Effectiveness of current digital cognitive behavioural therapy for insomnia interventions for adolescents with insomnia symptoms: A systematic review and meta‐analysis

Summary

Sleep problems occur in up to 20%–45% of adolescents. This systematic review and meta‐analysis examined the effectiveness of digital sleep interventions, based on cognitive behavioural therapy for insomnia, for adolescents with insomnia symptoms. The objective was to synthesise and quantify, through meta‐analyses, changes in sleep following completion of a digital sleep‐based intervention. MEDLINE, PubMed, PsycINFO, Scopus, EMBASE, CENTRAL, and Web of Science databases were searched from January 2012 to March 2024. Within‐subject studies or randomized–controlled trials reporting the effects of digital cognitive behavioural therapy for insomnia were included. Risk of bias was assessed using the integrated quality criteria for the review of multiple study designs. Random‐effects meta‐analyses estimated pooled standardised within‐subject mean differences to assess effectiveness. Nine studies involving 486 adolescents were included. Digital cognitive behavioural therapy for insomnia interventions were effective in reducing insomnia symptoms (Hedges’ g = 1.40), subjective sleep‐onset latency (Hedges’ g = 0.72) and waking after sleep onset (Hedges’ g = 0.47), and increasing subjective and objective total sleep time (Hedges’ g = −0.29 and −0.23, respectively). Other objective measures of sleep did not improve. All studies met the minimum ICROMS score and were considered to be of sufficient quality. Seven within‐subject studies failed to satisfy all mandatory criteria. These results suggest that digital cognitive behavioural therapy for insomnia interventions are effective in improving adolescent's perceptions of their sleep, but are less effective at improving some objective measures of sleep. To achieve a clear understanding of how digital cognitive behavioural therapy for insomnia interventions compare with other behavioural interventions, additional high‐quality randomized–controlled trials comparing digital cognitive behavioural therapy for insomnia interventions with traditional in‐person modalities are needed. (PROSPERO;CRD42021287479).

1. INTRODUCTION

Sleep problems are common in adolescents, with prevalence estimates varying between 20% and 45% depending on the country being assessed (Malhi et al., 2018). Sleep problems include difficulty initiating sleep, frequent night waking, daytime sleepiness and frequent nightmares, which lead to poor‐quality sleep (Moore, 2012). Sleep problems are associated with externalising behaviours such as aggression, low mood, emotional dysregulation and difficulty maintaining attention (Meijer et al., 2010; Shochat et al., 2013). Furthermore, sleep problems in adolescents may be associated with school refusal behaviours, such as not attending school, difficulties socialising at school, and anxiety in the classroom (Hochadel et al., 2014; Hysing et al., 2015).

Behavioural interventions, which target dysfunctional habits and routines around sleep, are used as first‐line treatments for sleep problems (Blake et al., 2019). They comprise multiple components, including strategies such as bedtime routines, positive reinforcement, sleep restriction therapy and stimulus fading. While most sleep interventions for young people involve parents as active participants (e.g. parental sleep education through workshops, consultations and pamphlets; Hiscock et al., 2015; Jones et al., 2013), by adolescence, most sleep interventions tend to involve the adolescent directly, through individual and group therapy (Blake et al., 2017).

Although clinician‐delivered behavioural sleep interventions have robust evidence supporting their efficacy (Harvey et al., 2014), there are more families needing treatment for sleep problems than there are health professionals who are adequately trained to deliver such interventions (Richardson et al., 2021). Clinician availability is not the only barrier to healthcare; adolescents and their parents can be hesitant to seek professional help, due to social factors such as stigma, lack of trust in healthcare profession, and personal financial situation (Cook et al., 2020; Lim et al., 2012; Radez et al., 2021). Digital interventions may overcome some of these barriers and provide an alternative treatment pathway to support early intervention for sleep problems. Being self‐delivered, digital interventions can be delivered remotely, completed in the home environment, and would be more cost effective, thereby mitigating some of the social and access barriers, provided they are effective (Musiat et al., 2014; Schueller et al., 2019).

Digital technologies have been used to deliver behaviour change interventions in paediatric populations with a range of health conditions, including obesity (Kouvari et al., 2022), anxiety (Christ et al., 2020) and diabetes (Deacon & Edirippulige, 2015), and have also been used to aid adolescents with neurodevelopmental disorders (Gallardo‐Montes et al., 2022; Khan et al., 2019). While the efficacy of digital sleep interventions, such as digital cognitive behavioural therapy for insomnia (dCBT‐I), in adult populations is well established (Hasan et al., 2022), there is less research in adolescent populations. Digital adolescent sleep interventions have burgeoned since the first study was published nearly a decade ago (de Bruin et al., 2014). An earlier systematic review in 2018 by Werner‐Seidler included three studies (Werner‐Seidler et al., 2018), but with growth in this field and preliminary searches identifying additional publications, a further review and meta‐analysis is now warranted. Further, evidence is needed to synthesise the acceptability of digital sleep‐based interventions so future research can attempt to address any limitations in this treatment method. The objectives of this review were to: (1) synthesise and quantify through meta‐analyses changes in adolescent sleep following completion of a digital sleep‐based intervention; and (2) assess the acceptability of digital sleep‐based interventions to adolescents.

2. METHODS

The protocol for this systematic review was registered in the international prospective register of systematic reviews, and can be viewed online (PROSPERO; CRD42021287479). This study was guided by the preferred reporting items for systematic reviews and meta‐analysis (PRISMA) guidelines.

2.1. Data sources and searches

Before searching databases, a PICO framework was used to establish search terms (Appendix A). An extensive search of MEDLINE, PUBMED, Psych info, Scopus, Embase, CENTRAL and Web of Science was conducted. The search strategy consisted of search terms including synonyms and descriptions of digital interventions, sleep problems and adolescents (Appendix A). Searches were limited from 2012 to March 2024 to identify current digital sleep interventions, and as an earlier systematic review by Werner‐Seidler did not identify any studies published prior to 2012 (Werner‐Seidler et al., 2018). Reference lists of identified studies were manually searched to identify other eligible articles.

2.2. Study selection

Studies were eligible if the following inclusion criteria were met: (1) the study delivered a self‐guided behavioural intervention to improve sleep; (2) the intervention was accessed via digital technology (e.g. computer, tablet, smart phone); (3) participants were adolescents aged 10–19 years; (4) the study design was a single‐arm study or a randomised–controlled trial (RCT); (5) the study reported at least one sleep outcome through a sleep questionnaire, or via self‐reported (e.g. sleep diary) or objectively measured sleep (e.g. actigraphy). Studies were excluded if: (1) they were delivered via telehealth (i.e. intervention was primarily delivered over a phone or internet call) and did not include any other digital aspect to the intervention (i.e. mobile application, website); and (2) they were not used directly by the adolescent (i.e. parents used the digital intervention to improve their child's sleep).

Study details were exported to EndNote X9, and two of the researchers (MC and RR) independently removed duplicate articles and reviewed the titles and abstracts against each inclusion and exclusion criteria to assess eligibility. After removing ineligible search results, full texts of each study were reviewed by two of the researchers (MC and RR) to determine eligibility for inclusion. Studies that appeared to share the same dataset were compared with their study protocol for reported study outputs to confirm it was the same dataset, and the study that was best suited to our review criteria was chosen to be included. Discrepancies between the reviewers were resolved through discussion.

2.3. Data extraction and risk of bias assessment

Data were extracted from the included studies by the first author (MC), including study characteristics (author, year published, study type, intervention, study duration, mode of delivery), participant characteristics (age, gender, underlying disorder), measurement tool (e.g. actigraphy, sleep diary, Insomnia Severity Index [ISI]) and sleep outcomes (e.g. sleep‐onset latency [SOL], wake after sleep onset [WASO], sleep efficiency [SE], total sleep time [TST] and time in bed [TIB]). Where data were not reported or needed clarification, authors were contacted (e.g. to provide the data).

Risk of bias was assessed using the integrated quality criteria for the review of multiple study designs (ICROMS), which involves reporting whether studies fulfil specific quality criteria relevant to the study design. For each criterion, studies were awarded two points if they met the criterion, one if it was unclear or zero if they did not meet the criterion. The sum of the points gave a global quality score. If the global quality score was < 60% of the maximum attainable score, the study was considered to have high risk of bias/low reliability. Some of the quality criteria are considered mandatory, depending on the study design where the study could be excluded based on methodological weakness. Given that our objective was to capture all the current relevant literature, we did not exclude data based on methodological weakness.

2.4. Treatment outcomes

To determine the efficacy of the various digital interventions, self‐reported and objective sleep outcomes, including SOL, WASO, SE, TST and TIB, were extracted. Total scores from validated sleep questionnaires were also extracted.

2.5. Data synthesis and analysis

Random‐effects meta‐analysis was conducted using “Meta” (Balduzzi et al., 2019) and “Metafor” (Viechtbauer, 2010) packages in R (Team, 2022a) and RStudio (Team, 2022b). Meta‐analyses were performed to show change between baseline and post‐intervention effects because RCTs eligible for inclusion did not measure the same sleep outcomes, a meta‐analysis using between‐group data was not possible. Instead, post‐intervention effect size of the control group and intervention group were reported separately to the within‐subject data.

Means, standard deviations and N values from baseline and post‐intervention within subjects were used to calculate Hedges’ g and 95% confidence interval values for each outcome in each study. The magnitude of the effect was interpreted using the criteria for small (0.2), medium (0.5) and large (0.8) effect sizes, as defined by Cohen (Cohen, 2013). For all studies, where required, the mean and standard deviation for some outcomes (e.g. TST, TIB, WASO, SOL) were converted into minutes.

As between‐study heterogeneity was expected, we used a random‐effects model to pool effect sizes. Pooled effect sizes were deemed statistically significant at p < 0.05. We used Hartung–Knapp adjustments (Knapp & Hartung, 2003) to calculate the confidence interval around the pooled effect. Forrest plots were created to show the individual and pooled effect sizes from the meta‐analysis. The restricted maximum likelihood estimator (Viechtbauer, 2005) was used to assess between‐study variance (tau²). Heterogeneity was assessed via the I ² statistic. I ² values of 25% are interpreted as low heterogeneity, and values of 50% and > 75% indicate moderate and high heterogeneity, respectively Sensitivity analyses to assess possible sources of heterogeneity were performed by excluding each study and rerunning the meta‐analysis, if three or more studies could be included in the meta‐analyses. Influential data points were identified using Cook's D, with a cut‐off value of 4/n. Publication bias was not assessed, as there were fewer than the recommended 10 studies required for assessment, because the power of the tests is too low to distinguish chance from real asymmetry.

To assess the acceptability of each digital intervention, the completion rate for each intervention was calculated. The number of participants in each intervention group who completed all necessary modules of the digital intervention, divided by the total number of participants that started the intervention (i.e. not including those who were allocated to the intervention but never commenced the intervention), was calculated for each study.

3. RESULTS

3.1. Search results

Of the 2118 potentially eligible articles, 1064 records were removed as duplicates and 994 were excluded after screening the titles and abstracts (Figure 1). A full‐text inspection of 60 articles followed, in which 14 articles met the inclusion criteria for this review. A further five studies were removed due to using the same population dataset (i.e. data were duplicated; de Bruin et al., 2018; de Bruin et al., 2020; de Bruin, Dewald‐Kaufmann, et al., 2015; de Bruin & Meijer, 2017; Li et al., 2021). Ultimately, nine articles were included in the data analysis (Åslund et al., 2023; Cliffe et al., 2020; de Bruin et al., 2014; de Bruin, Bögels, et al., 2015a; Georén et al., 2022; Mathews et al., 2022; Werner‐Seidler et al., 2019; Werner‐Seidler et al., 2023; Zetterqvist et al., 2020).

FIGURE 1.

PRISMA flow chart of included studies.

3.2. Overview of included studies

A total of 486 adolescents participated in the nine included studies. Study sample sizes ranged from 6 to 131 participants (69.2% were female), with a mean age of 15.01 years (range 12–19 years; Table 1). Two of the studies were RCTs (de Bruin, Bögels, et al., 2015a; Werner‐Seidler et al., 2023), and seven were single‐arm trials. Characteristics of the digital interventions are shown in Table 1. Intervention periods ranged from 6 to 8 weeks for most studies, with two studies having the intervention available for a maximum of 6 weeks and no minimum timeframe (Werner‐Seidler et al., 2019; Werner‐Seidler et al., 2023). Among the studies analysed, two used an app‐based sleep intervention without a therapist (Werner‐Seidler et al., 2019; Werner‐Seidler et al., 2023), and seven involved web‐based CBT‐I with therapist support via phone calls or text messages. Five studies reported longer‐term follow‐up assessment of sleep outcomes after the post‐intervention assessment, with follow‐up duration ranging from 2 to 6 months (de Bruin et al., 2014; de Bruin, Bögels, et al., 2015a; Georén et al., 2022; Werner‐Seidler et al., 2023; Zetterqvist et al., 2020). Sleep diaries were a part of all of the interventions, but two studies did not report the sleep diary data in their results (Cliffe et al., 2020; Werner‐Seidler et al., 2023). Insomnia symptoms were the most common sleep problem targeted, with all studies requiring the presence of insomnia symptoms for participation through sleep questionnaires. To the best of the authors knowledge, the Athens Insomnia Scale (Chung et al., 2011), Chronic Sleep Reduction Questionnaire (CSRQ; Dewald et al., 2012), Dysfunctional Beliefs and Attitudes about Sleep Scale for Children (Blunden et al., 2013), Epworth Sleepiness Scale for Children and Adolescents (Janssen et al., 2017), ISI (Chung et al., 2011), Pediatric Daytime Sleepiness Scale (Drake et al., 2003) and Pittsburgh Sleep Quality Index (Raniti et al., 2018) have been validated for adolescents, while the Flinders Fatigue Scale (Gradisar et al., 2007), Holland Sleep Disorders Questionnaire (HSDQ; Kerkhof et al., 2013), Pre‐Sleep Arousal Scale (Nicassio et al., 1985), Sleep Condition Indicator (Espie et al., 2014) and Sleep Related Behaviours Questionnaire (Ree & Harvey, 2004) are not formally validated, but used extensively to measure adolescent sleep.

TABLE 1.

Characteristics of included studies.

Study ID	Participants (male:female)	Participant age (years)	Control/comparison	Primary complaint or diagnosis	Intervention	Treatment components	Sleep measures	Sleep outcomes of interest
Åslund et al. (2023)	6:21	13–17	Baseline assessment	Insomnia	ISnooze: dCBT‐I intervention 6 × weekly modules Therapist support 3 × per week via messages on the app	Psychoeducation, relapse prevention, relaxation, sleep hygiene, sleep restriction, stimulus control	ISI b , PDSS b , sleep diary	Symptoms of insomnia, SOL, WASO, TST, SE
Cliffe et al. (2020)	12:37	14–17	Baseline assessment	Mental health problems, symptoms of insomnia	Sleepio: dCBT‐I for computer, smartphone or tablet 20‐min session each week for 6 weeks Weekly support telephone call (< 15 min) from a trained Sleepio assistant	Cognitive restructuring, mindfulness, paradoxical intention, positive imagery, psychoeducation, relaxation, sleep hygiene, sleep restriction, stimulus control	ISI b , SCI, sleep diary	Symptoms of insomnia
de Bruin et al. (2014)	5:21	13–19	Baseline assessment	Insomnia symptoms	Internet‐delivered CBT‐I 6 × weekly sessions One 15‐min phone call with therapist offered Booster session 2 months after the 6th session	Cognitive therapy, psychoeducation, relaxation, sleep restriction, sleep hygiene, stimulus control	Actigraphy, CSRQ b , HSDQ, sleep diary	SOL, WASO, TST, TIB, SE
de Bruin, Bögels, et al. (2015a)	29:87	12–19	Waiting list	Insomnia symptoms	Internet‐delivered CBT‐I 6 × weekly sessions Two 15‐min phone calls with therapist offered during week 2	Cognitive therapy, psychoeducation, relaxation, sleep restriction, sleep hygiene, stimulus control	Actigraphy, CSRQ b , HSDQ, sleep diary	SOL, WASO, TST, TIB, SE
Georén et al. (2022)	4:2	13–17	Baseline assessment	ASD, insomnia diagnosis	Internet‐delivered CBT‐I Psychoeducational videos for 8 weeks (~1 hr per week) 2 telephone calls per week with a practitioner (psychologist and psychotherapist)	Cognitive restructuring, home assignment, problem solving, psychoeducation, relaxation, safety behaviour, sleep hygiene, sleep restriction therapy, stimulus control	ISI b , AIS b , sleep diary	Symptoms of insomnia, SOL, WASO, TST, SE
Mathews et al. (2022)	25:34	13–17	Baseline assessment	Anxiety and/or depression, insomnia symptoms	Sleepio: dCBT‐I for computer, smartphone or tablet 20‐min session per week for 6 weeks Weekly support telephone call (< 15 min) from a trained Sleepio assistant	Cognitive restructuring, mindfulness, paradoxical intention, positive imagery, psychoeducation, relaxation, sleep hygiene, sleep restriction, stimulus control	ISI b , SCI, sleep diary	Symptoms of insomnia, SOL, SE
Werner‐Seidler et al. (2019)	17:33	12–16	Baseline assessment	Insomnia symptoms	Sleep ninja: dCBT‐I intervention for mobile phone 6 × training sessions Sleep tracking, recommended bedtimes and wind down routines	Psychoeducation, sleep focused cognitive therapy, sleep hygiene, stimulus control	ISI b , PSQI b , sleep diary	Symptoms of insomnia, SOL, WASO, TIB, SE
Werner‐Seidler et al. (2023)	40:89 a	12–16	Active control	Insomnia symptoms	Sleep ninja: dCBTi intervention for mobile phone 6 training sessions, sleep tracking, recommended bedtimes and wind down routines	Psychoeducation, sleep focused cognitive therapy, sleep hygiene, stimulus control	DBAS‐C b , ESS b , FFS, ISI b , PSAS, PSQI b , SRBQ	Symptoms of insomnia
Zetterqvist et al. (2020))	2:19	13–17	Baseline assessment	Psychiatric condition, insomnia symptoms	dCBT‐I delivered through computer or smartphone 7 × weekly modules with therapist support (through messages and comments)	Bedtime restriction, problem solving, psychoeducation, relapse prevention, relaxation, sleep hygiene, stimulus control	ISI b , sleep diary	Symptoms of insomnia, SOL, WASO, TST, SE

Abbreviations: AIS, Athens Insomnia Scale; ASD, autism spectrum disorder; CSRQ, Chronic Sleep Reduction Questionnaire; DBAS, Dysfunctional Beliefs and Attitudes About Sleep Scale for Children; dCBT‐I, digital cognitive behavioural therapy for insomnia; ESS, Epworth Sleepiness Scale for Children and Adolescents; FFS, Flinders Fatigue Scale; HSDQ, Holland Sleep Disorder Questionnaire; ISI, Insomnia Severity Index; N/A, not applicable; PDSS, Pediatric Daytime Sleepiness Scale; PSAS, Pre‐Sleep Arousal Scale; PSQI, Pittsburgh Sleep Quality Index; SCI, sleep condition indicator; SE, sleep efficiency; SOL, sleep‐onset latency; SRBQ, Sleep‐Related Behaviours Questionnaire; TIB, time in bed; TST, total sleep time; WASO, wake after sleep onset.

^a Three participants identified as other.

^b Questionnaire has been validated in adolescents.

For the purposes of this study, the only sleep questionnaire included in the meta‐analysis was the ISI, as it was most consistently used across the studies (7/9 studies).

3.3. Effects of interventions on sleep outcomes

3.3.1. Insomnia Severity Index

Seven studies measured change in ISI scores from baseline to post‐intervention. ISI scores decreased from baseline following the intervention (mean difference [MD] = 6.57; Hedges’ g = 1.40, 95% confidence interval [CI] [1.06; 1.74], p = < 0.001; Figure 2).

FIGURE 2.

Effect of digital sleep interventions on change in Insomnia Severity Index (ISI) scores from baseline to post‐intervention.

3.3.2. Sleep‐onset latency

Six studies measured change in SOL from baseline to post‐intervention. Objective measures of SOL did not change from baseline following the intervention (MD = 10.25 min; Hedges’ g = 0.59, 95% CI [–4.73; 5.91], p = 0.392; Figure 3), although subjective measures of SOL (from sleep diary) were decreased from baseline (MD = 37.95 min; Hedges’ g = 0.72, 95% CI [0.35; 1.09], p = 0.004; Figure 3).

FIGURE 3.

Effect of digital sleep interventions on change in sleep‐onset latency (SOL) from baseline to post‐intervention recorded via (a) actigraphy and (b) sleep diary.

3.3.3. Wake after sleep onset

Six studies measured change in WASO from baseline to post‐intervention. Objective measures of WASO did not change from baseline following the intervention (MD = −1.05 min; Hedges’ g = 0.00, 95% CI [−0.63; 0.64], p = 0.957; Figure 4), although there was a decrease in subjective measures of WASO (from sleep diary) from baseline (MD = 9.26 min; Hedges’ g = 0.47, 95% CI [0.14; 0.81], p = 0.014; Figure 4).

FIGURE 4.

Effect of digital sleep interventions on change in wake after sleep onset (WASO) from baseline to post‐intervention recorded via (a) actigraphy and (b) sleep diary.

3.3.4. Sleep efficiency

Seven studies measured change in SE from baseline to post‐intervention. Objective measures of SE did not change from baseline following the intervention (MD = −3.12%; Hedges’ g = −0.59, 95% CI [−6.31; 5.12], p = 0.412; Figure 5), and neither did subjective measures of SE (from sleep diary; MD = −10.12%; Hedges’ g = −0.60, 95% CI [−1.27; −0.06], p = 0.070; Figure 5).

FIGURE 5.

Effect of digital sleep interventions on change in sleep efficiency (SE) from baseline to post‐intervention recorded via (a) actigraphy and (b) sleep diary.

3.3.5. Total sleep time

Six studies measured change in TST from baseline to post‐intervention. Objective measures of TST increased from baseline following the intervention (MD = −14.50 min; Hedges’ g = −0.23, 95% CI [−0.39; −0.07], p = 0.034; Figure 6), as did subjective measures of TST (from sleep diary; MD = −21.05; Hedges’ g = −0.29, 95% CI [−0.51; −0.06], p = 0.021; Figure 6).

FIGURE 6.

Effect of digital sleep interventions on change in total sleep time (TST) from baseline to post‐intervention recorded via (a) actigraphy and (b) sleep diary.

3.3.6. Time in bed

Three studies measured change in TIB from baseline to post‐intervention. Objective measures of TIB did not change from baseline following the intervention, (MD = 2.50 min, Hedges’ g = 0.11, 95% CI [−2.16; 2.39], p = 0.637; Figure 7), and neither did subjective measures of TIB (from sleep diary; MD = 13.67 min, Hedges’ g = 0.19, 95% CI [−0.25; 0.62], p = 0.205; Figure 7).

FIGURE 7.

Effect of digital sleep interventions on change in time in bed (TIB) from baseline to post‐intervention recorded via (a) actigraphy and (b) sleep diary.

3.3.7. Between‐group analysis of RCTs

One of the two RCTs measured differences in change in ISI scores from baseline to post‐intervention between the intervention and control group (Werner‐Seidler et al., 2023). There was a greater reduction in post‐intervention ISI scores in those in the intervention group compared with the control group (Hedges’ g = −0.43).

The other RCT compared objective and subjective measures of SOL, WASO, SE, TST and TIB between the intervention and control group (de Bruin, Bögels, et al., 2015a). Relative to baseline, the intervention group had decreased objective (Hedges’ g = 1.08) and subjective (Hedges’ g = 0.63) measures of SOL, compared with the control group. The intervention group also had increased objective (Hedges’ g = −1.09) and subjective (Hedges’ g = –0.81) measures of SE and subjective (Hedges’ g = –0.36) measures of TST compared with the control group. Both objective and subjective measures of WASO and TIB and objective measures of TST did not differ at post‐treatment between the intervention and control group.

3.3.8. Within‐group analysis of sleep questionnaires

Across eight of the studies, 11 validated sleep questionnaires were used to measure within‐subject change in sleep from baseline to post‐intervention. Two studies measured change in sleep with the Pittsburgh Sleep Quality Index, and identified a reduction in scores from baseline to post‐intervention (Hedge's g = 0.58 and 5.02). Two studies measured change in sleep with the Sleep Condition Indicator, and also identified a reduction in scores from baseline to post‐intervention (Hedge's g = 2.47 and 1.66). Two studies measured change in sleep with both the HSDQ and the CSRQ; one did not identify a reduction in scores from baseline to post‐intervention (HSDQ Hedge's g = 0.91; CSRQ Hedge's g = 0.56), while the statistical significance of the change in the other study could not be determined (HSDQ Hedge's g = 0.85¹; CSRQ Hedge's g = 0.71¹). One study measured change in sleep with the Athens Insomnia Scale, with a reduction in scores from baseline to post‐intervention observed (Hedge's g = 1.85). One study measured change in daytime functioning with the Pediatric Daytime Sleepiness Scale, and did not observe a change in scores from baseline to post‐intervention (Hedge's g = 0.42). One study measured change in fatigue, pre‐sleep arousal and daytime sleepiness using three different questionnaires. They identified a reduction in scores from baseline to post‐intervention for two measures (Flinders Fatigue Scale, Hedge's g = 4.57; and Pre‐Sleep Arousal Scale, Hedge's g = 5.63), but not the Epworth Sleepiness Scale (Hedge's g = 1.28). One study measured change in sleep behaviours and beliefs with two questionnaires, and observed a reduction in scores from baseline to post‐intervention in both measures (Sleep‐Related Behaviours Questionnaire, Hedge's g = 2.08; and Dysfunctional Beliefs and Attitudes about Sleep Scale, Hedge's g = 4.99).

3.4. Risk of bias

All studies met the minimum total ICROMS score and were considered to be of sufficient quality (Table 2). For the two RCTs, risk of bias was low, with the studies satisfying all mandatory (and non‐mandatory) criteria. While blinding of participants was a specialised criterion, blinding was not possible in these studies, as participants were required to attend internet therapy, group therapy or be on a waitlist.

TABLE 2.

ICROMS risk of bias tool.

	Åslund et al. (2023)	Cliffe et al. (2020)	Georén et al. (2022)	de Bruin et al. (2014)	de Bruin, Bögels, et al. (2015a)	Mathews et al. (2022)	Werner‐Seidler et al. (2019)	Werner‐Seidler et al. (2023)	Zetterqvist et al. (2020)
Study design	NCBA	NCBA	NCBA	NCBA	RCT	NCBA	NCBA	RCT	NCBA
Clear statement of the research aims*, **	+	+	+	+	+	+	+	+	+
Adequate baseline measurement	+	+	+	+	N/A	+	+	N/A	+
Explanation for lack of control group	−	+	−	+	N/A	+	+	N/A	+
Sequence generation**	N/A	N/A	N/A	N/A	+	N/A	N/A	+	N/A
Allocation concealment**	N/A	N/A	N/A	N/A	+	N/A	N/A	+	N/A
Justification of sample choice*	+	+	+	+	N/A	+	+	N/A	+
Blinding**	N/A	N/A	N/A	N/A	+	N/A	N/A	+	N/A
Blinded assessment of primary outcome measures	+	+	+	+	+	+	+	+	+
Reliable primary outcome measures	+	+	+	+	+	+	+	+	+
Follow‐up of subjects (protection against exclusion bias)	N/A	N/A	N/A	N/A	+	N/A	N/A	+	N/A
Follow‐up of patients or episodes of care	N/A	N/A	N/A	N/A	+	N/A	N/A	+	N/A
Incomplete outcome data assessed	+	+	+	+	+	+	+	+	+
Intervention unlikely to affect data collection	+	+	+	+	+	+	+	+	+
Attempts to mitigate effects of no control*	−	−	−	−	N/A	−	−	N/A	−
Analysis sufficiently rigorous/free from bias	+	+	+	+	+	+	+	+	+
Free of selective outcome reporting	+	+	+	+	+	+	+	+	+
Limitations addressed	+	+	+	+	+	+	+	+	+
Conclusions clear and justified	+	+	+	+	+	+	+	+	+
Free of other bias	−	−	−	+	+	−	−	+	−
Ethics issues addressed	+	+	+	+	+	+	+	+	+
Total score	26	28	26	30	32	28	30	32	28

Note: +, yes; −, no; N/A, not applicable; NCBA, non‐controlled before and after; RCT, randomized–controlled trial; total score: ≥ 22, sufficient quality.

^*Mandatory criteria for NCBA.

^**Mandatory criteria for RCT.

The seven within‐subject studies all failed to satisfy the mandatory criterion of attempting to mitigate the effect of the lack of control group. While the authors did not explicitly state the reasons for the lack of control group, all did state that they were conducting a feasibility trial and for those purposes it can be interpreted that they did not require a control group. All of the within‐subject studies were deemed to have other biases; they all required participants to self‐report their sleep through sleep diaries. Sleep diaries are a subjective measure, which introduces reporter bias that was not mitigated with the inclusion of additional objective measure(s).

3.5. Treatment completion

The number of participants who completed the digital sleep interventions was reported in all studies, and ranged from 28% to 100%, with an average of 65% (Table 3). One study retained all participants, albeit with a small sample size of six participants (Georén et al., 2022). One of the studies included reasons for non‐adherence to the sleep intervention (Werner‐Seidler et al., 2019).

TABLE 3.

Number of participants that started and completed the digital sleep intervention.

Study ID	Started intervention	Completed intervention (%)	Reason for non‐adherence
Åslund et al. (2023)	27	23 b (85%)	NR
Cook et al. (2020)	30	13 (43%)	NR
de Bruin et al. (2014)	13	11 (85%)	NR
de Bruin, Bögels, et al. (2015a)	39	38 (97%)	NR
Georén et al. (2022)	6	6 (100%)	N/A
Mathews et al. (2022)	59	29 (49%)	NR
Werner‐Seidler et al. (2019)	45	15 (33%)	Took too long to work through, too much text to read, too repetitive
Werner‐Seidler et al. (2023)	96	27 (28%)	NR
Zetterqvist et al. (2020)	20	14 a (70%)	NR

Abbreviations: N/A, not applicable; NR, not reported.

^a Zetterqvist specified that if four of the seven modules were finished, then participants were counted as completed.

^b Aslund specified that if four of the six modules were finished, then participants were counted as completed.

3.6. Sensitivity analysis

We removed studies from each meta‐analysis to test the robustness of the pooled results and to explore the source of heterogeneity. For subjective SOL, heterogeneity was decreased after exclusion of the Werner‐Seidler et al. (2019) study (Hedges’ g = 0.83, 95% CI [0.43; 1.23], I ² = 0%) and the Zetterqvist et al. (2020) study (Hedges’ g = 0.60, 95% CI [0.28; 0.93], I ² = 0%). For subjective SE, heterogeneity was decreased after exclusion of the Åslund et al. (2023) study (Hedges’ g = −0.82, 95% CI [−1.13; −0.51], I ² = 2.7%). Despite these studies being identified as sources of heterogeneity, Cook's D revealed only Åslund et al. (2023) to be an influential data point in the meta‐analysis of subjective SE. After removing this study, the increase in subjective SE became significant (p = < 0.01).

4. DISCUSSION

This systematic review and meta‐analysis advances understanding of the effectiveness and acceptability of digital sleep interventions in adolescents aged 12–19 years. Our findings demonstrate that digital sleep interventions are effective in reducing ISI scores post‐intervention, decreasing subjective measures of SOL and WASO and increasing objective and subjective measures of TST. These results are important as they demonstrate that adolescents perceive that their sleep is improved after a digital sleep intervention, even if improvement was not always identified from objective measures.

The degree of reduction in ISI scores from baseline to post‐intervention in the present analysis is in line with a previous meta‐analysis of change in insomnia symptoms after digital CBT‐I in adult populations (SMD = 1.34 versus 1.36, respectively; Soh et al., 2020). In terms of comparing the adolescent and adult dCBT‐I fields (i.e. comparing characteristics from this meta‐analysis with the work of Soh et al., 2020), treatment components across the studies were comparable, with most studies involving sleep restriction, sleep hygiene education and stimulus control. Intervention duration was also similar, with digital interventions typically ranging from 6 to 8 weeks in both adolescent and adult studies. In terms of critical differences between the adolescent and adult fields, much of the data from adult studies have been derived from RCTs with control cohorts, not within‐subject studies like in the current meta‐analysis. Non‐randomised interventional studies have greater potential for bias compared with RCTs, and cannot establish a causal relationship between the intervention and sleep outcome (Reeves et al., 2019). In within‐subject studies, improvements could be due to regression to the mean or the passage of time. As such, despite a call to action for more adolescent CBT‐I research nearly a decade ago (Gradisar & Richardson, 2015), it appears that the adolescent dCBT‐I field still lags far behind the adult field.

In the present analysis, improvements in SOL, WASO and SE were observed when measured via subjective measures, but these findings were not supported by actigraphy results. This discrepancy between subjective and objective measures of sleep is consistent with the findings of a meta‐analysis of tech‐based sleep interventions for children younger than 12 years (Zhu et al., 2022). Studies have shown that children overestimate their TST, SOL and WASO when completing sleep diaries, compared with actigraphy (Mazza et al., 2020; Tremaine et al., 2010), which may be due to the two methods capturing different dimensions of sleep. These patterns of findings could also suggest that the interventions used might have improved the adolescent's perceptions of their sleep (e.g. due to psychoeducation about sleep stages and architecture, etc.), without improving their objective sleep. These results may also be due to response biases as subjective measures rely on a person's recollection of sleep and are subject to several response biases (e.g. social desirability and acquiescent responding; Kreitchmann et al., 2019), while actigraphy reflects motor activity (Aili et al., 2017). Actigraphy has been shown to be comparable to polysomnography for sleep duration, SOL and SE in children (Waldon et al., 2016), although it must be noted that actigraphy is not a perfect measure of sleep, having relatively poor detection of wakefulness during sleep (Sadeh, 2011). A number of recent studies have identified that CBT‐I does not significantly increase objective and subjective measures of TST; possibly due to the short‐term aim of CBT‐I being to reduce SOL and WASO, and increase SE but not TST (Lancee et al., 2016; Paine & Gradisar, 2011; Schlarb et al., 2018). However, TST has been shown to increase significantly by 3 to 24 months post‐intervention (Scott et al., 2022; Scott et al., 2023). Interestingly, the current meta‐analysis showed an increase in objective TST post‐intervention (12.0–26.6 min) despite the intervention durations being approximately 6–8 weeks. This increase in objective TST may be due to the age of the participant cohort. Adolescents commonly achieve significantly less than the 9.35 hr of sleep recommended for optimal cognitive performance and mood (Fuligni et al., 2019; Short et al., 2018), and may have greater capacity to increase TST. It should be noted that due to the small number of studies included in the objective TST analysis, confirmation of this outcome in future studies is warranted. Furthermore, due to a low number of studies with long‐term follow‐up data, we are unable to draw conclusions about long‐term changes in subjective and objective TST in response to digital interventions targeting sleep.

While these results are important, it must be noted that the participants in the included studies did not have a clinical diagnosis of insomnia, but displayed insomnia symptoms, and therefore the results need to be interpreted with this in mind. However, this does raise the idea that dCBT‐I interventions could be used in a stepped care model whereby adolescents are treated for insomnia before there is a worsening of symptoms (i.e. surpassing the diagnostic threshold) and greater need for more intensive clinician support. Another factor that warrants noting is that insomnia can co‐occur with many neurodevelopmental or health conditions (Nunes & Bruni, 2015; Shelton & Malow, 2021). Participants in the included studies did not have comorbidities, therefore precluding extrapolation of these findings to other populations where there are comorbidities.

One aim of this study was to assess acceptability of digital sleep interventions but, due to the lack of standardised acceptability measures used in the included studies, the best indication of this was through treatment completion. Completion rates of the interventions included in these meta‐analyses were surprisingly low. Only Georén et al. (2022) with a very small sample size (n = 6) had 100% completion. While treatment completion rates may provide insight into acceptability, there are other reasons that participants may not adhere to all necessary modules of an intervention. Personal factors such as well‐being, motivation and perceived lack of time, as well as digital factors such as personal device problems, internet accessibility and intervention design features can effect adherence to digital interventions (Renfrew et al., 2021). Only one study (Werner‐Seidler et al., 2019) reported reasons for participant non‐adherence through a questionnaire that included 23 items about technical issues, personal issues, intervention‐general and intervention‐specific issues. Future studies should include multi‐component measurement of acceptability, including measures such as the Acceptability of Intervention Measure (Weiner et al., 2017) and qualitative interviews, to get a full scope of the acceptability of an intervention.

Participant dropout also impacts treatment completion rates, and tends to be greater in those involving digital interventions (Torous et al., 2020). A previous systematic review that assessed the dropout rates of participants using digital interventions for depressive symptoms reported that study retention rates were higher in interventions that offered human feedback (Torous et al., 2020). Six of the studies included in this review included human feedback, commonly provided through optional telephone calls with a therapist. Two of the three studies that did not include human feedback recorded the lowest completion rates. It must be noted that the effect of human feedback on intervention completion rates in digital sleep applications has not previously been reported, so whether the lack of human feedback in these digital sleep interventions contributes to the low completion rates is yet to be confirmed. To increase engagement and retention rates, some researchers have incorporated the concept of co‐design into the development of the intervention (Martin et al., 2020; Werner‐Seidler et al., 2017). Utilising co‐design methods when developing a digital intervention for adolescents may increase motivation and encourage prolonged use of the intervention. Only two studies in this review, both by Werner‐Seidler et al. (Werner‐Seidler et al., 2019; Werner‐Seidler et al., 2023) and using the same intervention, incorporated the concept of co‐design into the development of their intervention.

4.1. Limitations

This review has limitations that should be acknowledged. First, the number of studies that met our eligibility criteria was small (n = 9). Furthermore, of the six of these that have follow‐up assessments, the duration of the follow‐up period was variable (2–6 months), which precluded evaluation of the long‐term efficacy of digital sleep interventions. Second, due to the small number of papers included in this review, we could not perform any subgroup analyses on the results due to the lack of statistical power (Cuijpers et al., 2021). Also, due to the small number of studies, appropriate statistical tests and funnel plots for publication bias were not performed as they would not be precise in detecting bias without the sufficient number of studies (≥ 10; Ioannidis & Trikalinos, 2007). Only two studies had a randomized–control design, and they reported different sleep outcomes, making it impossible to meta‐analyse effects. Due to the use of single‐arm studies, estimates of the cause–effect relationships between the digital sleep interventions and sleep outcomes may not be able to be ascertained (Reeves et al., 2019). As sleep can be affected by different medications and psychotherapy, RCTs can help to determine post‐treatment differences through the use of a control group. Not all studies in this review stated whether they excluded participants who were taking sleep medication or receiving psychotherapy, and therefore some of the changes in sleep may have been in response to other treatments. The use of a randomized–controlled design would mitigate these impacts, as having a control group with similar characteristics would mean any post‐treatment differences should be due to the intervention and not potential cofounding variables. It should also be noted that not all sleep outcome measures were measured across all studies, for example, the most frequently reported measures were SE and the ISI (n = 6 each). The inclusion of standard sleep outcomes (e.g. SOL, WASO, SE, TST, TIB) in future studies would improve the ability to compare between studies and confidently evaluate pooled data.

4.2. Future research

Digital intervention to treat sleep problems in adolescents is still an emerging field. To achieve a clear understanding of how digital sleep interventions compare with other behavioural interventions, additional high‐quality RCTs comparing digital interventions with traditional in‐person modalities are needed. Future studies should also try to include adolescents in the design process of any digital intervention, to encourage engagement and retention. While there are advantages and disadvantages to the inclusion of subjective and objective sleep measures, future studies should aim to use both measures to ensure a multi‐dimensional picture of sleep is captured.

5. CONCLUSION

The results of this systematic review and meta‐analysis suggest that dCBT‐I is effective in improving adolescents’ perceptions of their sleep, yet is less effective at objectively improving sleep. Due to the increasing digital interests of adolescents, and low completion rates of existing digital interventions, future studies should consider providing some form of human support during the intervention. Given the low number of studies involving digital sleep interventions, more RCTs are needed to evaluate the effectiveness of digital sleep interventions, especially compared with other available behavioural interventions.

AUTHOR CONTRIBUTIONS

Melissa A. Cleary: Conceptualization; investigation; writing – original draft; methodology; validation; visualization; writing – review and editing; software; formal analysis; data curation; resources. Cele Richardson: Conceptualization; writing – review and editing; visualization; methodology; project administration; supervision. Ruby J. Ross: Investigation; writing – review and editing; methodology; formal analysis; data curation. Helen S. Heussler: Conceptualization; writing – review and editing; supervision; project administration; visualization. Andrew Wilson: Conceptualization; writing – review and editing; project administration; supervision. Jenny Downs: Conceptualization; writing – review and editing; visualization; project administration; supervision; methodology. Jennifer Walsh: Conceptualization; writing – review and editing; visualization; supervision; project administration; methodology.

CONFLICT OF INTEREST STATEMENT

The authors declare there are no conflicts of interest to report.

APPENDIX A. PICO FRAMEWORK AND SEARCH STRATEGIES USED FOR EACH DATABASE SEARCHED: ALL DATABASES WERE LAST SEARCHED ON 29 MARCH 2024

PICO FRAMEWORK
Population	Adolescents (10–19 years) with insomnia
Intervention	Self‐guided digital behavioural intervention to improve sleep
Comparison	No restriction
Outcome	At least one sleep outcome reported through a validated sleep questionnaire, or via self‐reported or objectively measured sleep

MEDLINE	Search terms
1	Mobile applications.mp. or (digital intervention or internet treatment or health intervention or mobile application or cognitive behavio?ral therap* or digital solution* or internet‐mediated or mobile health or online therap* or e?intervention or digital treatment or support app or internet‐delivered or iCBT or CBTi or digital therap* or e?health).ti,ab.
2	sleep wake disorders/
3	sleep problem* or sleep disorder* or sleep difficult* or insomnia or sleep quality or sleep challenges or sleep disturbance* or sleep wake disorder* or sleep pattern* or sleep disruption or "sleep initiation and maintenance disorder*").ti,ab.
4	2 or 3
5	child/ or adolescent/ or (child* or adolescent* or p?ediatric or teenager* or young people).ti,ab.
6	1 and 4 and 5
7	limit 6 to (english language and yr="2012 ‐Current")

PsycInfo	Search terms
1	digital interventions/ or mobile applications/
2	(digital intervention or internet treatment or health intervention or mobile application or cognitive behavio?ral therap* or digital solution* or internet‐mediated or mobile health or online therap* or e?intervention or digital treatment or support app or internet‐delivered or iCBT or CBTi or digital therap* or e?health).ti,ab.
3	1 or 2
4	insomnia/ or sleep wake disorder/ or (sleep problem* or sleep disorder* or sleep difficult* or insomnia or sleep quality or sleep challenges or sleep disturbance* or sleep wake disorder* or sleep pattern* or sleep disruption or "sleep initiation and maintenance disorder*").ti,ab.
5	(child* or adolescent* or p?ediatric or teenager* or young people).ti,ab.
6	3 and 4 and 5
7	limit 6 to (english language and yr="2012 ‐Current")

Embase	Search terms
1	mobile application/
2	(digital intervention or internet treatment or health intervention or mobile application or cognitive behavio?ral therap* or digital solution* or internet‐mediated or mobile health or online therap* or e?intervention or digital treatment or support app or internet‐delivered or iCBT or CBTi or digital therap* or e?health).ti,ab.
3	1 or 2
4	sleep disorder/
5	Insomnia/
6	(sleep problem* or sleep disorder* or sleep difficult* or insomnia or sleep quality or sleep challenges or sleep disturbance* or sleep wake disorder* or sleep pattern* or sleep disruption or "sleep initiation and maintenance disorder").ti,ab.
7	4 or 5 or 6
8	child/
9	adolescent/
10	(child* or adolescent* or p?ediatric or teenager* or young people).ti,ab.
11	8 or 9 or 10
12	3 and 7 and 11
13	limit 12 to (english language and yr="2012 ‐Current")

PubMed

Search terms

1

(((“Mobile Applications”[MeSH Terms] AND “Mobile Applications”[All Fields]) OR “digital intervention”[All Fields] OR “internet treatment”[All Fields] OR “health intervention”[All Fields] OR “mobile application”[All Fields] OR “cognitive behavioral therap*”[All Fields] OR “cognitive behavioural therap*”[All Fields] OR “digital solution*”[All Fields] OR “internet‐mediated”[All Fields] OR “mobile health”[All Fields] OR “online therap*”[All Fields] OR “eintervention”[All Fields] OR “digital treatment”[All Fields] OR “support app”[All Fields] OR “internet‐delivered”[All Fields] OR “iCBT”[All Fields] OR “CBTi”[All Fields] OR “digital therap*”[All Fields] OR “ehealth”[All Fields]) AND (“sleep initiation and maintenance disorder”[All Fields] OR “Sleep wake disorder”[All Fields] OR “sleep initiation and maintenance disorders”[MeSH Terms] OR “sleep initiation and maintenance disorders”[All Fields] OR “insomnia*”[All Fields] OR “sleep problem*”[All Fields] OR “sleep disorder”[All Fields] OR “sleep difficult*”[All Fields] OR “sleep quality”[All Fields] OR “sleep challenges”[All Fields] OR “sleep disturbance*”[All Fields] OR “Sleep wake disorder”[All Fields] OR “sleep pattern”[All Fields] OR “sleep disruption”[All Fields]) AND (“child”[MeSH Terms] OR “child*”[All Fields] OR “children”[All Fields] OR “adolescent”[MeSH Terms] OR “adolescent*”[All Fields] OR “adolescence”[All Fields] OR “pediatric*”[All Fields] OR “teenager*”[All Fields] OR “young people”[All Fields])) AND (2012/1/1:2024/3/29[pdat]) AND (english[Filter]))

Scopus

Search terms

1

(TITLE‐ABS‐KEY ("digital intervention" OR "internet treatment" OR "health intervention" OR "mobile application" OR "cognitive behavio?ral therap*" OR "digital solution*" OR "internet‐mediated" OR "mobile health" OR "online therap*" OR "e?intervention" OR "digital treatment" OR "support app" OR "internet‐delivered" OR icbt OR cbti OR "digital therap*" OR "e?health") AND TITLE‐ABS‐KEY ("sleep problem*" OR "sleep disorder*" OR "sleep difficult*" OR insomnia OR "sleep quality" OR "sleep challenges" OR "sleep disturbance*" OR "sleep wake disorder*" OR "sleep pattern*" OR "sleep disruption" OR "sleep initiation and maintenance disorder*") AND TITLE‐ABS‐KEY (child* OR adolescent* OR p?ediatric OR teenager* OR "young people")) AND PUBYEAR > 2011 AND PUBYEAR < 2025 AND (LIMIT‐TO (LANGUAGE, "English"))

Web of Science	Search terms
1	(TI=("digital intervention" or "internet treatment" or "health intervention" or "mobile application" or "cognitive behavio$ral therap*" or "digital solution*" or "internet‐mediated" or "mobile health" or "online therap*" or "e$intervention" or "digital treatment" or "support app" or "internet‐delivered" or iCBT or CBTi or "digital therap*" or "e$health"))
2	(AB=("digital intervention" or "internet treatment" or "health intervention" or "mobile application" or "cognitive behavio$ral therap*" or "digital solution*" or "internet‐mediated" or "mobile health" or "online therap*" or "e$intervention" or "digital treatment" or "support app" or "internet‐delivered" or iCBT or CBTi or "digital therap*" or "e$health"))
3	#2 OR #1
4	(TI=("sleep problem*" or "sleep disorder*" or "sleep difficult*" or insomnia or "sleep quality" or "sleep challenges" or "sleep disturbance*" or "sleep wake disorder*" or "sleep pattern*" or "sleep disruption" or "sleep initiation and maintenance disorder*"))
5	(AB=("sleep problem*" or "sleep disorder*" or "sleep difficult*" or insomnia or "sleep quality" or "sleep challenges" or "sleep disturbance*" or "sleep wake disorder*" or "sleep pattern*" or "sleep disruption" or "sleep initiation and maintenance disorder*"))
6	#5 OR #4
7	(TI=(child* or adolescent* or p$ediatric or teenager* or "young people"))
8	(AB=(child* or adolescent* or p$ediatric or teenager* or "young people"))
9	#8 OR #7
10	#3 AND #6 AND #9
11	#10 and English (Languages) and 2024 or 2023 or 2022 or 2021 or 2020 or 2019 or 2018 or 2017 or 2016 or 2015 or 2014 or 2013 or 2012 (Publication Years)

CENTRAL	Search terms
1	Mobile application/
2	(digital intervention or internet treatment or health intervention or mobile application or cognitive behavio?ral therap* or digital solution* or internet‐mediated or mobile health or online therap* or e?intervention or digital treatment or support app or internet‐delivered or iCBT or CBTi or digital therap* or e?health).ti,ab.
3	1 or 2
4	sleep disorder/
5	insomnia/
6	(sleep problem* or sleep disorder* or sleep difficult* or insomnia or sleep quality or sleep challenges or sleep disturbance* or sleep wake disorder* or sleep pattern* or sleep disruption or "sleep initiation and maintenance disorder*").ti,ab.
7	4 or 5 or 6
8	child/
9	adolescent/
10	(child* or adolescent* or p?ediatric or teenager* or young people).ti,ab.
11	8 or 9 or 10
12	3 and 7 and 11
13	limit 12 to (yr="2012 ‐Current" and english language)

Cleary, M. A. , Richardson, C. , Ross, R. J. , Heussler, H. S. , Wilson, A. , Downs, J. , & Walsh, J. (2025). Effectiveness of current digital cognitive behavioural therapy for insomnia interventions for adolescents with insomnia symptoms: A systematic review and meta‐analysis. Journal of Sleep Research, 34(6), e14466. 10.1111/jsr.14466

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.