Correlates of short sleep duration among adolescents

Journal of Adolescence 77 (2019) 163–167

Contents lists available at ScienceDirect

Journal of Adolescence journal homepage: www.elsevier.com/locate/adolescence

Brief report

Correlates of short sleep duration among adolescents Rachel Widomea,∗, Aaron T. Bergera, Kathleen M. Lenka, Darin J. Ericksona, Melissa N. Laskaa, Conrad Iberb, Gudrun Kiliana, Kyla Wahlstromc a b c

T

Division of Epidemiology and Community Health, University of Minnesota School of Public Health, MN, USA Division of Pulmonary, Allergy, Critical Care and Sleep Medicine, Department of Medicine, University of Minnesota Medical School, MN, USA Department of Organizational Leadership, Policy and Development, College of Education and Human Development, University of Minnesota, USA

ARTICLE INFO

ABSTRACT

Keywords: Sleep Time use Adolescents Screen use Mental health Healthy youth development

Introduction: Short sleep duration is exceedingly common among adolescents and has implications for healthy youth development. We sought to document associations between adolescents’ sleep duration and characteristics of their schedules, behaviors, and wellbeing. Methods: We used data from the baseline wave (9th grade year) of the START study, a cohort of 2134 students in five Minnesota high schools to assess how self-reported sleep duration was associated with the prevalence of time-use characteristics (i.e. activity schedules, screen use), sleep-wake problems (i.e. trouble waking in the morning, falling asleep in class, etc.), and risk of depression. Results: Shorter sleep duration was associated with various behaviors including greater computer/screen time and screen use after bed, a lower probability of doing homework, participation in sports doing chores on school nights, and reporting that it takes at least 20 min to fall asleep on school days (p < 0.05). Suboptimal sleep duration was also associated with a higher probability of all reported sleep-wake problems as well as higher risk of depressive symptoms (p < 0.05). Conclusions: Given that getting an optimal amount of sleep can protect youth from risk and promote healthy youth development, it is critical that we gain a greater understanding of correlates and consequences of short sleep duration in order to develop a sleep-friendly culture for youth.

1. Introduction Sleep deprivation for American teens is the norm, rather than exception. Few US adolescents (7.6%) report getting an optimal amount of nighttime sleep (Eaton et al., 2010). Among adolescents, insufficient sleep has been shown to correlate with poorer school performance (Dewald, Meijer, Oort, Kerkhof, & Bögels, 2010; Pagel, Forister, & Kwiatkowki, 2007), depression (Berger, Widome, & Troxel, 2018; McMakin et al., 2016; Wahlstrom, Berger, & Widome, 2017), obesity risk, and injury (Owens et al., 2014). Of concern, certain groups face an amplified risk of sleep abbreviation; girls, non-whites, and those from households that face socioeconomic disadvantage are more likely to report short sleep compared to others (Owens et al., 2014). There are numerous intertwined drivers related to the adolescent brain, social environment, use of electronic devices, and other aspects of modern society that create a “perfect storm” (Crowley, Wolfson, Tarokh, & Carskadon, 2018) in adolescence that chronically hampers sleep. Previous literature has pointed to young people's social pressures to stay up late, time demands like homework or chores that might add stress and/or encroach on sleeping time, secondary schools that start at an early hour that is severely out of sync with typical adolescent circadian ∗

Corresponding author. Division of Epidemiology and Community Health, 1300 South 2nd Street, Suite 300, Minneapolis, MN, 55454, USA. E-mail address: [email protected] (R. Widome).

https://doi.org/10.1016/j.adolescence.2019.10.011 Received 8 July 2019; Received in revised form 18 October 2019; Accepted 30 October 2019 0140-1971/ © 2019 The Foundation for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.

Journal of Adolescence 77 (2019) 163–167

R. Widome, et al.

cycles, and others (Crowley et al., 2018; Owens et al., 2014). Associations between sleep duration and daily activities may be mutually reinforcing. How adolescents spend time during their waking hours (i.e., studying, working for pay, socializing, family activities) may relate to the amount of time they can successfully devote to sleeping. Also, teens' sleep adequacy, fueled by sleep quantity and quality, may determine how adolescents are willing and able to spend their time. Given this, we aimed to build upon previous work by examining these issues in a population-based sample of adolescents. For this descriptive analysis we hypothesized that sleep duration would be related to specific time-use characteristics of adolescents’ schedules, sleep-wake problems, and mood. These are important potential connections, as gaining knowledge of how schedules can be aligned for adolescents to have maximum access to sleep can inform actions intended to enhance wellbeing. 2. Methods The START study was designed to test whether high school start times have an impact on weight and weight-related behaviors by following a cohort of students in five Minnesota schools over three years. For this report, we analyzed data from student surveys during START's baseline period, when the participants were in the ninth grade (Spring 2016) and when START's schools started at 7:30 or 7:45am. 2.1. Study population and recruitment Letters were sent to the parents/guardians of all ninth grade students in five Minneapolis, MN metro-area high schools describing the voluntary nature of the study with directions for opting out of the research. Surveys were completed by students during the school day in the 2016 Spring term. From the schools’ class lists totaling 2362 students, 2134 or 90% returned surveys. 2.2. Measures Sleep duration was calculated from two items adapted from the Teen Sleep Habits Survey (“Sleep for Science,” n.d.): “About what time do you usually go to bed on school days?” and “About what time do you usually wake up on school days?” (response options were in 15-min intervals). Difference between bedtime and wake-up time yielded sleep duration, which was categorized into a fivelevel sleep duration measure (< 6.0, 6.0–6.75, 7.0–7.5, 7.75–8.25, and 8.5–10 h) based both on the frequency distribution and guided by research literature (for instance the 8.5–10 h is considered an “optimal” amount of sleep for adolescents (Eaton et al., 2010; Perlus, O'Brien, Haynie, & Simons-Morton, 2018)). We measured several aspects of the student's time-use during the last week including studying/doing homework, working at a paying job outside of the home, and participating in organized sports or physical activities, organized extracurricular activities and in household chores or family duties (e.g., mowing lawn, farm duties). Respondents could indicate if they either did these activities “in the morning, before school” and/or “in the evening on days that I had school.” Two dichotomized screen-time measures were: nonschoolwork computer time > 2 h/day and use of a device with a screen after getting into a bed. Six symptoms of sleep-wake behavior problems were derived from a 10-item scale of erratic sleep/wake behaviors over the last two weeks (Wolfson & Carskadon, 1998). Risk of depression was calculated using the six-item Kandel-Davies depressive symptoms scale (Kandel & Davies, 1982), which was developed for an adolescent population and has high test-retest reliability and internal reliability (Brunet et al., 2014; Kandel & Davies, 1982). A total score of 23 or higher indicates high risk of depression (Kandel & Davies, 1982). The START survey included demographic items: age, sex (male vs. female), and race/ethnicity (white non-Hispanic vs. non-white and/or Hispanic, dichotomized due to few respondents identifying as racial or ethnic minorities). To characterize socioeconomic status we used highest level of parent's/guardian's education (five level variable dichotomized to finished college vs. high school/ some college) and whether the participant qualified for free or reduced priced lunch (yes/no/don't know) (see https://frac.org/ school-meal-eligibility-reimbursements). 2.3. Analyses We first used confidence intervals to compare the 5-level sleep duration measure by demographics. We then computed regression models to examine associations between sleep duration and each time use, sleep-wake problems and depression measure for: (1) full sample, unadjusted; and (2) full sample, adjusted for demographics (student sex, race/ethnicity, parents education, free/reduced lunch eligibility, and age). A separate regression model was computed for each measure. We report prevalences of each characteristic of interest within each of the five sleep duration categories. All models included the participants’ school as a random effect to account for nesting of students within schools. Analyses were conducted using SAS 9.4 (SAS/STAT Inc., Cary, NC). 3. Results The START cohort participants are majority white non-Hispanic (79.2%), with at least one parent who completed college (79.5%), and not eligible for free or reduced-price lunch (68.6%) (Table 1). There are notable disparities in sleep duration in the cohort, as girls (vs. boys), non-white and/or Hispanic participants (vs. white non-Hispanic participants), and those with less socio-economic advantage (lesser parent education and qualifying for free or reduced-price lunch) were more likely to report fewer hours of sleep on school nights. 164

Journal of Adolescence 77 (2019) 163–167

R. Widome, et al.

Table 1 School night sleep duration by demographics. Hours of sleep (in 15 min intervals)

Total (n)

Sex Male Female Race/Ethnicity White non-Hispanic Non-white and/or Hispanic Parent Education Finished college High school or some college Qualify for free/reduced lunch Yes No I don't know

<6

6–6.75

7–7.5

7.75–8.25

8.5–10

190

408

582

582

323

Hours of sleep (in 15 min intervals) < 6 6–6.75

7–7.5

7.75–8.25

8.5–10

n%

% (95% CI)

% (95% CI)

% (95% CI)

% (95% CI)

% (95% CI)

1066 (50.9%) 1028 (49.1%)

39.5% (32.4, 46.5) 60.5 (53.5, 67.6)

47.9% (43.0, 52.8) 52.1 (47.2 57.0)

50.7% (46.6, 54.8) 49.3 (45.2, 53.4)

51.8% (47.7, 55.9) 48.2 (44.1, 52.3)

57.8% (52.4, 63.2) 42.2 (36.8, 47.6)

1659 (79.2%) 436 (20.8%)

72.4 (66.0, 78.9) 27.6 (21.1, 34.0)

74.6% (70.4, 78.9) 25.4 (21.1, 29.6)

78.8 (75.5, 82.2) 21.2 (17.8, 24.5)

80.4 (77.1, 83.6) 19.6 (16.4, 22.9)

86.4 (82.7, 90.2) 13.6 (9.8, 17.3)

1503 (79.5%) 388 (20.5%)

69.3 (62.3, 76.4) 30.7 (23.6, 37.8)

77.8% (73.5, 82.1) 22.2 (17.9, 26.5)

83.0 (79.8, 86.2) 17.0 (13.8, 20.2)

80.3 (76.9, 83.7) 19.7 (16.4, 23.1)

80.7 (76.1, 85.3) 19.3 (14.7, 23.9)

258 (12.8%) 1383 (68.6%) 374 (18.6%)

23.2 (17.1, 29.4) 57.5 (50.3, 64.7) 19.3 (13.6, 25.1)

13.7 (10.2, 17.1) 66.2 (61.5, 70.9) 20.1 (16.1 24.1)

10.6 (8.1, 13.2) 71.4 (67.6, 75.1) 18.0 (14.8, 21.2)

11.7 (9.1, 14.4) 70.4 (66.6, 74.2) 17.9 (14.7, 21.1)

10.6 (7.1, 14.1) 71.9 (66.8, 76.9) 17.6 (13.3, 21.8)

2134

Table 2 Unadjusted prevalencesa of time use characteristics, sleep-wake problems, and depression risk. START 2016. p-valuea (df = 4)

Hours of sleep (in 15 min intervals) <6 Time use characteristics Computer non-school screen time > 2 h/day Ever use device with screen after getting into bed for night Does homework before school Does homework on school nights Has job before school Has job on school nights Plays sports before school Plays sports on school nights Extracurriculars before school Extracurriculars on school nights Does chores before school Does chores on school nights Sleep-wake problems 20 + minutes to fall asleep on school days In the last 2 weeks, has: Ever woke too early and couldn't get back to sleep Ever needed to be told more than once to get up in morning Ever slept later than noon Ever arrived late to class because they overslept Ever fallen asleep in class Felt tired or sleepy during day, every day every day Kandel-Davies depressive symptoms scale At high risk of depression (depressive symptoms score ≥ 23)

6–6.75

7–7.5

7.75–8.25

8.5–10

71.1% 58.6% 94.0 95.1 18.4 18.1 46.0 52.9 Essentially 0% 6.0 10.1 8.1 4.1 38.3 46.7 5.8 4.5 31.9 30.6 14.3 11.7 52.5 62.4

49.1% 90.8 13.7 60.6

41.8% 85.5 15.8 63.2

39.3% 73.6 10.5 63.7

6.9 4.6 56.6 3.8 36.4 10.0 60.8

5.6 4.2 50.1 4.1 33.2 10.3 58.1

6.4 3.5 45.8 3.0 32.4 9.3 59.6

< .0001 < .0001 .0292 < .0001 N/A .100 .227 < .0001 .626 .392 .425 .199

42.0

26.9

23.4

22.4

25.9

< .0001

57.4 67.4 66.3 25.8 51.2 44.1

48.0 61.6 49.1 19.7 40.4 30.6

40.6 55.0 41.2 11.7 33.3 23.5

42.8 49.5 30.7 9.0 24.5 22.8

48.6 48.9 23.6 7.7 20.7 22.2

.0006 < .0001 < .0001 < .0001 < .0001 < .0001

32.6

17.4

11.9

10.8

9.0

< .0001

a

Marginal means from generalized linear mixed models (unadjusted except school ID included as random effect due to students nested within schools).

165

Journal of Adolescence 77 (2019) 163–167

R. Widome, et al.

Table 3 Adjusted prevalencesa of time use characteristics, sleep-wake problems, and depression risk by sleep duration category. START 2016. p-valuea (df = 4)

Hours of sleep (in 15 min intervals) <6 Time use characteristics Computer non-school screen time > 2 h/day Ever use device with screen after getting into bed for night Does homework before school Does homework on school nights Has job before school Has job on school nights Plays sports before school Plays sports on school nights Extracurriculars before school Extracurriculars on school nights Does chores before school Does chores on school nights Sleep-wake problems 20 + minutes to fall asleep on school days In the last 2 weeks, has: Woke too early and couldn't get back to sleep Needed to be told more than once to get up in morning Slept later than noon Arrived late to class because they overslept Fallen asleep in class Felt tired or sleepy during day, every day Kandel-Davies depressive symptoms scale At high risk of depression (depressive symptoms score ≥ 23)

6–6.75

7–7.5

7.75–8.25

8.5–10

66.4% 60.3% 93.8 96.9 18.5 18.4 45.1 53.3 Essentially 0% 5.4 9.6 6.3 2.7 38.3 48.9 7.6 4.5 32.7 31.5 12.7 11.3 47.9 63.6

49.9% 90.9 14.3 61.7

41.1% 84.7 16.8 65.5

41.4% 73.1 11.0 65.7

5.8 4.1 57.6 3.2 37.1 10.1 61.1

5.3 3.4 50.8 3.9 34.5 10.9 58.3

5.4 3.4 49.6 3.0 33.8 9.3 58.4

< .0001 < .0001 .089 < .0001 N/A .11 .45 .0016 .191 .57 .84 .025

41.0

26.8

22.8

23.3

27.9

.0003

56.2 68.9 64.5 22.0 48.0 41.2

45.0 62.5 46.2 17.6 39.8 28.9

40.2 54.9 40.7 10.8 33.0 23.9

42.7 49.8 30.3 8.4 24.0 21.0

48.6 48.6 20.9 5.7 20.7 21.7

.0097 < .0001 < .0001 < .0001 < .0001 < .0001

27.2

14.5

10.5

9.6

8.4

< .0001

a Marginal means from generalized linear mixed models adjusted for sex (male vs. female), age, free/reduced lunch eligibility (yes, no, don't know), parents' highest educational attainment (college completion vs. some college or less), and race/ethnicity (white non-Hispanic vs. non-white and/or Hispanic); school ID included as random effect due to students nested within schools.

Analyses of association between sleep and time-use, sleep-wake problems, and risk of depression reveal several associations in unadjusted models (Table 2) and adjusted models (Table 3). Though both non-school computer use and use of devices after bed time were common among all participants, those who reported the least sleep were most likely to report engaging in these activities. Evening homework timing was associated with longer sleep duration in adjusted models, but morning homework time was not. Overall, 45% of those who reported less than 6 h of sleep per night reported doing homework on school nights compared to 66% of those who reported getting 8.5–10 h of sleep per night. Students who reported less than 6 h of sleep per night (vs. longer sleep duration) were less likely to play sports or do chores on school nights. In unadjusted and adjusted models, all sleep-wake problems were more common among those reporting the shortest sleep duration and shorter sleep duration was associated with a “high risk” depression screener score. 4. Discussion In this analysis, doing tasks in the evening, whether homework, sports, or chores, was associated with significantly longer sleep duration, while morning activity tended to be more prevalent among teens with short sleep duration. Students who reported doing homework in the morning, before school starts, may potentially do this because they were not able to find time for it in the preceding night. Sacrificing sleep to study is counterproductive with regard to academic outcomes (Gillen-O’Neel, Huynh, & Fuligni, 2013). When adolescents force themselves to wake even earlier than necessary to try to complete homework, the impact of sleep loss may be compounded by the early wake up which is in conflict with adolescents' biological circadian tendency. Similar to previous research, we found screen time to be associated with less sleep (Harbard, Allen, Trinder, & Bei, 2016; Lemola, Perkinson-Gloor, Brand, Dewald-Kaufmann, & Grob, 2015; National Sleep Foundation, 2014; Safron, Schulenberg, & Bachman, 2001). Participants who slept less were more likely to have used a screen after getting into bed at night, and they were also more likely to have spent more than 2 h a day using a screen for non-school activities. It is concerning that participants who reported shorter sleep duration were more likely to be at high risk for depression and report sleep-wake problems, issues that could be impacting multiple areas including health, learning, and well-being. This study was a cross-sectional analysis so the temporal order of associations is not certain. However, this study has notable strengths including high participation from all schools and a large sample which enabled analysis of relatively uncommon characteristics, such as school night employment. Our findings indicate potential targets for those seeking to extend adolescent sleep duration, including by protecting morning hours for sleep-time, addressing screen time and other inhibitors of sleep onset. Fortunately, adolescent sleep is amenable to interventions such as creating school schedules that allow for sleeping later (Owens et al., 2014). 166

Journal of Adolescence 77 (2019) 163–167

R. Widome, et al.

Acknowledgements The Authors would like to thank the adolescents participating in the START study, the districts that welcomed us to do research in their schools, the START data collectors, and Bill Baker for his work to manage the data. Thank you to Kate Bauer for sharing your great ideas. This study is supported by funding from the National Institutes of Health’s (NIH) Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) (R01 HD088176). Additionally, the authors gratefully acknowledge support from the Minnesota Population Center (P2C HD041023) funded through a grant from NICHD. References Berger, A. T., Widome, R., & Troxel, W. M. (2018). School start time and psychological health in adolescents. Current Sleep Medicine Reports, 4(2), 110–117. https://doi. org/10.1007/s40675-018-0115-6. Brunet, J., Sabiston, C. M., Chaiton, M., Low, N. C. P., Contreras, G., Barnett, T. A., et al. (2014). Measurement invariance of the depressive symptoms scale during adolescence. BMC Psychiatry, 14, 95. https://doi.org/10.1186/1471-244X-14-95. Crowley, S. J., Wolfson, A. R., Tarokh, L., & Carskadon, M. A. (2018). An update on adolescent sleep: New evidence informing the perfect storm model. Journal of Adolescence, 67, 55–65. https://doi.org/10.1016/j.adolescence.2018.06.001. Dewald, J. F., Meijer, A. M., Oort, F. J., Kerkhof, G. A., & Bögels, S. M. (2010). The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta-analytic review. Sleep Medicine Reviews, 14(3), 179–189. https://doi.org/10.1016/j.smrv.2009.10.004. Eaton, D. K., McKnight-Eily, L. R., Lowry, R., Perry, G. S., Presley-Cantrell, L., & Croft, J. B. (2010). Prevalence of insufficient, borderline, and optimal hours of sleep among high school students - United States, 2007. Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine, 46(4), 399–401. https:// doi.org/10.1016/j.jadohealth.2009.10.011. Gillen‐O’Neel, C., Huynh, V. W., & Fuligni, A. J. (2013). To study or to sleep? The academic costs of extra studying at the expense of sleep. Child Development, 84(1), 133–142. https://doi.org/10.1111/j.1467-8624.2012.01834.x. Harbard, E., Allen, N. B., Trinder, J., & Bei, B. (2016). What’s Keeping Teenagers Up? Prebedtime Behaviors and Actigraphy-Assessed Sleep Over School and Vacation. Journal of Adolescent Health, 58(4), 426–432. https://doi.org/10.1016/j.jadohealth.2015.12.011. Kandel, D. B., & Davies, M. N. O. (1982). Epidemiology of depressive mood in adolescents. An empirical study. Archives of General Psychiatry, 39(10), 1205. https://doi. org/10.1001/archpsyc.1982.04290100065011. Lemola, S., Perkinson-Gloor, N., Brand, S., Dewald-Kaufmann, J. F., & Grob, A. (2015). Adolescents’ Electronic Media Use at Night, Sleep Disturbance, and Depressive Symptoms in the Smartphone Age. Journal of Youth and Adolescence, 44(2), 405–418. https://doi.org/10.1007/s10964-014-0176-x. McMakin, D. L., Dahl, R. E., Buysse, D. J., Cousins, J. C., Forbes, E. E., Silk, J. S., et al. (2016). The impact of experimental sleep restriction on affective functioning in social and nonsocial contexts among adolescents. The Journal of Child Psychology and Psychiatry and Allied Disciplines, 57(9), 1027–1037. https://doi.org/10.1111/ jcpp.12568. National Sleep Foundation (2014). 2014 Sleep in America Poll: Sleep in the Modern Family. Retrieved from http://sleepfoundation.org/sites/default/files/2014-NSFSleep-in-America-poll-summary-of-findings–FINAL-Updated-3-26-14-.pdf. Owens, J., Au, R., Carskadon, M., Millman, R., Wolfson, A., Braverman, P. K., et al. (2014). Insufficient sleep in adolescents and young adults: An update on causes and consequences. Pediatrics, 134(3), e921–e932. https://doi.org/10.1542/peds.2014-1696. Pagel, J. F., Forister, N., & Kwiatkowki, C. (2007). Adolescent sleep disturbance and school performance: The confounding variable of socioeconomics. Journal of Clinical Sleep Medicine: JCSM: Official Publication of the American Academy of Sleep Medicine, 3(1), 19–23. Perlus, J. G., O'Brien, F., Haynie, D. L., & Simons-Morton, B. G. (2018). Adolescent sleep insufficiency one year after high school. Journal of Adolescence, 68, 165–170. https://doi.org/10.1016/j.adolescence.2018.07.016. Sleep for Science. (n.d.). Retrieved August 30, 2015, from http://www.sleepforscience.org/contentmgr/showdetails.php/id/93. Safron, D. J., Schulenberg, J. E., & Bachman, J. G. (2001). Part-Time Work and Hurried Adolescence: The Links among Work Intensity, Social Activities, Health Behaviors, and Substance Use. Journal of Health and Social Behavior, 42(4), 425–449. https://doi.org/10.2307/3090188. Wahlstrom, K. L., Berger, A. T., & Widome, R. (2017). Relationships between school start time, sleep duration, and adolescent behaviors. Sleep Health, 3(3), 216–221. https://doi.org/10.1016/j.sleh.2017.03.002. Wolfson, A. R., & Carskadon, M. A. (1998). Sleep schedules and daytime functioning in adolescents. Child Development, 69(4), 875–887.

167