[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
CONSORT Diagrams of Participant Flow in Phase 1 and Phase 2
CONSORT Diagrams of Participant Flow in Phase 1 and Phase 2

CBT indicates cognitive behavioral therapy.

Figure 2.
Protocol Diagram
Protocol Diagram

Timing of the different questionnaires (Adolescent Sleep-Wake Scale [ASWS], Cleveland Adolescent Sleepiness Questionnaire [CASQ], Child and Adolescent Sleep Checklist for parents [CASC-P], Credibility/Expectancy Questionnaire [CEQ], and Morningness-Eveningness Scale for Children [MESC]) given during the baseline and end assessments, as well as the ASWS and CASQ given during segments 1 through 4, which were given either over the telephone (phase 1) or in person (phase 2).

Figure 3.
Change in Total Sleep Time (TST)
Change in Total Sleep Time (TST)

Change in TST from ad libitum baseline through the 4 weeks of treatment in the light plus cognitive behavioral therapy (CBT) and sham plus CBT groups. By the end of treatment, those in the group receiving light plus CBT gained more than 40 minutes of extra sleep per night. Data were averaged within individuals by the week and then across individuals and are presented as mean (SD). Vertical lines indicate SD.

Table 1.  
Demographic and Clinical Information for Study Samples at Baseline
Demographic and Clinical Information for Study Samples at Baseline
Table 2.  
Phase 1 and Phase 2 Outcomes by Randomization Condition at Baseline (Week 1) and End-Treatment (Week 5)a
Phase 1 and Phase 2 Outcomes by Randomization Condition at Baseline (Week 1) and End-Treatment (Week 5)a
1.
Keyes  KM, Maslowsky  J, Hamilton  A, Schulenberg  J.  The great sleep recession: changes in sleep duration among US adolescents, 1991-2012.  Pediatrics. 2015;135(3):460-468. doi:10.1542/peds.2014-2707PubMedGoogle ScholarCrossref
2.
Carskadon  MA, Mindell  JA, Drake  C.  Sleep in America Poll. Washington, DC: The National Sleep Foundation; 2006.
3.
Calamaro  CJ, Mason  TB, Ratcliffe  SJ.  Adolescents living the 24/7 lifestyle: effects of caffeine and technology on sleep duration and daytime functioning.  Pediatrics. 2009;123(6):e1005-e1010. doi:10.1542/peds.2008-3641PubMedGoogle ScholarCrossref
4.
Wolfson  AR, Carskadon  MA.  Understanding adolescents’ sleep patterns and school performance: a critical appraisal.  Sleep Med Rev. 2003;7(6):491-506. doi:10.1016/S1087-0792(03)90003-7PubMedGoogle ScholarCrossref
5.
Wolfson  AR, Carskadon  MA.  Sleep schedules and daytime functioning in adolescents.  Child Dev. 1998;69(4):875-887. doi:10.1111/j.1467-8624.1998.tb06149.xPubMedGoogle ScholarCrossref
6.
Knutson  KL, Spiegel  K, Penev  P, Van Cauter  E.  The metabolic consequences of sleep deprivation.  Sleep Med Rev. 2007;11(3):163-178. doi:10.1016/j.smrv.2007.01.002PubMedGoogle ScholarCrossref
7.
Weiss  A, Xu  F, Storfer-Isser  A, Thomas  A, Ievers-Landis  CE, Redline  S.  The association of sleep duration with adolescents’ fat and carbohydrate consumption.  Sleep. 2010;33(9):1201-1209. doi:10.1093/sleep/33.9.1201PubMedGoogle ScholarCrossref
8.
McGlinchey  EL, Harvey  AG.  Risk behaviors and negative health outcomes for adolescents with late bedtimes.  J Youth Adolesc. 2015;44(2):478-488. doi:10.1007/s10964-014-0110-2PubMedGoogle ScholarCrossref
9.
Killgore  WD.  Effects of sleep deprivation and morningness-eveningness traits on risk-taking.  Psychol Rep. 2007;100(2):613-626. doi:10.2466/pr0.100.2.613-626PubMedGoogle ScholarCrossref
10.
Talbot  LS, McGlinchey  EL, Kaplan  KA, Dahl  RE, Harvey  AG.  Sleep deprivation in adolescents and adults: changes in affect.  Emotion. 2010;10(6):831-841. doi:10.1037/a0020138PubMedGoogle ScholarCrossref
11.
McGlinchey  EL, Talbot  LS, Chang  KH, Kaplan  KA, Dahl  RE, Harvey  AG.  The effect of sleep deprivation on vocal expression of emotion in adolescents and adults.  Sleep. 2011;34(9):1233-1241. doi:10.5665/SLEEP.1246PubMedGoogle ScholarCrossref
12.
Dagys  N, McGlinchey  EL, Talbot  LS, Kaplan  KA, Dahl  RE, Harvey  AG.  Double trouble? the effects of sleep deprivation and chronotype on adolescent affect.  J Child Psychol Psychiatry. 2012;53(6):660-667. doi:10.1111/j.1469-7610.2011.02502.xPubMedGoogle ScholarCrossref
13.
Iglowstein  I, Jenni  OG, Molinari  L, Largo  RH.  Sleep duration from infancy to adolescence: reference values and generational trends.  Pediatrics. 2003;111(2):302-307. doi:10.1542/peds.111.2.302PubMedGoogle ScholarCrossref
14.
Breslau  N, Roth  T, Rosenthal  L, Andreski  P.  Daytime sleepiness: an epidemiological study of young adults.  Am J Public Health. 1997;87(10):1649-1653. doi:10.2105/AJPH.87.10.1649PubMedGoogle ScholarCrossref
15.
Mercer  PW, Merritt  SL, Cowell  JM.  Differences in reported sleep need among adolescents.  J Adolesc Health. 1998;23(5):259-263. doi:10.1016/S1054-139X(98)00037-8PubMedGoogle ScholarCrossref
16.
Carskadon  MA, Acebo  C, Jenni  OG.  Regulation of adolescent sleep: implications for behavior.  Ann N Y Acad Sci. 2004;1021:276-291. doi:10.1196/annals.1308.032PubMedGoogle ScholarCrossref
17.
Van den Bulck  J.  Television viewing, computer game playing, and Internet use and self-reported time to bed and time out of bed in secondary-school children.  Sleep. 2004;27(1):101-104. doi:10.1093/sleep/27.1.101PubMedGoogle ScholarCrossref
18.
Moseley  L, Gradisar  M.  Evaluation of a school-based intervention for adolescent sleep problems.  Sleep. 2009;32(3):334-341. doi:10.1093/sleep/32.3.334PubMedGoogle ScholarCrossref
19.
Kira  G, Maddison  R, Hull  M, Blunden  S, Olds  T.  Sleep education improves the sleep duration of adolescents: a randomized controlled pilot study.  J Clin Sleep Med. 2014;10(7):787-792.PubMedGoogle Scholar
20.
Hansen  M, Janssen  I, Schiff  A, Zee  PC, Dubocovich  ML.  The impact of school daily schedule on adolescent sleep.  Pediatrics. 2005;115(6):1555-1561. doi:10.1542/peds.2004-1649PubMedGoogle ScholarCrossref
21.
Gradisar  M, Dohnt  H, Gardner  G,  et al.  A randomized controlled trial of cognitive-behavior therapy plus bright light therapy for adolescent delayed sleep phase disorder.  Sleep. 2011;34(12):1671-1680. doi:10.5665/sleep.1432PubMedGoogle ScholarCrossref
22.
Qaseem  A, Kansagara  D, Forciea  MA, Cooke  M, Denberg  TD; Clinical Guidelines Committee of the American College of Physicians.  Management of chronic insomnia disorder in adults: a clinical practice guideline from the American College of Physicians.  Ann Intern Med. 2016;165(2):125-133. doi:10.7326/M15-2175PubMedGoogle ScholarCrossref
23.
Flynn-Evans  EE, Shekleton  JA, Miller  B,  et al.  Circadian phase and phase angle disorders in primary insomnia.  Sleep. 2017;40(12):zsx163. doi:10.1093/sleep/zsx163PubMedGoogle ScholarCrossref
24.
Crowley  SJ, Eastman  CI.  Human adolescent phase response curves to bright white light.  J Biol Rhythms. 2017;32(4):334-344. doi:10.1177/0748730417713423PubMedGoogle ScholarCrossref
25.
Khalsa  SBS, Jewett  ME, Cajochen  C, Czeisler  CA.  A phase response curve to single bright light pulses in human subjects.  J Physiol. 2003;549(pt 3):945-952. doi:10.1113/jphysiol.2003.040477PubMedGoogle ScholarCrossref
26.
Zeitzer  JM, Ruby  NF, Fisicaro  RA, Heller  HC.  Response of the human circadian system to millisecond flashes of light.  PLoS One. 2011;6(7):e22078. doi:10.1371/journal.pone.0022078PubMedGoogle ScholarCrossref
27.
Berson  DM, Dunn  FA, Takao  M.  Phototransduction by retinal ganglion cells that set the circadian clock.  Science. 2002;295(5557):1070-1073. doi:10.1126/science.1067262PubMedGoogle ScholarCrossref
28.
Najjar  RP, Zeitzer  JM.  Temporal integration of light flashes by the human circadian system.  J Clin Invest. 2016;126(3):938-947. doi:10.1172/JCI82306PubMedGoogle ScholarCrossref
29.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
30.
Carskadon  MA, Vieira  C, Acebo  C.  Association between puberty and delayed phase preference.  Sleep. 1993;16(3):258-262. doi:10.1093/sleep/16.3.258PubMedGoogle ScholarCrossref
31.
LeBourgeois  MK, Giannotti  F, Cortesi  F, Wolfson  AR, Harsh  J.  The relationship between reported sleep quality and sleep hygiene in Italian and American adolescents.  Pediatrics. 2005;115(1)(suppl):257-265. doi:10.1542/peds.2004-0815HPubMedGoogle ScholarCrossref
32.
Spilsbury  JC, Drotar  D, Rosen  CL, Redline  S.  The Cleveland Adolescent Sleepiness Questionnaire: a new measure to assess excessive daytime sleepiness in adolescents.  J Clin Sleep Med. 2007;3(6):603-612.PubMedGoogle Scholar
33.
Borkovec  TD, Nau  SD.  Credibility of analog therapy rationales.  J Behav Ther Exp Psychiatry. 1972;3:257-260. doi:10.1016/0005-7916(72)90045-6Google ScholarCrossref
34.
Carney  CE, Buysse  DJ, Ancoli-Israel  S,  et al.  The consensus sleep diary: standardizing prospective sleep self-monitoring.  Sleep. 2012;35(2):287-302. doi:10.5665/sleep.1642PubMedGoogle ScholarCrossref
35.
Buysse  DJ, Thompson  W, Scott  J,  et al.  Daytime symptoms in primary insomnia: a prospective analysis using ecological momentary assessment.  Sleep Med. 2007;8(3):198-208. doi:10.1016/j.sleep.2006.10.006PubMedGoogle ScholarCrossref
36.
Oka  Y, Horiuchi  F, Tanigawa  T.  Development and validation of a child sleep screening questionnaire: Child and Adolescent Sleep Checklist (CASC).  Jap J Sleep Med. 2009;3(2):404-408.Google Scholar
37.
Gayes  LA, Steele  RG.  A meta-analysis of motivational interviewing interventions for pediatric health behavior change.  J Consult Clin Psychol. 2014;82(3):521-535. doi:10.1037/a0035917PubMedGoogle ScholarCrossref
38.
Emmons  RA.  Personal strivings, daily life events, and psychological and physical well-being.  J Pers. 1991;59(3):453-472. doi:10.1111/j.1467-6494.1991.tb00256.xPubMedGoogle ScholarCrossref
39.
Cohen  J.  Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
40.
Feingold  A.  A regression framework for effect size assessments in longitudinal modeling of group differences.  Rev Gen Psychol. 2013;17(1):111-121. doi:10.1037/a0030048PubMedGoogle ScholarCrossref
41.
Package ‘lmerTest.’ https://cran.r-project.org/web/packages/lmerTest/lmerTest.pdf. Published February 11, 2019. Accessed August 28, 2019.
42.
Zeitzer  JM, Fisicaro  RA, Ruby  NF, Heller  HC.  Millisecond flashes of light phase delay the human circadian clock during sleep.  J Biol Rhythms. 2014;29(5):370-376. doi:10.1177/0748730414546532PubMedGoogle ScholarCrossref
43.
Dijk  D-J, Czeisler  CA.  Paradoxical timing of the circadian rhythm of sleep propensity serves to consolidate sleep and wakefulness in humans.  Neurosci Lett. 1994;166(1):63-68. doi:10.1016/0304-3940(94)90841-9PubMedGoogle ScholarCrossref
44.
Crowley  SJ, Acebo  C, Carskadon  MA.  Sleep, circadian rhythms, and delayed phase in adolescence.  Sleep Med. 2007;8(6):602-612. doi:10.1016/j.sleep.2006.12.002PubMedGoogle ScholarCrossref
45.
Kaplan  KA, Hardas  PP, Redline  S, Zeitzer  JM; Sleep Heart Health Study Research Group.  Correlates of sleep quality in midlife and beyond: a machine learning analysis.  Sleep Med. 2017;34:162-167. doi:10.1016/j.sleep.2017.03.004PubMedGoogle ScholarCrossref
46.
Kaplan  KA, Hirshman  J, Hernandez  B,  et al; Osteoporotic Fractures in Men (MrOS), Study of Osteoporotic Fractures SOF Research Groups.  When a gold standard isn’t so golden: lack of prediction of subjective sleep quality from sleep polysomnography.  Biol Psychol. 2017;123:37-46. doi:10.1016/j.biopsycho.2016.11.010PubMedGoogle ScholarCrossref
47.
Short  MA, Gradisar  M, Lack  LC, Wright  H, Carskadon  MA.  The discrepancy between actigraphic and sleep diary measures of sleep in adolescents.  Sleep Med. 2012;13(4):378-384. doi:10.1016/j.sleep.2011.11.005PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 3,310
    Original Investigation
    Psychiatry
    September 25, 2019

    Effect of Light Flashes vs Sham Therapy During Sleep With Adjunct Cognitive Behavioral Therapy on Sleep Quality Among Adolescents: A Randomized Clinical Trial

    Author Affiliations
    • 1Stanford Center for Sleep Sciences and Medicine, Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, California
    • 2Palo Alto University, Palo Alto, California
    JAMA Netw Open. 2019;2(9):e1911944. doi:10.1001/jamanetworkopen.2019.11944
    Key Points español 中文 (chinese)

    Question  Can adjustment of circadian timing through light flash therapy during sleep increase total sleep time in teenagers?

    Findings  In this double-blind, placebo-controlled, randomized clinical trial of 102 adolescents, light flash therapy alone was ineffective. When delivered in combination with a cognitive behavioral therapy meant to encourage an earlier bedtime, the combination of light flash therapy and cognitive behavioral therapy significantly and stably increased total sleep time by nearly 45 minutes per night.

    Meaning  Combination light flash therapy and cognitive behavioral therapy is an effective tool to increase sleep time and combat sleep loss in teenagers.

    Abstract

    Importance  Owing to biological, behavioral, and societal factors, sleep duration in teenagers is often severely truncated, leading to pervasive sleep deprivation.

    Objective  To determine whether a novel intervention, using both light exposure during sleep and cognitive behavioral therapy (CBT), would increase total sleep time in teenagers by enabling them to go to sleep earlier than usual.

    Design, Setting, and Participants  This double-blind, placebo-controlled, randomized clinical trial, conducted between November 1, 2013, and May 31, 2016, among 102 adolescents enrolled full-time in grades 9 to 12, who expressed difficulty going to bed earlier and waking up early enough, was composed of 2 phases. In phase 1, participants were assigned to receive either 3 weeks of light or sham therapy and were asked to try to go to sleep earlier. In phase 2, participants received 4 brief CBT sessions in addition to a modified light or sham therapy. All analyses were performed on an intent-to-treat basis.

    Interventions  Light therapy consisted of receiving a 3-millisecond light flash every 20 seconds during the final 3 hours of sleep (phase 1) or final 2 hours of sleep (phase 2). Sham therapy used an identical device, but delivered 1 minute of light pulses (appearing in 20-second intervals, for a total of 3 pulses) per hour during the final 3 hours of sleep (phase 1) or 2 hours of sleep (phase 2). Light therapy occurred every night during the 4-week intervention. Cognitive behavioral therapy consisted of four 50-minute in-person sessions once per week.

    Main Outcomes and Measures  Primary outcome measures included diary-based sleep times, momentary ratings of evening sleepiness, and subjective measures of sleepiness and sleep quality.

    Results  Among the 102 participants (54 female [52.9%]; mean [SD] age, 15.6 [1.1] years), 72 were enrolled in phase 1 and 30 were enrolled in phase 2. Mixed-effects models revealed that light therapy alone was inadequate in changing the timing of sleep. However, compared with sham therapy plus CBT alone, light therapy plus CBT significantly moved sleep onset a mean (SD) of 50.1 (27.5) minutes earlier and increased nightly total sleep time by a mean (SD) of 43.3 (35.0) minutes. Light therapy plus CBT also resulted in a 7-fold greater increase in bedtime compliance than that observed among participants receiving sham plus CBT (mean [SD], 2.21 [3.91] vs 0.29 [0.76]), as well as a mean 0.55-point increase in subjective evening sleepiness as compared with a mean 0.48-point decrease in participants receiving sham plus CBT as measured on a 7-point sleepiness scale.

    Conclusions and Relevance  This study found that light exposure during sleep, in combination with a brief, motivation-focused CBT intervention, was able to consistently move bedtimes earlier and increase total sleep time in teenagers. This type of passive light intervention in teenagers may lead to novel therapeutic applications.

    Trial Registration  ClinicalTrials.gov identifier: NCT01406691

    Introduction

    Insufficient sleep in adolescence is increasingly recognized as a public health concern, with implications for academic performance, health, safety, and psychological well-being. Across the previous 2 decades, sleep durations in adolescents have been declining.1 Nearly half of US teenagers report regular insufficient sleep and excessive daytime sleepiness, particularly on school days.2,3 In one sample of 100 middle-school and high-school students, 80% slept fewer than 8 hours on school nights, with 44% of the students reporting difficulty staying awake during school and 85% reporting using caffeine to combat daytime fatigue.3 The implications of insufficient sleep are far reaching. Associations have been established between sleep duration and teenagers’ grade point average,4,5 physical health,6,7 substance use and risk-taking behaviors,8,9 and psychological well-being.10-12

    Multiple factors contribute to insufficient sleep in adolescence. Empirical data indicate that adolescents sleep less with age,13,14 even as the need for sleep remains constant or increases.15 This decrement in sleep is likely owing to a combination of later bedtimes, coupled with fixed, earlier school start times,2,16 which results in insufficient time for sleep. Later bedtimes are thought to be caused by a set of reinforcing influences: a naturally occurring delay in the timing of the circadian clock16 and increasing psychosocial activities at night (eg, work or homework, socialization, and use of electronics).3,17 These nocturnal activities can laten bedtime both directly and indirectly, owing to the concomitant evening light exposure that further delays the timing of the circadian clock.

    Previous interventions have attempted to address this latening of bedtime using a combination of early morning phototherapy (to move the timing of the circadian clock to an earlier clock hour) and behavioral or education-based interventions (to target psychosocial and environmental factors), with mixed results. Classroom-based interventions, designed to improve sleep through education18,19 or through administration of in-class morning light therapy,20 have not been successful in changing weekday sleep durations. One trial using light therapy and cognitive behavioral therapy (CBT) improved the sleep of adolescents meeting criteria for delayed sleep phase disorder,21 but its applicability to a broader population of teenagers with sleep difficulties is unknown. Although the use of CBT for insomnia is a widely validated method,22 it is unlikely to be useful as a stand-alone treatment for delayed sleep, which is not equivalent to insomnia, as there are innate biological reasons (ie, delay in circadian timing) that would be difficult to overcome with behavioral efforts. Use of behavioral therapy to promote sleep at an earlier circadian phase (without shifting the clock to an earlier phase) may, in fact, lead to the induction of insomnia.23

    Traditional bright light phototherapy in the morning is designed to offset the natural delay in circadian timing (and bedtime) by evoking an advance in the timing of the circadian clock (ie, events on subsequent days are set to an earlier time). In adolescents, exposure to bright light prior to their habitual wake time, and to a lesser degree after awakening, will advance the timing of the circadian clock.24,25 In theory, this exposure would realign bedtime to a time more amenable to getting sufficient sleep. Such a change in circadian timing, however, would be temporary and would need to occur each day to offset the naturally occurring biological delay associated with adolescence. Also, traditional light therapy often requires teenagers to awaken earlier than their habitual wake time, which engenders even greater sleep loss. However, one study has shown that the human circadian system can respond to sequences of extraordinarily brief, millisecond-length flashes of light while people are sleeping (flash therapy), and that such light does not disrupt sleep.26 The physiological circuit underlying the circadian responses to light flashes is not completely understood, but the ability of the neurons that project from the retina to the suprachiasmatic nucleus to continue their electrical response for several minutes after cessation of light27 is likely to be in part responsible for the integration of light flashes over time (temporal integration). Using flashes of light is a powerful tool with which one can manipulate the circadian system, as it is more than 2- to 3-fold stronger than continuous light.28 Thus, flash therapy delivered during the last hours of habitual sleep could set the human circadian pacemaker to an earlier hour, allowing teenagers to go to sleep earlier, and without the burden of having to wake earlier than usual.

    Our study consisted of 2 phases. Phase 1 examined whether, independent of any behavioral adjustments, flash therapy during sleep could move sleep timing earlier and increase total sleep in teenagers. Phase 2 examined whether flash therapy in combination with brief CBT could move sleep timing earlier and increase total sleep time in teenagers. In both phases, participants had 1 week of ad libitum baseline sleep monitoring at home with a written diary followed by 4 weeks of treatment. The primary outcome measures were the timing of the onset of sleep and the total amount of sleep time.

    Methods
    Participants

    Phase 1 included 72 adolescents aged 14 to 18 years (recruited from November 1, 2013, to April 30, 2015), while phase 2 included 30 adolescents aged 14 to 18 years (recruited from October 1, 2015, to May 31, 2016) (Figure 1) (trial protocol is available in Supplement 1). Recruitment was accomplished via ongoing partnerships with local schools, pediatricians, and the Stanford Sleep Clinic, as well as posts to online bulletin boards and high school e-newsletters. Inclusion criteria were intentionally broad: students who were enrolled full-time in grades 9 to 12 and who expressed difficulty going to bed earlier and waking up early enough were eligible for the study. Exclusion criteria included taking medication specifically for sleep disorders (including over-the-counter medications or supplements such as melatonin), sleeping exclusively in the prone position, meeting criteria for bipolar disorder, meeting criteria for a sleep disorder other than delayed sleep phase disorder or insomnia disorder, or currently receiving treatment for a sleep disorder. All procedures were approved by the Stanford University Institutional Review Board and conformed to the principles laid out in the Declaration of Helsinki.29 Assent was obtained from the adolescent participants and consent was obtained from at least 1 parent of each participant. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

    Procedure

    Study procedures are illustrated in Figure 2. In general, both phases followed the same pattern: 1 week of baseline followed by 4 weeks of intervention. Phase 1 had 2 groups: light alone (light) and sham light alone (sham). Phase 2 also had 2 groups: light plus CBT and sham light plus CBT. In each phase, after ensuring that eligibility criteria were met, participants were randomly assigned to 1 of the 2 groups. Block randomization (n = 15, to account for possible seasonality effects) was completed double-blind using a computer-generated random sequence. All participants were correctly told they would receive light at night and that the purpose of the investigation was to advance bedtimes to increase sleep duration.

    Participants visited the laboratory with a parent for a baseline evaluation, followed by 1 week of ad libitum baseline sleep monitoring at home with a written sleep diary. At the end of this week, a technician visited the participant’s home to install the light device in the participant’s bedroom. Once the light was placed, participants were instructed to go to bed 1 hour earlier than their usual bedtime for the remainder of the study, to allow opportunity for the light to take effect. In phase 1, participants were called weekly to fill out questionnaires on sleepiness and sleep quality, and to report whether they noticed the light and whether the light interfered with their sleep. In phase 2, instead of weekly telephone calls, participants came to the laboratory for weekly therapy sessions (CBT, described below) during which the same questionnaires were completed. At the end of the 4-week intervention, participants returned to the laboratory with a parent to complete an end-of-treatment evaluation. Participants completed all portions of the study during the academic year and not during scheduled winter and spring breaks. If a teenager slept in 2 households, the light was placed in each household at approximately the same position relative to the bed.

    Measures

    At a baseline visit, participants completed the following questionnaires: Morningness-Eveningness Scale for Children (morning vs evening preference),30 Adolescent Sleep-Wake Scale (subjective sleep quality),31 Cleveland Adolescent Sleepiness Questionnaire (daytime sleepiness),32 Credibility/Expectancy Questionnaire (participant belief in and expectation of results from intervention),33 and a single question inquiring about the reasons for sleep difficulty. During the course of the study, participants completed a sleep diary34 every morning, answered a question about sleepiness delivered via text message 2 hours prior to baseline bedtime every evening (Ecological Momentary Analysis),35 and repeated the Adolescent Sleep-Wake Scale and Cleveland Adolescent Sleepiness Questionnaire each week of the intervention and again at the end of treatment. The Credibility/Expectancy Questionnaire was also repeated at the end of treatment. At baseline and the end of treatment, a parent completed the Child and Adolescent Sleep Checklist for parents (parent’s impression of participant’s sleep).36

    Intervention
    Flash Therapy

    Participants were randomized to receive an active or sham light stimulus for the duration of the treatment. In the active condition (light), daily light therapy (ie, all days of the week) in the form of brief light pulses began 3 hours (in phase 1) or 2 hours (in phase 2) before the participant’s target weekday wake time. Approximately 4000 lux of broad-spectrum white light (approximately 200-600 lux at the cornea after eyelid filtration) was delivered in each 3-millisecond flash, occurring 20 seconds apart. Light was generated by a custom-designed xenon flash bulb (Moflash Signaling). The sham group received 1 minute of light pulses (appearing in 20-second intervals, for a total of 3 pulses) each hour; this dose is believed to be insufficient to shift the circadian system.28 The light timers for both groups were preprogrammed by one of us (J.M.Z.) who was not involved in assessment at any point. The placement of the experimental light was done to minimize the distance between the light and normal sleeping position and was measured for each participant.

    CBT Intervention

    Consistent with other brief behavioral treatments for sleep disturbances, the CBT intervention consisted of four 50-minute, in-person sessions once per week delivered by a licensed clinical psychologist or trained doctoral student (K.A.K., H.B., and L.S.-G.). The CBT intervention had several core components, including (1) education on the circadian system, the effect of light, and physiological processes; (2) information about the role of sleep in domains relevant to adolescents (eg, athletic performance, physical appearance, weight management, and academics); (3) sleep hygiene and stimulus control, well-established insomnia treatment components designed to improve sleep and associated contextual cues; and (4) activity scheduling to wake up earlier on weekends and anticipate obstacles.

    Motivational interviewing and values-based clarification were central features of the CBT intervention. Given that motivational interviewing to promote behavior change may be strengthened by parent participation,37 teenagers and parents, together with the study therapist, set collaborative goals for the teenagers’ sleep at the baseline evaluation. To further promote change, teenagers were asked at the baseline evaluation to reflect the importance of sleep and their level of confidence and readiness to make sleep changes. Motivational interviewing was woven into each subsequent treatment session, including exploring ambivalence about change where resistance was met. At baseline, teenagers completed a worksheet on personal strivings,38 listing domains that were important to them (eg, “improve at softball,” “be happy,” and “get into college”). The treatment was subsequently tailored to adolescents’ individual goals, and the therapist could highlight relationships between adequate sleep and athletic performance, mood, or academic pursuits, as dictated by the adolescent.

    Statistical Analysis

    Group differences in demographic variables were analyzed using t tests for continuous data and χ2 tests for categorical data. Differences in baseline measures were also evaluated by calculating the standardized mean difference (SMD); an SMD of 0.2 is considered small.39 All analyses were based on the intent-to-treat model. For investigating changes in sleep diary variables, Ecological Momentary Analysis, and weekly sleep questionnaires, linear mixed-effects models allowing for random intercepts were used. The differential effect of treatment was also expressed as Cohen d, calculated by dividing model-estimated difference in mean change by observed pooled SD at baseline.40 Mixed-effects models were run in R, version 3.1.2 (R Project for Statistical Computing), using lme4, version 1.1-7. Two-tailed P values were generated using Satterthwaite’s approximations41 for degrees of freedom in lmerTest, version 2.0-29. P < .05 was considered significant.

    To examine variability across the intervention period, we calculated for each individual the proportion of days in which adolescents were compliant with prescribed bedtimes, defined by attempts to sleep at or before the target bedtime. As target bedtimes and wake times were collaboratively set in treatment sessions in phase 2, working toward an advance of 60 minutes, we further calculated a ratio of bedtime compliance relative to wake time compliance. For example, an individual who woke at the target wake time on 20 days and went to sleep at the target bedtime on 10 days would have a ratio of 0.5. This ratio would reflect compliance with sleep initiation (thought to be under circadian and self-motivated control) relative to compliance with sleep termination (more fixed through school start time). Ratios were log-transformed prior to statistical analyses.

    Results
    Sample Characteristics

    Among the 102 participants (54 female [52.9%]; mean [SD] age, 15.6 [1.1] years), 72 were enrolled in phase 1 and 30 were enrolled in phase 2. Study flow is depicted in Figure 1. Groups within each phase were mostly similar (Table 1). There were no differences in age or sex between those who withdrew (phase 1, 4 of 72 [5.6%]; and phase 2, 2 of 30 [6.7%]) and those who completed the studies. There also were no significant differences in baseline measures between individuals who completed vs those who withdrew in terms of daytime sleepiness (mean [SD] Cleveland Adolescent Sleepiness Questionnaire score, 40.3 [9.1] vs 42.9 [13.7]; SMD = −0.27; P = .52), subjective sleep quality (mean [SD] Adolescent Sleep-Wake Scale score, 18.2 [2.8] vs 17.6 [3.5]; SMD = 0.21; P = .63), and parents’ impression of their teenager’s sleep (mean [SD] Child and Adolescent Sleep Checklist for parents score, 16.9 [5.8] vs 17.2 [11.1]; SMD = −0.046; P = .95). There was no singular reason teenagers gave for being unable to go to sleep earlier than their current bedtime. One teenager in phase 2, randomized to the sham plus CBT group, was excluded from analyses after reporting that an illness during both the baseline and posttreatment periods yielded unrepresentative sleep.

    Sleep Diary Outcomes

    In phase 1, application of the light therapy alone was not associated with a change in sleep timing (Table 2). Sleep efficiency (ie, total sleep time divided by time in bed) improved 3% in the light group from baseline to the end of treatment (mean [SD], 89% [8.5%] vs 92% [8.9%]; P < .001), which was different from the 1% change in the sham group (mean [SD], 92% [5.0%] vs 93% [4.5%]; P = .049), although groups differed on this variable by 3% at baseline (mean [SD]: light group, 89% [8.5%] vs sham group, 92% [5.0%]; P = .04) (Table 2). In phase 1, however, adolescents in both groups were often not compliant with instructions to go to bed 1 hour earlier (111 of 896 days’ compliance [12.4%] in light group vs 74 of 931 days’ compliance [7.9%] in sham), resulting in considerable bedtime variability.

    In phase 2, compared with the sham plus CBT group, the light plus CBT group showed greater improvement in sleep. The light plus CBT group had significantly earlier sleep onsets (mean [SD], 50.1 [27.5] minutes earlier), greater total sleep time (mean [SD], 43.3 [35.0] minutes longer), and earlier attempts to initiate sleep (mean [SD], 41.5 [32.5] minutes earlier) (Table 2, Figure 3). As an exploratory measure, we examined compliance in maintaining a regular bedtime, relativized to compliance with maintaining a regular wake time. In the sham plus CBT group, relative compliance was low (mean [SD], 0.29 [0.76]). In comparison, relative compliance in maintaining a regular bedtime in the light plus CBT group was 7-fold greater (mean [SD], 2.21 [3.91]; t25 = −3.13; P = .006). This finding was also reflected in significantly greater variability in bedtimes in the sham plus CBT group (11.5 minutes) compared with the light plus CBT group (21.5 minutes; P = .04, F test). Wake time variability did not differ between the groups (sham plus CBT, 24.3 minutes; and light plus CBT, 17.2 minutes; P = .25, F test).

    Ecological Momentary Analysis

    In phase 1, both light and sham groups reported increased mean (SD) evening sleepiness ratings (light group: baseline, 3.62 [0.85]; and end of treatment, 4.01 [1.23]; and sham group: baseline, 3.14 [1.06]; and end of treatment, 3.39 [1.33]; P = .01) (Table 2). In phase 2, there was a significant time × treatment interaction present for evening sleepiness (P = .02) with increasing sleepiness in the light plus CBT group (+0.55; P = .07) and a nonsignificant decrease in sleepiness in the sham plus CBT group (−0.48; P = .15) during the evening (Table 2).

    Questionnaire Outcomes

    All groups from both phases reported reductions in daytime sleepiness (Cleveland Adolescent Sleepiness Questionnaire) and increases in sleep quality (Adolescent Sleep-Wake Scale) (Table 2). There were no differences between treatments within phases. In phase 2, parents reported a 4-point reduction in their child’s sleep problems in the light plus CBT group (mean [SD] score: baseline, 18.3 [7.17]; and end of treatment, 14.3 [9.51]; P = .007) but not the sham plus CBT group (mean [SD] score: baseline, 17.9 [8.56]; and end of treatment, 14.8 [8.04]; P = .09) (Table 2, Child and Adolescent Sleep Checklist for parents).

    Light Placement and Tolerability

    Light placement in participants’ bedrooms was a mean (SD) distance of 91.3 (62.2) cm (phase 1) and 111 (71.2) cm (phase 2) from participants’ pillows. There was no correlation between light placement and change in sleep onset time in either phase 1 (Pearson correlation, r = 0.11; P = .53) or phase 2 (Pearson correlation, r = 0.22; P = .46). In phase 1, more individuals in the light group relative to the sham group reported either noticing the light (29 vs 20; χ22 = 5.92; P = .02; n = 68) or having their sleep disturbed by the light (19 vs 7; χ22 = 8.97; P = .003; n = 68). To account for this disruption, the duration of the light exposure was reduced in phase 2. Subsequently, we found no differences in the number of teenagers in the light plus CBT group and in the sham plus CBT group who reported either seeing the light (14 vs 12; P > .99, Fisher exact test) or having their sleep disturbed by the light (6 vs 2; P = .22, Fisher exact test), although the low number of participants preclude us from concluding that there was no effect of the light on sleep.

    Treatment Equivalence

    In phase 2, the intervention was rated favorably, with both groups demonstrating high mean (SD) Credibility/Expectancy Questionnaire scores at the end of segment 1 (light plus CBT group, 33.6 [6.1]; and sham plus CBT group, 34.7 [7.0]) and at the end of segment 4 (light plus CBT group, 36.0 [5.3]; and sham plus CBT group, 35.9 [6.7]). There were no differences in Credibility/Expectancy Questionnaire scores between the groups at the end of segment 1 (t25 = −0.41; P = .69) or at the end of segment 4 (t23 = 0.30; P = .98).

    Discussion

    When given without supportive CBT, a sequence of 3-millisecond pulses of light delivered at the end of the nocturnal sleep episode was insufficient to bring about a change in sleep timing. When flash therapy was combined with a sleep-focused CBT intervention, however, we observed a robust advance in bed timing, decreases in time to fall asleep, and an increased nightly sleep duration of nearly 45 minutes. Without flash therapy, CBT alone did not significantly increase sleep and participant bedtimes were highly unstable, as teenagers were more than 7 times less compliant with earlier bed times as compared with their wake times. Our data are consistent with the theory that flash therapy advanced the timing of the circadian clock, enabling teenagers to initiate sleep at an earlier clock hour.24,42 The circadian clock elicits a strong wake-promoting signal in the hours immediately before normal bedtime.43 It is possible that adolescents in the sham plus CBT group attempted to initiate sleep when the circadian clock was signaling for wake, making it difficult to initiate sleep, thus creating lower compliance with the protocol and a potential to induce insomnia.23 In the light plus CBT group, circadian timing was likely moved earlier such that the adolescents in this group were not trying to go to sleep at a time during which the circadian clock was signaling for wake. This is supported by the Ecological Momentary Analysis sleepiness data: those in the light plus CBT group had greater sleepiness in the late evening, consistent with having their internal clock moved to an earlier time. Future experiments should include a direct measure of circadian timing to confirm this mechanism of action.

    In phase 1, light flashes were both ineffective and somewhat disruptive to sleep. In response to this shortcoming, we decreased the duration of the flash sequences from 3 hours to 2 hours, which appeared to alleviate the negative effects of the light on sleep. Not only was there no greater notice or disruption of the light flashes, but teenagers also slept to the same clock hour in both conditions, indicating that the light flashes did not specifically disrupt the end of sleep (when the flashes were being delivered). It is also possible that shortening the flash sequence increased its effectiveness by reducing the likelihood that the flashes had either no effect or had a phase-delaying effect on circadian timing (ie, the light was more likely to elicit only circadian phase advances).44 Given that the light flashes in phase 1 (3 hours) and phase 2 (2 hours) are not identical, it would be important to examine whether 2 hours of light flashes were sufficient to change the time of sleep onset without the administration of adjunctive CBT. The calendar dates of therapy administration were not identical between phase 1 and phase 2, precluding direct comparison.

    Future studies should include objective measurements of sleep (eg, actigraphy), given the possibility of reporting bias, as well as long-term monitoring of the light intervention. This latter improvement would be an important next step to determine ongoing acceptance of the light therapy and to determine whether the beneficial effects of light plus CBT could be maintained for months or years. The light sequence that we used (one 3-millisecond flash every 20 seconds for 2-3 hours) was selected based on early data.42 We have found more recently that a flash sequence occurring every 8 seconds is able to produce a change in circadian timing that is twice as large as the flash sequence used in the current study. If needed, bedtimes could be moved even earlier and more sleep could be obtained using an optimized light flash sequence, although convincing teenagers of the need for a further truncation of evening wake time could be difficult.

    Limitations

    The light flashes in this study were delivered through a customized beacon. Although we did not observe any associations between the placement of the beacon in the adolescents’ rooms and any of our measures, it is possible that this is not an ideal form because the beacon is not commercially available and could affect other individuals sleeping in the same room (eg, siblings) who might not want to be affected by the light. Development of an alternative light delivery system (eg, eye mask based or pillow based) might be necessary in such conditions.

    As previously mentioned, the use of objective sleep monitoring in a larger number of individuals will be critical in future studies of the effect of this intervention. Subjective sleep monitoring, actigraphic sleep monitoring, and polysomnographic sleep monitoring have some overlapping sources of variance but can be independent of one another, especially under conditions of poor sleep.45-47 There are shortcomings to each of these recording modalities, and the use of a combination of them may be necessary, especially in examination of the daytime sequelae of sleep deficits (eg, in academic performance), which is also an important area of follow-up.

    Conclusions

    Despite the potential limitations of the light delivery mechanism, when combined with CBT, light flashes at the end of sleep enabled an earlier bedtime and increased nightly sleep by approximately 45 minutes in teenagers. This type of passive light intervention in teenagers may lead to novel therapeutic applications.

    Back to top
    Article Information

    Accepted for Publication: August 5, 2019.

    Published: September 25, 2019. doi:10.1001/jamanetworkopen.2019.11944

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Kaplan KA et al. JAMA Network Open.

    Corresponding Author: Jamie M. Zeitzer, PhD, Stanford Center for Sleep Sciences and Medicine, Department of Psychiatry and Behavioral Sciences, Stanford University, 3801 Miranda Ave (151Y), Palo Alto, CA 94304 (jzeitzer@stanford.edu).

    Author Contributions: Drs Kaplan and Zeitzer had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Kaplan, Zeitzer.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Kaplan, Williams, Starr-Glass, Zeitzer.

    Critical revision of the manuscript for important intellectual content: Kaplan, Mashash, Batchelder, Zeitzer.

    Statistical analysis: Kaplan, Mashash, Williams, Zeitzer.

    Obtained funding: Kaplan, Zeitzer.

    Administrative, technical, or material support: Kaplan, Mashash, Williams, Starr-Glass.

    Supervision: Kaplan, Batchelder, Zeitzer.

    Conflict of Interest Disclosures: Dr Zeitzer reported receiving grants from the National Institute of Child Health and Human Development during the conduct of the study; and serving as an unpaid, uncompensated scientific advisor to a company (LumosTech) that is attempting to commercialize the flash technology; Stanford University holds the patent on this technology with Craig Heller, PhD (Stanford University), and Anthony van den Pol, PhD (Yale University); Dr Zeitzer is not on this patent and has no financial connection to the patent. No other disclosures were reported.

    Funding/Support: This study was supported by grant R21 HD073095-01 from the National Institute of Child Health and Human Development (Dr Zeitzer), grant UL1TR001085 from the Lucille Packard Foundation for Children’s Health, and Stanford Child Health Research Institute (Dr Kaplan).

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

    Data Sharing Statement: See Supplement 2.

    Additional Contributions: Booil Jo, PhD, Stanford University, provided biostatistical consultation. Rachel Manber, PhD, Stanford University, provided consultation on the cognitive behavioral therapy treatment. Shannon Sullivan, MD, and Rafael Pelayo, MD, Stanford University, provided medical supervision. They were not compensated for their contributions.

    References
    1.
    Keyes  KM, Maslowsky  J, Hamilton  A, Schulenberg  J.  The great sleep recession: changes in sleep duration among US adolescents, 1991-2012.  Pediatrics. 2015;135(3):460-468. doi:10.1542/peds.2014-2707PubMedGoogle ScholarCrossref
    2.
    Carskadon  MA, Mindell  JA, Drake  C.  Sleep in America Poll. Washington, DC: The National Sleep Foundation; 2006.
    3.
    Calamaro  CJ, Mason  TB, Ratcliffe  SJ.  Adolescents living the 24/7 lifestyle: effects of caffeine and technology on sleep duration and daytime functioning.  Pediatrics. 2009;123(6):e1005-e1010. doi:10.1542/peds.2008-3641PubMedGoogle ScholarCrossref
    4.
    Wolfson  AR, Carskadon  MA.  Understanding adolescents’ sleep patterns and school performance: a critical appraisal.  Sleep Med Rev. 2003;7(6):491-506. doi:10.1016/S1087-0792(03)90003-7PubMedGoogle ScholarCrossref
    5.
    Wolfson  AR, Carskadon  MA.  Sleep schedules and daytime functioning in adolescents.  Child Dev. 1998;69(4):875-887. doi:10.1111/j.1467-8624.1998.tb06149.xPubMedGoogle ScholarCrossref
    6.
    Knutson  KL, Spiegel  K, Penev  P, Van Cauter  E.  The metabolic consequences of sleep deprivation.  Sleep Med Rev. 2007;11(3):163-178. doi:10.1016/j.smrv.2007.01.002PubMedGoogle ScholarCrossref
    7.
    Weiss  A, Xu  F, Storfer-Isser  A, Thomas  A, Ievers-Landis  CE, Redline  S.  The association of sleep duration with adolescents’ fat and carbohydrate consumption.  Sleep. 2010;33(9):1201-1209. doi:10.1093/sleep/33.9.1201PubMedGoogle ScholarCrossref
    8.
    McGlinchey  EL, Harvey  AG.  Risk behaviors and negative health outcomes for adolescents with late bedtimes.  J Youth Adolesc. 2015;44(2):478-488. doi:10.1007/s10964-014-0110-2PubMedGoogle ScholarCrossref
    9.
    Killgore  WD.  Effects of sleep deprivation and morningness-eveningness traits on risk-taking.  Psychol Rep. 2007;100(2):613-626. doi:10.2466/pr0.100.2.613-626PubMedGoogle ScholarCrossref
    10.
    Talbot  LS, McGlinchey  EL, Kaplan  KA, Dahl  RE, Harvey  AG.  Sleep deprivation in adolescents and adults: changes in affect.  Emotion. 2010;10(6):831-841. doi:10.1037/a0020138PubMedGoogle ScholarCrossref
    11.
    McGlinchey  EL, Talbot  LS, Chang  KH, Kaplan  KA, Dahl  RE, Harvey  AG.  The effect of sleep deprivation on vocal expression of emotion in adolescents and adults.  Sleep. 2011;34(9):1233-1241. doi:10.5665/SLEEP.1246PubMedGoogle ScholarCrossref
    12.
    Dagys  N, McGlinchey  EL, Talbot  LS, Kaplan  KA, Dahl  RE, Harvey  AG.  Double trouble? the effects of sleep deprivation and chronotype on adolescent affect.  J Child Psychol Psychiatry. 2012;53(6):660-667. doi:10.1111/j.1469-7610.2011.02502.xPubMedGoogle ScholarCrossref
    13.
    Iglowstein  I, Jenni  OG, Molinari  L, Largo  RH.  Sleep duration from infancy to adolescence: reference values and generational trends.  Pediatrics. 2003;111(2):302-307. doi:10.1542/peds.111.2.302PubMedGoogle ScholarCrossref
    14.
    Breslau  N, Roth  T, Rosenthal  L, Andreski  P.  Daytime sleepiness: an epidemiological study of young adults.  Am J Public Health. 1997;87(10):1649-1653. doi:10.2105/AJPH.87.10.1649PubMedGoogle ScholarCrossref
    15.
    Mercer  PW, Merritt  SL, Cowell  JM.  Differences in reported sleep need among adolescents.  J Adolesc Health. 1998;23(5):259-263. doi:10.1016/S1054-139X(98)00037-8PubMedGoogle ScholarCrossref
    16.
    Carskadon  MA, Acebo  C, Jenni  OG.  Regulation of adolescent sleep: implications for behavior.  Ann N Y Acad Sci. 2004;1021:276-291. doi:10.1196/annals.1308.032PubMedGoogle ScholarCrossref
    17.
    Van den Bulck  J.  Television viewing, computer game playing, and Internet use and self-reported time to bed and time out of bed in secondary-school children.  Sleep. 2004;27(1):101-104. doi:10.1093/sleep/27.1.101PubMedGoogle ScholarCrossref
    18.
    Moseley  L, Gradisar  M.  Evaluation of a school-based intervention for adolescent sleep problems.  Sleep. 2009;32(3):334-341. doi:10.1093/sleep/32.3.334PubMedGoogle ScholarCrossref
    19.
    Kira  G, Maddison  R, Hull  M, Blunden  S, Olds  T.  Sleep education improves the sleep duration of adolescents: a randomized controlled pilot study.  J Clin Sleep Med. 2014;10(7):787-792.PubMedGoogle Scholar
    20.
    Hansen  M, Janssen  I, Schiff  A, Zee  PC, Dubocovich  ML.  The impact of school daily schedule on adolescent sleep.  Pediatrics. 2005;115(6):1555-1561. doi:10.1542/peds.2004-1649PubMedGoogle ScholarCrossref
    21.
    Gradisar  M, Dohnt  H, Gardner  G,  et al.  A randomized controlled trial of cognitive-behavior therapy plus bright light therapy for adolescent delayed sleep phase disorder.  Sleep. 2011;34(12):1671-1680. doi:10.5665/sleep.1432PubMedGoogle ScholarCrossref
    22.
    Qaseem  A, Kansagara  D, Forciea  MA, Cooke  M, Denberg  TD; Clinical Guidelines Committee of the American College of Physicians.  Management of chronic insomnia disorder in adults: a clinical practice guideline from the American College of Physicians.  Ann Intern Med. 2016;165(2):125-133. doi:10.7326/M15-2175PubMedGoogle ScholarCrossref
    23.
    Flynn-Evans  EE, Shekleton  JA, Miller  B,  et al.  Circadian phase and phase angle disorders in primary insomnia.  Sleep. 2017;40(12):zsx163. doi:10.1093/sleep/zsx163PubMedGoogle ScholarCrossref
    24.
    Crowley  SJ, Eastman  CI.  Human adolescent phase response curves to bright white light.  J Biol Rhythms. 2017;32(4):334-344. doi:10.1177/0748730417713423PubMedGoogle ScholarCrossref
    25.
    Khalsa  SBS, Jewett  ME, Cajochen  C, Czeisler  CA.  A phase response curve to single bright light pulses in human subjects.  J Physiol. 2003;549(pt 3):945-952. doi:10.1113/jphysiol.2003.040477PubMedGoogle ScholarCrossref
    26.
    Zeitzer  JM, Ruby  NF, Fisicaro  RA, Heller  HC.  Response of the human circadian system to millisecond flashes of light.  PLoS One. 2011;6(7):e22078. doi:10.1371/journal.pone.0022078PubMedGoogle ScholarCrossref
    27.
    Berson  DM, Dunn  FA, Takao  M.  Phototransduction by retinal ganglion cells that set the circadian clock.  Science. 2002;295(5557):1070-1073. doi:10.1126/science.1067262PubMedGoogle ScholarCrossref
    28.
    Najjar  RP, Zeitzer  JM.  Temporal integration of light flashes by the human circadian system.  J Clin Invest. 2016;126(3):938-947. doi:10.1172/JCI82306PubMedGoogle ScholarCrossref
    29.
    World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
    30.
    Carskadon  MA, Vieira  C, Acebo  C.  Association between puberty and delayed phase preference.  Sleep. 1993;16(3):258-262. doi:10.1093/sleep/16.3.258PubMedGoogle ScholarCrossref
    31.
    LeBourgeois  MK, Giannotti  F, Cortesi  F, Wolfson  AR, Harsh  J.  The relationship between reported sleep quality and sleep hygiene in Italian and American adolescents.  Pediatrics. 2005;115(1)(suppl):257-265. doi:10.1542/peds.2004-0815HPubMedGoogle ScholarCrossref
    32.
    Spilsbury  JC, Drotar  D, Rosen  CL, Redline  S.  The Cleveland Adolescent Sleepiness Questionnaire: a new measure to assess excessive daytime sleepiness in adolescents.  J Clin Sleep Med. 2007;3(6):603-612.PubMedGoogle Scholar
    33.
    Borkovec  TD, Nau  SD.  Credibility of analog therapy rationales.  J Behav Ther Exp Psychiatry. 1972;3:257-260. doi:10.1016/0005-7916(72)90045-6Google ScholarCrossref
    34.
    Carney  CE, Buysse  DJ, Ancoli-Israel  S,  et al.  The consensus sleep diary: standardizing prospective sleep self-monitoring.  Sleep. 2012;35(2):287-302. doi:10.5665/sleep.1642PubMedGoogle ScholarCrossref
    35.
    Buysse  DJ, Thompson  W, Scott  J,  et al.  Daytime symptoms in primary insomnia: a prospective analysis using ecological momentary assessment.  Sleep Med. 2007;8(3):198-208. doi:10.1016/j.sleep.2006.10.006PubMedGoogle ScholarCrossref
    36.
    Oka  Y, Horiuchi  F, Tanigawa  T.  Development and validation of a child sleep screening questionnaire: Child and Adolescent Sleep Checklist (CASC).  Jap J Sleep Med. 2009;3(2):404-408.Google Scholar
    37.
    Gayes  LA, Steele  RG.  A meta-analysis of motivational interviewing interventions for pediatric health behavior change.  J Consult Clin Psychol. 2014;82(3):521-535. doi:10.1037/a0035917PubMedGoogle ScholarCrossref
    38.
    Emmons  RA.  Personal strivings, daily life events, and psychological and physical well-being.  J Pers. 1991;59(3):453-472. doi:10.1111/j.1467-6494.1991.tb00256.xPubMedGoogle ScholarCrossref
    39.
    Cohen  J.  Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
    40.
    Feingold  A.  A regression framework for effect size assessments in longitudinal modeling of group differences.  Rev Gen Psychol. 2013;17(1):111-121. doi:10.1037/a0030048PubMedGoogle ScholarCrossref
    41.
    Package ‘lmerTest.’ https://cran.r-project.org/web/packages/lmerTest/lmerTest.pdf. Published February 11, 2019. Accessed August 28, 2019.
    42.
    Zeitzer  JM, Fisicaro  RA, Ruby  NF, Heller  HC.  Millisecond flashes of light phase delay the human circadian clock during sleep.  J Biol Rhythms. 2014;29(5):370-376. doi:10.1177/0748730414546532PubMedGoogle ScholarCrossref
    43.
    Dijk  D-J, Czeisler  CA.  Paradoxical timing of the circadian rhythm of sleep propensity serves to consolidate sleep and wakefulness in humans.  Neurosci Lett. 1994;166(1):63-68. doi:10.1016/0304-3940(94)90841-9PubMedGoogle ScholarCrossref
    44.
    Crowley  SJ, Acebo  C, Carskadon  MA.  Sleep, circadian rhythms, and delayed phase in adolescence.  Sleep Med. 2007;8(6):602-612. doi:10.1016/j.sleep.2006.12.002PubMedGoogle ScholarCrossref
    45.
    Kaplan  KA, Hardas  PP, Redline  S, Zeitzer  JM; Sleep Heart Health Study Research Group.  Correlates of sleep quality in midlife and beyond: a machine learning analysis.  Sleep Med. 2017;34:162-167. doi:10.1016/j.sleep.2017.03.004PubMedGoogle ScholarCrossref
    46.
    Kaplan  KA, Hirshman  J, Hernandez  B,  et al; Osteoporotic Fractures in Men (MrOS), Study of Osteoporotic Fractures SOF Research Groups.  When a gold standard isn’t so golden: lack of prediction of subjective sleep quality from sleep polysomnography.  Biol Psychol. 2017;123:37-46. doi:10.1016/j.biopsycho.2016.11.010PubMedGoogle ScholarCrossref
    47.
    Short  MA, Gradisar  M, Lack  LC, Wright  H, Carskadon  MA.  The discrepancy between actigraphic and sleep diary measures of sleep in adolescents.  Sleep Med. 2012;13(4):378-384. doi:10.1016/j.sleep.2011.11.005PubMedGoogle ScholarCrossref
    ×