COVID-19 Pandemic Restrictions on Sleep Timing, Sleepiness, and Sleep Quality in Early Adolescents: Insights from a Longitudinal Study

 

Abstract

COVID-19 pandemic restrictions, including school closures, caused dramatic changes to sleep behaviors, such as sleep duration, wake and bedtime, and the associated outcomes such as daytime sleepiness and sleep quality, some of which have been positive. Studies in young children have found that sleep duration increased during the height of the COVID -19 pandemic, however, it is unknown if daytime sleepiness was reduced. In 2019, 288 students were recruited from 15 middle schools in a Southeastern U.S. metropolitan area (Mage = 12.09 years; 54% female; 48% Black; 37% White; 10% Hispanic, 5% Other). Wave 2 in 2020 was interrupted by the COVID-19 pandemic, with 198 students completing the assessment before the pandemic and 82 after the pandemic began in fall 2020. Youth were provided Actigraph watches to wear for 7 consecutive nights and self-reported sleepiness via in person or online surveys. Sleep duration, bedtime, waketime, and number of nighttime awakenings were calculated from Actigraph data. Results show that children who participated during the pandemic had later bedtimes and waketimes while also reporting less sleepiness. There was no difference in sleep duration or nighttime awakenings between children who participated before the pandemic and during the pandemic. Future studies should examine other factors besides sleep duration that may have contributed to lower levels of sleepiness (stress due to at home learning) in children during the COVID-19 pandemic. 

Keywords: COVID-19, Sleep duration, Sleepiness, Sleep quality

 

Although sleep is important for all age groups, sufficient and good quality sleep during adolescence is crucial for development and wellbeing (Bruce et al., 2017). However, 40% of US adolescents report getting less than 7 hours of sleep on most nights (Paruthi et al., 2016). Furthermore, almost one-third of children report excessive daytime sleepiness that contributes to poor emotion regulation and lower academic performance (Wesley et al., 2022). COVID-19 pandemic restrictions, including school closures, caused dramatic changes to sleep behaviors, some of which have been later bed and waketimes, and longer sleep duration (Paterson et al., 2021). Studies in young children have found that sleep duration increased during the height of the COVID -19 pandemic (Kaditis et al., 2021), however, it is unknown if daytime sleepiness was reduced. Additionally, bedtimes and waketimes occurred later, likely due to implementation of remote learning and the elimination of travel time. In contrast to young children, adults experienced lower sleep quality as a result of the pandemic (Bruni et al., 2022), but fewer studies have examined the effect of the pandemic on sleep behaviors and associated outcomes in adolescents. The current study will examine sleep behaviors pre and during COVID-19 pandemic in young adolescents. 

Development

Studies have reported that adolescents actually require more sleep than preadolescents, to support physical and cognitive changes during this transition (Carskadon, 1990). However, adolescents typically get less than the optimal amount of sleep due to both psychosocial and biological changes. Regarding psychosocial development, sleep in adolescence is altered by screen time, extracurricular activities, socialization, academic demands like homework (Galván, 2020), and excessive caffeine use (Owens et al., 2020). Biologically, adolescents face changes in melatonin production which causes a shift in their circadian rhythm. The delayed release of melatonin introduces a preference for evening chronicity in adolescents (Crowley et al., 2007). The preference for evening chronicity combined with early school start times facilitates the development of chronic sleep debt, sleeping less than perceived sleep need (Alfonsi et al., 2020; Shen et al., 2021). Such sleep debt produces poor quality of life, sleepiness, fatigue, anxiety, and poor academic performance (Touitou, 2013). Adolescent evening chronicity occurs in tandem with the developmental increase in risk-taking and poor decision making (Telzer et al., 2013). This developmentally appropriate increase in risk-taking and evening chronicity increases the importance of sufficient sleep in adolescents to counteract behavioral, psychological, or academic issues. 

Sleep duration and timing

The National Sleep Foundation’s sleep guidelines are published at 9 to 11 hours per day for school aged children (6-13 years) (Hirshkowitz et al., 2015). However, 12-18-year-old adolescents tend to sleep for about 7 hours each night (Galland et al., 2018). Today’s adolescents typically have more variable sleep compared to adults due to different demands between school days and weekends (Gradisar et al., 2011). Such insufficient, variable sleep in adolescents has negative impacts on mood, attention, school performance, depression/suicidal ideation, anxious feelings, fatigue, obesity, and car crashes linked to drowsiness (Fuligni et al., 2018; Owens et al., 2020) and plays an important role in learning and cognitive performance (Perkinson-Gloor et al., 2013). Many studies have reported longer sleep duration in adolescents during the COVID-19 pandemic (Sharma et al., 2021; Temkin et al., 2018). 

As adolescents transition from middle school to high school, school schedules affect sleep patterns, which is most commonly seen in early rise times as the school day begins earlier (Carskadon et al., 1993). This trend has been noted for over 20 years, but the COVID-19 produced an opportunity for a widespread transition to delayed start times. A national survey of U.S. adolescents found an association between anxiety, depression, and a decline in physical activity and delayed bed and waketimes during COVID-19 (Wesley et al., 2022). In contrast, a study in Singapore found delayed school start times resulted in sustained benefits in sleep duration, daytime alertness, and mental well-being (Lo et al., 2018). There are inconclusive reports on the positive or negative outcomes related to sleep behaviors in adolescents during COVID-19. Further, it is unknown if the reported effects on adolescent sleep and well-being have persisted after school reopening.

Daytime sleepiness

Adolescents with chronic short sleep duration report feeling sleepy and having difficulties staying awake during daytime activities (Drake et al., 2003). Pediatric excessive daytime sleepiness is associated with behavioral problems, mood disturbances, depression, dysregulation of affect, impairments in neurocognitive function, increased risk for alcohol and drug use in teenagers, declining academic performance, and safety concerns (Owens et al., 2020). Additionally, increased sleepiness has been reported to have the strongest relation to school performance (Dewald et al., 2010). 

Prior to the onset of the COVID-19 pandemic, research had reported inconsistent findings on daytime sleepiness. One study of Brazilian adolescents found an association between longer sleep duration and higher scores of daytime sleepiness (Alves et al., 2020) suggesting that other factors are involved in sleepiness other than short sleep duration. Few studies have examined the effect of COVID-19’s increase in duration on daytime sleepiness. 

Sleep quality

Sleep quality is commonly defined through multiple quantitative aspects of sleep, such as sleep duration, sleep latency, or number of arousals, in addition to subjective measures, such as restfulness or depth (Buysee et al., 1989). Poor sleep quality in adolescence is associated with high-risk behaviors, diminished attentional and behavioral control, and poor emotional regulation (Telzer et al., 2013). One study reported worsened sleep quality during the pandemic despite increases in sleep duration and more appropriate bed and waketimes (Silva et al., 2022). Similar studies are limited in their causal inferences by assessments beginning after the onset of the COVID-19 pandemic but report that adolescent sleep quality during COVID-19 has been reported as worsening (Rocha & Fuligni, 2023). 

Sleep quality is often assessed through self-report measures by the child or parent. These reports are limited by subjectivity and reporting biases. Parents tend to overestimate child sleep and underestimate the number of awakenings (Short et al., 2013). Note that self-report is problematic due to restful sleep being defined as unconscious (Dewald et al., 2010). However, objective measures of sleep quality (e.g., number of awakenings) can be assessed using actigraphy and has been reported as substantially more reliable than subjective measures (Short et al., 2012). Data on healthy adolescent sleep is difficult to record with guaranteed validity due to the limitations of each measure. Thus, the combination of subjective and objective measures ensures the most accuracy (Sadeh, 2011). 

COVID-19

The COVID-19 pandemic and the associated economic, health, and educational ramifications caused a large-scale disruption to adolescent development. In October of 2020, it was reported that 23 million were infected globally with a death toll of 1,074,732 deaths (Benner & Mistry, 2020). Families were at risk for familial disruption due to death or infection, financial instability, job loss, and educational disruptions caused by transitions to online learning (Benner & Mistry, 2020). Adults in the U.S., China, Greece, and Italy reported poorer sleep quality and more sleep disturbances during the COVID-19 (G. Hisler et al., 2020). While sleep quality was disturbed, there is inconclusive evidence on the significant difference in adult sleep duration during COVID-19 (G. C. Hisler & Twenge, 2021). Adults have reported negative sleep behaviors and outcomes due to the COVID-19 pandemic, however adolescents may have experienced improved sleep. Youth reported waketimes up to 2.1-2.9 hours later and averaged 7.9-8.7 hours of sleep during the COVID-19 pandemic, possibly due to the transition to remote instruction (Weingart et al., 2021). A study of Canadian adolescents reported greater sleep duration was associated with higher SES scores (Duncan et al., 2023). The trending increase in sleep duration during the COVID-19 pandemic varies by sub-population with racial-ethnic minorities reaping less benefits (Duncan et al., 2023). More investigation into the relationships between sleep behaviors and COVID-19 in adolescents and how they differ from adults is needed. 

Current study

Adolescent sleep health is crucial to emotional well-being and academic performance, however existing research focuses on the COVID-19 pandemic’s impact on adult sleep behaviors and outcomes. There remains a need for longitudinal data on adolescent sleep behaviors and outcomes, in particular using multiple, objective measures of sleep. The purpose of this study was to investigate sleepiness, sleep duration, bedtimes, waketimes, and number of awakenings (measure of sleep quality) in early adolescents pre and during the COVID-19 pandemic. It is expected that youth experienced longer sleep duration, later bedtimes and waketimes, fewer nighttime awakenings, and less daytime sleepiness during the pandemic compared to Pre-Pandemic. 

Methods

Sample

The sample is comprised of 288 students recruited from 15 middle schools in the Southeastern U.S. (Mage = 12.09 years; 54% female; 48% Black; 37% White; 10% Hispanic, 5% Other). The fifteen middle schools were spread across six school districts, from both urban and rural areas. Wave 1 data was collected from January 2019 to December 2019. Wave 2 began in January 2020 but was halted due to COVID-19 school restrictions in March 2020. Data collection transitioned to a home and lab-based format in Fall 2020. In Wave 2, 198 students completed study activities before the pandemic and 82 after the pandemic in the Fall of 2020. 

Procedure

Participants were recruited for the Wave 1 assessment during regular school days from one homeroom of a teacher that was selected by the principal to serve as the primary contact in each participating school. Information about the study was presented to students and packets were distributed with consent forms and more information. One week later, the students returned signed parental consent and student assent forms to staff. All data collection for the baseline assessment took place at school from January to November 2019. Students received an activity (ActiGraph GT9X Link) tracker and were instructed to wear the watch for 8 consecutive days and 7 nights. Self-reported measures were administered at school on electric tablets during a nonacademic class session at the end of the week. After the pandemic, study materials were delivered to the home and picked up one week later. Participants completed self-reported measures via a confidential online questionnaire. For both waves, students and parents were compensated with gift cards for their time. The University Institutional Review Board approved all study procedures. 

Measures 

Sleep duration and timing

During each study week, participants wore Actigraph GT9X Link activity trackers for 8 days and 7 consecutive nights. Sleep duration, bedtimes, and waketimes were calculated and using Actilife desktop analysis (v. 6.13.3) and validated with self-reported sleep diaries. Validated sleep scoring algorithms are then applied to differentiate wakefulness from sleep and thus calculate total duration of sleep (Sadeh et al., 2009). Actigraphy underestimates adolescent sleep time, therefore. sleep duration was calculated as the average time spent in bed each night across seven nights (Short et al., 2012). 

Sleepiness

In both waves, participants self-reported sleepiness using the Pediatric Daytime Sleepiness Scale (PDSS) that was administered at school (pre pandemic) or via an online survey (post pandemic) (Drake et al., 2003). The Pediatric Daytime Sleepiness Scale (PDSS) has 8 items (e.g., How often do you fall asleep or get drowsy during class periods?) and is rated from 1 (never) to 5 (very often/always). Items were averaged with higher scores indicating higher levels of sleepiness. 

Number of Awakenings

For the current study, sleep quality was defined as the average number of nighttime awakenings per night across the week and is commonly used as a measure of sleep quality (Goelema et al., 2019). Nighttime awakenings per night was calculated using Actilife desktop analysis (v. 6.13.3) and averaged across the week with more nighttime awakenings indicating lower sleep quality.

Demographic characteristics 

Youth self-reported sex (0 = male; 1 = female) whereas parents reported their child’s race/ethnicity, which was dichotomized as 0 (non-minority status) or 1 (racial-ethnic minority). Parents also reported annual household income (13-point scale from 1 <$5,000 to 13 >$90,000) and educational attainment (1 = Less than 12th grade/no diploma to 7 = Graduate or professional degree). 

Analyses

All statistical analyses were performed with SPSS version 27. Sleep duration, timing, and number of awakenings were calculated using Actilife v. 6.13.3.  Descriptives and correlations were conducted for all key variables. Multivariate regression analyses tested relationships between COVID-19 participation (pre COVID-19 pandemic vs participating after March 2020) and Wave 2 sleep variables (sleep duration, bedtime and waketime, number of awakenings, and daytime sleepiness). Demographics included child sex, minority status, parent education, household income, COVID-19 participation, and Wave 1 sleep behaviors. Independent sample t-tests were conducted to test differences in sleep variables between Wave 1 and Wave 2. Paired samples t-tests were conducted to examine differences in sleep variables between Wave 2 Pre-Pandemic and Wave 2 During-Pandemic.

Preliminary analyses

The descriptive statistics are presented in Tables 1-2. In Wave 1, children slept 6.70 hours per night, went to bed at 10:37 PM and awoke at 6:34 AM, had 20.13 awakenings per night, and a daytime sleepiness score of 1.95. Across all participants in Wave 2, children slept 6.77 hours per night, went to bed at 11:20 PM and awoke at 6:34 AM, had fewer awakenings (14.48 per night) compared to Wave 1, and a daytime sleepiness score of 1.97. In Wave 2 Pre-Pandemic, youth slept slightly longer than in Wave 1 with 6.55 hours per night, went to bed at 11:12 PM, awoke at 6:33 AM, experienced 14.16 awakenings per night and a daytime sleepiness score of 2.06. Wave 2 During-Pandemic youth slept of 7.07 hours per night, went to bed at 11:32 PM, awoke almost one hour later than those who participated Pre-Pandemic (7:22 AM) had 14.94 awakenings per night a daytime sleepiness score of 1.85.

Bivariate Analyses

Paired sample t-tests determined that youth in Wave 1 were found to have more frequent awakenings compared to youth in Wave 2. Bedtimes at Wave 2 were later compared to Wave 1. Similarly, Wave 2 waketimes were later than Wave 1 waketimes. There was no significant difference between sleep duration and daytime sleepiness between Wave 1 or Wave 2 (Table 1). 

Independent sample t-test examined differences between the Wave 2 Pre-Pandemic and During-Pandemic groups (Table 2). Youth who participated During-Pandemic had later waketimes compared to Pre-Pandemic. However, there was no significant difference in sleep duration between Pre- and During-Pandemic youth. During-Pandemic youth had longer sleep duration compared to Pre-Pandemic. Sleep quality and daytime sleepiness also had no significant difference between Pre- and During-Pandemic youth.

Pearson product moment correlations found that bedtimes, waketimes, sleep duration and sleep quality were significantly related within and across waves. Wave 1 sleep duration was not associated with Wave 1 or Wave 2 daytime sleepiness. However, longer sleep duration at Wave 2 was related to less Wave 2 daytime sleepiness. Overall, longer sleep duration was related to later bedtimes, later waketimes, and lower scores of daytime sleepiness but only within waves. 

Multivariate regression analyses

After controlling for demographics, Wave 1 sleep variables, and COVID-19 participation, students who participated in Wave 2 During-Pandemic had later bedtimes and waketimes, less sleepiness, and lower number of awakenings compared to children who participated Pre-Pandemic. Covid-19 participant had no unique effect on sleep duration at Wave 2. Regarding demographics, minority status participants had shorter sleep duration and later bedtimes and waketimes at Wave 2. Females reported higher more daytime sleepiness at Wave 2. Youth with parents with lower educational attainment reported lower sleep quality and more daytime sleepiness at Wave.

Discussion

The current study examined sleepiness, sleep duration, bedtimes, waketimes, and number of awakenings in early adolescents pre and during the COVID-19 pandemic. It was hypothesized that youth would experience longer sleep duration, later bedtimes and waketimes, fewer nighttime awakenings (better sleep quality), and less daytime sleepiness During-Pandemic compared to Pre-Pandemic. 

Sleep duration

Contrary to hypotheses, COVID-19 participation did not have an effect on sleep duration. Preliminary results showed that overall sleep duration increased from Wave 1 to Wave 2, but this relationship was not significant in the adjusted models. One reason for this could be that the effect of racial-ethnic minority status attenuated the overall effect that pandemic participation had on sleep duration. Previous studies in racial ethnic minorities found that daily discrimination was linked to shorter sleep duration that night, possibly due to increased allostatic load that disrupts the body’s stress response (Brody et al., 2014; Chen et al., 2022). Despite school closures that allowed students to wake up later, it could not compensate for the effect of discrimination on sleep duration. Future studies should examine racial ethnic group differences in the relationship between COVID-19 and sleep duration and how discrimination moderates this relationship. 

Sleep timing

Consistent with hypotheses, results demonstrate that children who participated during the pandemic had later bedtimes and waketimes, which is consistent with other findings that show increases in sleep duration and wake and bedtimes (Rocha & Fuligni, 2023) during the pandemic. Although youth have a preference for evening chronicity, it is possible that stayed awake later at night due to increased screen time and reduced daily structure, physical activity, and mental health (Weingart et al., 2021). As a result of later bedtimes, waketime also increased, as children slept longer from staying up later. Racial-ethnic minorities experienced later bedtimes and waketimes during the pandemic. This is consistent with other findings reporting insufficient sleep in black adolescents compared to their white peers (Mitchell et al., 2020). This is suspected to be caused by social factors, bedroom environment context, and school factors (Mitchell et al., 2020). Future studies should examine the social and environmental factors related to racial-ethnic minority sleep timing and COVID-19.

Daytime sleepiness

Contrary to hypotheses, During-Pandemic youth reported more daytime sleepiness compared to Pre-Pandemic youth, possibly due to later waketimes. This finding was inconsistent with previous studies showing that delayed school start times decreased sleepiness (Rocha & Fuligni, 2023; Weingart et al., 2021). This might be related to changes in sleep quality as there was an increase in the number of awakenings during Wave 2 During-Pandemic, but the mechanism remains unclear. Although decreased sleep duration is strongly associated with greater daytime sleepiness (Alves et al., 2020), more studies should examine other factors such as chronic health problems (e.g., asthma) or mental health problems and their impact on daytime sleepiness. 

Regarding sociodemographic characteristics, females had higher scores of daytime sleepiness, which could be due to earlier pubertal timing compared to their male peers (Carskadon et al., 1993) and is also consistent with other studies reporting gendered differences in daytime sleepiness (Lewien et al., 2021; Meyer et al., 2019). Children with less educated parents also reported higher scores of daytime sleepiness which could be due to inconsistent bedtime routines and environmental factors like number of household members and room-sharing (Lewien et al., 2021). Future studies should investigate the moderating role of individual sociodemographic characteristics in relationships between daytime sleepiness and COVID-19. 

Number of Awakenings

Contrary to hypotheses, the number of awakenings in Wave 2 During-Pandemic was greater compared to Wave 2 Pre-Pandemic, which is consistent with many studies citing worse sleep quality During-Pandemic (I Gusti Ayu Trisna Windiani et al., 2023; Richter et al., 2023; Silva et al., 2022). The decrease in sleep quality during the pandemic despite increases in sleep duration contradicts many findings (Gruber et al., 2021; Ramos Socarras et al., 2021). While the current study only utilized one indicator of sleep quality, which may not have been a good indicator with this sample. Future studies should use multiple indicators of sleep quality when examining COVID-19’s effect on adolescent sleep. 

Children with less educated parents had a higher average of awakenings and therefore lower sleep quality, which could be caused by the COVID-19 pandemic’s disproportionate effect on families with low socioeconomic status (SES) (Collier Villaume et al., 2021) due to the increase in household stressors such as a change in job status or decreased income faced During-Pandemic (Yoshikawa et al., 2020). 

Strengths

This study has several strengths. While recording sleep measures, we utilized both subjective and objective measures. These measures enhanced the reliability and validity of sleep behaviors. Subjective measurements of daytime sleepiness and sleep measures were self-reported from the adolescents instead of parental report. The study timeline allowed for Wave 1 to be complete before the onset of the COVID-19 pandemic. This design provided a valuable Pre-Pandemic reference point, which is essential for understanding the pandemic’s effect. 

Limitations 

The current study was not without limitations. First, participants from this study were recruited from public schools (one private school) in a large metropolitan area in the Southeast United States, so results may not be generalizable in other geographic regions or cultures. Second, inclusion criteria required the adolescents to speak English and mentally and physically complete the study tasks, and therefore excluded non-English speaking students and those with disabilities. Third, attrition due to COVID-19 contributed to a 31% decrease in Wave 2 participation may have affected generalizability. Fourth, the objective measures of sleep (e.g., actigraphy) were limited by its accuracy in reporting sleep wake cycles (Ancoli-Israel et al., 2003), however, Actigraph data was validated with self-reported sleep diaries. Fifth, clinical sleep symptomology and other variables were not assessed but may have impacted sleep behaviors, sleepiness, or sleep quality. Sixth, inconsistent reintroductions to in-person learning During-Pandemic across schools comprised on of many disruptions to school and family routines along with poorer mental well-being and increased stress brought on by the COVID-19 pandemic which likely impacted sleep. 

Implications

The implications of this study suggest that adolescents, as a distinct developmental group, are particularly vulnerable to sleep problems, with the COVID-19 pandemic likely contributing to issues such as compromised sleep quality and increased daytime sleepiness (Becker & Gregory, 2020). This vulnerability is attributed to the unique stressors associated with the pandemic, which, while not comparable to natural disasters, can be considered a traumatic experience for adolescents. The significant disruption to family routines may have intensified stress levels, consequently impacting sleep patterns in adolescents.

In light of these findings, there is potential for interventions to mitigate the impact of such stressful disruptions on adolescent sleep. Drawing insights from the positive outcomes observed in a similar study involving adolescents affected by a tornado in Jiangsu, mindfulness interventions could be considered. The study indicated that adolescents with higher levels of mindfulness exhibited a greater capacity for establishing a safe inner environment conducive to trauma recovery (Liu et al., 2022). Therefore, implementing mindfulness interventions tailored to the unique needs of adolescents facing disruptions, such as those caused by the pandemic, may offer a valuable avenue for promoting sleep health and overall well-being in this population.

Conclusion 

In conclusion, we found youth who participated During-Pandemic had less sleepiness, longer sleep duration, and a higher number of awakenings compared to students who participated Pre-Pandemic. The unique decrease of sleep quality during the pandemic as sleep duration increased opens the door for future research questions. This is important as it adds to our current knowledge of adolescent sleep in the face of intense stressors. Overall, these results expand our current knowledge of youth sleep characteristics during the height of the pandemic. The results support the need for more research on what tools could help adolescents sleep health during stressful periods. Future studies should examine the interactive effect of sleep duration and sleep timing on sleepiness to inform intervention efforts aiming to improve sleep outcomes in adolescents. Particularly, informing our understanding of early adolescent sleep behaviors in the face of immense stressors like COVID-19 or natural disasters. 

 

Acknowledgement

The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R305A180074 to the University of Alabama at Birmingham. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.