Abstract
Objective:
To determine the micro-longitudinal effects of duration and timing of screen-based activities on sleep within and between adolescents.
Study Design:
Daily survey and actigraphy data from the age 15 wave of the Future of Families and Child Wellbeing Study were analyzed using multilevel modeling. 475 adolescents provided three or more days of valid daily survey and nighttime sleep data.
Results:
Within-person results showed that on days when adolescents played video games more than their daytime average±SE (1.3±1.2 hours), sleep onset (6±2 min, p<0.01) and midpoint (4±2 min, p<0.02) were delayed for each additional hour of gaming. Between-person results showed that for each hour adolescents used screens to communicate with friends across the day, sleep onset was later (11±3 min, p<0.01), sleep midpoint was later (8±3 min, p<0.01), and sleep duration was shorter (−5±2 min, p<0.03). Adolescents who used screens to communicate with friends or play video games in the hour before bed had later sleep onset (30±14 min, p<0.03) and midpoint (25±13 min, p<0.05).
Conclusions and Relevance:
Among adolescents, passive screen use such as browsing the Internet or watching videos may not affect sleep timing or duration, but limiting interactive screen-based activities could protect adolescent sleep health and well-being.
INTRODUCTION
Adolescence is a critical developmental period for health behaviors including sleep. Both the National Sleep Foundation1 and the American Academy of Sleep Medicine2 recommend that adolescents regularly sleep 8 to 10 hours per day for optimal physical and mental well-being. Adolescents without adequate sleep are at increased risk of obesity3 as well as impaired cognition, emotion regulation, and mental health4. There are several broad reasons why adolescents may not get enough sleep for optimal daily functioning, or even the minimum recommended amount. First, physiological changes that occur during puberty lead to a circadian phase delay and a preference for eveningness5. Second, there are structural constraints for most adolescents whereby early morning schedules truncate morning sleep6. Finally, adolescence is marked by a growth in autonomy linked with changes in social, academic, and extracurricular activities. Some of these activities involve the use of screen-based digital media7.
Approximately 95% of adolescents in the United States engage in screen-based activities daily8,9. Adolescents with smartphones are also more likely to take the device into the bedroom and use the device right before bed10. Literature reviews consistently note that general screen use among adolescents is associated with both later sleep onset and shorter sleep duration11–14. This relationship may be due to the light emitted by the device15 or due to other factors such as time displacement, notifications, and physiological or psychological arousal while interacting with the screen14,16. Adolescents often use smartphones and other personal devices to access social media, play video games, and watch television and videos8,9, all of which have been associated with worse sleep11–14. However, while all such screen-based activities are vehicles for entertainment, interactive activities may be more stimulating than passive activities17.
Most screen-based activities fall along a continuum of interactivity, from necessarily interactive engagement that requires active input (e.g., video games) to exclusively passive entertainment (e.g., watching videos). The degree of interactivity versus passivity may be the greatest predictor of whether a screen-based activity affects sleep. Activities that are inherently more interactive, such as video games and social media, are more likely to affect sleep than passive activities10. Adolescents who play video games are likely to have difficulties falling asleep17–19, shorter sleep duration17,20–22, worse sleep quality20, greater daytime sleepiness21, and greater variability in sleep timing and sleep duration23. Adolescents who use social media or messenger applications often have greater sleep onset latency17, later sleep onset19, shorter sleep duration16,17,20,21,24,25, worse sleep quality20,26,27, and more daytime sleepiness during the next day21. Texting has also been associated with later bedtimes, shorter sleep duration, and greater daytime sleepiness28–30. Conversely, while video-watching has been associated with difficulties in falling asleep17,19, shorter sleep duration17,20,22,31, worse sleep quality17,20,26, and greater sleep variability23, many other studies have found either no association with sleep14,21,24,30,32,33 or that video-watching may instead improve sleep health34,35.
Many studies demonstrate a connection between greater daytime and evening screen use and worse sleep health14. Additionally, concurrent engagement in multiple screen-based activities may exacerbate the effect of screen use on sleep36. However, most studies are cross-sectional, examining associations between adolescents rather than within the same adolescent across multiple nights, and therefore do not establish temporal precendence14,37. Moreover, most studies do not objectively measure sleep outcomes, with recent exceptions21,23,24,28,30,36,38. Fewer studies have evaluated how passive and interactive screen-based activities differently affect actigraphic sleep timing and sleep duration21,23 both within and between adolescents.
This study aims to expand our current understanding of how different screen-based activities predict sleep health. Specifically, we used multilevel modeling to investigate the within- and between-person micro-longitudinal associations of daytime and evening activities with subsequent sleep timing and duration using repeated measures of both self-reported screen time and actigraphic sleep over one week. We hypothesized that the time spent engaging in interactive screen-based activities (i.e., video games, social media), but not more passive activities (i.e., visiting websites or shopping on the Internet, watching television, videos, and movies), throughout the day would be associated with later sleep timing (sleep onset, midpoint, and offset) and shorter sleep duration17,23,28,30. We also hypothesized that using interactive screen-based activities, but not passive screen-based activities, in the hour before bed would be associated with later sleep timing and shorter sleep duration22.
METHODS
Participants
We analyzed actigraphic and survey data from the Future of Families and Child Wellbeing Study (FFCWS), a longitudinal birth cohort from 20 large United States cities39. Nonmarital births were oversampled, which resulted in a larger proportion of families of low socioeconomic status and minority mothers. Written consent was provided by mothers upon recruitment (N=4898; see 39, for more details). Mothers responded to surveys at the time of their child’s birth and when the child was ages 1, 3, 5, 9, and 15 years. At age 15 years (between February 2014 and February 2016), a randomly selected subset of the participants (N=1,049) assented (and parents consented) to participate in the actigraphy substudy, for which they completed daily surveys and wore wrist actigraphy devices for one week. Of these adolescents, 475 provided three or more days/nights of valid daily survey and that night’s sleep data. Princeton University provided IRB approval for the overall study, and Stony Brook University provided IRB approval for the actigraphy substudy.
Measures
Daytime screen-based activity predictors
Adolescent daytime screen-based activities were assessed with daily surveys. Adolescents were asked about how many hours that day they spent (1) communicating with friends by email, instant messaging, texting on the phone, or through social media sites, such as Facebook or Twitter, (2) playing games on the computer, television, or a handheld device, (3) visiting websites or shopping on the Internet, and (4) watching television, videos, and movies on any device. Adolescents endorsed 0, 1, 2, 3, 4, or “5 or more hours” in response to how long they engaged in each activity. Responses were treated as continuous integer hours40. We calculated person mean values to examine interindividual variability as well as person-mean-centered values to examine intraindividual variability.
Screen-based activity before bed predictors
Adolescent screen-based activities before bed were also assessed with daily surveys. Adolescents were asked regarding the previous night “Did you do any of the following activities in the hour before you went to bed?”. Adolescents endorsed either “yes” or “no” to whether they (1) talked, texted, or played games on a phone, computer, or tablet, and (2) watched television or movies. We calculated an adolescent’s mean likelihood to endorse either “yes” or “no” to examine their interindividual variability as well as person-mean-centered values to examine intraindividual variability. Both person means and person-mean-centered responses were treated as continuous values.
Actigraphic sleep outcomes
Sleep outcomes were derived from wrist-worn actigraphy data (Actiwatch Spectrum; Philips-Respironics, Murrysville, PA). Participants were asked to wear the actigraphy devices on their non-dominant hand for one week. Data were downloaded using Philips Actiware software (Version 6.0.4, Philips Respironics, 2017), and at least two scorers independently determined the validity of each day and set sleep intervals using a validated algorithm41. The scorers adjudicated any discrepancies among the number of valid days, cut-points, sleep intervals, and any differences of sleep duration and wake after sleep onset that were greater than 15 minutes. Trained scorers used low wrist activity levels as well as detected light levels to determine sleep intervals. Compared to the gold standard of polysomnography, accuracy of actigraphy is over 95% on a 30-second epoch basis41. Sleep measures were calculated for the main nighttime sleep interval (i.e., the longest nighttime duration). Shorter sleep periods more than one hour before or after that main nighttime sleep interval were not included.
Sleep onset was the start of the actigraphic sleep period. Sleep midpoint was the middle of the participant’s nighttime sleep period, which was calculated by dividing nighttime sleep duration in half and adding it to sleep onset. Sleep offset was the end of the actigraphic sleep period. Sleep duration was calculated as the length of the nighttime actigraphic sleep period and was reported in minutes. All actigraphic sleep variables were treated as continuous outcomes.
Covariates
School nights
Adolescents were asked daily whether they went to school that day (yes/no). If adolescents endorsed “yes”, then the previous evening was considered to be a school night, with endorsements of “no” considered to be free nights.
Bedtime routines
Adolescents were asked whether they had a regular bedtime routine, which they rated as “not true”, “sometimes true”, or “often true”. Bedtime routines were treated as a categorical variable.
Sociodemographic factors
Sociodemographic factors included age in years (centered at age 15) when adolescents completed the daily surveys and wore the actigraphy device, birth sex (male or female), self-identified race (non-Hispanic White, non-Hispanic Black/African American, Hispanic/Latino, or Multiracial/Other; missing data were coded as Multiracial/Other), primary caregiver’s education level (less than high school, high school or equivalent, some college or technical school, or college or graduate school), household income-to-poverty ratio (<49%, 50–99%, 100–199%, 200–299%, or >300%), and family structure at home (mother and father, mother and new partner, mother only, or other type of family structure). Age was a continuous variable, with all other sociodemographic variables treated as categorical.
Statistical analysis
The MIXED procedure (SAS 9.4) was used to examine the within- and between-person associations between daytime (number of hours) and evening (yes/no) screen-based activities and subsequent sleep timing and duration. Model 1 was an empty, intercept-only model to examine the initial fit of all sleep outcomes within a multilevel model. Model 2 adjusted for covariates (school nights, bedtime routines, and sociodemographic factors). Model 3 tested the micro-longitudinal association of each individual screen-based activity with sleep, separated into within- and between-person components, and adjusted for covariates. The within-person component consisted of person-mean-centered responses, and the between-person component consisted of person mean responses. Models using complete case analysis were estimated using restricted maximum likelihood. The intercept was treated as a random variable nested within each adolescent using an unstructured covariance matrix. The between-within option was used to divide the residual degrees of freedom into between- and within-person components. Within-person reliability of measurements was assessed through intraclass correlation coefficients (ICCs) and Cronbach’s alpha, and normality of the residual distribution was examined (skewness<|3|; kurtosis<|10|).
In addition to the primary analyses, we conducted sensitivity analyses to assess the differences between the initial FFCWS sample, the sleep substudy sample, and the final analyzed sample in this manuscript. Compared to the initial sample, we found that participants in the sleep substudy sample were more likely to be female, live with their mother, and less likely to be multiracial/other race/ethnicity. Compared to the initial or substudy samples, participants from marginalized groups were less likely to be included in the final analyzed sample, including adolescents who were not White, adolescents whose mothers did not complete high school, adolescents whose family’s household income-to-poverty ratio was lower, and adolescents with single parents. We also retested Model 3 while adjusting for the effect of summertime data collection on sleep outcomes.
RESULTS
Participant demographics are shown in Table 1. Adolescents in the analytical sample (N=475; M=15.4±0.5 years old) provided an average of 4.8 days of daily survey and sleep data, with a median of 5 days. 56% of data were collected on free nights when adolescents did not have school the next day. Mean sleep onset (clocktime±SD) of the sample was 00:26±01:42, sleep midpoint was 04:21±01:41, sleep offset was 08:20±01:44, and mean sleep duration was 7.8±1.1 hours (see Table A1 for separation by school vs. free nights). Adolescents reported that they used screens an average of 2.0±1.5 hours per day communicating with friends by email, instant messaging, texting on the phone, or through social media; 1.3±1.2 hours per day playing games on the computer, television, or a handheld device; 0.7±0.9 hours per day visiting websites or shopping on the Internet; and 1.7±1.1 hours per day watching television, videos, and movies on any device. Adolescents also endorsed that in the hour before bed they talked, texted, or played games on a phone, computer, or tablet 77±32% of the total nights and watched television or movies 69±34% of the total nights.
Table 1.
Participant demographics (N=475)
| N | % | |
|---|---|---|
| Birth sex | ||
| Female | 253 | 53.3% |
| Male | 222 | 46.7% |
| Race | ||
| Black/African American | 191 | 40.2% |
| Hispanic/Latino | 118 | 24.8% |
| White | 94 | 19.8% |
| Multiracial/Other | 72 | 15.2% |
| Household income-to-poverty ratio | ||
| <49% | 42 | 8.8% |
| 50–99% | 85 | 17.9% |
| 100–199% | 123 | 25.9% |
| 200–299% | 77 | 16.2% |
| >300% | 148 | 31.2% |
| Caregiver’s education | ||
| Less than high school | 66 | 13.9% |
| High school or equivalent | 82 | 17.3% |
| Some college or technical school | 223 | 47.0% |
| College or graduate school | 104 | 21.9% |
| Family structure | ||
| Mother + father | 160 | 33.7% |
| Mother + new partner | 116 | 24.4% |
| Mother only | 165 | 34.7% |
| Father or other | 34 | 7.1% |
According to ICCs for the screen-based activity predictors, differences across adolescents accounted for approximately 66% of the variability in communicating with friends, suggesting that 34% of the variability was observed within adolescents across days. According to Cronbach’s alpha, the reliability of assessing communicating with friends for a median of 5 days per adolescent was 0.91. Differences across adolescents accounted for approximately 45–54% of variability in playing video games, browsing the Internet, and watching television, videos, and movies on any device (reliability=0.81–0.86). ICCs and Cronbach’s alpha for the sleep outcomes indicated that differences across adolescents accounted for approximately 52–54% of the variability in sleep onset and midpoint (reliability=0.85–0.86), 39% of variability in sleep offset (reliability=0.76), and 16% of variability in sleep duration (reliability=0.49). Residuals were normally distributed.
Covariates
Models were adjusted for all covariates (Model 2), including school nights, bedtime routines, and sociodemographic factors, shown in Table 2. School nights were associated with earlier sleep onset (−1.2 hours), midpoint (−1.9 hours), and offset (−2.6 hours) and shorter sleep duration (−1.3 hours), p<0.001. Male adolescents had sleep onsets and midpoints that were 20–32 minutes later, p≤0.012, and slept 24 fewer minutes, p<0.001, than female adolescents. Compared to White adolescents, Black/African American adolescents slept 18 fewer minutes, p=0.039, whereas Hispanic/Latino adolescents had a later sleep midpoint and offset by 27 minutes, p≤0.041. Adolescents whose household income-to-poverty ratio was <49% had later sleep onset (1.0 hours), later sleep midpoint (0.8 hours), and shorter sleep duration (−0.5 hours), p≤0.028. Finally, the sleep offset of adolescents who lived with only their mother was 24 minutes later than adolescents living with both parents, p=0.018.
Table 2.
Multilevel models of associations between covariates and sleep outcomes (min)
| Sleep Onset (b±SE min)a |
Sleep Midpoint (b±SE min)a |
Sleep Offset (b±SE min)a |
Sleep Duration (b±SE min) |
|
|---|---|---|---|---|
| Intercept | 00:14±00:20 | 04:34±00:19*** | 08:56±00:18*** | 517.3±13.9*** |
| School night (ref: Free night) | −70.1±4.9*** | −111.2±4.3*** | −157.4±5.2*** | −77.5±4.8*** |
| Bedtime routine (ref: Not true) | ||||
| Sometimes true | 6.5±12.7 | 1.0±11.6 | 0.2±11.2 | −10±8.7 |
| Often true | −10.1±11.3 | −8.8±10.4 | −12.9±10.0 | 2.7±7.8 |
| Age (15-years-centered) | −10.9±8.4 | −6.8±7.7 | 5.8±7.3 | 8.3±5.7 |
| Birth sex (ref: Female) | 32.1±8.8*** | 20.3±8.0* | 15±7.7† | −24.1±6.0*** |
| Race (ref: White) | ||||
| Black/African American | 20.6±13.0 | 11.4±11.9 | 4.4±11.4 | −18.4±8.9* |
| Hispanic/Latino | 18.1±14.2 | 26.7±13.0* | 27.2±12.4* | 17.9±9.6† |
| Multiracial/Other | 4.8±15.2 | 4.9±14.0 | 2.1±13.4 | 0.2±10.4 |
| Household income-to-poverty ratio (ref: >300%) | ||||
| <49% | 60.2±18.8** | 45.7±17.2** | 23.0±16.5 | −28.3±12.9* |
| 50–99% | −2.8±14.7 | −7.1±13.4 | −17.8±12.9 | −9.0±10 |
| 100–199% | 17.7±13.0 | 13.9±11.9 | 4.6±11.4 | −8.0±8.9 |
| 200–299% | −9.0±14.2 | −13.7±13.0 | 1.4±12.5 | −11.1±9.7 |
| Caregiver’s education (ref: <High school) | ||||
| High school or equivalent | 15.4±16.1 | 13.1±14.8 | 13.4±14.1 | −4.1±10.9 |
| Some college or technical school | 11.4±14.5 | 8.9±13.3 | 6.9±12.7 | −3.7±9.9 |
| College or graduate school | 7.7±17.6 | 9.1±16.1 | 4.6±15.4 | 3.8±12.0 |
| Family structure (ref: Mother + father) | ||||
| Mother + new partner | −4.1±11.9 | 3.9±11.0 | 11.0±10.5 | 15.5±8.1† |
| Mother only | 9.6±11.4 | 12.2±10.5 | 23.7±10.0* | 5.2±7.8 |
| Father or other | 20.3±18.1 | 22.5±16.6 | 19.2±15.9 | 4.7±12.3 |
| Participants | 475 | |||
| Observations | 2259 | |||
Note.
p<.001
p<.01
p<.05
p<.1
Intercept units are clocktime±SE
Interactive screen-based activities
The within- and between-person associations of each screen-based activity with each sleep outcome are shown in Tables 3 and 4, adjusted for all covariates.
Table 3.
Adjusted multilevel models of associations between daytime activity (hours) and sleep outcomes (min)
| Sleep Onset (b±SE min) |
Sleep Midpoint (b±SE min) |
Sleep Offset (b±SE min) |
Sleep Duration (b±SE min) |
|
|---|---|---|---|---|
| Communicating with friends a | ||||
| Person-mean-centered | −0.1±2.0 | −1.3±1.8 | −2.3±2.3 | −1.9±2.2 |
| Person mean | 10.5±3.0*** | 8.2±2.8** | 5.3±2.7† | −4.8±2.1* |
| Playing video games b | ||||
| Person-mean-centered | 5.5±2.0** | 4.2±1.7* | −0.4±2.3 | −2.3±2.2 |
| Person mean | 9.3±3.9* | 6.6±3.6† | 4.9±3.4 | −5.3±2.7* |
| Browsing the Internet c | ||||
| Person-mean-centered | 4.6±2.5† | 3.4±2.2 | −0.9±2.8 | −2.4±2.8 |
| Person mean | 9.1±4.9† | 7.2±4.5 | 8.3±4.3† | −4.1±3.3 |
| Watching videos d | ||||
| Person-mean-centered | 1.7±1.9 | 2.3±1.8 | 3.6±2.1† | 0.3±2.1 |
| Person mean | 1.9±4.1 | 3.6±4.4 | 1.9±3.6 | −1.6±2.8 |
| Participants | 475 | |||
| Observations | 2259 | |||
Note.
p<.001
p<.01
p<.05
p<.1
Communicating with friends by email, instant messaging, texting on your phone, or through social media sites, such as Facebook or Twitter.
Playing games on the computer, TV, or a handheld device.
Visiting websites or shopping on the Internet.
Watching TV, videos, and movies (on any device).
All models controlled for school nights, bedtime routines, age, birth sex, race, household income-to-poverty ratio, caregiver’s education level, and family structure.
Table 4.
Adjusted multilevel models of associations between activity before bed (yes/no) and sleep outcomes (min)
| Sleep Onset (b±SE min) |
Sleep Midpoint (b±SE min) |
Sleep Offset (b±SE min) |
Sleep Duration (b±SE min) |
|
|---|---|---|---|---|
| Communicating & playing video games a | ||||
| Person-mean-centered | 5.4±7.1 | 5.1±6.1 | 4.0±7.9 | −1.0±7.7 |
| Person mean | 30.4±13.7* | 25.2±12.6* | 21.4±12.0† | −10.1±9.4 |
| Watching videos b | ||||
| Person-mean-centered | 7.7±6.0 | 6.2±5.2 | 4.9±6.8 | −3.4±6.6 |
| Person mean | −7.0±13.1 | −3.5±12.0 | 9.6±11.5 | 7.7±8.9 |
| Participants | 475 | |||
| Observations | 2259 | |||
Note.
p<.001
p<.01
p<.05
p<.1
Talked, texted, or played games on a phone, computer, or tablet.
Watched television or movies.
All models controlled for school nights, bedtime routines, age, birth sex, race, household income-to-poverty ratio, caregiver’s education level, and family structure.
Daytime screen-based activities
Within-person results showed that on days when adolescents played one more hour of video games than their usual, sleep onset was delayed by 6±2 mins (p=0.007) and sleep midpoint was delayed by 4±2 mins (p=0.015) that night. There were no other within-person associations between screen-based activities during the day and sleep that night. Between-person results are shown in Figure 1. Every hour throughout the day that adolescents used screens to communicate with friends was associated with delayed sleep onset (11±3 mins, p<0.001), delayed sleep midpoint (8±3 mins, p=0.004), and shorter sleep duration (−5±2 mins, p=0.023). Similarly, every hour adolescents used screens to play video games was associated with delayed sleep onset (9±4 mins, p=0.017) and shorter sleep duration (−5±3 mins, p=0.046), although these between-person effects were nonsignificant when adjusting for changes in sleep during the summer.
Figure 1.
Between-person associations of communicating with friends (person mean, hours) with (A) sleep onset, (B) sleep midpoint, and (C) sleep duration, and between-person associations of playing video games (person mean, hours) with (D) sleep onset, (E) sleep midpoint, and (F) sleep duration. All models controlled for school nights, bedtime routines, age, birth sex, race, caregiver’s education level, household income-to-poverty ratio, and family structure at home. Shaded bands show 95% confidence interval of sleep outcome predicted by each daytime screen-based activity.
Screen-based activities before bed
There were no significant within-person associations between engaging in activities in the hour before bed and subsequent sleep outcomes. Between-person results revealed that adolescents who talked, texted, or played games on a device in the hour before bed had later sleep onset (30±14 mins, p=0.027) and sleep midpoint (25±13 mins, p=0.046) versus those who did not.
Passive screen-based activities
We found no significant associations between passive screen-based activities and subsequent sleep. Browsing the Internet and watching television, videos, and movies during the day were not significantly associated with sleep that night. Likewise, watching television or movies in the hour before bed was not associated with sleep timing or duration.
DISCUSSION
We investigated how several screen-based activities affected sleep timing and duration, both within and between adolescents. We found that daytime and evening screen-based activities that were more interactive significantly affected sleep, compared to passive screen-based activities. Within adolescents, playing video games during the day was the only screen-based activity to delay sleep timing that night (about five minutes for every hour adolescents spent playing more than their own average). Adolescents who used screens to communicate with friends had later sleep onset and midpoint regardless of whether the activity was during the day or right before bed. Additionally, more hours spent communicating with friends during the day was associated with shorter sleep duration. By comparison, passive activities requiring minimal interaction during use, such as browsing the Internet or watching videos, were not associated with subsequent sleep. Adolescents may only need to limit interactive screen-based activities to protect sleep health and well-being.
In general, these findings largely support prior studies that found social media and video games to be associated with worse sleep as measured by actigraphy in adolescents21,23,24,28,30,36,38. For example, one study found that time spent on social media and video games at night was associated with shorter sleep duration21, and other studies found that social media use24 and texting30 were associated with decreased sleep duration among adolescents. These studies, however, did not examine sleep timing outcomes. By comparison, our analyses examined multiple sleep timing outcomes (i.e., sleep onset, midpoint, and offset). Furthermore, we investigated both within- and between-person associations that showed daily relationships between screen-based activities and subsequent sleep, which provide novel contributions to the literature.
Limitations
We did not find any associations between screen-based activities in the hour before bed and subsequent sleep within adolescents. This null finding may be due to several limitations related to how survey items assessed each activity before bed. One, the survey items asked only about engagement in each activity during the one hour right before bed. Two, responses to each survey item were binary, which precluded examination of whether the time spent on each screen-based activity in the hour before bed predicted subsequent sleep. Three, communicating with friends and playing video games were included in a single survey item. In future studies, activities may be asked as separate survey items to independently assess the impact of different screen-based activities on sleep.
There were some limitations that should be addressed in future studies. First, adolescents self-reported the number of integer hours they spent throughout the day engaging in each screen-based activity and they endorsed whether they used the device in the hour before bed. Responses may not accurately reflect the amount of time spent in each screen-based activity. Future studies should measure screen time objectively. Second, the specific content or type of media within each screen-based activity was not assessed. For example, video games are inherently interactive7,11 and may require regular input to control gameplay. However, there are many different types of video games. Some games are puzzles that only need decisional input, while other games are fast-paced shooters that necessitate constant vigilance to make progress. Future studies should ask about the content of media consumed during screen-based activities with different degrees of vigilance or rapid-response violence required. Third, adolescents responded to survey items that did not ask about which devices were used, thereby conflating whether the screen-based activity was performed on a smartphone, television, or both simultaneously. The type of device used could have implications as the amount of light exposure during typical use varies among screens. For example, smartphones are typically smaller devices but are often held closer to one’s face, whereas televisions are typically much larger yet are viewed from a distance. Fourth, engaging in multiple screen-based activities simultaneously was not assessed. Future studies should ask about concurrent screen use, which has been associated with shorter sleep duration36.
There are additional limitations of the present study that may preclude generalizability. Data were collected from adolescents between 2014 and 2016 and may not accurately represent how modern adolescents spend time on screen-based activities; results of this study may not reflect the current effects of screen-based activities on sleep. The results may also not be representative of adolescents from marginalized groups or disadvantaged households, who were more likely to be excluded from the final analyzed sample. Additionally, the direction of causality of between-person findings may be reversed, with worse sleep outcomes modifying how adolescents spend time on each screen-based activity. Finally, we did not adjust for potentially confounding health status information. Prior studies using these data have found associations of sleep with physical activity42, depressive symptomatology16, mood43,44, and neighborhood safety45, but we opted not to adjust for all of these factors given the possibility of overadjustment and multicollinearity.
Conclusion
This study substantially contributes to the literature by examining the separate within- and between-person effects of duration and timing of different screen-based activities on actigraphic sleep, with a focus on sleep timing and duration. For every hour adolescents spent on interactive activities such as social media or video games, sleep timing was about 10 minutes later and sleep duration was about 5 minutes shorter, and engagement in these activities right before bed was associated with 25–30 minutes later sleep timing between adolescents. Passive screen-based activities were not associated with sleep timing or sleep duration. Reducing the time spent on recreational screen-based activities that are inherently interactive, particularly close to bedtime, or advancing activities to be earlier in the evening may protect adolescent sleep health and well-being46.
Supplementary Material
IMPLICATIONS AND CONTRIBUTION.
In this cohort study, adolescents who engaged in interactive screen-based activities throughout the day and in the hour before bed exhibited later sleep timing and shorter sleep duration. Passive screen use among adolescents may not affect sleep timing or duration, but limiting interactive screen-based activities could protect adolescent sleep health and well-being.
Acknowledgements
We thank the families who participated in the study, the FFCWS team, and the Actigraphy Data Coordinating Center at Penn State.
Statement of Financial Support:
The study was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health under award numbers R01HD073352 (to LH), R01HD36916, R01HD39135, and R01HD40421, as well as a consortium of private foundations. David A. Reichenberger was supported by the National Aeronautics and Space Administration (80NSSC20M0097) issued through the PA Space Grant Consortium and by the Prevention and Methodology Training Program (T32 DA017629) with funding from the National Institute on Drug Abuse.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures of Interest: Orfeu M. Buxton received subcontract grants to Penn State from Proactive Life LLC (formerly Mobile Sleep Technologies) doing business as SleepSpace (NSF/STTR #1622766, NIH/NIA SBIR R43-AG056250, R44-AG056250), received honoraria/travel support for lectures from Boston University, Boston College, Tufts School of Dental Medicine, New York University, University of Miami, University of Utah, University of South Florida, University of Arizona, Eric H. Angle Society of Orthodontists, Spencer Study Club, Harvard Chan School of Public Health and Allstate, and receives an honorarium for his role as the Editor-in-Chief of Sleep Health. Lauren Hale previously received an honorarium from the National Sleep Foundation for her role as Editor-in-Chief of the journal Sleep Health, in addition to honoraria for lectures/consultation from University of Miami, University of Utah, Baylor College of Medicine, Auburn University, Columbia University, Idorsia, and the National Sleep Foundation. The remaining authors have no financial relationships relevant to this article to disclose., and receives an honorarium from the National Sleep Foundation for his role as the Editor-in-Chief of Sleep Health. Lauren Hale previously received an honorarium from the National Sleep Foundation for her role as Editor-in-Chief of the journal Sleep Health, in addition to honoraria for lectures/consultation from University of Miami, University of Utah, Baylor College of Medicine, Auburn University, Columbia University, Idorsia, and the National Sleep Foundation. The remaining authors have no financial relationships relevant to this article to disclose.
REFERENCES
- 1.Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation’s updated sleep duration recommendations: final report. Sleep Health J Natl Sleep Found 2015;1(4):233–243. doi: 10.1016/j.sleh.2015.10.004 [DOI] [PubMed] [Google Scholar]
- 2.Paruthi S, Brooks LJ, D’Ambrosio C, et al. Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement of the American Academy of Sleep Medicine. J Clin Sleep Med JCSM Off Publ Am Acad Sleep Med 2016;12(6):785–786. doi: 10.5664/jcsm.5866 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ogilvie RP, Patel SR. The Epidemiology of Sleep and Obesity. Sleep Health 2017;3(5):383–388. doi: 10.1016/j.sleh.2017.07.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tarokh L, Saletin JM, Carskadon MA. Sleep in adolescence: Physiology, cognition and mental health. Neurosci Biobehav Rev 2016;70:182–188. doi: 10.1016/j.neubiorev.2016.08.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Saxvig IW, Evanger LN, Pallesen S, et al. Circadian typology and implications for adolescent sleep health. Results from a large, cross-sectional, school-based study. Sleep Med Published online April 22, 2021. doi: 10.1016/j.sleep.2021.04.020 [DOI] [PubMed] [Google Scholar]
- 6.Nahmod NG, Lee S, Master L, Chang AM, Hale L, Buxton OM. Later high school start times associated with longer actigraphic sleep duration in adolescents. Sleep 2019;42(2). doi: 10.1093/sleep/zsy212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lissak G Adverse physiological and psychological effects of screen time on children and adolescents: Literature review and case study. Environ Res 2018;164:149–157. doi: 10.1016/j.envres.2018.01.015 [DOI] [PubMed] [Google Scholar]
- 8.Rideout V, Robb MB. The Common Sense census: Media use by tweens and teens, 2019. Common Sense Media Published online 2019. [Google Scholar]
- 9.Vogels EA, Gelles-Watnick R, Massarat N. Teens, Social Media and Technology 2022 Published online 2022. [Google Scholar]
- 10.Gradisar M, Wolfson AR, Harvey AG, Hale L, Rosenberg R, Czeisler CA. The Sleep and Technology Use of Americans: Findings from the National Sleep Foundation’s 2011 Sleep in America Poll. J Clin Sleep Med 2013;09(12):1291–1299. doi: 10.5664/jcsm.3272 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bartel KA, Gradisar M, Williamson P. Protective and risk factors for adolescent sleep: A meta-analytic review. Sleep Med Rev 2015;21:72–85. doi: 10.1016/j.smrv.2014.08.002 [DOI] [PubMed] [Google Scholar]
- 12.Carter B, Rees P, Hale L, Bhattacharjee D, Paradkar MS. Association Between Portable Screen-Based Media Device Access or Use and Sleep Outcomes: A Systematic Review and Meta-analysis. JAMA Pediatr 2016;170(12):1202–1208. doi: 10.1001/jamapediatrics.2016.2341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hale L, Kirschen GW, LeBourgeois MK, et al. Youth screen media habits and sleep: sleep-friendly screen-behavior recommendations for clinicians, educators, and parents. Child Adolesc Psychiatr Clin N Am 2018;27(2):229–245. doi: 10.1016/j.chc.2017.11.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hale L, Guan S. Screen Time and Sleep among School-Aged Children and Adolescents: A Systematic Literature Review. Sleep Med Rev 2015;21:50–58. doi: 10.1016/j.smrv.2014.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bowler J, Bourke P. Facebook use and sleep quality: Light interacts with socially induced alertness. Br J Psychol 2019;110(3):519–529. doi: 10.1111/bjop.12351 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Li X, Buxton OM, Lee S, Chang AM, Berger LM, Hale L. Sleep mediates the association between adolescent screen time and depressive symptoms. Sleep Med 2019;57:51–60. doi: 10.1016/j.sleep.2019.01.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hisler G, Twenge JM, Krizan Z. Associations between screen time and short sleep duration among adolescents varies by media type: evidence from a cohort study. Sleep Med 2020;66:92–102. doi: 10.1016/j.sleep.2019.08.007 [DOI] [PubMed] [Google Scholar]
- 18.Higuchi S, Motohashi Y, Liu Y, Maeda A. Effects of playing a computer game using a bright display on presleep physiological variables, sleep latency, slow wave sleep and REM sleep. J Sleep Res 2005;14(3):267–273. doi: 10.1111/j.1365-2869.2005.00463.x [DOI] [PubMed] [Google Scholar]
- 19.Varghese NE, Santoro E, Lugo A, et al. The Role of Technology and Social Media Use in Sleep-Onset Difficulties Among Italian Adolescents: Cross-sectional Study. J Med Internet Res 2021;23(1):e20319. doi: 10.2196/20319 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Galland BC, de Wilde T, Taylor RW, Smith C. Sleep and pre-bedtime activities in New Zealand adolescents: differences by ethnicity. Sleep Health 2020;6(1):23–31. doi: 10.1016/j.sleh.2019.09.002 [DOI] [PubMed] [Google Scholar]
- 21.Perrault AA, Bayer L, Peuvrier M, et al. Reducing the use of screen electronic devices in the evening is associated with improved sleep and daytime vigilance in adolescents. Sleep 2019;42(9):zsz125. doi: 10.1093/sleep/zsz125 [DOI] [PubMed] [Google Scholar]
- 22.Pillion M, Gradisar M, Bartel K, Whittall H, Kahn M. What’s “app”-ning to adolescent sleep? Links between device, app use, and sleep outcomes. Sleep Med 2022;100:174–182. doi: 10.1016/j.sleep.2022.08.004 [DOI] [PubMed] [Google Scholar]
- 23.Hrafnkelsdottir SM, Brychta RJ, Rognvaldsdottir V, et al. Less screen time and more physical activity is associated with more stable sleep patterns among Icelandic adolescents. Sleep Health 2020;6(5):609–617. doi: 10.1016/j.sleh.2020.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lee PH, Tse ACY, Wu CST, Mak YW, Lee U. Temporal association between objectively measured smartphone usage, sleep quality and physical activity among Chinese adolescents and young adults. J Sleep Res Published online 2021:e13213. doi: 10.1111/jsr.13213 [DOI] [PubMed] [Google Scholar]
- 25.Sampasa-Kanyinga H, Colman I, Goldfield GS, Hamilton HA, Chaput JP. Sex differences in the relationship between social media use, short sleep duration, and body mass index among adolescents. Sleep Health 2020;6(5):601–608. doi: 10.1016/j.sleh.2020.01.017 [DOI] [PubMed] [Google Scholar]
- 26.Hoyt LT, Maslowsky J, Olson JS, Harvey AG, Deardorff J, Ozer EJ. Adolescent Sleep Barriers: Profiles within a Diverse Sample of Urban Youth. J Youth Adolesc 2018;47(10):2169–2180. doi: 10.1007/s10964-018-0829-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Woods HC, Scott H. #Sleepyteens: Social media use in adolescence is associated with poor sleep quality, anxiety, depression and low self-esteem. J Adolesc 2016;51:41–49. doi: 10.1016/j.adolescence.2016.05.008 [DOI] [PubMed] [Google Scholar]
- 28.Fobian AD, Avis K, Schwebel DC. The Impact of Media Use on Adolescent Sleep Efficiency. J Dev Behav Pediatr JDBP 2016;37(1):9–14. doi: 10.1097/DBP.0000000000000239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Garmy P, Ward TM. Sleep Habits and Nighttime Texting Among Adolescents. J Sch Nurs 2018;34(2):121–127. doi: 10.1177/1059840517704964 [DOI] [PubMed] [Google Scholar]
- 30.Tavernier R, Heissel JA, Sladek MR, Grant KE, Adam EK. Adolescents’ technology and face-to-face time use predict objective sleep outcomes. Sleep Health 2017;3(4):276–283. doi: 10.1016/j.sleh.2017.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Mireku MO, Barker MM, Mutz J, et al. Night-time screen-based media device use and adolescents’ sleep and health-related quality of life. Environ Int 2019;124:66–78. doi: 10.1016/j.envint.2018.11.069 [DOI] [PubMed] [Google Scholar]
- 32.Kubiszewski V, Fontaine R, Rusch E, Hazouard E. Association between electronic media use and sleep habits: an eight-day follow-up study. Int J Adolesc Youth 2014;19(3):395–407. doi: 10.1080/02673843.2012.751039 [DOI] [Google Scholar]
- 33.Lemola S, Perkinson-Gloor N, Brand S, Dewald-Kaufmann JF, Grob A. Adolescents’ Electronic Media Use at Night, Sleep Disturbance, and Depressive Symptoms in the Smartphone Age. J Youth Adolesc 2015;44(2):405–418. doi: 10.1007/s10964-014-0176-x [DOI] [PubMed] [Google Scholar]
- 34.Fuligni AJ, Hardway C. Daily Variation in Adolescents’ Sleep, Activities, and Psychological Well-Being. J Res Adolesc 2006;16(3):353–378. doi: 10.1111/j.1532-7795.2006.00498.x [DOI] [Google Scholar]
- 35.Twenge JM, Hisler GC, Krizan Z. Associations between screen time and sleep duration are primarily driven by portable electronic devices: evidence from a population-based study of U.S. children ages 0–17. Sleep Med 2019;56:211–218. doi: 10.1016/j.sleep.2018.11.009 [DOI] [PubMed] [Google Scholar]
- 36.Harrington DM, Ioannidou E, Davies MJ, et al. Concurrent screen use and cross-sectional association with lifestyle behaviours and psychosocial health in adolescent females. Acta Paediatr Published online March 3, 2021:apa.15806. doi: 10.1111/apa.15806 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Snyder CK, Chang AM. Mobile technology, sleep, and circadian disruption. In: Grandner MA, ed. Sleep and Health Academic Press; 2019:159–170. doi: 10.1016/B978-0-12-815373-4.00013-7 [DOI] [Google Scholar]
- 38.Kim Y, Umeda M, Lochbaum M, Sloan RA. Examining the day-to-day bidirectional associations between physical activity, sedentary behavior, screen time, and sleep health during school days in adolescents. PLOS ONE 2020;15(9):e0238721. doi: 10.1371/journal.pone.0238721 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Reichman NE, Teitler JO, Garfinkel I, McLanahan SS. Fragile Families: sample and design. Child Youth Serv Rev 2001;23(4–5):303–326. doi: 10.1016/S0190-7409(01)00141-4 [DOI] [Google Scholar]
- 40.Rhemtulla M, Brosseau-Liard PÉ, Savalei V. When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychol Methods 2012;17(3):354–373. doi: 10.1037/a0029315 [DOI] [PubMed] [Google Scholar]
- 41.Marino M, Li Y, Rueschman MN, et al. Measuring sleep: accuracy, sensitivity, and specificity of wrist actigraphy compared to polysomnography. Sleep 2013;36(11):1747–1755. doi: 10.5665/sleep.3142 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Master L, Nye RT, Lee S, et al. Bidirectional, Daily Temporal Associations between Sleep and Physical Activity in Adolescents. Sci Rep 2019;9(1):7732. doi: 10.1038/s41598-019-44059-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Mathew GM, Reichenberger DA, Master L, Buxton OM, Chang AM, Hale L. Actigraphic Sleep Variability is Associated With Lower Positive Mood in Adolescents. J Adolesc Health 2023. doi: 10.1016/j.jadohealth.2023.04.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Master L, Nahmod NG, Mathew GM, Hale L, Chang AM, Buxton OM. Why so slangry (sleepy and angry)? Shorter sleep duration and lower sleep efficiency predict worse next-day mood in adolescents. J Adolesc 2023;95(6):1140–1151. doi: 10.1002/jad.12182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Nahmod NG, Master L, McClintock HF, Hale L, Buxton OM. Neighborhood Disadvantage Is Associated with Lower Quality Sleep and More Variability in Sleep Duration among Urban Adolescents. J Urban Health 2022;99(1):102–115. doi: 10.1007/s11524-021-00570-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Twenge JM, Martin GN. Gender differences in associations between digital media use and psychological well-being: Evidence from three large datasets. J Adolesc 2020;79:91–102. doi: 10.1016/j.adolescence.2019.12.018 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.