Skip to main content
PMC home page
HHS Author Manuscripts logoLink to HHS Author Manuscripts
. Author manuscript; available in PMC: 2019 Dec 1.
Published in final edited form as: Fam Syst Health. 2018 Aug 23;36(4):439–450. doi: 10.1037/fsh0000367

Examining Family and Neighborhood Level Predictors of Sleep Duration in Urban Youth

Nancy White Street a,b, Marie C McCormick b, S Bryn Austin b,c, Natalie Slopen d, Rima Habre e, Beth E Molnar f
PMCID: PMC6310096  NIHMSID: NIHMS980589  PMID: 30137999

Abstract

Introduction:

Shortened sleep duration in adolescence has been found to be associated with adverse health outcomes. While several studies have explored individual predictors, few have examined the role of neighborhood-level factors, family and peer contexts as predictors of sleep among adolescents.

Methods:

We examined contextual factors of sleep duration in a sample of 1,614 urban, public high school students from the 2008 Boston Youth Survey. Neighborhood data came from the 2008 Boston Neighborhood Survey of 1,710 adult Boston residents, the 2009 American Community Survey Census (ACS), and Boston Police.

Results:

Using multi-level linear regression, adjusting for neighborhood and school clustering, age, race and sex, we found concentrated neighborhood poverty to be positively associated with sleep duration (β=0.09, p=0.03). Family context was significantly associated with longer sleep duration: >1–3 hours of homework per night reported longer sleep compared to students reporting ≤1 hour per night (β=0.20, p=0.005). Students reporting lower levels of positive parenting influence had shorter sleep duration (0–25th percentile: β=−0.25, p=0.01; 26th-50th β=−0.24, p=0.03), compared to students in the highest percentile. Students who never ate dinner with family had shorter sleep duration as compared to those having dinner with family five or more times per week (β=−0.22, p=0.05).

Discussion:

Our findings focusing on neighborhood and family context represent potentially modifiable practices. These finding are important for public health advocates and health care providers as they seek to curb the epidemic of sleep deprivation in youth.

Keywords: Sleep, Youth, Context, Neighborhoods, Family

INTRODUCTION

Optimal sleep takes up a third of a person’s day, and is known to be an important component of good health (Roberts, Roberts, & Duong, 2009). The National Sleep Foundation recommends adolescents get 9 hours of sleep per night (Hirshkowitz et al., 2015); however, sleep deprivation amongst U.S. youth is highly prevalent. The Youth Risk Behavior Surveillance System (YRBSS) indicates that ~73% of teenagers nationwide report insufficient nighttime sleep, defined as 7 hours or less of sleep per night (Kann, 2016). Shortened sleep duration in adolescence increases risk for adverse health outcomes including overweight and mood disorders (Berkey CS, 2008; Carskadon, 2004; Dinges et al., 1997; Gordon-Larsen, Adair, Nelson, & Popkin, 2003), decreases attention and focus (Fallone, Acebo, Arnedt, Seifer, & Carskadon, 2001), and increases risk for accidents (Brown, Finn, DeSautel, & Mezzanotte, 1996).

Recent research has focused on identification of risk factors for sleep deprivation. Studies find that shorter sleep duration is associated with individual level characteristics including older age (Hitze et al., 2009), minority ethnicity, obesity/overweight status (Knutson, 2005; Nixon et al., 2008), and male sex (Eaton et al., 2010; Moore et al., 2011; Spilsbury et al., 2004). Additionally, bedtimes and sleep routines are found to affect sleep quality and duration in youth (Lund, Reider, Whiting, & Prichard, 2010; Mindell, 2009; Nixon et al., 2008; Spilsbury et al., 2004). Limited studies have examined the role of broader contextual factors such as family and neighborhood conditions predicting sleep (McHale, Kim, Kan, & Updegraff, 2011; Spilsbury et al., 2005; Spilsbury et al., 2006).

Social-ecological theory outlines the multiple levels affecting individual behavior, illustrating the need to investigate individual, interpersonal (peer), family, and neighborhood factors as influences of sleep characteristics (Bronfenbrenner, 1986) and served as the theoretical framework for our study.

Neighborhood factors, including poverty, crime, collective efficacy, physical and social disorder, have been linked to children’s health and well-being (Browning, Burrington, Leventhal, & Brooks-Gunn, 2008; Browning, Leventhal, & Brooks-Gunn, 2004; Dupere, Leventhal, Crosnoe, & Dion, 2010). Few studies have explored the association between neighborhood level characteristics and sleep duration in youth. Neighborhood disadvantage (including area-level socioeconomic and built environment) has been associated with higher rates of sleep apnea (Spilsbury et al., 2006), shorter sleep duration (Moore et al., 2011), and higher odds of serious sleep problems (Singh & Kenney, 2013). Neighborhood characteristics including high crime rates (McHale et al., 2011) and greater social fragmentation (Pabayo, Molnar, Street, & Kawachi, 2014) have been associated with inadequate sleep and shorter sleep duration. Additionally, timing and consistency of sleep were associated with demographic, behavioral (e.g., number of household members, sleep hygiene behaviors), and neighborhood (e.g., type of housing, quality of public space) factors (Marco, Wolfson, Sparling, & Azuaje, 2011).

Family characteristics have been associated with children’s health and well-being in a limited number of studies. Positive parenting and family meals have predicted improved health outcomes in youth, most notably overweight/obese (Gordon-Larsen, Adair, Nelson, & Popkin, 2004; Gordon-Larsen, Adair, & Popkin, 2003; Ritchie, Welk, Styne, Gerstein, & Crawford, 2005) and mental health (Eisenberg, Olson, Neumark-Sztainer, Story, & Bearinger, 2004; Gillman, 2000; Sallis & Nader, 1988). A few studies support the impact of family environment on sleep characteristics. Positive characteristics of home environments were found to be significantly associated with increased sleep time for African American children (Spilsbury et al., 2005). A study of young children found a positive family environment to be significantly associated with fewer sleep problems (Lumeng et al., 2007). Studies find increased marital conflict (El-Sheikh, Buckhalt, Mize, & Acebo, 2006), absent fathers, and distant or high-conflict mother-child relationships (Bell & Belsky, 2008) were associated with disruptions in quantity and quality of children’s sleep.

Peers represent a proximal influence for high school students, as they spend a majority of waking hours with classmates and friends (Elkington, Bauermeister, & Zimmerman, 2004). Peer influence in adolescence is linked to health behaviors (Buhi & Goodson, 2007; Rai et al., 2003) and subjective well-being (Morgan et al., 2009), with positive peer influence specifically being associated with positive affect and healthy behaviors (Buhi & Goodson, 2007).

Recognizing that sleep is important to health and growing evidence that health is linked to neighborhood context, we explored whether neighborhood level characteristics predicted variation in duration of sleep amongst a sample of urban adolescent students. We hypothesized that students residing in neighborhoods with higher poverty rates, more social and physical disorder, and higher reports of violent crime would report less sleep.

Further, we explored whether family factors and peer influences predicted variation in sleep duration. We hypothesized that students with one or more family and peer supports would report more sleep.

METHODS

Sample

Data for this analysis came from multiple sources. The adolescent data came from the 2008 administration of the Boston Youth Survey (BYS), a biennial paper survey of Boston Public High School students (grades 9–12) (Dunn, Johnson, & Green, 2012). The survey included demographic characteristics, neighborhood resources, and household factors, emphasizing violence perpetration, witnessing and victimization (Pabayo et al., 2014). Of the 2,725 students from 22 participating schools selected to participate, 68.9% completed the survey (n=1878; non-completers included those who chose not to take the survey (n=99, 3.6%), those without parent/caregiver permission to take the survey (n= 24, <1%), and those absent for survey administration (n=724, 27%)). The BYS was administered in 2008 by trained Harvard School of Public Health staff. A statement requesting assent from students was read prior to survey administration; students could decline participation.

The neighborhood level data came from 2009 American Community Survey Census (ACS) data, crime reports from Boston Police, and the Boston Neighborhood Survey (BNS), a telephone survey of 1,710 adults (age 18+). Key informants from Boston identified 38 neighborhoods, each comprising multiple contiguous census blocks. Stratified sampling of neighborhoods was proportional to neighborhood population size. The BNS survey was administered in 2008 in English and Spanish by trained staff from Fact Finders, Inc., a professional research firm. Of individuals who spoke with an interviewer, 31% completed the survey.

The data sets were merged based on data obtained by the BYS, including neighborhood where student lives, street of residence, nearest cross street and zip-codes.

The Office of Human Research Administration at the Harvard School of Public Health approved data collection procedures.

Measures

Sleep Duration.

Hours of sleep per school night were assessed in the BYS using a two–question format. Students reported the closest hour and minutes for 1) their typical bedtime on a school night, and 2) their typical wake up on a school morning. Average school-night sleep duration was calculated from these two questions, taken from the Sleep Timing Questionnaire, with established reliability (α=0.80) and validity (r=0.7–0.8 to two-week sleep diary) (Monk et al., 2003).

Neighborhood-level Concentrated Poverty.

Poverty level in each neighborhood was represented by an index value combining three measures. ACS data were used to calculate 1) percent of persons living below the federal poverty line, 2) percent of residents 16 years or older unemployed, and 3) percent of households on public assistance. We conducted a principle component analysis of these three variables and created the index of concentrated neighborhood poverty (Molnar, 2004; Sampson, Raudenbush, & Earls, 1997).

Neighborhood Violent Crime Rate.

For each neighborhood, count data on robberies, aggravated assaults, and homicides in 2008 was provided by the Boston Police Department.

Neighborhood Disorder.

Information on neighborhood-level disorder was drawn from the BNS. Six items were used from physical and social disorder scales from previously published surveys: Project of Human Development in Chicago Neighborhoods (Molnar, 2004) and Measuring violence related attitudes, behaviors and influences among youths (Dahlberg, Toal, Swahn, & Behrens, 2005). Items included aspects of physical or social disorder in one’s neighborhood (response options: not a problem, kind of a problem, a big problem). The mean score of these six items was calculated (α=0.84).

Collective Efficacy: Social Cohesion and Control.

Collective efficacy, a measure of neighborhood cohesion and informal social control, was assessed in the BNS using a 10-item scale with established reliability and validity with adults (Sampson et al., 1997). Individual responses were used to calculate neighborhood-specific mean scores. Higher scores reflect higher reports of neighborhood cohesion and control (α=0.87).

Hours of Homework.

A question on the BYS asked students to report on average number of hours spent on homework daily (≤1 hour, >1–3 hours, >3 hours).

Positive Peer Influence.

Peer influence was measured using three questions from the BYS capturing positive influences (i.e., my friends stay out of trouble, follow parent rules, and do well in school) (α=0.85), presented as percentiles ranging from low to high scores from the Likert scale responses.

Positive Parent/Guardian Influence.

Three questions (i.e., an adult in my household tries to understand my point of view, tells me that he/she loves me and/or wants good things for me, and is available for me to talk with about my problems) from the BYS were combined to assess positive parent/guardian influence. Four response options (ranging from strongly disagree to strongly agree) were summed for a 12-point range. This variable is presented in percentiles, where higher scores reflected a high level of parent/guardian influence (α=0.79).

Frequency of Family Dinner.

The BYS asked students how frequently they sat down for dinner with their family per week (none, 1–2 times, 3–4 times, ≥5).

Statistical Analysis

We conducted univariate descriptive analyses of sleep time, social/ physical disorder score, percent poverty, collective efficacy and violent crime reports by neighborhood (n=38) and calculated Spearman correlation coefficients between neighborhood level factors. We fit multi-level random intercept models (also known as mixed effects models) to estimate the relationship between individual and neighborhood level independent variables and sleep duration. These multi-level models allow for assessment of multiple levels within a single model and account for the clustered structure of the data where students are situated within neighborhoods and schools.

We fit four models to analyze variation in sleep duration by individual, family, and neighborhood factors. The first model adjusted for age, sex, and race/ethnicity. The second model additionally included hours of homework and peer influence, and the third model additionally included positive parental influence and frequency of family dinners. The final model further adjusted for concentrated poverty. All analyses were conducted using SAS version 9.3 (SAS, 2011) with PROC MIXED for multi-level models, accounting for neighborhood and school clusters using random intercepts.

Missing Data and Sample Size

Sleep duration information was missing for 304 students. Sensitivity analysis revealed significant differences in missing sleep duration data by race/ethnicity and age. Therefore, we imputed values for sleep duration based on predictors including age, sex, and grade level using the Iveware program (n=1878) (Raghunathan & Elliott, 2011). We had useable residential information for geocoding from 88% of students, resulting in a final sample of 1,614 students.

RESULTS

Characteristics of Youth Sample

Of the sample of 1,614 youth residing in 38 Boston neighborhoods, 46% percent were male, 42.8% identified themselves as Black non-Hispanic and 33% Hispanic. The sample age ranged from 14 to 19 years (mean=16.7). The mean sleep time of the youth on a weeknight was 6.8 hours (range: 2.25–11.75 hours; Table 1).

Table 1.

Sleep Duration by Individual and Family Contextual Factors Predictors, Boston Youth Survey, 2008 (N=1614)a

n (%) Sleep time (hours) Mean (SE) p-value
Full Sample 1614(100) 6.79 (1.30)
Sex
 Male 739 (45.8) 6.78 (0.07) 0.83
 Female 875 (54.2) 6.80 (0.07) Ref
Age
 ≤15 135 (8.4) 7.16 (0.14) <0.001
 >15 and <=16 315 (19.5) 6.83 (0.12) 0.05
 >16 and<=17 436 (27.0) 6.89 (0.11) 0.01
 >17 and <= 18 425 (26.3) 6.71 (0.10) 0.30
 ≥18 303 (18.8) 6.60 (0.09) Ref
Race/Ethnicity
 White, non-Hispanic 144 (8.9) 7.16 (0.13) Ref
 Black, non-Hispanic 691 (42.8) 6.75 (0.14) 0.004
 Hispanic 535 (33.1) 6.82 (0.14) 0.01
 Asian 130 (8.1) 6.75 (0.17) 0.07
 Other 114 (7.1) 6.53 (0.19) 0.001
Time spent on homework daily
 ≤ 1 hour 745 (46.2) 6.74 (0.07) Ref
 >1–3 hours 690 (42.8) 6.96 (0.07) 0.001
 >3 hours 179 (11.1) 6.41 (0.11) 0.01
Positive parent influence (percentile)
 0–25th 526 (32.6) 6.68 (0.09) 0.0003
 26–50th 468 (29.0) 6.76 (0.10) 0.02
 51–75th 295 (18.3) 6.84 (0.11) 0.13
 76–100th 325 (20.1) 7.01 (0.09) Ref
Positive peer influence (percentile)
 0–25th 478 (29.6) 6.67 (0.09) 0.44
 26–50th 306 (19.0) 6.79 (0.13) 0.70
 51–75th 475 (29.4) 6.97 (0.10) 0.02
 76–100th 355 (22.0) 6.75 (0.09) Ref
Frequency of Family Dinners
 None 590 (36.6) 6.61 (0.96) 0.001
 1–2 451 (27.9) 6.85 (0.11) 0.50
 3–4 211 (13.1) 7.02 (0.12) 0.43
 5+ 362 (22.4) 6.92 (0.90) Ref
Open in a new tab
a

Data was imputed using IVEware. Bivariate analysis adjusted for school level clusters using SAS 9.3, ProcMixed. SE=standard error. Ref=reference.

Neighborhood Characteristics

The percent of families living below poverty within the neighborhoods ranged from 5% to 54% across the 38 neighborhoods (mean=21%). Neighborhood percent unemployment ranged from 3% to 22%. Neighborhood collective efficacy values ranged from 3.20 to 4.11 (mean=3.7). Violent crime in 2008 ranged from 45 to 386 events (mean=192) (Table 2). The index of neighborhood concentrated poverty ranged from 1.7 to 2.7.

Table 2.

Neighborhood Cluster-level Characteristics, Boston Neighborhood Survey (BNS), 2008 (N=38)

N Mean Median Min Max SD
Physical and social disorder 38 2.86 2.83 2.06 3.98 0.47
Social cohesion and control 38 3.70 3.66 3.20 4.11 0.23
% below poverty line 38 20.70 21.16 5.12 54.66 9.85
% employeda 38 88.54 89.93 77.95 97.02 4.69
% of households on public assistance 38 6.45 5.34 0.17 18.61 4.25
Violent crime (count, 2008 data) 38 191.77 145.00 45.00 386.00 107.68
Open in a new tab
a

Reflects employment among individuals ages 16 and older.

Sleep Duration by Neighborhood

Multi-level regression analyses revealed a statistically significant increase in sleep duration by concentrated poverty, adjusted for demographic characteristics (β=0.09, p=0.03). Neighborhood collective efficacy, physical/social disorder and violent crime incidents were not significantly associated with sleep duration in adjusted models (Table 3).

Table 3.

Multi-level Multivariable Parameter Estimates (SE) between Neighborhood Factors (controlling for individual factors) and Sleep Duration in Boston Youth Survey/Boston Neighborhood Survey, 2008 (n=1614)a

Neighborhood Predictors Model 1 Model 2 Model 3 Model 4
Concentrated Poverty 0.09 (0.04) 0.10 (0.07) 0.07 (0.04) 0.06 (0.04)
Collective Efficacy 0.07 (0.30)
Social and Physical Disorder 0.08 (0.09)
Violent Crime (percentile)
 0–25th −0.11 (0.12)
 26–50th −0.01 (0.12)
 51–75th 0.05 (0.12)
 76–100th Ref
Open in a new tab
a

Separate models were created for neighborhood factors, adjusted for neighborhood level concentrated poverty and age, race and ethnicity.

Collective efficacy was highly correlated with concentrated poverty; therefore we tested it without concentrated poverty in a separate model not shown (parameter estimate −0.26 (SE=0.18, p>.05).

Statistically significant differences (p<.05) are in bold. All models were adjusted for neighborhood and school level clustering using SAS 9.3, Proc Mixed. Ref=reference.

Sleep Duration by Individual and Family Factors

In bivariate analyses examining individual characteristics (Table 1), Whites reported a longer average duration of sleep (7.16 hours), relative to Black non-Hispanic (6.75 hours), Hispanic (6.82 hours) and bi- or multi-racial students (6.53 hours). Sleep duration was inversely associated with age; oldest students reported shortest average sleep duration (6.6 hours). There was no significant difference in sleep duration between boys and girls. Students in lower quartiles of positive parent influence reported less sleep (0–25th percentile: 6.68 hours; 26–50th percentile: 6.76 hours) compared to those is the highest quartile (76–100th: 7.01 hours). Students in the 51–75 percentile of positive peer influence reported more sleep time as compared to those in the highest percentile (6.97 hours versus 6.75 hours). Further, students who reported never having family dinners had significantly less sleep as compared to those who reported five or more family dinners each week (6.6 hours versus 6.9 hours). All p-values were <0.05.

In analyses controlling for school level clustering and gender, age and race/ethnicity were significant predictors of sleep duration (Model 1), as were family contextual factors (Table 4). Students who reported >1–3 hours spent on homework daily had significantly longer sleep duration compared to those who spent ≤1 (β=0.20, p=0.005); students reporting >3 hours spent on homework had shorter sleep duration than those who spent ≤1 (β=−0.34, p=0.004; Model 2). The effect of hours of homework on sleep duration remained significant when adjusting for positive parental influence and frequency of family dinner (Model 3). Students reporting lower positive parent influence had less sleep than those in the highest percentile (0–25th percentile: β=−0.25, p=0.01; 26th-50th β=−0.24, p=0.03). Frequency of family dinners significantly predicted sleep duration, with those never eating family dinner reporting less sleep than those who reported five or more family dinners each week (β=−0.22, p=0.05). In Model 4, which included both individual and family predictors, neighborhood concentrated poverty was significantly associated with sleep duration, with those living in higher concentrated poverty having longer sleep duration (β=0.10, p=0.009). Hours of homework, positive parenting influence and frequency of family dinner remained statistically significant when adjusting for neighborhood concentrated poverty.

Table 4.

Multi-level Parameter Estimates (SE) of Sleep Duration (hours) by Demographics, Peers, Hours of Homework and Family Context in Boston Neighborhood Survey / Boston Youth Survey, 2008 (n=1614)

Model 1 Model 2 Model 3 Model 4
Sex
 Male −0.04(0.07) −0.05(0.07) −0.06(0.07) −0.05(0.06)
 Female Ref Ref Ref Ref
Age
 ≤15 0.53(0.14) 0.53(0.14) 0.50(0.14) 0.51(0.13)
 >15 and <=16 0.21(0.12) 0.23(0.12) 0.20(0.12) 0.20 (0.11)
 >16 and<=17 0.28(0.11) 0.29(0.12) 0.28(0.12) 0.29(0.11)
 >17 and <= 18 0.08(0.10) 0.09(0.10) 0.07(0.10) 0.07(0.10)
 ≥18 Ref Ref Ref Ref
Race/Ethnicity
 White, non-Hispanic Ref Ref Ref Ref
 Black, non-Hispanic −0.39(0.13) −0.37(0.13) −0.31(0.13) −0.40(0.13)
 Hispanic −0.35(0.14) −0.32(0.13) −0.28(0.13) −0.34(0.13)
 Asian −0.35(0.17) −0.28(0.18) −0.21(0.17) −0.23(0.17)
 Other −0.65(0.19) −0.61(0.19) −0.54(0.19) −0.61(0.19)
Time on homework daily
 ≤ 1 hour Ref Ref Ref
 >1–3 hours 0.20 (0.07) 0.17(0.07) 0.17(0.07)
 >3 hours −0.34 (0.12) −0.37(0.12) −0.36(0.12)
Positive peer influence
 0–25th percentile −0.13(0.10)
 26–50th percentile −0.01(0.13)
 51–75th percentile 0.18(0.10)
 76–100th percentile Ref
Positive parental influence
 0–25th percentile −0.25(0.09) −0.26(0.09)
 26–50th percentile −0.24(0.11) −0.24(0.11)
 51–75th percentile −0.17(0.11) −0.17(0.11)
 76–100th percentile Ref Ref
Freq. of Family Dinners
 None −0.22(0.11) −0.22(0.11)
 1–2 −0.02(0.12) −0.02(0.12)
 3–4 0.09(0.12) 0.09(0.12)
 5+ Ref Ref
Neighborhood Concentrated Povertya 0.10(0.04)
Open in a new tab

Data was imputed using IVEware.

a

Neighborhood concentrated poverty index ranged from 2.7 to −1.7. Bold indicates significant p<0.05. Ref=reference.

DISCUSSION

In this representative sample of Boston public high school students, higher neighborhood concentrated poverty was associated with longer sleep duration, controlling for age, race and sex. Three family contextual factors (hours of homework, positive parental influence, and participating in family dinner) were significant predictors of sleep duration, even after adjusting for age, race, sex and concentrated poverty. Among individual-level characteristics, increasing age was associated with shorter sleep duration. White students reported significantly longer sleep durations compared to all other racial groups; bi- and multi-racial students report the shortest sleep duration.

Concentrated neighborhood poverty was associated with sleep duration in our analyses in the reverse direction to our hypothesis. Students living in neighborhoods with higher concentration of poverty slept longer than those from neighborhoods will lower concentrations. Our findings lead us to speculate that children from higher socioeconomic neighborhoods may have greater access to electronic devices (e.g., laptops and smart phones) and may be up later at night using devices. These children may also be engaged in organized afterschool and evening activities, leading to later bedtimes.

Our hypothesis, that higher levels of neighborhood collective efficacy would predict longer sleep duration, was not supported in this analysis. Higher collective efficacy is thought to support community members through enhanced communication and greater community engagement (Kawachi & Berkman, 2000). Our neighborhood findings differ from previous research that suggests neighborhood social control and safety are linked to increases in healthy behaviors among youth (Cradock, Kawachi, Colditz, Gortmaker, & Buka, 2009; Molnar, 2004). Further, neighborhood physical and social disorder and violent crime were not associated with sleep duration in our adjusted analyses. This finding is in keeping with prior work with urban youth (Moore et al., 2011) yet differs from a study of Mexican-American youth that reports a marginal association between neighborhood crime and shorter sleep duration (McHale et al., 2011).

Our neighborhood-level assessment utilized is not exhaustive of the multiple influences of youth health behaviors. The association between neighborhood environments and sleep is likely complex and may relate to cultural and other factors we did not include. Future research should include additional neighborhood-level factors that may be targets for intervention for improving sleep, including environmental noise levels (traffic, sirens, etc.). Additional consideration of measuring quality of sleep amongst youth may address sleep hygiene. Sleep duration represents a component of sleep hygiene, while quality of sleep and other sleep distracting factors may be important considerations in analyses of contextual-level factors on adolescents’ sleep.

We recognize that sleep is a behavior occurring within the home and may be primarily driven by family context, or the proximal household environment, rather than neighborhood environment. In this study, family context most consistently predicted sleep duration, where 1–3 hours of homework per night, positive parenting influence, and frequency of family dinners were all significant predictors of longer sleep duration. These findings are consistent with findings noting the role of family in promoting healthy behaviors, including proper nutrition and sleep hygiene (Ritchie et al., 2005; Sallis & Nader, 1988; Spilsbury et al., 2005), and set the stage for parent education by health professionals and educators.

Regarding individual characteristics, older age was associated with less sleep, consistent with a large body of evidence (Crowley, Acebo, & Carskadon, 2007; McHale et al., 2011; Moore et al., 2011; Smaldone, Honig, & Byrne, 2007). Older children may have less parental oversight than younger children, thus have more control over their bedtime. Furthermore, research on physiological changes in circadian rhythms across adolescence suggests biological process may contribute to this pattern, in addition to the social patterning of later bedtimes (Carskadon, 1999, 2005; Crowley et al., 2007). Racial differences in sleep duration documented in this study are consistent with other studies and may reflect a variety of environmental conditions that require further investigation (Moore et al., 2011; Smaldone et al., 2007).

Our study has several limitations. Data comes from a sample of urban youth residing in a densely populated, segregated city. Our sample of students attending public high schools in Boston, a school system of mainly low-income students, truncated our sample’s socioeconomic range. Notably, those absent from school were not represented. Even though data used in this analysis came from Boston adolescents in 2008–09, it is likely that the impact of family context remains a strong predictor of sleep duration now, approximately ten years later, although more recent data would be needed to formally investigate this. Further, a causal relationship between neighborhood-level and family contextual variables with sleep duration cannot be established as data were cross-sectional. We relied on students’ self-report of sleep duration and family context. Finally, we lack data on students’ family socio-economic status and were unable to adjust for this factor.

Despite these limitations, a major strength is the large, multi-level design permitting examination the joint effects of neighborhood, individual and family predictors of sleep. We incorporated data from multiple sources, including Census data and Police reports. Neighborhood characteristic variables were derived from adults, providing independent reports of collective efficacy and disorder. Information on sleep duration and family context was obtained from adolescents, which strengthens our work (compared with parental reports). Furthermore, we imputed missing data in the youth survey to attenuate risk for potential bias related to survey completion.

Our findings focusing on family context represent potentially modifiable practices, most notably family dinners at least one or two times per week as well as expressions of parental support, such as being understanding, being available to talk about problems, and telling youth they are loved. Primary care providers and nurses may want to incorporate these findings into their health education and counselling of families with young children. Further, these findings are important for educators and public health advocates working to lower sleep deprivation in youth and can easily translate into a campaign for improved youth health.

To our knowledge, this is one of few studies examining the joint effects of neighborhood level and family factors on sleep duration in youth. We encourage research in this area to understand the association of contextual characteristics and sleep duration among adolescents. Longitudinal studies are needed to investigate the causal relationship between neighborhood and family characteristics with sleep. More socioeconomically diverse samples may afford researchers the opportunity to explore these associations further.

Acknowledgements.

The Boston Youth Survey (BYS) was conducted in collaboration with the Boston Public Health Commission (Barbara Ferrer, Director), Boston’s Office of Human Services (Larry Mayes, Chief), Boston Public Schools (Carol Johnson, Superintendent), and the Office of The Honorable Mayor Thomas M. Menino. The survey would not have been possible without the participation of the faculty, staff, administrators, and students of Boston Public Schools. This work was supported by a grant from the Center for Disease Control and Prevention (CDC) National Center for Injury Prevention and Control (NCIPC) (U49CE00740) to the Harvard Youth Violence Prevention Center (David Hemenway, Principal Investigator). The authors thank Renee Johnson, PhD for her helpful guidance and insight on this study. Nancy White Street is a recipient of support from the Maternal and Child Health Bureau (MCHB) training grant (5T76 MC 00001) and the MCHB Epidemiological MPH/SPH Institute grant (T03MC07648).

Appendix Table. Description of study measures

Measure Description Scale or Items Source
Sleep Duration Hours of sleep per school night Derived hours based on closest hour and minutes for typical bedtime on a school night and typical wake up time on a school morning; Sleep Timing Questionnaire BYS
Neighborhood-level Concentrated Poverty Combined neighborhood poverty level index score Percent of persons living below the federal poverty line, Percent of residents 516 years or older unemployed, and 3) percent of households on public assistance; Principle component analysis used to create index score ACS
Neighborhood Violent Crime Rate Neighborhood count of crimes Count data for 2008 of robberies, aggravated assaults, and homicides Boston Police Dept.
Neighborhood Disorder Neighborhood physical and social disorder 6-items reporting about aspects of physical or social disorder; response options included not a problem, kind of a problem, a big problem; neighborhood-specific mean scores derived BNS
Collective Efficacy: Social Cohesion and Control Neighborhood cohesion and informal social control 10-item scale of collective efficacy (Sampson et al., 1997); neighborhood-specific mean scores derived BNS
Hours of Homework Daily average hours of homework Average number of hours spent on homework per day (≤1 hour, >1–3 hours, >3 hours) BYS
Positive Peer Influence Presence of positive peer influences Reported agreement for the following items: my friends stay out of trouble, my friends follow parent rules, and my friends do well in school; derived percentiles based on responses BYS
Positive Parent/Guardian Influence Presence of positive parent/guardian influences Reported agreement for the following items: an adult in my household tries to understand my point of view, tells me that he/she loves me and/or wants good things for me, and is available for me to talk with about my problems; derived percentiles based on responses BYS
Frequency of Family Dinner Weekly average family dinner Frequency of dinner with family per week (none, 1–2 times, 3–4 times, 5+ times) BYS
Open in a new tab

BYS = Boston Youth Survey; ACS = American Community Survey Census; BNS = Boston Neighborhood Survey

REFERENCES

  1. Bell BG, & Belsky J (2008). Parents, parenting, and children’s sleep problems: Exploring reciprocal effects. British Journal of Developmental Psychology, 26(4),579–593. doi: [DOI] [Google Scholar]
  2. Berkey CS RH, Colditz GA. . (2008). Weight gain in older adolescent females: The internet, sleep, coffee and alcohol. Journal of Pediatrics, 153(5), 635–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bronfenbrenner U (1986). ECOLOGY OF THE FAMILY AS A CONTEXT FOR HUMAN DEVELOPMENT -RESEARCH PERSPECTIVES. Developmental Psychology, 22(6), 723–742. doi: 10.1037//0012-1649.22.6.723 [DOI] [Google Scholar]
  4. Brown WD, Finn PE, DeSautel CL, & Mezzanotte WS (1996). Adolescence, sleepiness and driving. Sleep Research, 25, 459. [Google Scholar]
  5. Browning CR, Burrington LA, Leventhal T, & Brooks-Gunn J (2008). Neighborhood structural inequality, collective efficacy, and sexual risk behavior among urban youth. Journal of Health and Social Behavior, 49(3), 269–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Browning CR, Leventhal T, & Brooks-Gunn J (2004). Neighborhood context and racial differences in early adolescent sexual activity. Demography, 41(4), 697–720. doi: 10.1353/dem.2004.0029 [DOI] [PubMed] [Google Scholar]
  7. Buhi ER, & Goodson P (2007). Predictors of Adolescent Sexual Behavior and Intention: A Theory-Guided Systematic Review. Journal of Adolescent Health, 40(1), 4–21. [DOI] [PubMed] [Google Scholar]
  8. Carskadon MA (1999). When worlds collide - Adolescent need for sleep versus societal demands. Phi Delta Kappan, 80(5), 348–353. [Google Scholar]
  9. Carskadon MA (2004). Sleep deprivation: health consequences and societal impact. Medical Clinics of North America, 88(3), 767-+. doi: 10.1016/j.mcna.2004.03.001 [DOI] [PubMed] [Google Scholar]
  10. Carskadon MA (2005). Sleep and circadian rhythms in children and adolescents: Relevance for athletic performance of young people. Clinics in Sports Medicine, 24(2), 319-+. doi: 10.1016/j.csm.2004.12.001 [DOI] [PubMed] [Google Scholar]
  11. Cradock AL, Kawachi I, Colditz GA, Gortmaker SL, & Buka SL (2009). Neighborhood social cohesion and youth participation in physical activity in Chicago. Social Science & Medicine, 68(3), 427–435. [DOI] [PubMed] [Google Scholar]
  12. Crowley SJ, Acebo C, & Carskadon MA (2007). Sleep, circadian rhythms, and delayed phase in adolescence. Sleep Med, 8(6), 602–612. doi: 10.1016/j.sleep.2006.12.002 [DOI] [PubMed] [Google Scholar]
  13. Dahlberg LL, Toal SB, Swahn M, & Behrens CB (Eds.). (2005). Measuring violence-related attitudes, behaviors, and influences among youths: A compenduim of assessment tools (2nd edition ed.). Alanta, GA: Centers for Disease Control, National Center for Injury Prevention and Control. [Google Scholar]
  14. Desantis AS, Diez Roux AV, Moore K, Baron KG, Mujahid MS, & Nieto FJ (2013). Associations of neighborhood characteristics with sleep timing and quality: the Multi-Ethnic Study Of Atherosclerosis. Sleep, 36(10), 1543–1551. doi: 10.5665/sleep.3054 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dinges DF, Pack F, Williams K, Gillen KA, Powell JW, Ott GE, … Pack AI (1997). Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4–5 hours per night. Sleep, 20(4), 267–277. [PubMed] [Google Scholar]
  16. Dunn EC, Johnson RM, & Green JG (2012). The Modified Depression Scale (MDS): A Brief, No-Cost Assessment Tool to Estimate the Level of Depressive Symptoms in Students and Schools. School Mental Health, 4(1), 34–45. doi: 10.1007/s12310-011-9066-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dupere V, Leventhal T, Crosnoe R, & Dion E (2010). Understanding the positive role of neighborhood socioeconomic advantage in achievement: the contribution of the home, child care, and school environments. Dev Psychol, 46(5), 1227–1244. doi: 10.1037/a0020211 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Eaton DK, Kann L, Kinchen S, Shnaklin S, Ross J, Harris WA, … Wechsler H (2010). Youth risk behavior surveillance - United States, 2009. [PubMed] [Google Scholar]
  19. Eisenberg ME, Olson RE, Neumark-Sztainer D, Story M, & Bearinger LH (2004). Correlations between family meals and psychosocial well-being among adolescents. Arch Pediatr Adolesc Med, 158(8), 792–796. doi: 10.1001/archpedi.158.8.792 [DOI] [PubMed] [Google Scholar]
  20. El-Sheikh M, Buckhalt JA, Mize J, & Acebo C (2006). Marital conflict and disruption of children’s sleep. Child Dev, 77(1), 31–43. doi: 10.1111/j.1467-8624.2006.00854.x [DOI] [PubMed] [Google Scholar]
  21. Elkington KS, Bauermeister JA, & Zimmerman MA (2004). Do parents and peers matter? A prospective socio-ecological examination of substance use and sexual risk among African American youth. Journal of Adolescence, 34(5), 1035–1047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fallone G, Acebo C, Arnedt JT, Seifer R, & Carskadon MA (2001). Effects of acute sleep restriction on behavior, sustained attention, and response inhibition in children. Percept Mot Skills, 93(1), 213–229. doi: 10.2466/pms.2001.93.1.213 [DOI] [PubMed] [Google Scholar]
  23. Gillman MW (2000). Family Dinner and Diet Quality Among Older Children and Adolescents. Archives of Family Medicine, 9(3), 235–240. doi: 10.1001/archfami.9.3.235 [DOI] [PubMed] [Google Scholar]
  24. Gordon-Larsen P, Adair L, Nelson M, & Popkin B (2003). Obesity incidence and prevalence during the transition period between adolescence and adulthood: The National Longitudinal Study of Adolescent Health. Obesity Research, 11, A28–A28. [DOI] [PubMed] [Google Scholar]
  25. Gordon-Larsen P, Adair LS, Nelson MC, & Popkin BM (2004). Five-year obesity incidence in the transition period between adolescence and adulthood: the National Longitudinal Study of Adolescent Health. American Journal of Clinical Nutrition, 80(3), 569–575. [DOI] [PubMed] [Google Scholar]
  26. Gordon-Larsen P, Adair LS, & Popkin BM (2003). Socio-demographic and environmental correlates of obesity in US adolescents: the national longitudinal study of adolescent health. American Journal of Physical Anthropology, 103–103. [Google Scholar]
  27. Hale L, Hill TD, & Burdette AM (2010). Does sleep quality mediate the association between neighborhood disorder and self-rated physical health? Prev Med, 51(3–4), 275–278. doi: 10.1016/j.ypmed.2010.06.017 [DOI] [PubMed] [Google Scholar]
  28. Hale L, Hill TD, Friedman E, Nieto FJ, Galvao LW, Engelman CD, … Peppard PE (2013). Perceived neighborhood quality, sleep quality, and health status: evidence from the Survey of the Health of Wisconsin. Soc Sci Med, 79, 16–22. doi: 10.1016/j.socscimed.2012.07.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hill TD, Burdette AM, & Hale L (2009). Neighborhood disorder, sleep quality, and psychological distress: testing a model of structural amplification. Health Place, 15(4), 1006–1013. doi: 10.1016/j.healthplace.2009.04.001 [DOI] [PubMed] [Google Scholar]
  30. Hirshkowitz M, Whiton K, Albert S, Alessi C, Bruni O, DonCarlos L, … Setters Belinda, V. MV, Ware J. Catesby (2015). National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health, 1(4), 233–243. [DOI] [PubMed] [Google Scholar]
  31. Hitze B, Bosy-Westphal A, Bielfeldt F, Settler U, Plachta-Danielzik S, Pfeuffer M, … Muller MJ (2009). Determinants and impact of sleep duration in children and adolescents: data of the Kiel Obesity Prevention Study. Eur J Clin Nutr, 63(6), 739–746. doi: 10.1038/ejcn.2008.41 [DOI] [PubMed] [Google Scholar]
  32. Kann L (2016). Youth risk behavior surveillance—United States, 2015. MMWR. Surveillance Summaries, 65. [DOI] [PubMed] [Google Scholar]
  33. Kawachi I, & Berkman LF (Eds.). (2000). Social cohesion, social capital and health New York: Oxford University Press. [Google Scholar]
  34. Knutson KL (2005). Sex differences in the association between sleep and body mass index in adolescents. J Pediatr, 147(6), 830–834. doi: 10.1016/j.jpeds.2005.07.019 [DOI] [PubMed] [Google Scholar]
  35. Lumeng J, Somashekar D, Appugliese D, Kaciroti N, Corwyn R, & Bradley R (2007). Shorter sleep duration is associated with overweight at ages 9 to 12 years. Pediatrics, 120(5), 1020–1029. [DOI] [PubMed] [Google Scholar]
  36. Lund HG, Reider BD, Whiting AB, & Prichard JR (2010). Sleep patterns and predictors of disturbed sleep in a large population of college students. J Adolesc Health, 46(2), 124–132. doi: 10.1016/j.jadohealth.2009.06.016 [DOI] [PubMed] [Google Scholar]
  37. Marco CA, Wolfson AR, Sparling M, & Azuaje A (2011). Family socioeconomic status and sleep patterns of young adolescents. Behav Sleep Med, 10(1), 70–80. doi: 10.1080/15402002.2012.636298 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. McHale SM, Kim JY, Kan M, & Updegraff KA (2011). Sleep in Mexican-American adolescents: social ecological and well-being correlates. J Youth Adolesc, 40(6), 666–679. doi: 10.1007/s10964-010-9574-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mindell JA, Owens JA. (2009). A Clinical Guide to Pediatric Sleep: Diagnosis and Management of Sleep Problems (2nd ed.). Philadelphia: Wolters Kluwer: Lippincott Williams & Wilkins. [Google Scholar]
  40. Molnar BE, Gortmaker SL, Bull FC Buka SL (2004). Unsafe to Play? Neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents. American Journal of Health Promotion, 18(5), 378–386. [DOI] [PubMed] [Google Scholar]
  41. Monk TH, Buysse DJ, Kennedy KS, Potts JM, DeGrazia JM, & Miewald JM (2003). Measuring sleep habits without using a diary: The sleep timing questionnaire. Sleep, 26(2), 208–212. [DOI] [PubMed] [Google Scholar]
  42. Moore M, Kirchner HL, Drotar D, Johnson N, Rosen C, & Redline S (2011). Correlates of adolescent sleep time and variability in sleep time: the role of individual and health related characteristics. Sleep Med, 12(3), 239–245. doi: 10.1016/j.sleep.2010.07.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Morgan ML, Zeldin S, Wilson D, Collura J, Vera EM, Gonzales RR, … Dick Coyle L (2009). Subjective Well-Being in Urban Adolescents: Interpersonal, Individual, and Community Influences. Youth &Society, 43(2), 609–634. doi: 10.1177/0044118x09353517 [DOI] [Google Scholar]
  44. Nixon GM, Thompson JM, Han DY, Becroft DM, Clark PM, Robinson E, … Mitchell EA (2008). Short sleep duration in middle childhood: risk factors and consequences. Sleep, 31(1), 71–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Pabayo R, Molnar BE, Street N, & Kawachi I (2014). The relationship between social fragmentation and sleep among adolescents living in Boston, Massachusetts. J Public Health (Oxf), 36(4), 587–598. doi: 10.1093/pubmed/fdu001 [DOI] [PubMed] [Google Scholar]
  46. Raghunathan TE, & Elliott M (2011). IVEware: Imputation and Variance Estimation Software. Retrieved from http://www.isr.umich.edu/src/smp/ive/ [Google Scholar]
  47. Rai AA, Stanton B, Wu Y, Li X, Galbraith J, Cottrell L, … Burns J (2003). Relative influences of perceived parental monitoring and perceived peer involvement on adolescent risk behaviors: an analysis of six cross-sectional data sets. J Adolesc Health, 33(2), 108–118. [DOI] [PubMed] [Google Scholar]
  48. Ritchie LD, Welk G, Styne D, Gerstein DE, & Crawford PB (2005). Family environment and pediatric overweight: What is a parent to do? Journal of American Dietetic Association, 105(suppl 1), S70–S79. [DOI] [PubMed] [Google Scholar]
  49. Roberts RE, Roberts CR, & Duong HT (2009). Sleepless in adolescence: prospective data on sleep deprivation, health and functioning. J Adolesc, 32(5), 1045–1057. doi: 10.1016/j.adolescence.2009.03.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sallis JF, & Nader PR (Eds.). (1988). Family determinants of health behaviors. New York: Plenum Press. [Google Scholar]
  51. Sampson RJ, Raudenbush SW, & Earls F (1997). Neighborhoods and violent crime: a multilevel study of collective efficacy. Science, 277(5328), 918–924. [DOI] [PubMed] [Google Scholar]
  52. SAS. (2011). SAS 9.3. Cary, NC. [Google Scholar]
  53. Singh GK, & Kenney MK (2013). Rising Prevalence and Neighborhood, Social, and Behavioral Determinants of Sleep Problems in US Children and Adolescents, 2003–2012. Sleep Disord, 2013, 394320. doi: 10.1155/2013/394320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Smaldone A, Honig JC, & Byrne MW (2007). Sleepless in America: inadequate sleep and relationships to health and well-being of our nation’s children. Pediatrics, 119 Suppl 1, S29–37. doi: 10.1542/peds.2006-2089F [DOI] [PubMed] [Google Scholar]
  55. Spilsbury JC, Storfer-Isser A, Drotar D, Rosen CL, Kirchner HL, & Redline S (2005). Effects of the home environment on school-aged children’s sleep. Sleep, 28(11), 1419–1427. [DOI] [PubMed] [Google Scholar]
  56. Spilsbury JC, Storfer-Isser A, Drotar D, Rosen CL, Kirchner LH, Benham H, & Redline S (2004). Sleep behavior in an urban US sample of school-aged children. Arch Pediatr Adolesc Med, 158(10), 988–994. doi: 10.1001/archpedi.158.10.988 [DOI] [PubMed] [Google Scholar]
  57. Spilsbury JC, Storfer-Isser A, Kirchner HL, Nelson L, Rosen CL, Drotar D, & Redline S (2006). Neighborhood disadvantage as a risk factor for pediatric obstructive sleep apnea. J Pediatr, 149(3), 342–347. doi: 10.1016/j.jpeds.2006.04.061 [DOI] [PubMed] [Google Scholar]

RESOURCES