Emerging adulthood is a time of risk for sleep problems. An epidemic of shortened sleep times and resultant sleepiness in teens and young adults has been reported worldwide [1,2]. Sleep problems are reported in two-thirds of adolescents and young adults (AYAs) [3-5] and are linked to academic and mental health problems and suicidality [6-11]. AYAs also experience a pubertal shift toward later or delayed circadian rhythmicity, such that they become tired later, resulting in later rise times, although they must maintain early rise times to attend school [12-15]. This biological shift toward a later sleep onset time is further compounded by the fact that AYAs are sensitive to evening light [12,16,17] and are exposed to academic (eg, early school start times) and social pressures (eg, peers reinforcing later activities), which can increase sleeplessness. Moreover, some AYAs exert control over what they eat and drink (eg, caffeine) without an adequate understanding of the consequences of particular behaviors on their sleep and general health [18]. As a result, AYAs have chronic sleep loss, resulting in pathological sleepiness, a condition associated with an increased risk of motor vehicle accidents and poor emotional, interpersonal, and academic functioning [19,20].

Providing accessible interventions for sleep problems is crucial to mitigate personal, social, and economic costs. Treatment options for most sleep problems typically consist of pharmacological and behavioral approaches [21,22]. Sleep medications are associated with significant concerns regarding use by AYAs [23]. Although cognitive behavioral therapy for insomnia (CBT-I) approaches are the most efficacious for helping individuals with insomnia with few troublesome side effects [24], few AYAs have access to behavioral sleep medicine treatments [25,26]. Furthermore, CBT-I protocols (when accessible) do not target the range of sleep problems that AYAs experience beyond insomnia symptoms, including hypersomnia, excessive daytime sleepiness (eg, from voluntary sleep restriction), delayed phase circadian rhythm, and often uninformed use of sleep-interfering substances, some of which may be experienced concurrently, highlighting the promise of transdiagnostic behavioral sleep treatments for AYAs [27-29]. There are multiple barriers to accessing nonpharmacological evidence-based sleep treatments, including poor dissemination of CBT-I and behavioral sleep treatments in real-world settings [30] and a shortage of providers who are trained in such treatments [31]. It is clear that AYAs have diverse sleep problems (ie, not just insomnia) and require tailored evidence-based strategies that comprehensively address these diverse needs, making transdiagnostic cognitive behavioral treatments [32] a more suitable option than nonadapted CBT-I.

Smartphone apps are uniquely suited to improve access for those who are not seeking treatment but are nevertheless at risk as well as treatment-seeking groups with poor access to evidence-based treatments, and reportedly increase users’ motivation, goal-setting, and confidence in and control over their ability to enact health-promoting behavior change [33]. AYAs are drawn to technology; therefore, technology provides an opportunity to directly reach this large invisible at-risk cohort for whom sleep disorders contribute to psychopathology and vice versa [34,35]. Although there are web-based programs for infant (aged 6-36 months) [36], pediatric (aged 1-8 years) [37,38], adolescent [39-41], and adult sleep disturbances [42], there are no evidence-based transdiagnostic programs that adequately address the multitude of sleep problems observed in AYAs and that span both adolescence and young adulthood. Put differently, as some of the sleep problems of AYAs are unique and go beyond insomnia (eg, voluntary sleep restriction, circadian phase delay, and poor sleep hygiene), pediatric and adult programs are not suitable for this age group.

In response to a dearth of appropriate treatment options for AYAs with a range of sleep difficulties beyond insomnia, we created a self-management app called DOZE (Delivering Online Zzz’s with Empirical Support), which was co-designed by AYAs in user consultation sessions conducted by the developers of the app. DOZE is based on the principles of cognitive behavioral sleep medicine [32] and allows AYAs access to effective, tailored strategies to address a range of sleep problems (eg, insomnia, daytime sleepiness, and delayed phase circadian rhythms), which also range from subclinical to clinical in terms of their severity. Importantly, consistent with their preferences for technology and autonomy and as revealed by the user consultation sessions, the self-management approach entrusts AYAs to set their own goals and make their own health behavior changes when provided with the tools and education to do so. This study reports on an open trial evaluating the app’s feasibility over a 4-week intervention period. The primary objective of this study is to evaluate whether AYAs regard evidence-based cognitive behavioral strategies in a self-management app (DOZE) as acceptable and credible. Toward the aim of evaluating feasibility, we also want to understand how AYAs use DOZE and which aspects of the app they find most valuable. As AYAs co-designed the app, we hypothesize that AYAs would rate the app as acceptable and credible and report high satisfaction. No a priori hypotheses are made regarding which aspects of the app would be most valuable to AYAs, and an exploratory analysis of app use patterns is conducted to understand which aspects of the app are most useful and to provide further insight into AYAs’ primary sleep complaints and treatment goals.

The secondary objective is to examine preliminary efficacy, acknowledging that these results may be quite modest because of the range of sleep problems addressed (eg, both insomnia and hypersomnia) and that, without sleep disturbance severity as part of the inclusion criteria, we might recruit a subclinical sample. Outcomes focus on whether DOZE positively affects health-related behavior change, improvements in daytime symptoms (ie, sleepiness, fatigue, energy, and psychological symptoms), and health-related quality of life. We hypothesize that AYAs who use DOZE would see improvements in sleep indices at end point. Furthermore, we hypothesize that the use of DOZE for 4 weeks may contribute to improvements in energy, mood, and health-related quality of life at end point compared with baseline.

Individuals between the ages of 15 and 24 years with self-reported dissatisfaction with their sleep were eligible to participate in this web-based study. Given that a clinical diagnosis of a sleep disorder is not required to use DOZE (ie, any AYA who wants to improve sleep can use it), this study did not have any other inclusion or exclusion criteria to be able to generalize the results of this study to the population that will use DOZE. Self-reported dissatisfaction with sleep was queried with a single question, “How do you feel about your sleep?” and participants could respond using an open textbox. Computer and internet literacy of participants was assumed. Participants were recruited using posters at Ryerson University and in the downtown Toronto community, dissemination of study information to mental health providers via listservs of professional organizations (eg, Canadian Association for Cognitive and Behavioural Therapy and Ontario Psychological Association), and via social media (eg, Twitter and Facebook). Recruitment materials advertised the opportunity for AYAs who were dissatisfied with their sleep or were experiencing difficulty sleeping to participate in a study testing a new sleep app for 4 weeks. Participation in this study was quasi-anonymous. Email confirmation was used to detect repeat participation. All participants provided informed consent and completed a quiz assessing their understanding of critical components of the consent form (detailed in the Procedure section).

DOZE [43] is a transdiagnostic, self-management, cognitive behavioral web-based app for sleep disturbance among AYAs. It is freely accessed from a web browser on any device (ie, it is not downloaded on a device). DOZE was designed using an iterative, experience-based co-design process that included feedback from AYA users and health care provider stakeholders [44-46]. Initial design consultation interviews conducted by an industry partner with AYA stakeholders revealed that, in addition to including a sleep diary for self-monitoring, AYAs wanted an app that provides feedback on their personal sleep habits (over and above feedback on their sleep), opportunities for goal-setting, and tips on how to make changes to their sleep. Following this initial phase, a user-informed redesign of DOZE was conducted with an industry partner (PIVOT Design), consistent with standard app development processes [47], including 6 user assessments with AYAs (aged 15-24 years) on iterative prototype versions of the app and consultations with health care provider stakeholders. Of note, AYAs who contributed to the co-design of DOZE were not participants in this study. Additional details about the co-design process are presented in Multimedia Appendix 1.

The final version of DOZE includes a sleep diary, adapted with permission from the Consensus Sleep Diary [48], comprising 8 questions that assess the following: bedtime (ie, time of getting into bed whether or not it was to initiate sleep), time of sleep attempt (ie, time of trying to sleep), time it took to fall asleep (sleep onset latency [SOL]), total time awake in the night (wake after sleep onset), time of final awakening (ie, time of waking in the morning), rise time (ie, time of getting out of bed), amount of time spent napping, and use of sleep-interfering substances within 2 hours of bedtime. DOZE’s onboarding screens instructed participants to complete their sleep diary every morning within 2 hours of waking, and they were reminded of this using a pop-up notification if they completed it that same day outside of the 2-hour window. Completing an entry for the previous day was not permitted. Participants also received an automated reminder email each morning to complete the diary. Sleep indices were calculated based on information provided by the sleep diary, including SOL, variability in rise time (range of earliest to latest rise time), variability in bedtime (range of earliest to latest bedtime), wake after sleep onset, morning lingering in bed (the difference between time of final awakening and rise time), total sleep time (TST; total time spent asleep), total wake time (TWT; total time spent awake during the sleep period), time in bed (TIB; bedtime to rise time), and sleep efficiency (SE; percent of TIB spent asleep). Participants were able to view their average sleep indices over a 2-week period on a personalized dashboard.

After completing 2 weeks of sleep diaries, DOZE provides personalized feedback on sleep indices (TST, SOL, TWT, TIB, and SE) relative to age-adjusted norms [4] by indicating whether their indices are normal, too high, or too low on the Progress screen. Users then have the opportunity to set goals to improve their sleep in problem areas that are identified (ie, indices are outside the normal range). Feedback areas, which all represent components of evidence-based treatment [32], include naps, too much/too little TIB, sleep-interfering substances, lingering in bed in the morning, sleepiness, and jetlag without traveling (ie, sleep schedule variability). AYAs have the opportunity to set personalized goals in these areas (eg, reduce jetlag by 1 hour, 2 hours, or 3 hours), which are added to their personal dashboard, or they may choose not to set goals in identified areas, accounting for various stages of readiness to change. Notably, if a feedback area is not relevant to the individual based on their sleep indices (eg, they are spending a normal amount of TIB), they are not provided the opportunity to set a goal to improve in that area (eg, they are not shown the feedback area too much/too little TIB). In addition, participants could access a Tips section that included quizzes and psychoeducation about hyperarousal (difficulty winding down), chronotype (night owl living in an early bird’s world), difficulty getting out of bed in the morning, daytime sleepiness (trouble staying awake during the day), and fatigue (exhausted). Tips can also be saved to participants’ dashboards for a personalized user experience, and all participants have access to all of the tips. After setting goals, participants then complete sleep diaries for an additional 2 weeks and, at the conclusion of the second 2-week period, they are able to see a comparison of their progress between weeks 1-2 and weeks 3-4 on their dashboard. Screenshots of the example progress, goal-setting, and tips screens are presented in Figure 1. Participants did not receive any other support or interventions in this study.

Outcomes were assessed using analysis of aggregate DOZE use during the study period and through the completion of web-based qualitative and quantitative questionnaires.

Interested AYAs were directed to the DOZE website, where they were provided with a consent form and details of our privacy policy. As the study took place on the web, a brief quiz followed the consent form to ensure that participants understood key aspects of informed consent, including the purpose of the study, risks and benefits of participation, and limits of confidentiality. After the provision of informed consent, participants received a randomly generated 5-digit study ID and completed a battery of questionnaires using Qualtrics, which included a demographic form and the ISI, FSS, CSM, CASQ, CESDR-10, STICSA State and Trait, SF-36, and TEQ. At the end of the questionnaires, participants received a link to access DOZE and create an account.

As indicated by the in-app onboarding process, participants were instructed to use DOZE for 2 weeks (ie, period 1), after which they would receive personalized feedback on their sleep, have the opportunity to set goals and access tips, and track their sleep for 2 more weeks (ie, period 2) to examine their progress. Participants were contacted via email by the study coordinator to maintain adherence with sleep tracking if they missed several consecutive days of completing the sleep diary and were sent a standardized email encouraging them to get back on track the next day. Following the completion of the 4-week study period, participants were emailed a link to complete the web-based posttest battery of questionnaires that included the ISI, FSS, CSM, CASQ, CESDR-10, STICSA State and Trait, SF-36, and TEQ, as well as the Acceptability Scale and the Qualitative Exit Questionnaire. Participants were compensated CAD $12.50 (US $9.79) for each study component that they completed; thus, participants who completed all 4 components of the study were emailed a CAD $50 (US $39.17) gift card honorarium for their participation. This study was approved by the research ethics board of Ryerson University, Toronto, Canada. Data collection took place between September 2019 and January 2020. There were no significant changes to the methods of the study or the intervention (eg, bug fixes, downtimes, or content changes) while this study was underway.

User characteristics, including self-reported demographic and baseline clinical characteristics as well as sleep indices from period 1, were evaluated using descriptive statistics for all participants who created a DOZE account. The demographic characteristics and scores on all self-report measures at baseline were compared between early dropouts (ie, before, during, and after the baseline questionnaires) and those who went on to create a DOZE account to assess for possible differences, as well as between all dropouts and those who completed the study (study finishers). To evaluate the feasibility of DOZE, including the characteristics and app use patterns of all DOZE users, analyses of app use (ie, number of days of sleep diary completion, use of goals and tips, and Google Analytics) included participants who created an app account regardless of whether or not they completed the study. Histograms and P-P plots were generated for visual inspection of normality, and extreme outliers on sleep indices were removed on the basis of visual inspection of boxplots. Descriptive statistics were used to analyze demographic and clinical characteristics. Patterns of app use were analyzed using descriptive statistics and Google Analytics.

Owing to insufficient sleep diary data from DOZE users who did not complete the study (32/83, 39%), sleep diary indices are only presented for study finishers. In addition, analyses evaluating acceptability, credibility, and satisfaction with DOZE, and examining the efficacy of DOZE in improving sleep indices and self-reported measures of sleep and daytime functioning were only conducted among study finishers. Paired sample 2-tailed t tests were used to analyze changes in satisfaction with DOZE from baseline to end point, and acceptability was analyzed using descriptive statistics. Changes in sleep indices from period 1 to period 2 were analyzed using paired sample t tests. Similarly, paired sample t tests with 1000 bootstrapped samples were used to evaluate differences in self-reported insomnia severity, daytime symptoms, and health-related quality of life from baseline to end point to account for slight deviations from normality. Effect sizes were calculated for all analyses of change from baseline to end point using Cohen d for repeated measures. Unless otherwise specified, all analyses were performed using SPSS for Mac version 26 (IBM Corporation).

In total, 154 participants consented to participate in this study (Figure 2). Of the 154 participants who consented to participate, 71 (46.1%) dropped out during the baseline questionnaire process. Over half of the participants (83/154, 53.9%) created an app account and completed some tracking. Approximately 33.1% (51/154) of participants completed all 4 components of the study (baseline questionnaires, first 2 weeks of sleep diaries, second 2 weeks of sleep diaries, and end point questionnaires).

Demographic characteristics were analyzed using data from all DOZE users (83/83, 100%; Table 1). Participants were on average 20.17 years (SD 2.49 years). Out of the 83 participants, 20 (24%) were adolescents (between the ages of 15 and 18 years) and 63 (76%) were young adults (between the ages of 19 and 24 years). The sample was predominantly female (57/83, 69%) and European Canadian (37/83, 45%). Most participants were not currently seeing a health care professional for the purpose of a sleep problem (79/83, 95%) or a mental disorder (69/83, 83%). Despite low levels of treatment-seeking, the average score on the ISI indicated sleep disturbance in the mild range (mean 12.48, SD 4.67), and 76% (63/83) of participants exceeded the clinical cutoff on the ISI (ie, score>10). In addition, the sample was characterized by high levels of fatigue, as indicated by the FSS (mean 4.73, SD 1.01) and the energy scale of the SF-36 (mean 35.71, SD 18.24) and daytime sleepiness (mean 45.14, SD 9.47) and on average had an intermediate chronotype on the CSM (mean 30.07, SD 7.16). Levels of state (mean 42.25, SD 12.73) and trait anxiety (mean 42.71, SD 13.15), as well as depression (mean 16.85, SD 8.30), were mild in this sample. Notably, early dropouts of the study (ie, before, during, and after the baseline questionnaires) did not differ from DOZE users on age (t₁₁₄=0.40; P=.69), gender (χ²₂=3.0; P=.22), ethnicity (χ²₉=16.3; P=.06), ISI (t₁₂₄=1.17; P=.25), FSS (t₁₂₄=1.08; P=.28), CMQ (t₁₁₈=0.63; P=.53), CASQ (t₁₁₃=−0.60; P=.55), STICSA State (t₁₁₁=0.74; P=.46), STICSA Trait (t₁₀₄=0.41; P=.69), CESDR-10 (t₁₁₅=−0.03; P=.98), or any of the SF-36 scales (P values range from .55 to .99). Similarly, study finishers did not differ from participants who dropped out at any point on any of these demographic or clinical characteristics (data not shown).

We analyzed sleep indices for study finishers at periods 1 and 2 (Table 2). After removing outliers (ranging from n=1 to n=6 depending on the sleep index), period 1 sleep indices indicated that despite an average SE in the normal range (ie, >85%), DOZE users spent a clinically meaningful amount of time awake at night. DOZE users also had variable sleep schedules, characterized by >4 hours of variability in their bedtime, sleep attempt, wake time, and rise time over a 2-week period. In addition, DOZE users obtained 7 hours 20 minutes of sleep on average despite spending 8 hours 21 minutes in bed; a one-sample t test indicated that this is significantly lower than 8.5, the minimum recommended amount of TST for AYAs (t₅₀=−8.76; P<.001). In addition, DOZE users spent an average of nearly half an hour lingering in bed in the morning.

We tested the feasibility, acceptability, and credibility of a self-management behavioral sleep web-based app designed by and for AYAs (aged 15-24 years). The results suggest that AYAs rated the intervention as logical, had confidence that the self-management approach would work, and had confidence in recommending DOZE to a friend. Willingness significantly decreased from baseline to end point, perhaps reflecting different stages of readiness to implement the strategies provided, environmental limitations that can influence their ability to implement strategies (eg, early school rise times and high academic demands), and intrinsic values that are a barrier to engaging in the intervention (eg, valuing sleeping in on weekends over maintaining a regular schedule). However, it is important to note that the TEQ was completed before participants received any detailed information about the intervention (they were only aware that it was a new sleep app for teens and young adults), representing only naïve expectations of DOZE, whereas at end point they were familiar with the intervention components. DOZE was rated as highly acceptable across indices of ease of use, understandability, time commitment, and satisfaction. Perhaps not surprisingly, ratings about DOZE being enjoyable or succinct in describing their particular symptoms or quality of life issues were acceptable but slightly more modest than ease of use, low time commitment, and overall satisfaction. In support of feasibility, participants engaged with DOZE often (indicated by the number of sessions), most frequently visiting pages pertaining to the sleep diary or their progress and did so using various electronic devices (ie, mobile phones, computers, and tablets). Moreover, although this was not a clinical sample, prospective sleep diary insomnia indices significantly improved from period 1 to period 2 (eg, time spent awake and SE), and the amount of sleep they obtained increased too. AYA users reported significant improvements from baseline to end point in insomnia symptom severity, chronotype, depression, anxiety, and energy. Despite not recruiting a clinical sample, participants in this study overwhelmingly were not treatment-seekers; however, they evidenced very poor health-related quality of life indices and were above clinical cutoffs for insomnia and fatigue, consistent with what we see in the literature [71].

These results are supportive of self-management. AYAs logged on frequently and completed the diary at the instructed time (ie, within 2 hours of waking up in the morning). Moreover, they worked hard at (1) tracking their sleep (ie, tracking for approximately 10 days in each 14-day period) and (2) setting goals that treatment providers would set (eg, limiting schedule variability and naps and decreasing lingering in bed). They accessed tips to help with their goals, particularly how to wind down before bed, how to help with being a night owl, how to get up regularly in the morning, how to help with fatigue, and how to stay out of bed early in the evening. Thus, we can trust AYAs to set their own goals and make behavioral changes to achieve these goals when they are provided with the tools and education to do so autonomously. Trusting patients to make good choices may not come naturally for providers, but the literature suggests that many patients want autonomy in their health care choices [72] and produce good results when given tools that support their autonomy. As a research group, this was an instructive part of the process, and it was driven by both the literature and collaboration with our AYA advisors and stakeholders throughout the co-design process. Perhaps these findings can inform the tailoring of in-person strategies for AYAs to achieve better outcomes.

Statistically significant differences in sleep indices from period 1 to period 2 seemed unlikely, as we recruited a sample that was not selected for clinical levels of sleep disturbance, and we assumed there would be too little variability in these indices to detect significant effects. Furthermore, detecting a change in variables, such as SOL, naps, or TIB, in a nonclinical or subclinical sample in which some have opposing problems (eg, hypersomnia vs insomnia) is also surprising. For example, those with hypersomnia or sleepiness have SOL values that are too low, and those with insomnia have elevated SOL values; use of DOZE could result in changes in opposite directions and, when averaged across participants, appear to have affected no change at all. These problems can also occur in the same AYA during the same week, for example, sleepiness on some nights and insomnia on others. Similarly, some AYAs are told to eliminate naps, and others are told to increase them, resulting in low variability. Therefore, it is encouraging that insomnia indices, such as lingering in bed in the morning, TWT, TST, TIB, and SE, improved. Although AYAs set goals to decrease the variability in their sleep schedules, the improvement did not reach statistical significance. Perhaps this is why they accessed tips about how to get up in the morning (ie, as it is difficult for them to rise at a regular time), and our plan is to expand tips and resources for AYA users on the DOZE website.

An important implication of this research is that this web-based app may be an important tool to increase access to evidence-based interventions. Some AYAs prefer to use technology for health-related behavior change [34,73,74], and smartphone apps appear to have a high uptake potential; 78.04% used a mobile device for DOZE (2,705/2659 sessions), whereas only 20% used a desktop computer (546/2,659 sessions, 20.5%) and 1.4% used a tablet (38/2,659 sessions). Incorporating AYAs in the co-design process is also a notable strength of this web-based app and may promote continued uptake. That said, future research will need to determine whether there would be health disparities based on access to technology, Wi-Fi, and geographical location. Clinicians report favorable attitudes toward CBT-I apps [75], suggesting the possibility of widespread access to evidence-based sleep treatments. Given the limited number of expert providers in this unique group (ie, not pediatric but not fully adult either) and the need for sleep solutions at the level of primary care and community mental health clinics, this web-based app holds the promise of increasing access to tailored, age-specific care. We intend to conduct a health economic study on the impact of DOZE in the future.

There was a 39% (32/83 participants) dropout rate in the self-management intervention portion of the study, and data were not collected on reasons for discontinuation, although this would certainly be valuable. Although we would prefer a lower rate, the rate was lower than that in other sleep app studies in the literature. For example, digital interventions for insomnia have reported rates >50% [76]. Across adolescent mobile health apps more generally, in a study using an internet-delivered, self-help CBT intervention for chronic pain among adolescents (aged 12-17 years), 52% of participants dropped out of the treatment portion of the study [77]. The dropout rate in our study is perhaps not surprising as they were not treatment-seekers, so if their sleep was not a problem, they might have little motivation except for the CAD $50 (US $39.17) compensation for participation. The brief intervention period in this study may also have contributed to lower rates of dropout compared with other studies. In addition, participants in this study were emailed by the research coordinator if they did not complete sleep diaries for several consecutive days; as such, adherence may be lower when disseminated in the real world. A future trial could evaluate whether dropout rates are lower in a clinical sample. Moreover, much of the dropout in this study occurred during the baseline questionnaire period, suggesting that the self-report battery was too burdensome. These questionnaires are not part of the app, so this barrier to participation is not present with regular, real-world app use, and future studies can verify if the app dropout rate is actually lower in real-world app use.

Although appropriate for feasibility trials, the use of an open trial suggests caution when interpreting positive outcomes. Without a control group, we could not ascertain whether the results were specific to the app or the increased sleep attention. In addition, we recruited participants using various strategies (eg, through a university, the community, and mental health providers, which are the intended disseminators of DOZE). Although this could be seen as a strength, given that our sample likely reflects those who will use DOZE in the real world, it may also be a limitation as these could be considered 3 distinct samples. Without knowing more about these samples, we could not test for possible differences. In the same vein, although our intended users are all AYAs, we are cautious about stating that this would be effective in patients, as this was a community sample; although most participants experienced symptoms that were above clinical cutoffs, they were neither clinically assessed nor were they treatment-seekers. We intend to study the properties of the app in those who are treatment-seeking. Our treatment provider stakeholders suggested that they would be most interested in using the diary as an adjunct to their treatment, much like sleep diary apps in adults [78]. DOZE could be used as a tracking and posttherapy relapse prevention tool, with providers prescribing behavioral change rather than the AYA deciding their own treatment course. Given the strong preferences for self-management in this group, it would be interesting to test outcomes with providers versus self-management.

Our findings may also have been influenced by extraneous factors. First, several items in our measure assessing credibility reference specifically to treatment for insomnia and may not have applied equally to all participants, thereby influencing their responses. Moreover, a small percentage of participants were seeing a health care provider for the purpose of a mental disorder or a sleep disorder, which may have also influenced our results given the bidirectional relationship between sleep and psychopathology. Finally, we did not query medication use and could not rule out the possibility that some participants were using sleep medication, which may have influenced our results; although low levels of treatment-seeking for a sleep disorder suggests that this is unlikely, and participants enrolled in the trial as they were dissatisfied with their sleep despite current treatments, it nevertheless remains a possibility among a small minority of our participants.

In conclusion, given that there were statistically and clinically significant improvements in a nonclinical sample, it is promising for the tool to be used as designed: as a transdiagnostic tool for young adults to engage in sleep self-management digitally. Even those without sleep disorders were interested and effective in making health-related changes, which can translate into improved sleep habits and potential prevention, as well as potentially improving mental wellness in AYAs. Web-based interventions are an acceptable and effective method of facilitating access to evidence-based care among invisible groups, with the ultimate effect of reducing the public health burden.