An Extra Hour Matters: Delaying School Start Time Improves Sleep, Inhibitory Control, and Anxiety in Early Adolescents

An Extra Hour Matters: Delaying School Start Time Improves Sleep, Inhibitory Control, and Anxiety in Early Adolescents | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article An Extra Hour Matters: Delaying School Start Time Improves Sleep, Inhibitory Control, and Anxiety in Early Adolescents Eve REYNAUD, Lucie MALEVERGNE, Alexandre GRELLET, Adrien PAWLIK, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7085731/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Early school start times (SST) force adolescents to wake earlier than biologically preferred, creating chronic sleep debt. While evidence supports delaying SSTs, few controlled studies have used objective measures to assess their impact. We conducted a controlled intervention in a French boarding school to evaluate the effects of a one-hour delay in SST on sleep (N=50) and cognitive functioning (N=73) in early adolescents (age 12.8 years [11.7-14.2], 66% girls). After a baseline period with 8 a.m. starts (T0), four classes were randomized: half remained at the early schedule (Control-SST), half switched to 9 a.m. (Delayed-SST). Sleep measured by actigraphy, cognitive performance and mental health were assessed at baseline (T0) and 6 months later (T1). Total sleep time had decreased in Control-SST but increased in Delayed-SST, resulting at T1 in a 26-minute difference (Cohen's d=0.93, p=0.007). Sleep onset time did not differ between groups. Sleepiness and anxiety decreased in Delayed-SST but increased in Control-SST (d=-0.52, p=0.042 and d=-0.44, p=0.037 respectively). Inhibitory control improved in Delayed-SST compared to Control-SST (d=-0.79, p=0.001), with trends toward better sustained attention (d=-0.40, p=0.051). Delaying SST by one hour effectively counteracts typical pubertal sleep loss, and benefits both cognitive functioning and mental health outcomes in early adolescence. Health sciences/Health care Health sciences/Medical research Biological sciences/Neuroscience Biological sciences/Psychology Social science/Psychology School start time sleep attention cognition adolescent actigraphy Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Adolescence is a developmental period marked by significant biological and psychosocial changes, notably in the regulation of sleep and circadian rhythms. There is a natural tendency for sleep patterns to shift later during adolescence, a phenomenon known as a delay in chronotype or preferred sleep timing, primarily driven by pubertal maturation and its associated biological changes. 1,2 This shift results from both a slower accumulation of sleep pressure, part of the homeostatic regulation of sleep, 3 and a postponed onset of melatonin secretion, which reflects a shift in the internal circadian clock. 4,5 In parallel, adolescents experience growing autonomy from parental supervision and increasing exposure to evening social and academic activities, as well as screen time, all of which contribute to delayed sleep onset. 6 Early school start times (SST) lead to a mismatch between adolescents’ biological sleep preferences and imposed social schedules. The interaction of these biological and social influences results in a consistent reduction in total sleep time, a phenomenon that Carskadon and colleagues 7 have described as “the perfect storm”. National surveys and longitudinal studies consistently confirm that sleep duration declines with age during adolescence. 8–10 Importantly, this reduction is not due to a decreased need for sleep but rather stems from the growing misalignment between delayed sleep onset and early school start. 7 As a result, adolescents frequently accumulate sleep debt during the week which they attempt to compensate on weekends, further disrupting their circadian rhythms. 11 This chronic sleep restriction is now recognized as a major public health concern. Insufficient sleep during adolescence has been linked to a host of negative outcomes, including increased daytime sleepiness, attention deficits, lower academic achievement, mood instability, behavioral issues, increased substance use, and even heightened risk for suicidal ideation and behaviors. 12–14 In this context, delaying school start has emerged as a promising intervention to extend sleep duration and better align educational schedules with adolescents’ biological circadian rhythms. 15–17 In 2016, Wheatons and colleagues 15 identified in their systematic review 38 articles investigating the effect of delaying school start on sleep, behavior, health and academic outcomes. They found consistent associations between later start times and longer sleep on weeknights, better school attendance, reduced tardiness, fewer instances of falling asleep in class, improved academic performance, and a decrease in motor vehicle accidents. However, most studies included were cross-sectional design, comparing schools with different start time, or before-after design without control groups, limiting causal inferences. More recent quasi-experimental studies have assessed changes in outcomes before and after the implementation of a delayed school start times and compared them to those of a control group with no schedule modification, thus assessing difference-in-differences. 18,19 This design strengthens causal inference by controlling for both pre-existing differences in sleep patterns and the natural developmental trajectories of adolescent sleep and cognition. For instance, Widome and colleagues 18 took advantage of a district-led delay of 50 and 65 minutes in school start time at two high schools, while including three control schools from the same district that maintained their original 7:30 a.m. start time. Using actigraphy, they found that compared to the change observed in the control schools, students in the delayed-start schools gained on average 41 minutes of additional sleep on school nights after one year (CI 95% [25;57]) and 43 minutes after 2 years (CI 95% [25 ; 61]). The authors reported minimal difference-in-differences regarding sleep onset time and sleep efficiency. This study provides strong evidence in favor of delaying school start, however the lack of randomization leaves open the possibility of residual confounding, such as baseline differences regarding sleep measures and demographic characteristics. However, conducting true randomized control trials remain logistically challenging. 20 Thus, despite this growing body of research, gaps remain regarding controlled trials using objective measures of sleep and cognitive function. The present study sought to address this need by conducting a controlled trial to evaluate the effect of a one-hour delay in school start time (from 8:00 a.m. to 9:00 a.m.) on: (1) sleep parameters, measured with actigraphy, (2) sleepiness, anxiety and depressive symptoms, assessed through validated self-assessed scales and (3) cognitive functioning, including sustained attention and inhibitory control, measured using standardized tests. The study was carried out amongst early adolescents, in a boarding school setting, which provided a unique methodological advantage by minimizing inter-individual variability related to home environment, parental routines, and external social factors. Crucially, the absence of daily commuting eliminated a major source of variability in sleep and wake times often encountered in day school studies and eased logistical constraints. Methods 1.1. Population and setting The study was conducted in a public boarding school located in a rural area of France. All students resided at the school from Monday to Friday and usually shared a room with 3 or 4 other students. The school is part of a specialized program designed to support students from disadvantaged socio-economic backgrounds by providing a living environment conducive to studying, along with access to athletic and cultural extracurricular activities. The present study focused on early adolescence, in students in grade 7 and 8 (corresponding to the 2 nd and 3 rd year middle school in France), who are typically aged between 12 and 14 years old. Students in the last year of middle school were not included to avoid confounding effects related to the increased academic pressure of the national exam that is administered that year. The sample size is described in figure 1. Of the 98 students enrolled in the relevant grade levels, 86 agreed to participate in the study, and provided a signed informed consent from both the student and a parent or legal guardian. Data from 50 students were included in the actigraphy analyses, and 73 in the analyses based on computerized cognitive tests (see “Measures” section). Inclusion in the actigraphy analyses required ≥3 days of valid recordings at both baseline (T0) and follow-up (T1). Actigraphy data were missing for 6 students at T0 (three from each group), due to device non-use (N = 5) or technical failure (N = 1). At T1, compliance substantially declined, resulting in the exclusion of 30 participants from the actigraphy dataset, including 18 from the control group. Reasons for exclusion at T1 were complete non-use (N = 13), insufficient wear time (N = 6), device loss (N = 6), and device malfunction (N = 4). One participant was excluded post hoc and treated as non-compliant in the analyses due to altered sleep patterns during Ramadan (pre-dawn waking for religious observance). Regarding the computerized cognitive tests, 5 students were absent during the T0 evaluation and 8 during the T1 evaluation. The flow chart of attrition is illustrated in Figure 1. FIGURE 1 1.2. Study design The study was conducted during the 2023–2024 school year and involved two 7th grade classes and two 8th grade classes. All methods were carried out in accordance with relevant guidelines and regulations. The study protocol was approved by the Paris School of Economics Institutional Review Board ethics committee (n° 2023-020). From September to the end of October, all students followed a standard schedule with classes starting at 8:00 a.m., Tuesday through Friday. Because all classes started at 10:00 a.m. on Mondays to accommodate students with long commutes, and this schedule was consistent across both groups and unrelated to the intervention, data from Sunday nights were excluded from the analyses. Of note, class ending remained the same, the delay in school start was compensated by limiting free periods. After a two-week vacation period, a delayed school start time intervention was implemented. Classes were randomly assigned by the school administrators to either the Delayed-SST group (for delayed school start time), in which school start times were delayed by one hour after the fall break (i.e. from early November), or the Control-SST group, which maintained an 8:00 a.m. start throughout the school year. Baseline assessments (T0) were conducted between September and October, prior to the delayed school start intervention. During this time, students were instructed to wear an actigraph for two consecutive weeks. At the end of this phase, they completed a computerized evaluation including self-report questionnaires and cognitive tasks. These assessments were administered simultaneously by grade level using an online platform (Psytoolkit 21,22 ) on the school’s computers. The same assessment procedure was repeated between February and March (T1). The study protocol is illustrated in Figure 2. FIGURE 2 1.3. Measures 1.3.1. Actigraphy Sleep was assessed using actigraphy (Actiwatch 2; Philips-Respironics, Murrysville, USA, Philips Actiware software version 6.0.1), a wrist-worn device equipped with a piezoelectric accelerometer. The device samples activity data at a frequency of 32 Hz, aggregated into 60-second epochs. The threshold for sleep detection was set at the standard 40 activity counts per epoch. 23 Participants were instructed to press a button on the device to indicate the moment they went in bed, ready to sleep (Bedtime, BT, in decimal hours), and the moment they got out of bed (Rising Time, in decimal hours). When button presses were missing, BT and rising time were estimated using a combination of light and movement data from the device, supplemented by information from a sleep diary completed by the participants throughout the recording period. Sleep onset time (SOT, in decimal hours) was defined as the time at which sleep began, while Wake Time (WT, in decimal hours) was defined as the time sleep ended. Sleep Onset Latency (SOL, in minutes) was calculated as the interval between BT and SOT. Total sleep time (TST, in minutes) was defined as the duration between SOT and WT, minus any periods of Wake After Sleep Onset (WASO, in minutes). Sleep Efficiency (SE, expressed as a percentage) was calculated as TST divided by the time spent in bed (duration between BT and RT). Although students were asked to wear the actigraph continuously for 2 weeks, analyses focused exclusively on school nights (i.e. Monday through Thursday nights). While extended sleep duration on weekends may indicate adequate rest, it may also reflect compensatory sleep in response to accumulated weekday sleep debt. 24 1.3.2. Questionnaires Daytime sleepiness was evaluated using the French Sleepiness Scale for Adolescents (FSSA), 25 an adaptation of the Epworth Sleepiness Scale for Children and Adolescents. 26 This self-report measure requires students to rate the likelihood of dozing off in eight different situations (e.g. while reading or sitting in a classroom). Each item is rated on a Likert scale ranging from 0 ("would never fall asleep") to 3 ("high chance of falling asleep"). The total score ranges from 0 to 24, with higher scores indicating greater levels of daytime sleepiness. A cutoff score strictly above 11 indicates acute sleepiness. 25 Symptoms of anxiety and depression were evaluated using the Hospital Anxiety and Depression Scale (HAD), 27 a validated 14-item self-report questionnaire. The scale is divided into two subscales, anxiety and depression, each comprising seven items. Scores for each subscale range from 0 to 21, with higher scores reflecting more severe symptomatology. A cutoff score strictly above 11 in the anxiety subscale indicates probable emotional disorder and one strictly above 9 in the depression scale indicate probable depression. 27 Preference towards morningness or eveningness was assessed using the Morningness–Eveningness Scale for Children (MESC). 28 It is a 10-item questionnaire, where participants indicate at what time they would rather engage towards various activities, such as waking up, taking an exam or doing sports. A higher score indicates a greater preference for morningness. Participants with a score strictly inferior to 23 were defined as having an evening chronotype, and others as a morning or neutral chronotype. 1.3.3. Cognitive functioning Sustained attention was assessed using the “Sustained Attention to Response Task” (SART), 29 a computerized go/no-go paradigm designed to measure impulsivity and ability to maintain attention over time. Participants are required to monitor visual displays acknowledging responses to frequent neutral signals (GO trials) but withholding response when detecting rare targets (NO–GO trials). Specifically, participants were instructed to press the space bar on the keyboard when a digit appeared on the screen (“go” trials) but to withhold their response when the presented digit was a “3” (“no-go” trials). Each digit was displayed for 250 milliseconds (ms), followed by a visual mask lasting 900 ms. The task included 18 practice trials, followed by 225 experimental trials, comprising 200 “go” trials and 25 “no-go” trials for a duration of approximatively 7 minutes. Omission errors were defined as failures to respond during a go trial, while commission errors were defined as responses made during no-go trials. Reaction times (RTs) were computed only for correct go trials. To minimize the influence of extreme values, which are likely fast guesses without cognitive processing, a within-subject low-pass filter was applied to the RT data. Specifically, RTs faster than two standard deviations below the individual’s mean RT were excluded (0.28% of RT data was removed). The task duration was in average 5.6 (SD = 0.8) minutes. Students also completed a computerized Stroop task, 30 which provides a measure of inhibitory control, selective attention and processing speed. 31 In this task, participants were instructed to identify the ink color in which a word was printed (“red,” “green,” “blue,” or “yellow”) and to ignore the semantic content of the word itself. Each response was required within 200 ms of stimulus presentation. Following a 10-item training session, the experimental phase consisted of 100 trials: 25 congruent trials, in which the ink color matched the word meaning, and 75 incongruent trials, in which the word's meaning corresponded to a different color. The Stroop effect is defined as the difference in mean reaction time (RT) between incongruent and congruent trials, with greater RT differences indicating lower inhibitory control. Only correct responses were included in RT analyses. Within-participant RT outliers, defined as RTs faster than two standard deviations below the participant’s mean, were excluded, accounting for approximately 1.2% of the trials. On average, this task lasted 3.7 (SD = 0.24) minutes. 1.3.4. Socio-demographic data The school provided information about the participants age, gender (also confirmed within the computerized questionnaire) and scholarship status, which was used as a proxy for socio-economic status. Scholarships are based on the income of the person or persons responsible for the student and the number of dependent children (not on academic performance). 1.4. Analyses All analyses were non-parametrical and conducted using R program. 32 First, to assess differential attrition, a comparison between included and excluded participants was conducted to understand selection bias regarding gender, age, socio-economic status and group allocation. Secondly, we assessed the comparability of the included participants between intervention groups (Control-SST vs Delayed-SST) at baseline, where we compared gender, age, socio-economic status, as well as all actigraphy, questionnaires and cognitive functioning. Then, the difference between baseline and follow-up (T1-T0) was calculated for all outcome measures and compared between groups and referred to as difference-in-differences analyses. This methodology allows to control for baseline differences between groups and for the natural change that occurs when no intervention is conducted. We hypothesized that the delay in SST would benefit more participants with an evening chronotype, and thus conducted a complementary analysis stratifying by chronotype (MESC < 23). For all comparisons, Mann-Whitney U-tests were used to compare the distribution of continuous variables and chi-square to compare binary outcomes, using a two-tailed p<0.05 for significance. As a measure of precaution since students are grouped in class, we calculated the Intraclass Correlation Coefficients (ICC), to estimate if considering clustering in the analyses was necessary. An ICC close to zero implies no clustering of the data. The ICC of sleep duration at T0 was equal to -0.033 CI 95% [-0.042 ; +0.904]. Due to the small sample size and number of clusters, the calculation of the ICC presents approximation, with a wide confidence interval and a negative ICC. Negative ICCs are conventionally treated as zero, 33 therefore, it was deemed unnecessary to account for within-group clustering in the analyses. Results 2.1. Population Questionnaires and cognitive measures were available for the pre and post intervention conditions in 73 students (85% of the participants). After applying criteria for actigraphy data validity (at least three days of usable recording), we retained usable sleep data from 50 participants with pre and post intervention measures. Importantly, participants with and without usable data did not differ significantly in terms of age, gender, socio-economic status, or experimental group (see supplementary table). 2.2. Actigraphy study Baseline At baseline (T0), groups were not significantly different regarding socio-demographic characteristics (Table 1), with an average age of 12.8 years (SD = 0.67), 68% were girls and 48% benefited from a scholarship. Similarly, no differences were observed in objectively measured sleep on school nights, with an average sleep onset occurring at 9:50 p.m. (SD = 12 min), a rising time at 6:05 a.m. (SD = 15 min) and a total sleep time of 7 hours and 20 minutes (SD = 37 min). All participants had an average TST below the 9-hour recommendation for children aged 12 years or less, 94% had an average TST below the 8-hour recommendation for adolescents aged 13 to 18, and 18% had an average TST below 7 hours. Table 1 Difference-in-differences Difference-in-differences analysis for sleep outcomes are presented in table 2. Between T0 and T1, total sleep time (TST) significantly decreased in the Control-SST group by 15.6 min (SD = 28.9, p = 0.046) while no significant change was observed in the Delayed-SST group (+7.2 min, SD = 21.1, p = 0.092). This resulted in a significant between group difference-in-differences of +22.8 min (CI 95% [+7.6 ; +37.9], Cohen’s d = 0.93, p = 0.007). At T1, adolescents in the Control-SST group had in average 7 hours and 4 minutes of sleep per night (SD = 28 min), while those in the Delayed-SST group slept 7 hours and 30 minutes (SD = 24 min), corresponding to a significant 26-minute difference (CI 95% [+10 ; +43], illustrated in Figure 3). In the Control-SST group, 50% of the students slept less than 7 hours per night, compared to 13% in the Delayed-SST Group (X² = 6.3, p = 0.012). Table 2, figure 3 The evolution of sleep onset time between T0 and T1 did not significantly differ between the 2 groups (difference-in-differences -6.0 min CI 95% [-22.3 ; +10.3], p = 0.806). In contrast, wake time was significantly delayed by 12.6 min in the Delayed-SST group compared to the Control-SST group (CI 95% [-4.8 ; +30.0], p = 0.031). Interestingly, when participants were stratified by chronotype, adolescents with an evening chronotype showed a significant TST difference-in-differences of 28.5 minutes between the Delayed-SST and Control-SST groups (95% CI [+8.8 ; +52.7], p = 0.004), while no significant difference was found for adolescents with a morning or neutral chronotype (9.6 minutes, 95% CI [-15.0 ; +35.2]; see Figure 4). These findings suggest that delaying school start time may be particularly beneficial for adolescents with biologically later sleep preferences. Figure 4 2.3. Questionnaires and cognitive tests Baseline At baseline, no significant differences were observed between the two groups in self-reported sleepiness (FSSA), anxiety (HAD-A), or depressive scores (HAD-D), see table 3. Overall, 25% of participants scored above the clinical cutoff for excessive sleepiness (FSSA), 29% for anxiety symptoms (HAD-A), and 18% for depressive symptoms (HAD-D). In the SART, students in the Control-SST group made significantly more commission errors than those in the Delayed-SST group at baseline (respectively 15.7, SD = 5.9 vs 12.7, SD = 5.6, p = 0.026). They also exhibited a tendency towards faster reaction time (181.62 ms, SD = 83.77 vs 221.44 ms, SD = 94.86) compared to the Delayed-SST group (p = 0.064). Regarding cognitive performances, no significant group differences were found on the STROOP test, assessing inhibitory control, selective attention and processing speed. Table 3 Difference-in-differences Difference-in-differences analyses for the questionnaires and cognitive tests are detailed in table 4. Between T0 and T1, students in the Control-SST group showed a significant increase in sleepiness score of 2 points on the FSSA scale (SD = 3.5) whereas no statistically significant change was observed in the Delayed-SST group (0.0, SD = 4.1), resulting in a medium effect size (Cohen’s d = 0.52, p = 0.042). The score of anxiety significantly decreased in the Delayed-SST between T0 and T1, whereas no significant change was observed in the Control-SST group (difference-in-differences -1.7; 95% CI [-3.6 ; +0.1], p = 0.037). The evolution of depressive scores was not significantly different between the 2 groups (difference-in-differences -1.3; 95% CI [-2.9 ; +0.2], p = 0.077). Performances on the Sustained Attention Response Task (SART) indicated a significant reduction in commission errors in the Delayed-SST group (-4.5 errors, SD = 4.8, p<0.001) but not in the control group (-2.2, SD = 6.7, p = 0.081). This group-by-time difference approached statistical significance (difference-in-differences = -2.3; 95% CI [-5.0 ; +0.4], p = 0.051) suggesting a trend toward improved impulsivity control following the intervention. No significant group differences were obtained in changes in omission errors or in reaction times (respectively d = 0.19, p = 0.899 and d = -0.07, p = 0.554). Regarding the Stroop task, performance did not significantly change in the Control-SST group between T0 and T1 (+17.9 ms, SD = 71.1, p = 0.158), whereas performance improved in the Delayed-SST group (-49.9 ms, SD = 99.5, p = 0.005). The resulting difference-in-differences was 67.9 ms (95% CI [-108.3 ; -27.4], Cohen’s d = -0.79, p = 0.001), reflecting a significant improvement in inhibitory control and selective attention following the delayed school start. Table 4 Discussion This controlled study evaluated the impact of a one-hour delay in school start time on sleep patterns, cognitive functioning and mental health in early adolescents. Our findings demonstrate significant benefits of this intervention on multiple outcomes, reinforcing the growing body of literature advocating for later school start times for adolescents. Objective measurements of sleep revealed that delaying school start times by one hour effectively counteracts the typical pubertal decline in sleep duration observed during adolescence. 9 While students in the Control-SST group exhibited the expected decline in sleep duration over the six-month period (-15.6 min), those in the Delayed-SST group showed a modest gain (+7.2 min), yielding a difference-in-differences of 23 minutes. This result closely aligns with the meta-analysis by Bowers et al. 17 which reported an average gain of 20 minutes in total sleep time across longitudinal studies implementing delayed school start times (weighted effect size of d = 0.38, p<0.001). Importantly, the magnitude of benefit appears to scale with the extent of the delay; longer school start delays allowed greater increase in sleep duration. However, Bowers et al. 17 were unable to evaluate potential moderators such as sex, age or the type of sleep measures (subjective vs objective) due to insufficient reporting or limited variability across studies. Given the impossibility of blinding school start time to participants, the reliance on self-reported sleep data may create expectancy effects and social desirability bias potentially inflating estimates of effect sizes. Our findings suggest that the benefits of later school start times are already seen beneficial in early adolescence, before the full onset of pubertal circadian shifts. This developmental period is characterized by the initial emergence of neurobiological changes, most notably a progressive delay in melatonin secretion timing 5 and a shift towards evening sleep–wake preferences. 34 This developmental trajectory makes early school start times increasingly misaligned with adolescents' sleep needs. Our data further supports this interpretation showing that early adolescents with an evening chronotype benefited most from the delayed start time. As circadian phase delay intensifies throughout puberty, the benefits of later start times are likely to become more pronounced with age. 35 Further studies are needed to better characterize how pubertal maturation, from early through late adolescence, modulates the effects of delayed start times on sleep and daytime functioning. Importantly, we observed that delayed school start time did not significantly affect sleep onset time. This finding is important as one of the main challenges to implementing later school start times is the common misconception that students will simply go to bed later, thereby negating any potential benefits. Instead, the preserved bedtime combined with later wake times resulted in extended sleep duration. This pattern has been consistently reported in prior studies. 18,36–38 In parallel with this objective increased in sleep duration, subjective sleepiness was also improved. While students in the control group reported increased sleepiness over time, those in the intervention group maintained stable levels, resulting in a medium effect size difference between groups. This results confirms the ubiquitously reported improved sleepiness in delayed school start studies. 17 Beyond improvements in sleep quantity and sleepiness our study demonstrated relevant effects on mental health. Anxiety symptoms significantly decreased in the Delayed-SST group, while remaining unchanged in controls. Changes in depressive symptoms did not reach significance, the observed trend was in the expected direction. Those results are in line with longitudinal studies linking insufficient sleep and poor sleep quality to increased anxiety and depressive symptoms in adolescents. 39 Previous studies have also reported improvement in mood, 40 well-being 41 and depressive symptoms following a delayed school start time. 37,38,42 Similarly, sleep-based interventions have also shown improvement in anxiety among adolescents. 43 For example, Blake and colleagues 44 conducted a randomized control trial implementing a cognitive-behavioral based sleep intervention in 123 adolescents, and found improvement in anxiety symptoms, although not in depressive symptoms. Few studies have directly assessed the effect of delayed school start times on standardized cognitive measures. Alfonsi and colleagues 45 demonstrated that students starting at 8:00 am exhibited progressive declines in psychomotor vigilance performance across the academic year, while those starting at 9:00 a.m. maintained stable performance, supporting the role of delayed school start in preserving sustained attention. Lufi et al 46 also reported improvement in sustained and selective attention after 2 weeks of delayed school start time. In our study, the intervention was also associated with improvement in cognitive functioning. We observed a significant reduction in the Stroop interference effect, with a robust difference-in-differences effect size (d = -0.79), indicating improved inhibitory control and selective attention in the Delayed-SST group. Similarly, a trend toward improved performance was also observed in the SART, with reduced commission errors, suggesting improved impulsivity control. These findings are consistent with established evidence that prefrontal cortex functions are particularly vulnerable to sleep restriction in adolescents. 47 These experimental findings are further reinforced by converging real-world behavioral evidence showing that adolescents with insufficient or poor sleep-quality are more susceptible to impulsivity-related outcomes. These include higher sensation-seeking tendencies, increased transport risk taking, greater involvement in violent or delinquent behaviors and increased substance use. 48 Taken together, this body of evidence highlights the critical role of sleep in supporting cognitive control and psychosocial development during adolescence. Therefore, by counteracting sleep deprivation delaying school start times could serve as a scalable, systemic intervention to mitigate sleep-related vulnerabilities in this age group. Although we did not assess the impact of the intervention on academic performances, some studies have observed positive associations between later school start times and improved academic performance. 19,49,50 However, the systematic review by Biller and colleagues, 51 concluded that no generalizable improvement in academic achievement can be drawn due to substantial heterogeneity in academic outcomes, study designs and delay durations. Focusing on peer-reviewed longitudinal studies with a control group, three reported no change in global academic outcomes 50,52,53 (such as grade point average), one found a GPA improvement after 2 years, 19 and others reported subject-specific improvements (such as reading or math). 49,50 Authors underscore the limitations of using academic test scores alone as indicators of the success or failure of such interventions. 51,52 Despite the demonstrated efficacy of the intervention, the average sleep duration in the delayed-SST group remained below the recommended 8 to 10 hour per night for adolescents. 54 This aligns with previous findings that even when structural changes are implemented, a substantial proportion of adolescents continue to obtain insufficient sleep. Bower et al. reported in their meta-analysis 17 a post intervention average sleep duration of 7 hours and 24 minutes. 54 The present intervention allowed to partially counter the prevalence of chronically sleep deprived students, with 50% in the Control-SST group sleeping less than 7 hours compared to 13% in the Delayed-SST Group at follow-up. However, even if this reduction in the prevalence of short sleep is meaningful, it indicates that later start times alone are insufficient. These results highlight the need to consider additional structural and cultural measures alongside start time adjustments to promote adequate sleep in adolescents. Complementary interventions, such as limiting screen electronic devices, 55 limiting late extracurricular activities and regulating electronic assignment submission times, may be necessary to foster a school environment that is more conducive to healthy sleep. A broader cultural shift that recognizes the importance of sleep for learning, emotional regulation and health is needed to help adolescents meet their biological sleep needs. The present study has several strengths, including the use of objective measures of sleep, standardized cognitive tests, a before/after study design with a control group and a randomized allocation ensuring baseline comparability. However, the relatively small sample size, particularly for actigraphy analyses due to compliance issues, should be acknowledged. The actigraphy sample size was reduced due to compliance issues, particularly among boys and participants in the Control-SST group, possibly reflecting lower engagement in the absence of intervention benefit. This attrition can introduce a possible selection bias, although baseline comparisons did not differ significantly between included and excluded participants. Additionally, subgroup analyses (e.g., by chronotype) lacked statistical power. As previous studies have experienced, recruiting schools to participate in a randomized study which implies modification of schedules is particularly difficult. 20 The controlled boarding school setting of our study, while controlling for environmental variability and commute facilities, may limit generalizability to typical day school populations, where evening routines are less regulated. Despite strong scientific consensus and recommendations from major professional organizations including the National Sleep Foundation, 56 the American Academy of Sleep Medicine, 57 the American Academy of Pediatrics, 58 implementation of delayed school start times remains limited. In the United States, 13 % of middle schools start at 9 a.m. or later, 59 while in France middle school uniformly starts at 8 a.m. Kelley et al 60 demonstrated that even later start times (e.g., 10:00 a.m.) can yield further benefits, reinforcing the argument that current schedules are misaligned with adolescent biology. The primary barriers to implementation are logistical rather than scientific, with school districts citing concerns about transportation, after-school activities and families schedules. 20,61 This gap between evidence and practice illustrates how operational constraints can impede the translation of sleep research into educational practice, leaving many adolescents with chronic sleep deprivation despite effective solutions. Interestingly, delayed start times have also been associated with improved sleep in parents 62 and teachers, 63 and parent-teacher associations are increasingly advocating a delay in school. 64 Whether this growing support from advocacy groups will be enough to overcome long-standing logistical challenges and lead to widespread policy change remains an open question. In conclusion, this controlled trial provides compelling evidence that a one-hour delay in school start time significantly improve adolescent sleep duration and daytime functioning. The intervention effectively counteracted the natural developmental decline in sleep without delaying bedtimes, resulting in both objective sleep improvements and enhanced cognitive performance, particularly for inhibitory control. These findings add to the growing scientific consensus supporting later school start times as a public health strategy to mitigate sleep deprivation in adolescents. Educational policymakers should consider these benefits when weighing the logistical challenges of implementing delayed start times against the potential improvements in student well-being and cognitive functioning. Declarations Data availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Acknowledgements We would like to acknowledge to all participants, as well as to the teachers and the school head, whose time, effort, and support in coordinating schedules and facilitating the assessment were instrumental in making this study possible. Funding We would also like to acknowledge the National Education Scientific Council (Conseil scientifique de l'Éducation nationale) for the financial support and the IDEE program (ANR-21-ESRE-0034) the administrative, technical and statistical support. Author contributions E.R. designed the study, carried out the analyses and drafted the initial manuscript. E.R., A.G., L.M. collected data., A.G., A.P and M.G. provided technical and statistical support. A.E.R. and S.M. coordinated and supervised study planification and conceptualization. All authors critically reviewed and revised the manuscript. Competing interests The authors declare no competing interests. References Feinberg, I., Davis, N. M., de Bie, E., Grimm, K. J. & Campbell, I. G. The maturational trajectories of NREM and REM sleep durations differ across adolescence on both school-night and extended sleep. Am. J. Physiol.-Regul. Integr. Comp. Physiol. 302 , R533–R540 (2012). Ohayon, M. M., Carskadon, M. A., Guilleminault, C. & Vitiello, M. V. Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. Sleep 27 , 255-1273 (2004). Campbell, I. et al. Adolescent changes in homeostatic regulation of EEG activity in the delta and theta frequency bands during NREM sleep. Sleep 34 , 83-91 (2011). Crowley, S. J. et al. A longitudinal assessment of sleep timing, circadian phase, and phase angle of entrainment across human adolescence. PloS One 9 (11) , e112199 (2014). Kennaway, D. J. The dim light melatonin onset across ages, methodologies, and sex and its relationship with morningness/eveningness. Sleep 46(5) , zsad033 (2023). Hale, L. & Guan, S. Screen time and sleep among school-aged children and adolescents: A systematic literature review. Sleep Med. Rev . 21 , 50–58 (2015). Carskadon, M. A. Sleep in adolescents: the perfect storm. Pediatr. Clin. North Am . 58 , 637–647 (2011). National Sleep Foundation. Sleep in the Modern Family. https://sleepfoundation.org/sleep-polls-data/sleep-in-america-poll/2014-sleep-and-family (2014). Sadeh, A., Dahl, R. E., Shahar, G. & Rosenblat-Stein, S. Sleep and the transition to adolescence: a longitudinal study. Sleep 32 , 1602–1609 (2009). Leger, D., Beck, F., Richard, J.-B. & Godeau, E. Total Sleep Time Severely Drops during Adolescence. PLoS ONE 7 , e45204 (2012). Crowley, S. J., Acebo, C. & Carskadon, M. A. Sleep, circadian rhythms, and delayed phase in adolescence. Sleep Med . 8 , 602–612 (2007). Owens, J., Adolescent Sleep Working Group, & Committee on Adolescence. Insufficient sleep in adolescents and young adults: an update on causes and consequences. Pediatrics 134 , e921-932 (2014). Rolling, J. et al. Sleep and circadian rhythms in adolescents with attempted suicide. Sci. Rep. 14 , 8354 (2024). Winsler, A., Deutsch, A., Vorona, R. D., Payne, P. A. & Szklo-Coxe, M. Sleepless in Fairfax: the difference one more hour of sleep can make for teen hopelessness, suicidal ideation, and substance use. J. Youth Adolesc. 44 , 362–378 (2015). Wheaton, A. G., Chapman, D. P. & Croft, J. B. School Start Times, Sleep, Behavioral, Health, and Academic Outcomes: a Review of the Literature. J. Sch. Health 86 , 363–381 (2016). Minges, K. E. & Redeker, N. S. Delayed school start times and adolescent sleep: A systematic review of the experimental evidence. Sleep Med. Rev . 28 , 86–95 (2016). Bowers, J. M. & Moyer, A. Effects of school start time on students’ sleep duration, daytime sleepiness, and attendance: a meta-analysis. Sleep Health 3 , 423–431 (2017). Widome, R. et al. The START study: An evaluation to study the impact of a natural experiment in high school start times on adolescent weight and related behaviors. JAMA Pediatr . 6 , 66 (2020). James, S. A., Erickson, D. J., Lammert, S. & Widome, R. School start time delays and high school educational outcomes: Evidence from the START/LEARN study. J. Adolesc. 95 , 751–763 (2023). Illingworth, G. et al. Challenges in implementing and assessing outcomes of school start time change in the UK: experience of the Oxford Teensleep study. Sleep Med. 60 , 89–95 (2019). Stoet, G. PsyToolkit: A software package for programming psychological experiments using Linux. Behav. Res. Methods 42 , 1096–1104 (2010). Stoet, G. PsyToolkit: A Novel Web-Based Method for Running Online Questionnaires and Reaction-Time Experiments. Teach. Psychol . 44 , 24–31 (2017). Cellini, N., Buman ,Matthew P., McDevitt ,Elizabeth A., Ricker ,Ashley A. & and Mednick, S. C. Direct comparison of two actigraphy devices with polysomnographically recorded naps in healthy young adults. Chronobiol. Int. 30 , 691–698 (2013). Yang, F. N., Picchioni, D. & Duyn, J. H. Effects of sleep-corrected social jetlag on measures of mental health, cognitive ability, and brain functional connectivity in early adolescence. Sleep 46, zsad259 (2023). Gustin, M.-P. et al. French Sleepiness Scale for Adolescents-8 items: A discriminant and diagnostic validation. L’Encéphale 49 , 109–116 (2023). Janssen, K. C., Phillipson, S., O’Connor, J. & Johns, M. W. Validation of the Epworth Sleepiness Scale for Children and Adolescents using Rasch analysis. Sleep Med. 33 , 30–35 (2017). White, D., Leach, C., Sims, R., Atkinson, M. & Cottrell, D. Validation of the Hospital Anxiety and Depression Scale for use with adolescents. Br. J. Psychiatry J. Ment. Sci . 175 , 452–454 (1999). Carskadon, M. A., Vieira, C. & Acebo, C. Association between puberty and delayed phase preference. Sleep 16 , 258–262 (1993). Robertson, I. H., Manly, T., Andrade, J., Baddeley, B. T. & Yiend, J. ‘Oops!’: Performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. Neuropsychologia 35 , 747–758 (1997). Stroop, J. R. Studies of interference in serial verbal reactions. J. Exp. Psychol. Gen. 121 , 15–23 (1992). MacLeod, C. M. Half a century of research on the Stroop effect: An integrative review. Psychol. Bull. 109 , 163–203 (1991). R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing (2023). Ntani, G., Inskip, H., Osmond, C. & Coggon, D. Consequences of ignoring clustering in linear regression. BMC Med. Res. Methodol . 21 , 139 (2021). Hagenauer, M. H., Perryman, J. I., Lee, T. M. & Carskadon, M. A. Adolescent Changes in the Homeostatic and Circadian Regulation of Sleep. Dev. Neurosci. 31 , 276–284 (2009). Dagys, N. et al. Double trouble? The effects of sleep deprivation and chronotype on adolescent affect. J. Child Psychol. Psychiatry 53 , 660–667 (2012). Adam, E. K., Snell, E. K. & Pendry, P. Sleep timing and quantity in ecological and family context: a nationally representative time-diary study. J. Fam. Psychol. JFP J. Div. Fam. Psychol. Am. Psychol. Assoc. Div . 43 21, 4–19 (2007). Boergers, J., Gable, C. J. & Owens, J. A. Later school start time is associated with improved sleep and daytime functioning in adolescents. J. Dev. Behav. Pediatr. 35 , 11–17 (2014). Wahlstrom, K. Changing Times: Findings From the First Longitudinal Study of Later High School Start Times. NASSP Bull. 86 , 3–21 (2002). McMakin, D. L. & Alfano, C. A. Sleep and anxiety in late childhood and early adolescence: Curr. Opin. Psychiatry 28 , 483–489 (2015). Whitaker, R. C., Dearth-Wesley, T., Herman, A. N., Oakes, J. M. & Owens, J. A. A quasi-experimental study of the impact of school start time changes on adolescents’ mood, self-regulation, safety, and health. Sleep Health 5 , 466–469 (2019). Lo, J. C. et al. Sustained benefits of delaying school start time on adolescent sleep and well-being. Sleep 41(6) , zsy052 (2018). Owens, J. A., Belon, K. & Moss, P. Impact of delaying school start time on adolescent sleep, mood, and behavior. A rch. Pediatr. Adolesc. Med. 164 , 608–614 (2010). Blake, M. J., Sheeber, L. B., Youssef, G. J., Raniti, M. B. & Allen, N. B. Systematic Review and Meta-analysis of Adolescent Cognitive–Behavioral Sleep Interventions. Clin. Child Fam. Psychol. Rev. 20 , 227–249 (2017). Blake, M. J. et al. A cognitive-behavioral and mindfulness-based group sleep intervention improves behavior problems in at-risk adolescents by improving perceived sleep quality. Behav. Res. Ther. 99, 147–156 (2017). Alfonsi, V. et al. The Association Between School Start Time and Sleep Duration, Sustained Attention, and Academic Performance. Nat. Sci. Sleep 12 , 1161–1172 (2020). Lufi, D., Tzischinsky, O. & Hadar, S. Delaying School Starting Time by One Hour: Some Effects on Attention Levels in Adolescents. J . Clin. Sleep Med . 07 , 137–143 (2011). Anastasiades, P. G., de Vivo, L., Bellesi, M. & Jones, M. W. Adolescent sleep and the foundations of prefrontal cortical development and dysfunction. Prog. Neurobiol . 218 , 102338 (2022). Short, M. A. & Weber, N. Sleep duration and risk-taking in adolescents: A systematic review and meta-analysis. Sleep Med. Rev. 41, 185–196 (2018). Jung, H. A late bird or a good bird? The effect of 9 o’clock attendance policy on student’s achievement. Asia Pac. Educ. Rev. 19 , 511–529 (2018). Kim, T. The Effects of School Start Time on Educational Outcomes: Evidence from the 9 O’clock Attendance Policy in South Korea. BE J. Econ. Anal. Policy 22 , 439–474 (2022). Biller, A. M., Meissner, K., Winnebeck, E. C. & Zerbini, G. School start times and academic achievement - A systematic review on grades and test scores. Sleep Med. Rev. 61 , 101582 (2022). Lenard, M., Morrill, M. S. & Westall, J. High school start times and student achievement: Looking beyond test scores. Econ. Educ. Rev. 76 , 101975 (2020). Rhie, S. & Chae, K. Y. Effects of school time on sleep duration and sleepiness in adolescents. PloS One 13 , e0203318 (2018). Paruthi, S. et al. Consensus Statement of the American Academy of Sleep Medicine on the Recommended Amount of Sleep for Healthy Children: Methodology and Discussion. J. Clin. Sleep Med . 12 , 1549–1561 (2016). Perrault, A. A. et al. Reducing the use of screen electronic devices in the evening is associated with improved sleep and daytime vigilance in adolescents. Sleep 42 , zsz125 (2019). Ziporyn, T. D. et al. Adolescent sleep health and school start times: Setting the research agenda for California and beyond. A research summit summary. Sleep Health J. Natl. Sleep Found . 8 , 11–22 (2022). Watson, N. F. et al. Delaying Middle School and High School Start Times Promotes Student Health and Performance: An American Academy of Sleep Medicine Position Statement. J. Clin. Sleep Med . 13 , 623–625 (2017). Adolescent Sleep Working Group, Committee on Adolescence, & Council on School Health. School start times for adolescents. Pediatrics 134 , 642–649 (2014). Taie, S. & Lewis, L. Characteristics of 2020–21 Public and Private K–12 Schools in the United States: Results From the National Teacher and Principal Survey First Look (NCES 2022-111). https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid = 2022111 (2022). Kelley, P., Lockley, S. W., Kelley, J. & Evans, M. D. R. Is 8:30 a.m. Still Too Early to Start School? A 10:00 a.m. School Start Time Improves Health and Performance of Students Aged 13-16. Front. Hum. Neurosci. 11 , 588 (2017). Fitzpatrick, J. M., Silva, G. E. & Vana, K. D. Perceived Barriers and Facilitating Factors in Implementing Delayed School Start Times to Improve Adolescent Sleep Patterns. J. Sch. Health 91.2 , 94-101 (2021) doi:10.1111/josh.12983. Meltzer, L. J., Wahlstrom, K. L., Plog, A. E. & McNally, J. Impact of changing school start times on parent sleep. Sleep Health 8 , 130–134 (2022). Wahlstrom, K. L., Plog, A. E., McNally, J. & Meltzer, L. J. Impact of Changing School Start Times on Teacher Sleep Health and Daytime Functioning. J. Sch. Health 93 , 128–134 (2023). Trevorrow, T., Zhou, E. S., Dietch, J. R. & Gonzalez, B. D. Position statement: start middle and high schools at 8:30 am or later to promote student health and learning. Transl. Behav. Med. 9 , 167–169 (2019). Tables Tables 1 to 4 are available in the Supplementary Files section Additional Declarations No competing interests reported. Supplementary Files Supplementarytable.docx Tables.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 06 Mar, 2026 Reviews received at journal 04 Mar, 2026 Reviews received at journal 01 Mar, 2026 Reviewers agreed at journal 22 Feb, 2026 Reviewers agreed at journal 19 Feb, 2026 Reviews received at journal 06 Jan, 2026 Reviewers agreed at journal 02 Jan, 2026 Reviewers invited by journal 26 Nov, 2025 Editor invited by journal 01 Aug, 2025 Editor assigned by journal 31 Jul, 2025 Submission checks completed at journal 16 Jul, 2025 First submitted to journal 16 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7085731","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":551348285,"identity":"4eaa4680-4951-421d-a736-fd4574dc8242","order_by":0,"name":"Eve REYNAUD","email":"","orcid":"","institution":"Université Claude Bernard Lyon 1, CNRS, INSERM, CRNL U1028 UMR 5292, FORGETTING Team","correspondingAuthor":false,"prefix":"","firstName":"Eve","middleName":"","lastName":"REYNAUD","suffix":""},{"id":551348286,"identity":"1a451f05-9152-40f4-aa66-b6e4913013a1","order_by":1,"name":"Lucie MALEVERGNE","email":"","orcid":"","institution":"Université Claude Bernard Lyon 1, CNRS, INSERM, CRNL U1028 UMR 5292, FORGETTING Team","correspondingAuthor":false,"prefix":"","firstName":"Lucie","middleName":"","lastName":"MALEVERGNE","suffix":""},{"id":551348287,"identity":"3422ed00-3b2a-4915-b39a-48bfdccaa31e","order_by":2,"name":"Alexandre GRELLET","email":"","orcid":"","institution":"Paris School of Economics","correspondingAuthor":false,"prefix":"","firstName":"Alexandre","middleName":"","lastName":"GRELLET","suffix":""},{"id":551348288,"identity":"3f821fdc-74ec-4959-bd5a-6c77bc2d0232","order_by":3,"name":"Adrien PAWLIK","email":"","orcid":"","institution":"Paris School of Economics","correspondingAuthor":false,"prefix":"","firstName":"Adrien","middleName":"","lastName":"PAWLIK","suffix":""},{"id":551348289,"identity":"baf32577-dc2a-4484-94a0-4734935fb9f3","order_by":4,"name":"Marc GURGAND","email":"","orcid":"","institution":"Paris School of Economics","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"","lastName":"GURGAND","suffix":""},{"id":551348290,"identity":"21a9c14c-4314-43ec-81dd-27a755a63f24","order_by":5,"name":"Amandine E. REY","email":"","orcid":"","institution":"Université Claude Bernard Lyon 1, CNRS, INSERM, CRNL U1028 UMR 5292, FORGETTING Team","correspondingAuthor":false,"prefix":"","firstName":"Amandine","middleName":"E.","lastName":"REY","suffix":""},{"id":551348291,"identity":"c1cd2509-9f12-4f4e-a5a5-4a01f37c81b9","order_by":6,"name":"Stéphanie MAZZA","email":"data:image/png;base64,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","orcid":"","institution":"Université Claude Bernard Lyon 1, CNRS, INSERM, CRNL U1028 UMR 5292, FORGETTING Team","correspondingAuthor":true,"prefix":"","firstName":"Stéphanie","middleName":"","lastName":"MAZZA","suffix":""}],"badges":[],"createdAt":"2025-07-09 16:08:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7085731/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7085731/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97181584,"identity":"f7170dce-9045-4599-aee1-560ab6c1e794","added_by":"auto","created_at":"2025-12-01 16:37:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":19908,"visible":true,"origin":"","legend":"\u003cp\u003eStudy protocol\u003c/p\u003e\n\u003cp\u003eAssessments were conducted at baseline (T0) and follow-up (T1) across both the delayed and control school start time groups. Measures were collected for 7\u003csup\u003eth\u003c/sup\u003e (dotted) and 8\u003csup\u003eth\u003c/sup\u003e (hashed) grade students in successive sessions. Actigraphy lasted 2 weeks and questionnaires and cognitive assessment were conducted at the end of the 2 weeks. In yellow are the period where students began class at 8 a.m. (both group at baseline), and in blue the period where they began at 9 a.m. (only for the delayed School Start Time group, after baseline)\u003c/p\u003e","description":"","filename":"Binder11.png","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/7a4c6271ffa5342d1115e44d.png"},{"id":97181589,"identity":"6178d9a3-6e82-4835-886f-7724eb8d6894","added_by":"auto","created_at":"2025-12-01 16:37:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":32386,"visible":true,"origin":"","legend":"\u003cp\u003eFlow chart\u003c/p\u003e\n\u003cp\u003eParticipant flowchart illustrating enrollment, group allocation, and follow-up. A total of 98 students were initially considered, with 86 providing informed consent and allocated to either the control school start time (SST), with an 8 a.m. start throughout the year, or the Delayed-SST group, with a 9 a.m. school start following baseline measures. The figure details group sizes for the actigraphy study and the computerized tests separately.\u003c/p\u003e","description":"","filename":"Binder12.png","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/72721cc74cbcc02d03a165f9.png"},{"id":97249258,"identity":"1e6e19b6-a733-4ea4-8994-7eb2f366d3ee","added_by":"auto","created_at":"2025-12-02 13:11:50","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":11798,"visible":true,"origin":"","legend":"\u003cp\u003eDifferences-in-differences in total sleep time by time and group Mean total sleep time (TST) decreased significantly from T0 to T1 for the Control-SST group (dashed line) while the increase of TST in the Delayed-SST (solid line) was not significant. At T1, TST differed by 26 minutes between groups. Error bars represent standard errors of the mean.\u003c/p\u003e","description":"","filename":"Binder13.png","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/251e6fba0b535a929383a34a.png"},{"id":97181585,"identity":"86179b4a-b313-46f5-b86b-97be2f112c9b","added_by":"auto","created_at":"2025-12-01 16:37:58","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":7805,"visible":true,"origin":"","legend":"\u003cp\u003eDifferences-in-differences in total sleep time by time, group and chronotype Panel A) Within adolescents with neutral or morning chronotype (MESC≥23), TST did not significantly change from T0 to T1 for neither group, the differences-in-differences was also not significant. Panel B) Within adolescents with an evening chronotype (MESC\u0026lt;23), Total sleep time (TST) significantly increased from T0 to T1 for students in the delay school start time, but not for those in the control-SST, resulting in a differences-in-differences of 28.5 minutes (p = 0.004).\u003c/p\u003e","description":"","filename":"Binder14.png","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/fa1e3d5c772eec712344a7ae.png"},{"id":97252484,"identity":"b991853c-6e4c-44b7-9069-b96238f5ac15","added_by":"auto","created_at":"2025-12-02 13:21:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":625012,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/d6123322-1c18-4767-b7a0-38ebe8a7f897.pdf"},{"id":97181587,"identity":"a077a076-eaa9-4e6f-aadf-749da74502b0","added_by":"auto","created_at":"2025-12-01 16:37:58","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15402,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarytable.docx","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/237de2d3eace47c4f9903df9.docx"},{"id":97181586,"identity":"2c22b895-cd94-4d9d-833e-b482c65b7491","added_by":"auto","created_at":"2025-12-01 16:37:58","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":23007,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7085731/v1/710794c7f10af09ca48fceb1.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"An Extra Hour Matters: Delaying School Start Time Improves Sleep, Inhibitory Control, and Anxiety in Early Adolescents","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAdolescence is a developmental period marked by significant biological and psychosocial changes, notably in the regulation of sleep and circadian rhythms. There is a natural tendency for sleep patterns to shift later during adolescence, a phenomenon known as a delay in chronotype or preferred sleep timing, primarily driven by pubertal maturation and its associated biological changes.\u003csup\u003e1,2\u003c/sup\u003e This shift results from both a slower accumulation of sleep pressure, part of the homeostatic regulation of sleep,\u003csup\u003e3\u003c/sup\u003e and a postponed onset of melatonin secretion, which reflects a shift in the internal circadian clock.\u003csup\u003e4,5\u003c/sup\u003e In parallel, adolescents experience growing autonomy from parental supervision and increasing exposure to evening social and academic activities, as well as screen time, all of which contribute to delayed sleep onset.\u003csup\u003e6\u003c/sup\u003e Early school start times (SST) lead to a mismatch between adolescents’ biological sleep preferences and imposed social schedules. The interaction of these biological and social influences results in a consistent reduction in total sleep time, a phenomenon that Carskadon and colleagues\u003csup\u003e7\u003c/sup\u003e have described as “the perfect storm”. National surveys and longitudinal studies consistently confirm that sleep duration declines with age during adolescence.\u003csup\u003e8–10\u003c/sup\u003e Importantly, this reduction is not due to a decreased need for sleep but rather stems from the growing misalignment between delayed sleep onset and early school start.\u003csup\u003e7\u003c/sup\u003e As a result, adolescents frequently accumulate sleep debt during the week which they attempt to compensate on weekends, further disrupting their circadian rhythms.\u003csup\u003e11\u003c/sup\u003e This chronic sleep restriction is now recognized as a major public health concern. Insufficient sleep during adolescence has been linked to a host of negative outcomes, including increased daytime sleepiness, attention deficits, lower academic achievement, mood instability, behavioral issues, increased substance use, and even heightened risk for suicidal ideation and behaviors.\u003csup\u003e12–14\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eIn this context, delaying school start has emerged as a promising intervention to extend sleep duration and better align educational schedules with adolescents’ biological circadian rhythms.\u003csup\u003e15–17\u003c/sup\u003e In 2016, Wheatons and colleagues\u003csup\u003e15\u003c/sup\u003e identified in their systematic review 38 articles investigating the effect of delaying school start on sleep, behavior, health and academic outcomes. They found consistent associations between later start times and longer sleep on weeknights, better school attendance, reduced tardiness, fewer instances of falling asleep in class, improved academic performance, and a decrease in motor vehicle accidents. However, most studies included were cross-sectional design, comparing schools with different start time, or before-after design without control groups, limiting causal inferences.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMore recent quasi-experimental studies have assessed changes in outcomes before and after the implementation of a delayed school start times and compared them to those of a control group with no schedule modification, thus assessing difference-in-differences.\u003csup\u003e18,19\u003c/sup\u003e This design strengthens causal inference by controlling for both pre-existing differences in sleep patterns and the natural developmental trajectories of adolescent sleep and cognition. For instance, Widome and colleagues\u003csup\u003e18\u003c/sup\u003e took advantage of a district-led delay of 50 and 65 minutes in school start time at two high schools, while including three control schools from the same district that maintained their original 7:30 a.m. start time. Using actigraphy, they found that compared to the change observed in the control schools, students in the delayed-start schools gained on average 41 minutes of additional sleep on school nights after one year (CI\u003csub\u003e95%\u003c/sub\u003e [25;57]) and 43 minutes after 2 years (CI\u003csub\u003e95%\u0026nbsp;\u003c/sub\u003e[25 ; 61]). The authors reported minimal difference-in-differences regarding sleep onset time and sleep efficiency. This study provides strong evidence in favor of delaying school start, however the lack of randomization leaves open the possibility of residual confounding, such as baseline differences regarding sleep measures and demographic characteristics. However, conducting true randomized control trials remain logistically challenging.\u003csup\u003e20\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThus, despite this growing body of research, gaps remain regarding controlled trials using objective measures of sleep and cognitive function. The present study sought to address this need by conducting a controlled trial to evaluate the effect of a one-hour delay in school start time (from 8:00 a.m. to 9:00 a.m.) on: (1) sleep parameters, measured with actigraphy, (2) sleepiness, anxiety and depressive symptoms, assessed through validated self-assessed scales and (3) cognitive functioning, including sustained attention and inhibitory control, measured using standardized tests. The study was carried out amongst early adolescents, in a boarding school setting, which provided a unique methodological advantage by minimizing inter-individual variability related to home environment, parental routines, and external social factors. Crucially, the absence of daily commuting eliminated a major source of variability in sleep and wake times often encountered in day school studies and eased logistical constraints.\u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e1.1.\u0026nbsp;\u0026nbsp;Population and setting\u003c/p\u003e\n\u003cp\u003eThe study was conducted in a public boarding school located in a rural area of France. All students resided at the school from Monday to Friday and usually shared a room with 3 or 4 other students. The school is part of a specialized program designed to support students from disadvantaged socio-economic backgrounds by providing a living environment conducive to studying, along with access to athletic and cultural extracurricular activities. The present study focused on early adolescence, in students in grade 7 and 8 (corresponding to the 2\u003csup\u003end\u003c/sup\u003e and 3\u003csup\u003erd\u003c/sup\u003e year middle school in France), who are typically aged between 12 and 14 years old. Students in the last year of middle school were not included to avoid confounding effects related to the increased academic pressure of the national exam that is administered that year.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe sample size is described in figure 1. Of the 98 students enrolled in the relevant grade levels, 86 agreed to participate in the study, and provided a signed informed consent from both the student and a parent or legal guardian. Data from 50 students were included in the actigraphy analyses, and 73 in the analyses based on computerized cognitive tests (see \u0026ldquo;Measures\u0026rdquo; section). Inclusion in the actigraphy analyses required \u0026ge;3 days of valid recordings at both baseline (T0) and follow-up (T1). Actigraphy data were missing for 6 students at T0 (three from each group), due to device non-use (N = 5) or technical failure (N = 1). At T1, compliance substantially declined, resulting in the exclusion of 30 participants from the actigraphy dataset, including 18 from the control group. Reasons for exclusion at T1 were complete non-use (N = 13), insufficient wear time (N = 6), device loss (N = 6), and device malfunction (N = 4). One participant was excluded post hoc and treated as non-compliant in the analyses due to altered sleep patterns during Ramadan (pre-dawn waking for religious observance). Regarding the computerized cognitive tests, 5 students were absent during the T0 evaluation and 8 during the T1 evaluation. The flow chart of attrition is illustrated in Figure 1.\u003c/p\u003e\n\u003cp\u003eFIGURE 1\u003c/p\u003e\n\u003cp\u003e1.2. \u0026nbsp;Study design\u003c/p\u003e\n\u003cp\u003eThe study was conducted during the 2023\u0026ndash;2024 school year and involved two 7th grade classes and two 8th grade classes. All methods were carried out in accordance with relevant guidelines and regulations. The study protocol was approved by the Paris School of Economics Institutional Review Board ethics committee (n\u0026deg; 2023-020). From September to the end of October, all students followed a standard schedule with classes starting at 8:00 a.m., Tuesday through Friday. Because all classes started at 10:00 a.m. on Mondays to accommodate students with long commutes, and this schedule was consistent across both groups and unrelated to the intervention, data from Sunday nights were excluded from the analyses. Of note, class ending remained the same, the delay in school start was compensated by limiting free periods. After a two-week vacation period, a delayed school start time intervention was implemented. Classes were randomly assigned by the school administrators to either the Delayed-SST group (for delayed school start time), in which school start times were delayed by one hour after the fall break (i.e. from early November), or the Control-SST group, which maintained an 8:00 a.m. start throughout the school year.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;Baseline assessments (T0) were conducted between September and October, prior to the delayed school start intervention. During this time, students were instructed to wear an actigraph for two consecutive weeks. At the end of this phase, they completed a computerized evaluation including self-report questionnaires and cognitive tasks. These assessments were administered simultaneously by grade level using an online platform (Psytoolkit\u003csup\u003e21,22\u003c/sup\u003e) on the school\u0026rsquo;s computers. The same assessment procedure was repeated between February and March (T1). The study protocol is illustrated in Figure 2.\u003c/p\u003e\n\u003cp\u003eFIGURE 2\u003c/p\u003e\n\u003cp\u003e1.3. \u0026nbsp;Measures\u003c/p\u003e\n\u003cp\u003e1.3.1.\u0026nbsp;Actigraphy\u003c/p\u003e\n\u003cp\u003eSleep was assessed using actigraphy (Actiwatch 2; Philips-Respironics, Murrysville, USA, Philips Actiware software version 6.0.1), a wrist-worn device equipped with a piezoelectric accelerometer. The device samples activity data at a frequency of 32 Hz, aggregated into 60-second epochs. The threshold for sleep detection was set at the standard 40 activity counts per epoch.\u003csup\u003e23\u003c/sup\u003e Participants were instructed to press a button on the device to indicate the moment they went in bed, ready to sleep (Bedtime, BT, in decimal hours), and the moment they got out of bed (Rising Time, in decimal hours). When button presses were missing, BT and rising time were estimated using a combination of light and movement data from the device, supplemented by information from a sleep diary completed by the participants throughout the recording period. Sleep onset time (SOT, in decimal hours) was defined as the time at which sleep began, while Wake Time (WT, in decimal hours) was defined as the time sleep ended. Sleep Onset Latency (SOL, in minutes) was calculated as the interval between BT and SOT. Total sleep time (TST, in minutes) was defined as the duration between SOT and WT, minus any periods of Wake After Sleep Onset (WASO, in minutes). Sleep Efficiency (SE, expressed as a percentage) was calculated as TST divided by the time spent in bed (duration between BT and RT). Although students were asked to wear the actigraph continuously for 2 weeks, analyses focused exclusively on school nights (i.e. Monday through Thursday nights). While extended sleep duration on weekends may indicate adequate rest, it may also reflect compensatory sleep in response to accumulated weekday sleep debt.\u003csup\u003e24\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e1.3.2.\u0026nbsp;Questionnaires\u003c/p\u003e\n\u003cp\u003eDaytime sleepiness was evaluated using the French Sleepiness Scale for Adolescents (FSSA),\u003csup\u003e25\u003c/sup\u003e an adaptation of the Epworth Sleepiness Scale for Children and Adolescents.\u003csup\u003e26\u003c/sup\u003e This self-report measure requires students to rate the likelihood of dozing off in eight different situations (e.g. while reading or sitting in a classroom). Each item is rated on a Likert scale ranging from 0 (\u0026quot;would never fall asleep\u0026quot;) to 3 (\u0026quot;high chance of falling asleep\u0026quot;). The total score ranges from 0 to 24, with higher scores indicating greater levels of daytime sleepiness. A cutoff score strictly above 11 indicates acute sleepiness.\u003csup\u003e25\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSymptoms of anxiety and depression were evaluated using the Hospital Anxiety and Depression Scale (HAD),\u003csup\u003e27\u003c/sup\u003e a validated 14-item self-report questionnaire. The scale is divided into two subscales, anxiety and depression, each comprising seven items. Scores for each subscale range from 0 to 21, with higher scores reflecting more severe symptomatology. A cutoff score strictly above 11 in the anxiety subscale indicates probable emotional disorder and one strictly above 9 in the depression scale indicate probable depression.\u003csup\u003e27\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003ePreference towards morningness or eveningness was assessed using the Morningness\u0026ndash;Eveningness Scale for Children (MESC).\u003csup\u003e28\u003c/sup\u003e It is a 10-item questionnaire, where participants indicate at what time they would rather engage towards various activities, such as waking up, taking an exam or doing sports. A higher score indicates a greater preference for morningness. Participants with a score strictly inferior to 23 were defined as having an evening chronotype, and others as a morning or neutral chronotype.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e1.3.3.\u0026nbsp;Cognitive functioning\u003c/p\u003e\n\u003cp\u003eSustained attention was assessed using the \u0026ldquo;Sustained Attention to Response Task\u0026rdquo; (SART),\u003csup\u003e29\u003c/sup\u003e a computerized go/no-go paradigm designed to measure impulsivity and ability to maintain attention over time. Participants are required to monitor visual displays acknowledging responses to frequent neutral signals (GO trials) but withholding response when detecting rare targets (NO\u0026ndash;GO trials). Specifically, participants were instructed to press the space bar on the keyboard when a digit appeared on the screen (\u0026ldquo;go\u0026rdquo; trials) but to withhold their response when the presented digit was a \u0026ldquo;3\u0026rdquo; (\u0026ldquo;no-go\u0026rdquo; trials). Each digit was displayed for 250 milliseconds (ms), followed by a visual mask lasting 900 ms. The task included 18 practice trials, followed by 225 experimental trials, comprising 200 \u0026ldquo;go\u0026rdquo; trials and 25 \u0026ldquo;no-go\u0026rdquo; trials for a duration of approximatively 7 minutes. Omission errors were defined as failures to respond during a go trial, while commission errors were defined as responses made during no-go trials. Reaction times (RTs) were computed only for correct go trials. To minimize the influence of extreme values, which are likely fast guesses without cognitive processing, a within-subject low-pass filter was applied to the RT data. Specifically, RTs faster than two standard deviations below the individual\u0026rsquo;s mean RT were excluded (0.28% of RT data was removed). The task duration was in average 5.6 (SD = 0.8) minutes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eStudents also completed a computerized Stroop task,\u003csup\u003e30\u003c/sup\u003e which provides a measure of inhibitory control, selective attention and processing speed.\u003csup\u003e31\u003c/sup\u003e In this task, participants were instructed to identify the ink color in which a word was printed (\u0026ldquo;red,\u0026rdquo; \u0026ldquo;green,\u0026rdquo; \u0026ldquo;blue,\u0026rdquo; or \u0026ldquo;yellow\u0026rdquo;) and to ignore the semantic content of the word itself. Each response was required within 200 ms of stimulus presentation. Following a 10-item training session, the experimental phase consisted of 100 trials: 25 congruent trials, in which the ink color matched the word meaning, and 75 incongruent trials, in which the word\u0026apos;s meaning corresponded to a different color. The Stroop effect is defined as the difference in mean reaction time (RT) between incongruent and congruent trials, with greater RT differences indicating lower inhibitory control. Only correct responses were included in RT analyses. Within-participant RT outliers, defined as RTs faster than two standard deviations below the participant\u0026rsquo;s mean, were excluded, accounting for approximately 1.2% of the trials. On average, this task lasted 3.7 (SD = 0.24) minutes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e1.3.4. Socio-demographic data\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe school provided information about the participants age, gender (also confirmed within the computerized questionnaire) and scholarship status, which was used as a proxy for socio-economic status. Scholarships are based on the income of the person or persons responsible for the student and the number of dependent children (not on academic performance).\u003c/p\u003e\n\u003cp\u003e1.4. \u0026nbsp;Analyses\u003c/p\u003e\n\u003cp\u003eAll analyses were non-parametrical and conducted using R program.\u003csup\u003e32\u003c/sup\u003e First, to assess differential attrition, a comparison between included and excluded participants was conducted to understand selection bias regarding gender, age, socio-economic status and group allocation. Secondly, we assessed the comparability of the included participants between intervention groups (Control-SST vs Delayed-SST) at baseline, where we compared gender, age, socio-economic status, as well as all actigraphy, questionnaires and cognitive functioning. Then, the difference between baseline and follow-up (T1-T0) was calculated for all outcome measures and compared between groups and referred to as difference-in-differences analyses. This methodology allows to control for baseline differences between groups and for the natural change that occurs when no intervention is conducted. We hypothesized that the delay in SST would benefit more participants with an evening chronotype, and thus conducted a complementary analysis stratifying by chronotype (MESC \u0026lt; 23). For all comparisons, Mann-Whitney U-tests were used to compare the distribution of continuous variables and chi-square to compare binary outcomes, using a two-tailed p\u0026lt;0.05 for significance. As a measure of precaution since students are grouped in class, we calculated the Intraclass Correlation Coefficients (ICC), to estimate if considering clustering in the analyses was necessary. An ICC close to zero implies no clustering of the data. The ICC of sleep duration at T0 was equal to -0.033 CI\u003csub\u003e95%\u003c/sub\u003e[-0.042 ; +0.904]. Due to the small sample size and number of clusters, the calculation of the ICC presents approximation, with a wide confidence interval and a negative ICC. Negative ICCs are conventionally treated as zero,\u003csup\u003e33\u003c/sup\u003e therefore, it was deemed unnecessary to account for within-group clustering in the analyses.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e2.1.\u0026nbsp;\u0026nbsp;Population\u003c/p\u003e\n\u003cp\u003eQuestionnaires and cognitive measures were available for the pre and post intervention conditions in 73 students (85% of the participants). After applying criteria for actigraphy data validity (at least three days of usable recording), we retained usable sleep data from 50 participants with pre and post intervention measures. Importantly, participants with and without usable data did not differ significantly in terms of age, gender, socio-economic status, or experimental group (see supplementary table).\u003c/p\u003e\n\u003cp\u003e2.2.\u0026nbsp;\u0026nbsp;Actigraphy study\u003c/p\u003e\n\u003cp\u003eBaseline\u003c/p\u003e\n\u003cp\u003eAt baseline (T0), groups were not significantly different regarding socio-demographic characteristics (Table 1), with an average age of 12.8 years (SD = 0.67), 68% were girls and 48% benefited from a scholarship. Similarly, no differences were observed in objectively measured sleep on school nights, with an average sleep onset occurring at 9:50 p.m. (SD = 12 min), a rising time at 6:05 a.m. (SD = 15 min) and a total sleep time of 7 hours and 20 minutes (SD = 37 min). All participants had an average TST below the 9-hour recommendation for children aged 12 years or less, 94% had an average TST below the 8-hour recommendation for adolescents aged 13 to 18, and 18% had an average TST below 7 hours.\u003c/p\u003e\n\u003cp\u003eTable 1\u003c/p\u003e\n\u003cp\u003eDifference-in-differences\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDifference-in-differences analysis for sleep outcomes are presented in table 2. Between T0 and T1, total sleep time (TST) significantly decreased in the Control-SST group by 15.6 min (SD = 28.9, p = 0.046) while no significant change was observed in the Delayed-SST group (+7.2 min, SD = 21.1, p = 0.092). This resulted in a significant between group difference-in-differences of +22.8 min (CI\u003csub\u003e95%\u003c/sub\u003e[+7.6 ; +37.9], Cohen’s d = 0.93, p = 0.007). At T1, adolescents in the Control-SST group had in average 7 hours and 4 minutes of sleep per night (SD = 28 min), while those in the Delayed-SST group slept 7 hours and 30 minutes (SD = 24 min), corresponding to a significant 26-minute difference (CI\u003csub\u003e95%\u003c/sub\u003e[+10 ; +43], illustrated in Figure 3). In the Control-SST group, 50% of the students slept less than 7 hours per night, compared to 13% in the Delayed-SST Group (X² = 6.3, p = 0.012).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2, figure 3\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe evolution of sleep onset time between T0 and T1 did not significantly differ between the 2 groups (difference-in-differences -6.0 min CI\u003csub\u003e95%\u003c/sub\u003e[-22.3 ; +10.3], p = 0.806). In contrast, wake time was significantly delayed by 12.6 min in the Delayed-SST group compared to the Control-SST group (CI\u003csub\u003e95%\u003c/sub\u003e[-4.8 ; +30.0], p = 0.031). Interestingly, when participants were stratified by chronotype, adolescents with an evening chronotype showed a significant TST difference-in-differences of 28.5 minutes between the Delayed-SST and Control-SST groups (95% CI [+8.8 ; +52.7], p = 0.004), while no significant difference was found for adolescents with a morning or neutral chronotype (9.6 minutes, 95% CI [-15.0 ; +35.2]; see Figure 4). These findings suggest that delaying school start time may be particularly beneficial for adolescents with biologically later sleep preferences.\u003c/p\u003e\n\u003cp\u003eFigure 4\u003c/p\u003e\n\u003cp\u003e2.3.\u0026nbsp;\u0026nbsp;Questionnaires and cognitive tests\u003c/p\u003e\n\u003cp\u003eBaseline\u003c/p\u003e\n\u003cp\u003eAt baseline, no significant differences were observed between the two groups in self-reported sleepiness (FSSA), anxiety (HAD-A), or depressive scores (HAD-D), see table 3. Overall, 25% of participants scored above the clinical cutoff for excessive sleepiness (FSSA), 29% for anxiety symptoms (HAD-A), and 18% for depressive symptoms (HAD-D). In the SART, students in the Control-SST group made significantly more commission errors than those in the Delayed-SST group at baseline (respectively 15.7, SD = 5.9 vs 12.7, SD = 5.6, p = 0.026). They also exhibited a tendency towards faster reaction time (181.62 ms, SD = 83.77 vs 221.44 ms, SD = 94.86) compared to the Delayed-SST group (p = 0.064). Regarding cognitive performances, no significant group differences were found on the STROOP test, assessing inhibitory control, selective attention and processing speed.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 3\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDifference-in-differences\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDifference-in-differences analyses for the questionnaires and cognitive tests are detailed in table 4. Between T0 and T1, students in the Control-SST group showed a significant increase in sleepiness score of 2 points on the FSSA scale (SD = 3.5) whereas no statistically significant change was observed in the Delayed-SST group (0.0, SD = 4.1), resulting in a medium effect size (Cohen’s d = 0.52, p = 0.042).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe score of anxiety significantly decreased in the Delayed-SST between T0 and T1, whereas no significant change was observed in the Control-SST group (difference-in-differences -1.7; 95% CI [-3.6 ; +0.1], p = 0.037). The evolution of depressive scores was not significantly different between the 2 groups (difference-in-differences -1.3; 95% CI [-2.9 ; +0.2], p = 0.077).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePerformances on the Sustained Attention Response Task (SART) indicated a significant reduction in commission errors in the Delayed-SST group (-4.5 errors, SD = 4.8, p\u0026lt;0.001) but not in the control group (-2.2, SD = 6.7, p = 0.081). This group-by-time difference approached statistical significance (difference-in-differences = -2.3; 95% CI [-5.0 ; +0.4], \u003cem\u003ep\u003c/em\u003e = 0.051) suggesting a trend toward improved impulsivity control following the intervention. No significant group differences were obtained in changes in omission errors or in reaction times (respectively d = 0.19, p = 0.899 and d = -0.07, p = 0.554).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRegarding the Stroop task, performance did not significantly change in the Control-SST group between T0 and T1 (+17.9 ms, SD = 71.1, p = 0.158), whereas performance improved in the Delayed-SST group (-49.9 ms, SD = 99.5, p = 0.005). The resulting difference-in-differences was 67.9 ms (95% CI [-108.3 ; -27.4], Cohen’s d = -0.79, p = 0.001), reflecting a significant improvement in inhibitory control and selective attention following the delayed school start.\u003c/p\u003e\n\u003cp\u003eTable 4\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis controlled study evaluated the impact of a one-hour delay in school start time on sleep patterns, cognitive functioning and mental health in early adolescents. Our findings demonstrate significant benefits of this intervention on multiple outcomes, reinforcing the growing body of literature advocating for later school start times for adolescents.\u003c/p\u003e\n\u003cp\u003eObjective measurements of sleep revealed that delaying school start times by one hour effectively counteracts the typical pubertal decline in sleep duration observed during adolescence.\u003csup\u003e9\u003c/sup\u003e While students in the Control-SST group exhibited the expected decline in sleep duration over the six-month period (-15.6 min), those in the Delayed-SST group showed a modest gain (+7.2 min), yielding a difference-in-differences of 23 minutes. This result closely aligns with the meta-analysis by Bowers et al.\u003csup\u003e17\u003c/sup\u003e which reported an average gain of 20 minutes in total sleep time across longitudinal studies implementing delayed school start times (weighted effect size of d = 0.38, p\u0026lt;0.001). Importantly, the magnitude of benefit appears to scale with the extent of the delay; longer school start delays allowed greater increase in sleep duration. However, Bowers et al.\u003csup\u003e17\u003c/sup\u003e were unable to evaluate potential moderators such as sex, age or the type of sleep measures (subjective vs objective) due to insufficient reporting or limited variability across studies. Given the impossibility of blinding school start time to participants, the reliance on self-reported sleep data may create expectancy effects and social desirability bias potentially inflating estimates of effect sizes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur findings suggest that the benefits of later school start times are already seen beneficial in early adolescence, before the full onset of pubertal circadian shifts. This developmental period is characterized by the initial emergence of neurobiological changes, most notably a progressive delay in melatonin secretion timing\u003csup\u003e5\u003c/sup\u003e and a shift towards evening sleep–wake preferences.\u003csup\u003e34\u003c/sup\u003e This developmental trajectory makes early school start times increasingly misaligned with adolescents' sleep needs. Our data further supports this interpretation showing that early adolescents with an evening chronotype benefited most from the delayed start time. As circadian phase delay intensifies throughout puberty, the benefits of later start times are likely to become more pronounced with age.\u003csup\u003e35\u003c/sup\u003e Further studies are needed to better characterize how pubertal maturation, from early through late adolescence, modulates the effects of delayed start times on sleep and daytime functioning.\u003c/p\u003e\n\u003cp\u003eImportantly, we observed that delayed school start time did not significantly affect sleep onset time. This finding is important as one of the main challenges to implementing later school start times is the common misconception that students will simply go to bed later, thereby negating any potential benefits. Instead, the preserved bedtime combined with later wake times resulted in extended sleep duration. This pattern has been consistently reported in prior studies.\u003csup\u003e18,36–38\u003c/sup\u003e In parallel with this objective increased in sleep duration, subjective sleepiness was also improved. While students in the control group reported increased sleepiness over time, those in the intervention group maintained stable levels, resulting in a medium effect size difference between groups. This results confirms the ubiquitously reported improved sleepiness in delayed school start studies.\u003csup\u003e17\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBeyond improvements in sleep quantity and sleepiness our study demonstrated relevant effects on mental health. Anxiety symptoms significantly decreased in the Delayed-SST group, while remaining unchanged in controls. Changes in depressive symptoms did not reach significance, the observed trend was in the expected direction. Those results are in line with longitudinal studies linking insufficient sleep and poor sleep quality to increased anxiety and depressive symptoms in adolescents.\u003csup\u003e39\u003c/sup\u003e Previous studies have also reported improvement in mood,\u003csup\u003e40\u003c/sup\u003e well-being \u003csup\u003e41\u003c/sup\u003e and depressive symptoms following a delayed school start time.\u003csup\u003e37,38,42\u003c/sup\u003e Similarly, sleep-based interventions have also shown improvement in anxiety among adolescents.\u003csup\u003e43\u003c/sup\u003e For example, Blake and colleagues\u003csup\u003e44\u003c/sup\u003e conducted a randomized control trial implementing a cognitive-behavioral based sleep intervention in 123 adolescents, and found improvement in anxiety symptoms, although not in depressive symptoms.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Few studies have directly assessed the effect of delayed school start times on standardized cognitive measures. Alfonsi and colleagues\u003csup\u003e45\u003c/sup\u003e demonstrated that students starting at 8:00 am exhibited progressive declines in psychomotor vigilance performance across the academic year, while those starting at 9:00 a.m. maintained stable performance, supporting the role of delayed school start in preserving sustained attention. Lufi et al\u003csup\u003e46\u003c/sup\u003e also reported improvement in sustained and selective attention after 2 weeks of delayed school start time. In our study, the intervention was also associated with improvement in cognitive functioning. We observed a significant reduction in the Stroop interference effect, with a robust difference-in-differences effect size (d = -0.79), indicating improved inhibitory control and selective attention in the Delayed-SST group. Similarly, a trend toward improved performance was also observed in the SART, with reduced commission errors, suggesting improved impulsivity control. These findings are consistent with established evidence that prefrontal cortex functions are particularly vulnerable to sleep restriction in adolescents.\u003csup\u003e47\u003c/sup\u003e These experimental findings are further reinforced by converging real-world behavioral evidence showing that adolescents with insufficient or poor sleep-quality are more susceptible to impulsivity-related outcomes. These include higher sensation-seeking tendencies, increased transport risk taking, greater involvement in violent or delinquent behaviors and increased substance use.\u003csup\u003e48\u003c/sup\u003e Taken together, this body of evidence highlights the critical role of sleep in supporting cognitive control and psychosocial development during adolescence. Therefore, by counteracting sleep deprivation delaying school start times could serve as a scalable, systemic intervention to mitigate sleep-related vulnerabilities in this age group.\u003c/p\u003e\n\u003cp\u003eAlthough we did not assess the impact of the intervention on academic performances, some studies have observed positive associations between later school start times and improved academic performance.\u003csup\u003e19,49,50\u003c/sup\u003e However, the systematic review by Biller and colleagues,\u003csup\u003e51\u003c/sup\u003e concluded that no generalizable improvement in academic achievement can be drawn due to substantial heterogeneity in academic outcomes, study designs and delay durations. Focusing on peer-reviewed longitudinal studies with a control group, three reported no change in global academic outcomes\u003csup\u003e50,52,53\u003c/sup\u003e (such as grade point average), one found a GPA improvement after 2 years,\u003csup\u003e19\u003c/sup\u003e and others reported subject-specific improvements (such as reading or math).\u003csup\u003e49,50\u003c/sup\u003e Authors underscore the limitations of using academic test scores alone as indicators of the success or failure of such interventions.\u003csup\u003e51,52\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eDespite the demonstrated efficacy of the intervention, the average sleep duration in the delayed-SST group remained below the recommended 8 to 10 hour per night for adolescents.\u003csup\u003e54\u003c/sup\u003e This aligns with previous findings that even when structural changes are implemented, a substantial proportion of adolescents continue to obtain insufficient sleep. Bower et al. reported in their meta-analysis\u003csup\u003e17\u003c/sup\u003e a post intervention average sleep duration of 7 hours and 24 minutes.\u003csup\u003e54\u003c/sup\u003e The present intervention allowed to partially counter the prevalence of chronically sleep deprived students, with 50% in the Control-SST group sleeping less than 7 hours compared to 13% in the Delayed-SST Group at follow-up. However, even if this reduction in the prevalence of short sleep is meaningful, it indicates that later start times alone are insufficient. These results highlight the need to consider additional structural and cultural measures alongside start time adjustments to promote adequate sleep in adolescents. Complementary interventions, such as limiting screen electronic devices,\u003csup\u003e55\u003c/sup\u003e limiting late extracurricular activities and regulating electronic assignment submission times, may be necessary to foster a school environment that is more conducive to healthy sleep. A broader cultural shift that recognizes the importance of sleep for learning, emotional regulation and health is needed to help adolescents meet their biological sleep needs.\u003c/p\u003e\n\u003cp\u003eThe present study has several strengths, including the use of objective measures of sleep, standardized cognitive tests, a before/after study design with a control group and a randomized allocation ensuring baseline comparability. However, the relatively small sample size, particularly for actigraphy analyses due to compliance issues, should be acknowledged. The actigraphy sample size was reduced due to compliance issues, particularly among boys and participants in the Control-SST group, possibly reflecting lower engagement in the absence of intervention benefit. This attrition can introduce a possible selection bias, although baseline comparisons did not differ significantly between included and excluded participants. Additionally, subgroup analyses (e.g., by chronotype) lacked statistical power. As previous studies have experienced, recruiting schools to participate in a randomized study which implies modification of schedules is particularly difficult.\u003csup\u003e20\u003c/sup\u003e The controlled boarding school setting of our study, while controlling for environmental variability and commute facilities, may limit generalizability to typical day school populations, where evening routines are less regulated.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDespite strong scientific consensus and recommendations from major professional organizations including the National Sleep Foundation,\u003csup\u003e56\u003c/sup\u003e the American Academy of Sleep Medicine,\u003csup\u003e57\u003c/sup\u003e the American Academy of Pediatrics,\u003csup\u003e58\u003c/sup\u003e implementation of delayed school start times remains limited. In the United States, 13 % of middle schools start at 9 a.m. or later,\u003csup\u003e59\u003c/sup\u003e while in France middle school uniformly starts at 8 a.m. Kelley et al\u003csup\u003e60\u003c/sup\u003e demonstrated that even later start times (e.g., 10:00 a.m.) can yield further benefits, reinforcing the argument that current schedules are misaligned with adolescent biology. The primary barriers to implementation are logistical rather than scientific, with school districts citing concerns about transportation, after-school activities and families schedules.\u003csup\u003e20,61\u003c/sup\u003e This gap between evidence and practice illustrates how operational constraints can impede the translation of sleep research into educational practice, leaving many adolescents with chronic sleep deprivation despite effective solutions. Interestingly, delayed start times have also been associated with improved sleep in parents\u003csup\u003e62\u003c/sup\u003e and teachers,\u003csup\u003e63\u003c/sup\u003e and parent-teacher associations are increasingly advocating a delay in school.\u003csup\u003e64\u003c/sup\u003e Whether this growing support from advocacy groups will be enough to overcome long-standing logistical challenges and lead to widespread policy change remains an open question.\u003c/p\u003e\n\u003cp\u003eIn conclusion, this controlled trial provides compelling evidence that a one-hour delay in school start time significantly improve adolescent sleep duration and daytime functioning. The intervention effectively counteracted the natural developmental decline in sleep without delaying bedtimes, resulting in both objective sleep improvements and enhanced cognitive performance, particularly for inhibitory control. These findings add to the growing scientific consensus supporting later school start times as a public health strategy to mitigate sleep deprivation in adolescents. Educational policymakers should consider these benefits when weighing the logistical challenges of implementing delayed start times against the potential improvements in student well-being and cognitive functioning.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eData availability\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe would like to acknowledge to all participants, as well as to the teachers and the school head, whose time, effort, and support in coordinating schedules and facilitating the assessment were instrumental in making this study possible.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eWe would also like to acknowledge the National Education Scientific Council (Conseil scientifique de l\u0026apos;\u0026Eacute;ducation nationale) for the financial support and the IDEE program (ANR-21-ESRE-0034) the administrative, technical and statistical support.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eE.R. designed the study, carried out the analyses and drafted the initial manuscript. E.R., A.G., L.M. collected data., A.G., A.P and M.G. provided technical and statistical support. A.E.R. and S.M. coordinated and supervised study planification and conceptualization. All authors critically reviewed and revised the manuscript.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":" References","content":"\u003col\u003e\n\u003cli\u003eFeinberg, I., Davis, N. M., de Bie, E., Grimm, K. J. \u0026amp; Campbell, I. G. The maturational trajectories of NREM and REM sleep durations differ across adolescence on both school-night and extended sleep. Am. \u003cem\u003eJ. Physiol.-Regul. Integr. Comp. Physiol.\u003c/em\u003e \u003cstrong\u003e302\u003c/strong\u003e, R533\u0026ndash;R540 (2012).\u003c/li\u003e\n\u003cli\u003eOhayon, M. M., Carskadon, M. A., Guilleminault, C. \u0026amp; Vitiello, M. V. Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 255-1273 (2004).\u003c/li\u003e\n\u003cli\u003eCampbell, I. et al. Adolescent changes in homeostatic regulation of EEG activity in the delta and theta frequency bands during NREM sleep. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 83-91 (2011).\u003c/li\u003e\n\u003cli\u003eCrowley, S. J. et al. A longitudinal assessment of sleep timing, circadian phase, and phase angle of entrainment across human adolescence. \u003cem\u003ePloS One\u003c/em\u003e \u003cstrong\u003e9 (11)\u003c/strong\u003e, e112199 (2014).\u003c/li\u003e\n\u003cli\u003eKennaway, D. J. The dim light melatonin onset across ages, methodologies, and sex and its relationship with morningness/eveningness. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e46(5)\u003c/strong\u003e, zsad033 (2023).\u003c/li\u003e\n\u003cli\u003eHale, L. \u0026amp; Guan, S. Screen time and sleep among school-aged children and adolescents: A systematic literature review. \u003cem\u003eSleep Med. Rev\u003c/em\u003e. \u003cstrong\u003e21\u003c/strong\u003e, 50\u0026ndash;58 (2015).\u003c/li\u003e\n\u003cli\u003eCarskadon, M. A. Sleep in adolescents: the perfect storm. \u003cem\u003ePediatr. Clin. North Am\u003c/em\u003e. \u003cstrong\u003e58\u003c/strong\u003e, 637\u0026ndash;647 (2011).\u003c/li\u003e\n\u003cli\u003eNational Sleep Foundation. Sleep in the Modern Family. https://sleepfoundation.org/sleep-polls-data/sleep-in-america-poll/2014-sleep-and-family (2014).\u003c/li\u003e\n\u003cli\u003eSadeh, A., Dahl, R. E., Shahar, G. \u0026amp; Rosenblat-Stein, S. Sleep and the transition to adolescence: a longitudinal study. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 1602\u0026ndash;1609 (2009).\u003c/li\u003e\n\u003cli\u003eLeger, D., Beck, F., Richard, J.-B. \u0026amp; Godeau, E. Total Sleep Time Severely Drops during Adolescence. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e45204 (2012).\u003c/li\u003e\n\u003cli\u003eCrowley, S. J., Acebo, C. \u0026amp; Carskadon, M. A. Sleep, circadian rhythms, and delayed phase in adolescence. \u003cem\u003eSleep Med\u003c/em\u003e. \u003cstrong\u003e8\u003c/strong\u003e, 602\u0026ndash;612 (2007).\u003c/li\u003e\n\u003cli\u003eOwens, J., Adolescent Sleep Working Group, \u0026amp; Committee on Adolescence. Insufficient sleep in adolescents and young adults: an update on causes and consequences. \u003cem\u003ePediatrics\u003c/em\u003e \u003cstrong\u003e134\u003c/strong\u003e, e921-932 (2014).\u003c/li\u003e\n\u003cli\u003eRolling, J. et al. Sleep and circadian rhythms in adolescents with attempted suicide. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 8354 (2024).\u003c/li\u003e\n\u003cli\u003eWinsler, A., Deutsch, A., Vorona, R. D., Payne, P. A. \u0026amp; Szklo-Coxe, M. Sleepless in Fairfax: the difference one more hour of sleep can make for teen hopelessness, suicidal ideation, and substance use.\u003cem\u003e J. Youth Adolesc. \u003c/em\u003e\u003cstrong\u003e44\u003c/strong\u003e, 362\u0026ndash;378 (2015).\u003c/li\u003e\n\u003cli\u003eWheaton, A. G., Chapman, D. P. \u0026amp; Croft, J. B. School Start Times, Sleep, Behavioral, Health, and Academic Outcomes: a Review of the Literature. \u003cem\u003eJ. Sch. Health\u003c/em\u003e \u003cstrong\u003e86\u003c/strong\u003e, 363\u0026ndash;381 (2016).\u003c/li\u003e\n\u003cli\u003eMinges, K. E. \u0026amp; Redeker, N. S. Delayed school start times and adolescent sleep: A systematic review of the experimental evidence. \u003cem\u003eSleep Med. Rev\u003c/em\u003e. \u003cstrong\u003e28\u003c/strong\u003e, 86\u0026ndash;95 (2016).\u003c/li\u003e\n\u003cli\u003eBowers, J. M. \u0026amp; Moyer, A. Effects of school start time on students\u0026rsquo; sleep duration, daytime sleepiness, and attendance: a meta-analysis. \u003cem\u003eSleep Health\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 423\u0026ndash;431 (2017).\u003c/li\u003e\n\u003cli\u003eWidome, R. et al. The START study: An evaluation to study the impact of a natural experiment in high school start times on adolescent weight and related behaviors. \u003cem\u003eJAMA Pediatr\u003c/em\u003e. \u003cstrong\u003e6\u003c/strong\u003e, 66 (2020).\u003c/li\u003e\n\u003cli\u003eJames, S. A., Erickson, D. J., Lammert, S. \u0026amp; Widome, R. School start time delays and high school educational outcomes: Evidence from the START/LEARN study. \u003cem\u003eJ. Adolesc.\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 751\u0026ndash;763 (2023).\u003c/li\u003e\n\u003cli\u003eIllingworth, G. et al. Challenges in implementing and assessing outcomes of school start time change in the UK: experience of the Oxford Teensleep study. \u003cem\u003eSleep Med.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 89\u0026ndash;95 (2019).\u003c/li\u003e\n\u003cli\u003eStoet, G. PsyToolkit: A software package for programming psychological experiments using Linux. \u003cem\u003eBehav. Res. Methods\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 1096\u0026ndash;1104 (2010).\u003c/li\u003e\n\u003cli\u003eStoet, G. PsyToolkit: A Novel Web-Based Method for Running Online Questionnaires and Reaction-Time Experiments. \u003cem\u003eTeach. Psychol\u003c/em\u003e. \u003cstrong\u003e44\u003c/strong\u003e, 24\u0026ndash;31 (2017).\u003c/li\u003e\n\u003cli\u003eCellini, N., Buman ,Matthew P., McDevitt ,Elizabeth A., Ricker ,Ashley A. \u0026amp; and Mednick, S. C. Direct comparison of two actigraphy devices with polysomnographically recorded naps in healthy young adults. \u003cem\u003eChronobiol. Int.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 691\u0026ndash;698 (2013).\u003c/li\u003e\n\u003cli\u003eYang, F. N., Picchioni, D. \u0026amp; Duyn, J. H. Effects of sleep-corrected social jetlag on measures of mental health, cognitive ability, and brain functional connectivity in early adolescence.\u003cem\u003e Sleep\u003c/em\u003e \u003cstrong\u003e46,\u003c/strong\u003e zsad259 (2023).\u003c/li\u003e\n\u003cli\u003eGustin, M.-P. et al. French Sleepiness Scale for Adolescents-8 items: A discriminant and diagnostic validation. \u003cem\u003eL\u0026rsquo;Enc\u0026eacute;phale\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 109\u0026ndash;116 (2023).\u003c/li\u003e\n\u003cli\u003eJanssen, K. C., Phillipson, S., O\u0026rsquo;Connor, J. \u0026amp; Johns, M. W. Validation of the Epworth Sleepiness Scale for Children and Adolescents using Rasch analysis. \u003cem\u003eSleep Med.\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 30\u0026ndash;35 (2017).\u003c/li\u003e\n\u003cli\u003eWhite, D., Leach, C., Sims, R., Atkinson, M. \u0026amp; Cottrell, D. Validation of the Hospital Anxiety and Depression Scale for use with adolescents. \u003cem\u003eBr. J. Psychiatry J. Ment. Sci\u003c/em\u003e. \u003cstrong\u003e175\u003c/strong\u003e, 452\u0026ndash;454 (1999).\u003c/li\u003e\n\u003cli\u003eCarskadon, M. A., Vieira, C. \u0026amp; Acebo, C. Association between puberty and delayed phase preference. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 258\u0026ndash;262 (1993).\u003c/li\u003e\n\u003cli\u003eRobertson, I. H., Manly, T., Andrade, J., Baddeley, B. T. \u0026amp; Yiend, J. \u0026lsquo;Oops!\u0026rsquo;: Performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. \u003cem\u003eNeuropsychologia\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 747\u0026ndash;758 (1997).\u003c/li\u003e\n\u003cli\u003eStroop, J. R. Studies of interference in serial verbal reactions.\u003cem\u003e J. Exp. Psychol. Gen.\u003c/em\u003e \u003cstrong\u003e121\u003c/strong\u003e, 15\u0026ndash;23 (1992).\u003c/li\u003e\n\u003cli\u003eMacLeod, C. M. Half a century of research on the Stroop effect: An integrative review. \u003cem\u003ePsychol. Bull. \u003c/em\u003e\u003cstrong\u003e109\u003c/strong\u003e, 163\u0026ndash;203 (1991).\u003c/li\u003e\n\u003cli\u003eR Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing (2023).\u003c/li\u003e\n\u003cli\u003eNtani, G., Inskip, H., Osmond, C. \u0026amp; Coggon, D. Consequences of ignoring clustering in linear regression. \u003cem\u003eBMC Med. Res. Methodol\u003c/em\u003e. \u003cstrong\u003e21\u003c/strong\u003e, 139 (2021).\u003c/li\u003e\n\u003cli\u003eHagenauer, M. H., Perryman, J. I., Lee, T. M. \u0026amp; Carskadon, M. A. Adolescent Changes in the Homeostatic and Circadian Regulation of Sleep. \u003cem\u003eDev. Neurosci.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 276\u0026ndash;284 (2009).\u003c/li\u003e\n\u003cli\u003eDagys, N. et al. Double trouble? The effects of sleep deprivation and chronotype on adolescent affect. \u003cem\u003eJ. Child Psychol. Psychiatry\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 660\u0026ndash;667 (2012).\u003c/li\u003e\n\u003cli\u003eAdam, E. K., Snell, E. K. \u0026amp; Pendry, P. Sleep timing and quantity in ecological and family context: a nationally representative time-diary study. \u003cem\u003eJ. Fam. Psychol. JFP J. Div. Fam. Psychol. Am. Psychol. \u003c/em\u003e\u003cem\u003eAssoc. Div\u003c/em\u003e. \u003cstrong\u003e43\u003c/strong\u003e 21, 4\u0026ndash;19 (2007).\u003c/li\u003e\n\u003cli\u003eBoergers, J., Gable, C. J. \u0026amp; Owens, J. A. Later school start time is associated with improved sleep and daytime functioning in adolescents. \u003cem\u003eJ. Dev. Behav. Pediatr.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 11\u0026ndash;17 (2014).\u003c/li\u003e\n\u003cli\u003eWahlstrom, K. Changing Times: Findings From the First Longitudinal Study of Later High School Start Times. \u003cem\u003eNASSP Bull.\u003c/em\u003e \u003cstrong\u003e86\u003c/strong\u003e, 3\u0026ndash;21 (2002).\u003c/li\u003e\n\u003cli\u003eMcMakin, D. L. \u0026amp; Alfano, C. A. Sleep and anxiety in late childhood and early adolescence: \u003cem\u003eCurr. Opin. Psychiatry\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 483\u0026ndash;489 (2015).\u003c/li\u003e\n\u003cli\u003eWhitaker, R. C., Dearth-Wesley, T., Herman, A. N., Oakes, J. M. \u0026amp; Owens, J. A. A quasi-experimental study of the impact of school start time changes on adolescents\u0026rsquo; mood, self-regulation, safety, and health. \u003cem\u003eSleep Health\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 466\u0026ndash;469 (2019).\u003c/li\u003e\n\u003cli\u003eLo, J. C. et al. Sustained benefits of delaying school start time on adolescent sleep and well-being. Sleep \u003cstrong\u003e41(6)\u003c/strong\u003e, zsy052 (2018).\u003c/li\u003e\n\u003cli\u003eOwens, J. A., Belon, K. \u0026amp; Moss, P. Impact of delaying school start time on adolescent sleep, mood, and behavior. A\u003cem\u003erch. Pediatr. Adolesc. Med.\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, 608\u0026ndash;614 (2010).\u003c/li\u003e\n\u003cli\u003eBlake, M. J., Sheeber, L. B., Youssef, G. J., Raniti, M. B. \u0026amp; Allen, N. B. Systematic Review and Meta-analysis of Adolescent Cognitive\u0026ndash;Behavioral Sleep Interventions. \u003cem\u003eClin. Child Fam. Psychol. Rev.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 227\u0026ndash;249 (2017).\u003c/li\u003e\n\u003cli\u003eBlake, M. J. et al. A cognitive-behavioral and mindfulness-based group sleep intervention improves behavior problems in at-risk adolescents by improving perceived sleep quality. \u003cem\u003eBehav. Res. Ther. \u003c/em\u003e\u003cstrong\u003e99,\u003c/strong\u003e 147\u0026ndash;156 (2017).\u003c/li\u003e\n\u003cli\u003eAlfonsi, V. et al. The Association Between School Start Time and Sleep Duration, Sustained Attention, and Academic Performance. \u003cem\u003eNat. Sci. Sleep\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1161\u0026ndash;1172 (2020).\u003c/li\u003e\n\u003cli\u003eLufi, D., Tzischinsky, O. \u0026amp; Hadar, S. Delaying School Starting Time by One Hour: Some Effects on Attention Levels in Adolescents. J\u003cem\u003e. Clin. Sleep Med\u003c/em\u003e.\u003cstrong\u003e 07\u003c/strong\u003e, 137\u0026ndash;143 (2011).\u003c/li\u003e\n\u003cli\u003eAnastasiades, P. G., de Vivo, L., Bellesi, M. \u0026amp; Jones, M. W. Adolescent sleep and the foundations of prefrontal cortical development and dysfunction. \u003cem\u003eProg. Neurobiol\u003c/em\u003e. \u003cstrong\u003e218\u003c/strong\u003e, 102338 (2022).\u003c/li\u003e\n\u003cli\u003eShort, M. A. \u0026amp; Weber, N. Sleep duration and risk-taking in adolescents: A systematic review and meta-analysis. \u003cem\u003eSleep Med. Rev. \u003c/em\u003e\u003cstrong\u003e41,\u003c/strong\u003e 185\u0026ndash;196 (2018).\u003c/li\u003e\n\u003cli\u003eJung, H. A late bird or a good bird? The effect of 9 o\u0026rsquo;clock attendance policy on student\u0026rsquo;s achievement. \u003cem\u003eAsia Pac. Educ. Rev.\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 511\u0026ndash;529 (2018).\u003c/li\u003e\n\u003cli\u003eKim, T. The Effects of School Start Time on Educational Outcomes: Evidence from the 9 O\u0026rsquo;clock Attendance Policy in South Korea. \u003cem\u003eBE J. Econ. Anal. Policy\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 439\u0026ndash;474 (2022).\u003c/li\u003e\n\u003cli\u003eBiller, A. M., Meissner, K., Winnebeck, E. C. \u0026amp; Zerbini, G. School start times and academic achievement - A systematic review on grades and test scores. \u003cem\u003eSleep Med. Rev.\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 101582 (2022).\u003c/li\u003e\n\u003cli\u003eLenard, M., Morrill, M. S. \u0026amp; Westall, J. High school start times and student achievement: Looking beyond test scores. \u003cem\u003eEcon. Educ. Rev.\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 101975 (2020).\u003c/li\u003e\n\u003cli\u003eRhie, S. \u0026amp; Chae, K. Y. Effects of school time on sleep duration and sleepiness in adolescents. \u003cem\u003ePloS One\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, e0203318 (2018).\u003c/li\u003e\n\u003cli\u003eParuthi, S. et al. Consensus Statement of the American Academy of Sleep Medicine on the Recommended Amount of Sleep for Healthy Children: Methodology and Discussion. \u003cem\u003eJ. Clin. Sleep Med\u003c/em\u003e. \u003cstrong\u003e12\u003c/strong\u003e, 1549\u0026ndash;1561 (2016).\u003c/li\u003e\n\u003cli\u003ePerrault, A. A. et al. Reducing the use of screen electronic devices in the evening is associated with improved sleep and daytime vigilance in adolescents. \u003cem\u003eSleep\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, zsz125 (2019).\u003c/li\u003e\n\u003cli\u003eZiporyn, T. D. et al. Adolescent sleep health and school start times: Setting the research agenda for California and beyond. A research summit summary. \u003cem\u003eSleep Health J. Natl. Sleep Found\u003c/em\u003e. \u003cstrong\u003e8\u003c/strong\u003e, 11\u0026ndash;22 (2022).\u003c/li\u003e\n\u003cli\u003eWatson, N. F. et al. Delaying Middle School and High School Start Times Promotes Student Health and Performance: An American Academy of Sleep Medicine Position Statement. \u003cem\u003eJ. Clin. Sleep Med\u003c/em\u003e. \u003cstrong\u003e13\u003c/strong\u003e, 623\u0026ndash;625 (2017).\u003c/li\u003e\n\u003cli\u003eAdolescent Sleep Working Group, Committee on Adolescence, \u0026amp; Council on School Health. School start times for adolescents. \u003cem\u003ePediatrics\u003c/em\u003e \u003cstrong\u003e134\u003c/strong\u003e, 642\u0026ndash;649 (2014).\u003c/li\u003e\n\u003cli\u003eTaie, S. \u0026amp; Lewis, L. Characteristics of 2020\u0026ndash;21 Public and Private K\u0026ndash;12 Schools in the United States: Results From the National Teacher and Principal Survey First Look (NCES 2022-111). https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid = 2022111 (2022).\u003c/li\u003e\n\u003cli\u003eKelley, P., Lockley, S. W., Kelley, J. \u0026amp; Evans, M. D. R. Is 8:30 a.m. Still Too Early to Start School? A 10:00 a.m. School Start Time Improves Health and Performance of Students Aged 13-16. \u003cem\u003eFront. Hum. Neurosci.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 588 (2017).\u003c/li\u003e\n\u003cli\u003eFitzpatrick, J. M., Silva, G. E. \u0026amp; Vana, K. D. Perceived Barriers and Facilitating Factors in Implementing Delayed School Start Times to Improve Adolescent Sleep Patterns. \u003cem\u003eJ. Sch. Health \u003cstrong\u003e91.2\u003c/strong\u003e, 94-101 \u003c/em\u003e(2021) doi:10.1111/josh.12983.\u003c/li\u003e\n\u003cli\u003eMeltzer, L. J., Wahlstrom, K. L., Plog, A. E. \u0026amp; McNally, J. Impact of changing school start times on parent sleep. \u003cem\u003eSleep Health\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 130\u0026ndash;134 (2022).\u003c/li\u003e\n\u003cli\u003eWahlstrom, K. L., Plog, A. E., McNally, J. \u0026amp; Meltzer, L. J. Impact of Changing School Start Times on Teacher Sleep Health and Daytime Functioning. \u003cem\u003eJ. Sch. Health\u003c/em\u003e \u003cstrong\u003e93\u003c/strong\u003e, 128\u0026ndash;134 (2023).\u003c/li\u003e\n\u003cli\u003eTrevorrow, T., Zhou, E. S., Dietch, J. R. \u0026amp; Gonzalez, B. D. Position statement: start middle and high schools at 8:30 am or later to promote student health and learning. \u003cem\u003eTransl. Behav. Med.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 167\u0026ndash;169 (2019).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 4 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"School start time, sleep, attention, cognition, adolescent, actigraphy","lastPublishedDoi":"10.21203/rs.3.rs-7085731/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7085731/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Early school start times (SST) force adolescents to wake earlier than biologically preferred, creating chronic sleep debt. While evidence supports delaying SSTs, few controlled studies have used objective measures to assess their impact. We conducted a controlled intervention in a French boarding school to evaluate the effects of a one-hour delay in SST on sleep (N=50) and cognitive functioning (N=73) in early adolescents (age 12.8 years [11.7-14.2], 66% girls). After a baseline period with 8 a.m. starts (T0), four classes were randomized: half remained at the early schedule (Control-SST), half switched to 9 a.m. (Delayed-SST). Sleep measured by actigraphy, cognitive performance and mental health were assessed at baseline (T0) and 6 months later (T1). Total sleep time had decreased in Control-SST but increased in Delayed-SST, resulting at T1 in a 26-minute difference (Cohen's d=0.93, p=0.007). Sleep onset time did not differ between groups. Sleepiness and anxiety decreased in Delayed-SST but increased in Control-SST (d=-0.52, p=0.042 and d=-0.44, p=0.037 respectively). Inhibitory control improved in Delayed-SST compared to Control-SST (d=-0.79, p=0.001), with trends toward better sustained attention (d=-0.40, p=0.051). Delaying SST by one hour effectively counteracts typical pubertal sleep loss, and benefits both cognitive functioning and mental health outcomes in early adolescence.","manuscriptTitle":"An Extra Hour Matters: Delaying School Start Time Improves Sleep, Inhibitory Control, and Anxiety in Early Adolescents","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-01 16:37:53","doi":"10.21203/rs.3.rs-7085731/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-06T13:25:53+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-04T17:37:31+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-01T18:54:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"182119347665866135133654902043259605944","date":"2026-02-22T20:04:23+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"69454636870505900743523553711962851790","date":"2026-02-19T22:54:00+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-06T09:21:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"290107096438628403694124828228873863926","date":"2026-01-02T12:54:18+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-26T10:26:13+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-01T06:17:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-31T18:37:07+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-16T16:38:31+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-07-16T15:15:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cd4228fa-42b1-443a-84cc-6ccd8aad6a1c","owner":[],"postedDate":"December 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":58670658,"name":"Health sciences/Health care"},{"id":58670659,"name":"Health sciences/Medical research"},{"id":58670660,"name":"Biological sciences/Neuroscience"},{"id":58670661,"name":"Biological sciences/Psychology"},{"id":58670662,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-04-24T06:10:22+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-01 16:37:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7085731","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7085731","identity":"rs-7085731","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}