The SENSE study (Sleep and Education: learning New Skills Early): postintervention effects of a randomised controlled trial of a cognitive-behavioural and mindfulness-based group sleep improvement intervention among at-risk adolescents

© 2016 Dr. Matthew John BlakeObjective: There is growing recognition that many adolescents obtain insufficient and/or poor quality sleep. Sleep problems are also a major risk factor for the emergence of mental health problems in adolescence. However, few studies have examined disturbed sleep as a potential mechanism in the treatment and prevention of mental health problems among adolescents. Adolescent sleep problems can be treated using a range of approaches. School-based sleep education programs, which are typically delivered to whole school classes, have been shown to have little impact on sleep behaviour or mental health. Cognitive-behavioural and mindfulness-based sleep programs, which are typically delivered to at-risk or already symptomatic adolescents, have been shown to be more effective in improving sleep and emotional distress, but studies evaluating their effectiveness have been limited in several ways, including small sample sizes and inadequate/lack of control groups. We conducted a systematic review and meta-analysis examining the efficacy of cognitive-behavioural sleep interventions among adolescents. Searches of PubMed, PsycINFO, CENTRAL, EMBASE, and MEDLINE were performed from inception to 1 May 2016. Eight trials were selected (n=234, mean age=15.24 years; female=63.18%). Main outcomes were subjective (sleep diary/questionnaire) and objective (actigraphy) total sleep time (TST), sleep onset latency (SOL), sleep efficiency (SE), and wake after sleep onset (WASO). There was a small number of randomised controlled trials (RCTs; n=3), and a high risk of bias across the RCTs; therefore within sleep condition meta-analyses were examined. At post-intervention, subjective TST improved by 29.47 minutes (95% CI = 17.18, 41.75), SOL by 21.44 minutes (95% CI = -30.78, -12.11), SE by 5.34% (95% CI = 2.64, 8.04), and WASO by a medium effect size (d = 0.59 [95% CI = 0.36, 0.82). Objective SOL improved by 16.15 minutes (95% CI = -26.13, -6.17), and SE by 2.82% (95% CI = 0.58, 5.07). Global sleep quality, daytime sleepiness, depression, and anxiety also improved. Gains were generally maintained over time. Our meta-analysis provides preliminary evidence that cognitive-behavioural sleep interventions are an effective treatment for adolescent sleep problems, producing clinically meaningful responses within active treatment conditions. Their efficacy is maintained over time, and results in significant alleviation of sleep problems and improvement in functional outcomes. However, further large-scale, high-quality RCTs are needed to confirm these findings. The aim of this thesis was to investigate the post-intervention effects of a cognitive-behavioural/mindfulness-based group sleep intervention on sleep, mental health, and cognitive style among at-risk adolescents. The study went beyond simply measuring treatment outcomes to also evaluate mechanisms of change. Based on the behavioural, cognitive, hyperarousal, and transdiagnostic models of insomnia, a number of specific mediators were hypothesised to account for therapeutic change in cognitive-behavioural and mindfulness-based sleep interventions for adolescents, including earlier bedtimes, more consistent bedtimes, increased sleep hygiene awareness, and decreased dysfunctional beliefs and attitudes about sleep, worry, rumination, pre-sleep arousal, anxiety, and depression. Method: A RCT was conducted across Victorian secondary schools in Melbourne, Australia. Adolescents (aged 12-17 years) were recruited using a two-stage procedure, consisting of an in-school screening (n=1491) followed by a diagnostic interview for those meeting screening criteria (n=218), to identify students with high levels of anxiety and sleeping difficulties, but without past or current major depressive disorder (n=144). Eligible participants were randomised into either a sleep improvement intervention (‘Sleep SENSE’) or an active control ‘study skills’ intervention (‘Study SENSE’). One hundred twenty three participants began the interventions (Female=60%; Mean Age=14.48, SD=0.95), with 60 in the sleep condition and 63 in the control condition. All participants were required to complete a battery of mood, sleep and cognitive style questionnaires, seven-days of wrist actigraphy (an objective measurement of sleep), and sleep diary entry at pre-and-post intervention. Results: The sleep intervention condition was associated with significantly greater improvements in subjective sleep (global sleep quality, sleep onset latency, daytime sleepiness), objective sleep onset latency, anxiety, pre-sleep arousal, and sleep knowledge compared with the control intervention condition, with small-medium effect sizes. Parallel multiple mediation models showed that there were bidirectional relationships between improvements in subjective sleep quality and pre-sleep arousal/global anxiety. Conclusion: The SENSE study is an efficacy trial of a selective group-based sleep intervention for the treatment and prevention of sleep and mental health problems among at-risk adolescents experiencing both sleep and anxiety disturbance. The study provides evidence, using a methodologically rigorous design, including an active control comparison condition, that a multi-component group sleep intervention that includes cognitive-behavioural and mindfulness-based therapies, can improve wakefulness in bed variables, daytime dysfunction, anxiety, pre-sleep arousal, and sleep knowledge among at-risk adolescents. The results also provide evidence that pre-sleep arousal and anxiety are particularly important for adolescents’ perceived sleep quality, and should be key targets for new treatments of adolescent sleep problems. Public Health Significance: Given the high prevalence of adolescent sleep and internalising problems, the implications of an effective adolescent sleep intervention for clinical practice and public policy are potentially significant. However, changing sleep behaviour, especially objective measures of sleep, in this age group, has been challenging. This thesis shows that the Sleep-SENSE program can improve objective and subjective indices of sleep, as well as anxiety symptoms, when compared to an active control intervention. The results also showed that reductions in pre-sleep hyperarousal represent a key psychophysiological mechanism for therapeutic improvements in subjective sleep problems among anxious adolescents, and that cognitive behavioural and mindfulness-based sleep interventions should be directed towards adolescents with vulnerability for hyperarousal. Sleep SENSE is one of the only interventions demonstrated to be efficacious in improving sleep and mental health amongst vulnerable adolescents. Furthermore, the program is likely to be cost-effective - it involves a simple screening process and a group intervention format - and could be disseminated to a wide range of clinical and non-clinical settings in primary care, mental health, adolescent health and sleep medicine, and may assist in the treatment and prevention of adolescent sleep and mental health problems. The intervention also lends itself to flexible modes of delivery (e.g., non-specialist practitioners, group settings, individual settings, school-based, internet and other e-health modes of delivery), further enhancing its translational potential