 5, and 9 years (N=1912). The consort diagram in Figure 1 depicts how we arrived at the final sample size. The original sample size was reduced by 38% due to attrition by ages 5 and 9 years; the remaining sample was further reduced by 37% due to missing predictor and outcome responses.
Figure 1.
Consort diagram depicting missingness through each step of data cleaning. The final sample included 1912 children after excluding missing sociodemographic information; cohort attrition (children not surveyed at ages 5 and 9 years); missing sleep duration at either age 5 or 9 years; missing secondhand smoke (SHS) exposure in utero or at ages 3, 5, or 9 years; or missing potential confounders (i.e., asthma diagnosis, regular bedtimes, regular bedtime routines).
Measures
Secondhand smoke exposure
The child’s exposure to SHS was assessed during caregiver interviews. Mothers were asked during the baseline survey at the child’s birth, “During the pregnancy, how many cigarettes did you smoke?” When the child was 3, 5, and 9 years old, mothers were asked “On average, how many hours a day does your child spend in the same room with someone who is smoking? Please include the time they spend with a babysitter or family member, or anyone else, who is smoking.” Fewer than 20% of the sample reported ≥1 hour of SHS exposure at each age (see Table 1). Based on the positively skewed (≥6) and leptokurtic (≥40) distribution of the continuous count data32, we defined SHS exposure as “1” if there was any exposure, regardless of duration, and “0” if mothers did not smoke any cigarettes during pregnancy or the child did not spend any time in the same room with someone who is smoking.
Table 1.
Participant demographics
| N | Total % | |
|---|---|---|
| N=1912 | ||
| SHS exposure | ||
| In utero | 353 | 18.5% |
| Age 3 | 371 | 19.4% |
| Age 5 | 303 | 15.9% |
| Age 9 | 227 | 11.9% |
| Sleep duration | ||
| Age 5 wave (SD) | 1912 | 9.4 (1.2) hours |
| Age 9 wave (SD) | 1912 | 9.0 (1.1) hours |
| Asthma | ||
| Diagnosis by age 5 | 409 | 21.4% |
| Diagnosis by age 9 | 472 | 24.7% |
| Bedtime at age 5 | ||
| Regular bedtime | 1769 | 92.5% |
| Regular bedtime routine | 1505 | 78.7% |
| Bedtime at age 9 | ||
| Regular bedtime | 1815 | 94.9% |
| Regular bedtime routine | ||
| Definitely untrue | 76 | 4.0% |
| Somewhat untrue | 47 | 2.5% |
| Not really true | 19 | 1.0% |
| Somewhat true | 270 | 14.1% |
| Definitely true | 1500 | 78.5% |
| Birth sex | ||
| Male | 986 | 51.6% |
| Female | 926 | 48.4% |
| Age | ||
| Age 5 wave (SD) | 1912 | 5.1 (0.2) years |
| Age 9 wave (SD) | 1912 | 9.2 (0.3) years |
| Mother’s race | ||
| White | 437 | 22.9% |
| Black/African American | 982 | 51.4% |
| Hispanic/Latino | 425 | 2.2% |
| Multiracial/Other | 68 | 3.6% |
| Household income-to-poverty ratio at age 5 | ||
| ≤49% | 380 | 19.9% |
| 50-99% | 400 | 20.9% |
| 100-199% | 492 | 25.7% |
| 200-299% | 270 | 14.1% |
| >300% | 370 | 19.4% |
| Household income-to-poverty ratio at age 9 | ||
| ≤49% | 317 | 16.6% |
| 50-99% | 366 | 19.1% |
| 100-199% | 571 | 29.9% |
| 200-299% | 262 | 13.7% |
| >300% | 396 | 20.7% |
| Mother’s education | ||
| Less than high school | 572 | 29.9 |
| High school or equivalent | 597 | 31.2 |
| Some college or technical school | 517 | 27.0 |
| College or graduate school | 226 | 11.8 |
| Mother’s relationship with father | ||
| Married | 476 | 24.9% |
| Cohabitating | 681 | 35.6% |
| Visiting | 529 | 27.7% |
| Friends | 100 | 5.2% |
| Minimal/no relationship | 126 | 6.6% |
Note. SHS exposure = secondhand smoke exposure; approximately 40.8% at age 5 and 35.7% at age 9 live in households with income below the poverty threshold
Sleep duration
Sleep duration was assessed at both ages 5 and 9 years. Parents were asked “How many hours of sleep a night does your child usually get during the week?” Parents responded with integer hours, which were treated as a continuous variable. Responses that were fewer than 6 hours and greater than 15 hours were treated as missing (N=18)33,34. In this sample, the range of sleep duration was 6-15 hours at age 5 and 6-14 hours at age 9. Parent-reported sleep duration was not assessed at other waves.
Sociodemographic factors
Sociodemographic variables were collected from mothers at the time of the child’s birth, including the child’s birth sex (male or female), the mother’s self-identified race (non-Hispanic White, non-Hispanic Black/African American, Hispanic/Latino, or Multiracial/Other; missing data were coded as Multiracial/Other), the mother’s education level at child’s birth (less than high school, high school or equivalent, some college or technical school, or college or graduate school), and the mother’s relationship with the father at child’s birth (married, cohabitating, visiting, friends, minimal/no relationship). Household income-to-poverty ratio (≤49%, 50-99%, 100-199%, 200-299%, or >300%) at ages 5 and 9 years was also included. Household income-to-poverty ratios were calculated using US Census Bureau definitions of poverty thresholds (https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html), with household income-to-poverty ratios ≤49% and 50-99% considered living below the poverty threshold. The child’s age in decimal years at each wave of interest was mean-centered and included in analyses as a continuous covariate. Mean-centered sleep duration at age 5 was also included as a continuous covariate in the age 9 analysis so that interpretations of sleep duration at age 9 account for average sleep duration at age 5. All other covariates were analyzed as categorical variables.
Asthma diagnosis
At ages 5 and 9, parents were asked about whether a doctor or health professional had ever diagnosed their child with asthma. Children who were not diagnosed at age 5 may have been diagnosed with asthma by age 9.
Bedtime routines
Presence of regular bedtimes and routines around bedtime were assessed using two questions at each wave of interest. At age 5, parents were asked whether their child had a regular bedtime during the week and whether there was a regular routine when their child was put to bed. At age 9, parents were asked whether their child had a regular bedtime during the week. Parents were also asked whether there was a regular bedtime routine, which they rated on a five-point Likert scale from “definitely untrue” to “definitely true” and was treated as a categorical variable. All other responses to regular bedtimes and bedtime routines were dichotomized.
Statistical analysis
Nested multivariable regression models were used to evaluate longitudinal and cross-sectional associations between SHS exposure at ages 3, 5, and 9 and sleep duration at ages 5 and 9. SHS exposure was the predictor and sleep duration in hours was the outcome variable for all models. Model 1 was unadjusted, and Model 2 adjusted for covariates excluding the child’s diagnosis of asthma, regular bedtime, and regular bedtime routines. Model 3 further included the child’s diagnosis of asthma, but not the child’s regular bedtime and bedtime routines. Model 4 adjusted for all covariates, asthma diagnosis, as well as the child’s regular bedtime and bedtime routine. General Linear Models (SAS 9.4) were used to calculate b estimates and an alpha level of p<.05 (two-tailed) was used to determine level of statistical significance for all analyses. In addition to the primary analyses, we also conducted sensitivity analyses to assess how the final analyzed sample differed from the original FFCWS sample.
Results
Descriptive statistics
Participant demographics, frequency of SHS exposure, and mean sleep duration are shown in Table 1. Approximately 51% of mothers of children in the sample were Black/African American, 23% were White, 22% were Hispanic/Latino, and 4% were Multiracial/Other. Approximately half of the children in the sample were male. Records with missing data were disproportionately children who were Hispanic/Latino or Multiracial/Other and children whose mother’s education level was less than high school.
Approximately 18% of children were reported to have SHS exposure in utero, 19% at age 3, 16% at age 5, and 12% at age 9. Children at age 5 were reported to sleep an average of 9.41 hours (SD=1.22) and 8.96 hours (SD=1.10) at age 9. Sleep duration at both ages was normally distributed. The children’s sleep duration at age 5 was positively associated with sleep duration at age 9 (r=0.26, p<0.001).
Associations of secondhand smoke exposure with sleep duration at age 5
Results of analyses examining longitudinal and cross-sectional associations between SHS exposure in utero, at age 3, and at age 5 and sleep duration at age 5 are shown in Table 2. SHS exposure at age 3 was longitudinally associated with shorter weekday sleep duration at age 5 (b= −0.32, CI=[−0.47, −0.16], p<0.001) in unadjusted analyses (Model 1), although SHS exposure in utero was not (p=0.170). SHS exposure at age 5 was not cross-sectionally associated with weekday sleep duration (p=0.618). This initial, unadjusted model explained approximately 1.0% of the variance in weekday sleep duration (F(3,1908)=6.70, p<0.001).
Table 2.
Associations with Age 5 sleep duration (hrs) linear regression
| Model 1 Unadjusted |
Model 2 Covariates |
Model 3 Asthma Diagnosis |
Model 4 Bedtime and Bedtime Routines |
|
|---|---|---|---|---|
| b | b | b | b | |
| Constant | 9.46*** | 9.86*** | 9.85*** | 9.12*** |
| SHS exposure (ref: no exposure) | ||||
| In utero | 0.10 | 0.07 | 0.07 | 0.07 |
| Age 3 | −0.32*** | −0.25** | −0.25** | −0.24** |
| Age 5 | −0.04 | 0.04 | 0.04 | 0.05 |
| Asthma (ref: no diagnosis) | - | - | 0.05 | 0.03 |
| Regular bedtime (ref: no bedtime) | - | - | - | 0.69*** |
| Regular bedtime routine (ref: no regular routine) | - | - | - | 0.07 |
| Birth sex (ref: Male) | - | −0.05 | −0.05 | −0.06 |
| Age (mean-centered) | - | 0.17 | 0.17 | 0.13 |
| Race (ref: White) | ||||
| Black/African American | - | −0.55*** | −0.56*** | −0.54*** |
| Hispanic/Latino | - | −0.26** | −0.26** | −0.23* |
| Multiracial/Other | - | −0.14 | −0.14 | −0.16 |
| Household income-to-poverty ratio (ref: ≤49%) at age 5 | ||||
| 50-99% | - | 0.02 | 0.02 | 0.01 |
| 100-199% | - | 0.02 | 0.02 | 0.01 |
| 200-299% | - | 0.03 | 0.03 | 0.03 |
| >300% | - | 0.15 | 0.15 | 0.16 |
| Mother’s education (ref: Less than high school) | ||||
| High school or equivalent | - | −0.10 | −0.10 | −0.10 |
| Some college or technical school | - | −0.10 | −0.10 | −0.08 |
| College or graduate school | - | 0.20† | 0.20† | 0.18 |
| Mother’s relationship with father (ref: Married) | ||||
| Cohabitating | - | −0.12 | −0.12 | −0.09 |
| Visiting | - | 0.02 | −0.02 | 0.02 |
| Friends | - | 0.05 | 0.05 | 0.05 |
| Minimal/no relationship | - | −0.10 | −0.10 | −0.08 |
| R 2 | 0.01 | 0.07 | 0.07 | 0.10 |
| F | 6.70*** | 8.04*** | 7.67*** | 9.42*** |
| N | 1912 | |||
Note.
p<.001;
p<.0.1;
p<.05;
p<.10
After adjustment for covariates (Model 2), SHS exposure at age 3 continued to be longitudinally associated with shorter weekday sleep duration at age 5 (b= −0.25, CI=[−0.40, −0.10], p=0.001), suggesting that any exposure to SHS at age 3 predicted approximately 15.0 fewer minutes of sleep duration at age 5. Neither SHS exposure in utero nor at age 5 was associated with weekday sleep duration at age 5. Associations of covariates with sleep duration at age 5 are listed in Table 2. The adjusted model explained approximately 7.5% (ΔR2=0.06) of the variance in weekday sleep duration.
We next tested whether the association was attenuated if the child was diagnosed with asthma at age 5 (Model 3) as well as whether the child had a regular bedtime or regular bedtime routine (Model 4). We found that the diagnosis of asthma was not associated with sleep duration and did not attenuate the association between SHS exposure and sleep duration. Whether the child had a regular bedtime was cross-sectionally associated with sleep duration at age 5 (b=0.69, CI=[0.49, 0.90], p<0.001) but whether the child had a regular bedtime routine was not. However, even in Model 4, SHS exposure at age 3 was still longitudinally associated with shorter weekday sleep duration at age 5 (b= −0.24, CI=[−0.39, −0.09], p=0.002). The final model including asthma diagnosis, regular bedtime, and regular bedtime routines explained approximately 9.9% (ΔR2=0.02) of the variance in weekday sleep duration.
Associations of secondhand smoke exposure with sleep duration at age 9
We then examined the longitudinal and cross-sectional associations between SHS exposure in utero and at ages 3, 5, and 9 and sleep duration at age 9, which can be found in Table 3. In the first, unadjusted model, we found that SHS exposure at ages 3 (b= −0.32, CI=[−0.46, −0.17], p<0.001) and 5 (b= −0.18, CI=[−0.34, −0.03], p=0.019) were longitudinally associated with shorter weekday sleep duration at age 9, but SHS exposure in utero was marginally associated with longer weekday sleep duration (p=0.051). SHS exposure at age 9 was cross-sectionally associated with shorter weekday sleep duration at age 9 (b= −0.30, CI=[−0.46, −0.14], p<0.001), explaining approximately 3.5% of the variance in weekday sleep duration (F(4,1907)=17.31, p<0.001) in unadjusted analyses (Model 1).
Table 3.
Associations with Age 9 sleep duration (hrs) linear regression
| Model 1 Unadjusted |
Model 2 Covariates |
Model 3 Asthma Diagnosis |
Model 4 Bedtime and Bedtime Routines |
|
|---|---|---|---|---|
| b | b | b | b | |
| Constant | 9.06*** | 9.37*** | 9.40*** | 8.82*** |
| SHS exposure (ref: no exposure) | ||||
| In utero | 0.13† | 0.09 | 0.09 | 0.07 |
| Age 3 | −0.32*** | −0.20** | −0.21** | −0.20** |
| Age 5 | −0.18* | −0.14† | −0.14† | −0.14† |
| Age 9 | −0.30*** | −0.24** | −0.24** | −0.20* |
| Asthma (ref: no diagnosis) | - | - | −0.09 | −0.11* |
| Regular bedtime (ref: no bedtime) | - | - | - | 0.44*** |
| Regular bedtime routine (ref: Definitely untrue) | ||||
| Somewhat untrue | - | - | - | −0.28 |
| Not really true | - | - | - | 0.09 |
| Somewhat true | - | - | - | −0.04 |
| Definitely true | - | - | - | 0.25† |
| Birth sex (ref: Male) | - | 0.03 | 0.02 | 0.01 |
| Age (mean-centered) | - | −0.06 | −0.06 | −0.06 |
| Age 5 sleep (mean-centered) | - | 0.18*** | 0.18*** | 0.17*** |
| Race (ref: White) | ||||
| Black/African American | - | −0.43*** | −0.42*** | −0.43*** |
| Hispanic/Latino | - | −0.22** | −0.21** | −0.19* |
| Multiracial/Other | - | −0.57*** | −0.58*** | −0.54*** |
| Household income-to-poverty ratio (ref: ≤49%) at age 9 | ||||
| 50-99% | - | −0.05 | −0.06 | −0.07 |
| 100-199% | - | −0.05 | −0.06 | −0.06 |
| 200-299% | - | −0.08 | −0.08 | −0.08 |
| >300% | - | −0.05 | −0.05 | −0.08 |
| Mother’s education (ref: Less than high school) | ||||
| High school or equivalent | - | 0.02 | 0.02 | 0.02 |
| Some college or technical school | - | 0.18** | 0.18** | 0.19** |
| College or graduate school | - | 0.17† | 0.17 | 0.17† |
| Mother’s relationship with father (ref: Married) | ||||
| Cohabitating | - | −0.11 | −0.10 | −0.12† |
| Visiting | - | −0.12 | −0.11 | −0.13 |
| Friends | - | −0.28* | −0.27* | −0.27* |
| Minimal/no relationship | - | −0.09 | −0.08 | −0.10 |
| R 2 | 0.04 | 0.14 | 0.14 | 0.17 |
| F | 17.31*** | 14.51*** | 13.99*** | 14.16*** |
| N | 1912 | |||
Note.
p<.001;
p<.01;
p<.05;
p<.10
This pattern of results remained similar after adjustment for covariates in Model 2. SHS exposure at age 3 predicted 12.3 fewer minutes of weekday sleep duration at age 9 (b= −0.20, CI=[−0.34, −0.07], p=0.003) whereas SHS exposure at age 5 was marginally associated with 8.4 fewer minutes of weekday sleep duration (b= −0.14, CI=[−0.29, −0.01], p=0.060). SHS exposure in utero was not longitudinally associated with sleep duration at age 9 after adjustment for covariates. SHS exposure at age 9 was cross-sectionally associated with 14.4 fewer minutes of sleep duration at 9 (b= −0.24, CI=[−0.40, −0.08], p=0.002). Figure 2 depicts the associations with sleep duration within the analysis framework, and associations of covariates with sleep duration at age 9 are listed in Table 3. The adjusted model explained approximately 13.9% (ΔR2=0.10) of the variance in weekday sleep duration.
Figure 2.
Analysis framework of the study design and significant associations. The bottom row is the predictor variable (secondhand smoke (SHS) exposure) and the top row is the outcome variable (sleep duration in hours). Solid black lines depict significant associations between the predictor and outcome variables, and dashed grey lines depict non-significant associations. Note. **p<.01; *p<.05; †<.10
Finally, as we did for age 5 analyses, we tested for effects of whether the child was diagnosed with asthma at age 9 (Model 3) and whether the child had a regular bedtime or regular bedtime routine (Model 4). We found that asthma diagnosis was not associated with sleep duration at age 9 in Model 3 but was associated with 6.5 fewer minutes of sleep duration (b= −0.11, CI=[−0.22, −0.00], p=0.049) when controlling for regular bedtimes and bedtime routines in Model 4. We also found in Model 4 that having a regular bedtime was cross-sectionally associated with 26.4 more minutes of sleep duration at age 9 (b=0.44, CI=[0.20, 0.68], p<0.001). SHS exposure at age 3 remained longitudinally associated with shorter weekday sleep duration at age 9 (b= −0.20, CI=[−0.33, −0.07], p=0.004). SHS exposure at age 9 was also still cross-sectionally associated with weekday sleep duration (b= −0.20, CI=[−0.36, −0.05], p=0.010), although the inclusion of regular bedtime routines attenuated this association (from 14.4 fewer minutes of sleep duration to 12.2 fewer minutes), suggesting partial mediation by regular bedtimes. The final model including regular bedtimes and bedtime routines explained approximately 16.9% (ΔR2=0.03) of the variance in weekday sleep duration.
Discussion
The goal of this study was to identify cross-sectional and longitudinal associations between SHS exposure and sleep duration in preschool-and school-aged children. In general, we found evidence that exposure to SHS during childhood was longitudinally associated with shorter weekday sleep duration during both developmental periods, irrespective of whether the child was diagnosed with asthma or had a regular bedtime or bedtime routine. Interestingly, SHS exposure was also cross-sectionally associated with shorter weekday sleep duration at age 9 but was not at age 5. Finally, whether mothers smoked during pregnancy (i.e., SHS exposure in utero) was not longitudinally associated with sleep duration at either age.
Our results were consistent with previous findings of preschoolers and adolescents that identified cross-sectional associations between SHS exposure and poor sleep9,17,21. Furthermore, we extended these findings by identifying that SHS exposure at age 3 was longitudinally associated with shorter sleep duration at ages 5 and 9, and that SHS exposure at age 5 was marginally associated with shorter sleep duration at age 9. Our findings persisted after controlling for the effects of SHS exposure at each wave, suggesting possible long-term consequences of SHS exposure on sleep duration.
Contrary to prior evidence in preschoolers that exposure to SHS in utero was longitudinally associated with shorter sleep duration within one year of age24 and poor sleep when the child was age 39, we found no association between SHS exposure in utero and sleep duration at ages 5 or 9. SHS exposure in utero was marginally associated with longer sleep duration at age 9 in Model 1 (see Table 3); however, this association became nonsignificant upon inclusion of sociodemographic covariates. It may be that SHS exposure in utero is longitudinally associated with sleep duration up to around age 3 but not with ages beyond then. Alternatively, SHS exposure in utero may be associated with poor sleep9 and other sleep disturbances but is unrelated to sleep duration beyond the first year. Unfortunately, sleep duration was not measured at age 3 in the FFCWS so we cannot draw any conclusion regarding the longitudinal effect of SHS exposure in utero on sleep at that age.
This study has multiple strengths that inform the literature, namely the longitudinal design and the testing of potential confounding factors, such as how sociodemographic factors, child’s asthma diagnosis, and regular bedtime and bedtime routines may reduce the relationship between SHS exposure on sleep duration at both ages. First, we adjusted for race and related sociodemographic factors. Race has been previously shown to be associated with both children’s bedtimes29 and sleep30, and we indeed found that the mother’s self-reported race was associated with sleep duration. Including race reduced the effect of SHS exposure between 20% and 40% at ages 3, 5, and 9, yet SHS exposure continued to significantly predict shorter sleep duration. Additionally, mother’s education level and socioeconomic status have been previously shown to predict sleep duration30, but we found few associations when accounting for race and other sociodemographic factors. Other than race, sociodemographic factors did not account for the variance in sleep duration explained by SHS exposure.
Second, we adjusted for the child’s diagnosis of asthma because exposure to particulates in SHS may negatively affect health7 and induce asthma in children, which could further affect sleep. Certainly, there is evidence that SHS exposure is associated with a greater risk of asthma in children14, and children with asthma tend to have worse sleep25,26. We found that children in this sample who had an asthma diagnosis tended to have shorter sleep at age 9 (but not at age 5). However, when we included asthma diagnosis as a predictor in the multivariable regression analyses, asthma diagnosis was not associated with sleep duration at either age (Model 3), except at age 9 when accounting for regular bedtimes and bedtime routines (Model 4). The trends between SHS exposure and sleep duration found in Model 2 were otherwise consistent in Model 3. Ultimately, at least in this study, our findings suggested that parent-reported asthma diagnosis did not explain the relationship between SHS exposure and sleep duration.
Finally, we considered that SHS exposure may instead be a proxy for some other factor that could affect sleep during childhood, such as inconsistency of bedtime routines. To test this possibility, we examined the use of regular bedtimes and bedtime routines to index whether there was consistency or structure protecting or undermining the child’s sleep. Having a regular bedtime during the week was strongly associated with longer sleep duration at both ages 5 and 9, although whether there was an attached bedtime routine had no bearing on sleep duration. The one exception to this pattern was that children whose bedtime routine was rated as “definitely true” slept longer at age 9 than children whose bedtime routine was rated as “definitely untrue”, suggesting that only the most consistent bedtime routines were associated with sleep duration. Regardless, the inclusion of regular bedtimes and bedtime routines did not overtly attenuate the relationship between SHS exposure and sleep duration at either age.
Still, another untested or unmeasured factor may mediate the pathway between SHS exposure during childhood and sleep duration. In adult daily smokers, early awakenings seem to mediate the relationship between nicotine addiction severity and sleep health35. Parents and caregivers, who wake up early to engage in smoking behaviors, may inadvertently wake their child and thereby shorten the child’s sleep. Or, because SHS exposure has been associated with longer sleep onset latency18, exposure to SHS may stimulate the child via nicotine exposure or some other relevant substance, thereby complicating initiation of sleep at night. Future studies should attempt to identify and verify variables that may mechanistically explain these associations.
There were several limitations of this study. First, parents reported the child’s SHS exposure, which was subjective and may not accurately reflect the length or concentration of exposure. We also dichotomized SHS exposure based on the distribution of these continuous count data32. Future studies should examine serum cotinine, a metabolite of nicotine, to physiologically assess the concentration of SHS exposure in childhood18, which will preserve the variation of SHS exposure among children32. Second, parents reported sleep duration in integers (e.g., 9 or 10 hours) and were not specifically asked about daytime sleeping. Responses were thus a coarse estimate of the number of hours of sleep, were subject to recall bias and misreporting, and may therefore deviate from objective measures of sleep. Instead, future studies could use open-ended survey items that report sleep duration in both hours and minutes. Alternatively, studies should use actigraphy to objectively estimate sleep duration, which is used extensively in pediatric sleep research36–38. Third, parents may have lacked the knowledge of what constitutes a regular bedtime or regular bedtime routine. Going forward, providing examples of regular bedtimes and regular bedtime routines may help parents more accurately report. Fourth, sleep duration at age 3 was not collected in the FFCWS, which limited our ability to analyze the cross-sectional association between SHS exposure and sleep duration at age 3 as well as the longitudinal effects of SHS exposure in utero on sleep during childhood. Future studies should build upon these findings by collecting data from time points throughout the developmental lifespan.
Finally, the results of this study predominantly represented children of urban households and nonmarital births, thereby limiting generalizability to children living in rural or suburban areas. Moreover, due to cohort attrition and variable selection, children from marginalized groups were more likely to have records with missing data, including children who were Hispanic/Latino or Multiracial/Other and children whose mothers did not complete high school. The results therefore may not be representative of children who are not White or Black/African American and children who live in more disadvantaged households.
Conclusion
Our results affirm an association between SHS exposure and sleep. While we do not know if there is a causal link, SHS exposure is a modifiable caregiver behavior that may have longer-term consequences for childhood sleep duration. Eliminating a child’s exposure to SHS may protect up to approximately half an hour of sleep per night or around 4 hours each week. Moreover, although tobacco use in the United States has declined in recent decades39, 21% of adults still use combustible tobacco products40 and may therefore contribute to decrements in childhood sleep duration. More research should further investigate acute and long-term effects of SHS exposure on sleep outcomes, including collecting objective measures of sleep and evaluating physiological correlates such as serum cotinine18,28 to accurately identify levels of SHS exposure. Other possible factors should be considered along with SHS exposure and may account for, or mechanistically explain, the variance in sleep duration throughout childhood.
Acknowledgements
We thank the families who participated in the study and the FFCWS team. We also thank Orfeu Buxton, PhD, and Michael Russell, PhD, Pennsylvania State University, who assisted with interpretation of study findings detailed in this manuscript.
Statement of Financial Support:
Research reported in this manuscript was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health under award numbers R01HD073352 (to LH), R01HD36916, R01HD39135, and R01HD40421, as well as a consortium of private foundations.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of Interest: Dr. Lauren Hale previously received an honorarium from the National Sleep Foundation for her role as Editor-in-Chief of the journal Sleep Health. The remaining authors have no financial relationships relevant to this article to disclose. The authors have no conflicts of interest to disclose.
References
- 1.Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health J Natl Sleep Found. 2015; 1(4):233–243. doi: 10.1016/j.sleh.2015.10.004 [DOI] [PubMed] [Google Scholar]
- 2.Paruthi S, Brooks LJ, D’Ambrosio C, et al. Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement of the American Academy of Sleep Medicine. J Clin Sleep Med JCSM Off Publ Am Acad Sleep Med. 2016;12(6):785–786. doi: 10.5664/jcsm.5866 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wheaton AG, Jones SE, Cooper AC, Croft JB. Short Sleep Duration Among Middle School and High School Students — United States, 2015. Morb Mortal Wkly Rep. 2018;67(3):85– 90. doi: 10.15585/mmwr.mm6703a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sadeh A Consequences of Sleep Loss or Sleep Disruption in Children. Sleep Med Clin. 2007;2(3):513–520. doi: 10.1016/j.jsmc.2007.05.012 [DOI] [Google Scholar]
- 5.Aronen ET, Paavonen EJ, Fjällberg M, Soininen M, Törrönen J. Sleep and Psychiatric Symptoms in School-Age Children. J Am Acad Child Adolesc Psychiatiy. 2000;39(4):502– 508. doi: 10.1097/00004583-200004000-00020 [DOI] [PubMed] [Google Scholar]
- 6.Hart CN, Cairns A, Jelalian E. Sleep and Obesity in Children and Adolescents. Pediatr Clin. 2011;58(3):715–733. doi: 10.1016/j.pcl.2011.03.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.National Center for Chronic Disease Prevention and Health Promotion (US) Office on Smoking and Health. Preventing Tobacco Use Among Youth and Young Adults: A Report of the Surgeon General. Centers for Disease Control and Prevention (US); 2012. http://www.ncbi.nlm.nih.gov/books/NBK99237/ [PubMed] [Google Scholar]
- 8.Hwang S-H, Hwang JH, Moon JS, Lee D-H. Environmental tobacco smoke and children’s health. Korean J Pediatr. 2012;55(2):35–41. doi: 10.3345/kjp.2012.55.2.35 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Johansson A, Ludvigsson J, Hermansson G. Adverse health effects related to tobacco smoke exposure in a cohort of three-year olds. Acta Pædiatrica. 2008;97(3):354–357. doi: 10.1111/j.1651-2227.2007.00619.x [DOI] [PubMed] [Google Scholar]
- 10.Silvestri M, Franchi S, Pistorio A, Petecchia L, Rusconi F. Smoke exposure, wheezing, and asthma development: A systematic review and meta-analysis in unselected birth cohorts. Pediatr Pulmonol. 2015;50(4):353–362. doi: 10.1002/ppul.23037 [DOI] [PubMed] [Google Scholar]
- 11.Burke H, Leonardi-Bee J, Hashim A, et al. Prenatal and Passive Smoke Exposure and Incidence of Asthma and Wheeze: Systematic Review and Meta-analysis. Pediatrics. 2012; 129(4):735–744. doi: 10.1542/peds.2011-2196 [DOI] [PubMed] [Google Scholar]
- 12.Bandiera FC, Richardson AK, Lee DJ, He J-P, Merikangas KR. Secondhand Smoke Exposure and Mental Health Among Children and Adolescents. Arch Pediatr Adolesc Med. 2011;165(4):332–338. doi: 10.1001/archpediatrics.2011.30 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Herrmann M, King K, Weitzman M. Prenatal tobacco smoke and postnatal secondhand smoke exposure and child neurodevelopment: Curr Opin Pediatr. 2008;20(2):184–190. doi: 10.1097/MOP.ObO13e3282f56165 [DOI] [PubMed] [Google Scholar]
- 14.Polanska K, Hanke W, Ronchetti R, et al. Environmental tobacco smoke exposure and children’s health. Acta Paediatr. 2006;95(s453):86–92. doi: 10.1080/08035320600886562 [DOI] [PubMed] [Google Scholar]
- 15.Montgomery-Downs HE, Gozal D. Snore-Associated Sleep Fragmentation in Infancy: Mental Development Effects and Contribution of Secondhand Cigarette Smoke Exposure. Pediatrics. 2006;117(3):e496–e502. doi: 10.1542/peds.2005-1785 [DOI] [PubMed] [Google Scholar]
- 16.Ng DK, Chow P-Y, Chan C-H, Kwok K-L, Cheung JM, Kong FY. An update on childhood snoring. Acta Paediatr. 2006;95(9):1029–1035. doi: 10.1080/08035250500499432 [DOI] [PubMed] [Google Scholar]
- 17.Schwartz J, Bottorff JL, Richardson CG. Secondhand smoke exposure, restless sleep, and sleep duration in adolescents. Sleep Disorders, 2014. (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Yolton K, Xu Y, Khoury J, et al. Associations Between Second Hand Smoke Exposure and Sleep Patterns in Children. Pediatrics. 2010;125(2):e261–e268. doi: 10.1542/peds.2009-0690 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jara SM, Benke JR, Lin SY, Ishman SL. The association between secondhand smoke and sleep-disordered breathing in children: A systematic review. The Laryngoscope. 2015;125(1):241–247. doi: 10.1002/lary.24833 [DOI] [PubMed] [Google Scholar]
- 20.Groner JA, Nicholson L, Huang H, Bauer JA. Secondhand Smoke Exposure and Sleep-Related Breathing Problems in Toddlers. Acad Pediatr. 2019;19(7):835–841. doi: 10.1016/j.acap.2019.03.008 [DOI] [PubMed] [Google Scholar]
- 21.Chang VC, Chaput J-P, Roberts KC, Jayaraman G, Do MT. Factors associated with sleep duration across life stages: results from the Canadian Health Measures Survey. Chronic Dis Inj Can. 2018;38(11). doi: 10.24095/hpcdp.38.11.02 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Subramanyam R, Tapia IE, Zhang B, et al. Secondhand Smoke exposure and risk of Obstructive Sleep Apnea in Children. Int J Pediatr Otorhinolaryngol. 2020; 130:109807. doi: 10.1016/j.yporl.2019.109807 [DOI] [PubMed] [Google Scholar]
- 23.O’Callaghan F, O’Callaghan M, Scott JG, Najman J, Mamun AA. Effect of maternal smoking in pregnancy and childhood on child and adolescent sleep outcomes to 21 years: a birth cohort study. BMC Pediatr. 2019; 19(1):1–12. doi: 10.1186/sl2887-019-1439-l [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Stéphan-Blanchard E, Telliez F, Léké A, et al. The Influence of In Utero Exposure to Smoking on Sleep Patterns in Preterm Neonates. Sleep. 2008;31(12):1683–1689. doi: 10.1093/sleep/31.12.1683 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Chugh IM, Khanna P, Shah A. Nocturnal symptoms and sleep disturbances in clinically stable asthmatic children. Asian Pac J Allergy Immunol. 2006;24(2-3):135–142. [PubMed] [Google Scholar]
- 26.Diette GB, Markson L, Skinner EA, Nguyen TTH, Algatt-Bergstrom P, Wu AW. Nocturnal Asthma in Children Affects School Attendance, School Performance, and Parents’ Work Attendance. Arch Pediatr Adolesc Med. 2000;154(9):923–928. doi: 10.1001/archpedi.154.9.923 [DOI] [PubMed] [Google Scholar]
- 27.Fagnano M, Bayer AL, Isensee CA, Hernandez T, Halterman JS. Nocturnal Asthma Symptoms and Poor Sleep Quality Among Urban School Children with Asthma. Acad Pediatr. 2011;11(6):493–499. doi: 10.1016/j.acap.2011.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Cornelius MD, Goldschmidt L, Dempsey DA. Environmental tobacco smoke exposure in low-income 6-year-olds: Parent report and urine cotinine measures. Nicotine Tob Res. 2003;5(3):333–339. doi: 10.1080/1462220031000094141 [DOI] [PubMed] [Google Scholar]
- 29.Hale L, Berger LM, LeBourgeois MK, Brooks-Gunn J. Social and Demographic Predictors of Preschoolers’ Bedtime Routines. J Dev Behav Pediatr JDBP. 2009;30(5):394–402. doi: 10.1097/DBP.ObO13e3181ba0e64 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hale L, James S. Sociodemographic, Psychosocial, and Contextual Factors in Children’s Sleep. In: McHale SM, King V, Buxton OM, eds. Family Contexts of Sleep and Health Across the Life Course. National Symposium on Family Issues. Springer International Publishing; 2017:153–173. doi: 10.1007/978-3-319-64780-7 [DOI] [Google Scholar]
- 31.Reichman NE, Teitler JO, Garfinkel I, McLanahan SS. Fragile Families: sample and design. Child Youth Serv Rev. 2001;23(4-5):303–326. doi: 10.1016/S0190-7409(01)00141-4 [DOI] [Google Scholar]
- 32.MacCallum RC, Zhang S, Preacher KJ, Rucker DD. On the practice of dichotomization of quantitative variables. Psychol Methods. 2002;7(1):19–40. doi: 10.1037/1082-989X.7.1.19 [DOI] [PubMed] [Google Scholar]
- 33.James S, McLanahan S, Brooks-Gunn J, et al. Sleep Duration and Telomere Length in Children. J Pediatr. 2017;187:247–252.e1. doi: 10.1016/j.jpeds.2017.05.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.James S, Hale L. Sleep Duration and Child Well-Being: A Nonlinear Association. J Clin Child Adolesc Psychol. 2017;46(2):258–268. doi: 10.1080/15374416.2016.1204920 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Branstetter SA, Horton WJ, Mercincavage M, Buxton OM. Severity of Nicotine Addiction and Disruptions in Sleep Mediated by Early Awakenings. Nicotine Tob Res. 2016; 18(12):2252–2259. doi: 10.1093/ntr/ntwl79 [DOI] [PubMed] [Google Scholar]
- 36.Meltzer LJ, Montgomery-Downs HE, Insana SP, Walsh CM. Use of actigraphy for assessment in pediatric sleep research. Sleep Med Rev. 2012;16(5):463–475. doi: 10.1016/j.smrv.2011.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Sadeh A The role and validity of actigraphy in sleep medicine: An update. Sleep Med Rev. 2011;15(4):259–267. doi: 10.1016/j.smrv.2010.10.001 [DOI] [PubMed] [Google Scholar]
- 38.Sadeh A, Acebo C. The role of actigraphy in sleep medicine. Sleep Med Rev. 2002;6(2): 113–124. doi: 10.1053/smrv.2001.0182 [DOI] [PubMed] [Google Scholar]
- 39.Garrett BE, Dube SR, Trosclair A, Caraballo RS, Pechacek TF, Centers for Disease Control and Prevention (CDC). Cigarette smoking - United States, 1965-2008. MMWR Suppl. 2011;60(1):109–113. [PubMed] [Google Scholar]
- 40.Cornelius ME. Tobacco Product Use Among Adults — United States, 2019. MMWR Morb Mortal Wkly Rep. 2020;69. doi: 10.15585/mmwr.mm6946a4 [DOI] [PMC free article] [PubMed] [Google Scholar]