dCBT indicates digital cognitive behavioral therapy; SHE, sleep hygiene education.
eTable 1. Descriptives for Participants Who Completed and Who Were Lost to Follow-up.
eTable 2. Mediation Analyses on PROMIS Subscales and GSII Total Score, and 2nd and 3rd Ranked Complaints
eTable 3. Mediation Effects for Secondary Outcomes
eTable 4. Frequency of Adverse Events by Treatment Group
Customize your JAMA Network experience by selecting one or more topics from the list below.
Espie CA, Emsley R, Kyle SD, et al. Effect of Digital Cognitive Behavioral Therapy for Insomnia on Health, Psychological Well-being, and Sleep-Related Quality of Life: A Randomized Clinical Trial. JAMA Psychiatry. 2019;76(1):21–30. doi:10.1001/jamapsychiatry.2018.2745
Can digital cognitive behavioral therapy for insomnia improve functional health, psychological well-being, and sleep-related quality of life, and does a reduction in insomnia symptoms mediate these potential improvements?
In a 2-arm, parallel-group randomized clinical trial that included 1711 persons, digital cognitive behavioral therapy significantly improved insomnia symptoms, functional health, psychological well-being, and sleep-related quality of life at 4, 8, and 24 weeks after initiation of treatment. Improvements at 8 and 24 weeks were mediated by improvements in insomnia at week 4 and 8, respectively.
Treating insomnia with digital cognitive behavioral therapy could be a therapeutic pathway for addressing self-reported health, well-being, and quality of life.
Digital cognitive behavioral therapy (dCBT) is a scalable and effective intervention for treating insomnia. Most people with insomnia, however, seek help because of the daytime consequences of poor sleep, which adversely affects quality of life.
To investigate the effect of dCBT for insomnia on functional health, psychological well-being, and sleep-related quality of life and to determine whether a reduction in insomnia symptoms was a mediating factor.
Design, Setting, and Participants
This online, 2-arm, parallel-group randomized trial comparing dCBT for insomnia with sleep hygiene education (SHE) evaluated 1711 participants with self-reported symptoms of insomnia. Participants were recruited between December 1, 2015, and December 1, 2016, and dCBT was delivered using web and/or mobile channels plus treatment as usual; SHE comprised a website and a downloadable booklet plus treatment as usual. Online assessments took place at 0 (baseline), 4 (midtreatment), 8 (posttreatment), and 24 (follow-up) weeks. Programs were completed within 12 weeks after inclusion.
Main Outcomes and Measures
Primary outcomes were scores on self-reported measures of functional health (Patient-Reported Outcomes Measurement Information System: Global Health Scale; range, 10-50; higher scores indicate better health); psychological well-being (Warwick-Edinburgh Mental Well-being Scale; range, 14-70; higher scores indicate greater well-being); and sleep-related quality of life (Glasgow Sleep Impact Index; range, 1-100; higher scores indicate greater impairment). Secondary outcomes comprised mood, fatigue, sleepiness, cognitive failures, work productivity, and relationship satisfaction. Insomnia was assessed with the Sleep Condition Indicator (range: 0-32; higher scores indicate better sleep).
Of the 1711 participants included in the intention-to-treat analysis, 1329 (77.7%) were female, mean (SD) age was 48.0 (13.8) years, and 1558 (91.1%) were white. Use of dCBT was associated with a small improvement in functional health compared with SHE (adjusted difference [95% CI] at week 4, 0.90 [0.40-1.40]; week 8, 1.76 [1.24-2.28]; week 24, 1.76 [1.22-2.30]) and psychological well-being (adjusted difference [95% CI] at week 4, 1.04 [0.28-1.80]; week 8, 2.68 [1.89-3.47]; week 24, 2.95 [2.13-3.76]), and with a large improvement in sleep-related quality of life (at week 4, −8.76 [−11.83 to −5.69]; week 8, –17.60 [−20.81 to −14.39]; week 24, −18.72 [−22.04 to −15.41]) (all P < .01). A large improvement in insomnia mediated these outcomes (range mediated, 45.5%-84.0%).
Conclusions and Relevance
Use of dCBT is effective in improving functional health, psychological well-being, and sleep-related quality of life in people reporting insomnia symptoms. A reduction in insomnia symptoms mediates these improvements. These results confirm that dCBT improves both daytime and nighttime aspects of insomnia, strengthening existing recommendations of CBT as the treatment of choice for insomnia.
isrctn.org identifier: ISRCTN60530898
Insomnia disorder, comprising reports of poor sleep with associated daytime effects occurring 3 or more nights per week for 3 or more months,1 presents in 10% to 12% of adults.2-4 In addition, insomnia is associated with mental health disorders,5 cardiovascular disease,6 and type 2 diabetes.7,8 Increased fatigue, impaired work productivity, reduced quality of life, and relationship dissatisfaction are also common in those with insomnia.9-11 Such impaired functioning is an important driver for help-seeking behavior.12
The recommended intervention for insomnia is cognitive behavioral therapy (CBT),13-16 a psychological treatment designed to break patterns of maladaptive thinking and behavior. Cognitive behavioral therapy comprises a behavioral component (stimulus control, sleep restriction, and relaxation) combined with a cognitive (managing sleep-related worries, racing mind, and intrusive thoughts) and an educational (sleep hygiene) component. Meta-analyses indicate that CBT has moderate to large and durable effects on sleep quality, sleep efficiency, sleep-onset latency, and wake time after sleep onset.17-19 Moreover, recent meta-analyses indicate that digital CBT (dCBT), delivered using automated web platforms or a mobile app,20 is also efficacious.21,22 The effects of CBT and dCBT on the nighttime symptoms of insomnia, therefore, appear robust. However, daytime symptoms are a core part of insomnia disorder, integral to its clinical presentation. Improving constructs such as functional health, psychological well-being, and quality of life may therefore be crucial to treating insomnia satisfactorily.
We investigated the attributable effect of a reduction in insomnia symptoms after receiving dCBT for insomnia on 3 key areas of quality of life: functional health status, psychological well-being, and patient-generated, sleep-related quality of life. Although there is evidence that CBT may yield generalized benefits in both the general population and patient groups23-29 and some primary data that CBT for insomnia may reduce depressive or anxiety symptoms,30-34 evidence is mixed,35,36 and an adequately powered, definitive trial investigating functional health status, psychological well-being, and a patient-generated assessment of quality of life has not yet been conducted. Moreover, we wanted to conduct a formal test of the mediating effect of improved insomnia symptoms on these outcomes.
Quiz Ref IDOur primary hypotheses were that dCBT for insomnia would improve functional health status and psychological well-being and would reduce sleep-related quality-of-life impairment at weeks 4, 8, and 24 (research question [RQ] 1) and that the effect of dCBT on these outcomes at weeks 8 and 24 would be mediated by a reduction in insomnia symptoms measured at weeks 4 and 8, respectively (RQ 2). Our secondary hypotheses were that dCBT would also improve domains of personal functioning (negative mood, fatigue, and relationship dysfunction) and daytime performance (sleepiness, concentration, and productivity) at weeks 4 and 8 (RQ 3), improvements would be maintained at the week 24 follow-up, and the effect of dCBT at weeks 8 and 24 would be mediated by a reduction in insomnia complaints at weeks 4 and 8, respectively (RQ 4).
Quiz Ref IDThe study was an online, 2-arm, single-blind, parallel-group, superiority randomized clinical trial of dCBT (Digital Insomnia Therapy to Assist Your Life as Well as Your Sleep [DIALS]) in addition to any treatment the participant was previously receiving (treatment as usual [TAU]) vs sleep hygiene education (SHE) in addition to TAU. We used simple randomization with an allocation ratio of 1:1 as recommended for large clinical trials.37 Randomization was carried out by an outside automated online system (surveygizmo.eu), ensuring that allocation could not be influenced by the research team. The trial design and progress of participants through the trial are summarized in the Figure. Screening, informed consent, assessments, allocation to condition, and delivery of the interventions were carried out entirely online. Participants were recruited between December 1, 2015, and December 1, 2016; programs were completed within 12 weeks after inclusion. The study received ethical approval from the University of Oxford Medical Sciences Inter-Divisional Ethics Committee (ref MS-IDREC-C2-2015-024), and participants provided online informed consent. The DIALS trial has been registered at http://www.isrctn.com (ISRCTN60530898); the protocol has been published and is available in Supplement 1.38
According to the original protocol, a sample size of 433 participants per treatment group was required to detect a standardized effect size of 0.25 with 90% power, assuming a significance level of P < .01667 (corrected for 3 primary outcomes), and to detect a large mediation effect with more than 80% power. To account for 13% attrition, we increased the sample size to 500 per treatment arm. During the trial, we further extended recruitment because the attrition rate was larger than expected. A total of 1718 participants enrolled, and the final sample comprised 1711 participants because 7 participants entered the trial twice. For each of these 7 participants, a single entry contributed to the analyses. When both entries were randomized to the same condition, the response with the most completed data was selected; when entries were randomized to different groups, the first response was selected unless the participant accessed treatment; if treatment was accessed, the response corresponding with the treatment allocation was used. Inclusion and exclusion criteria have been reported earlier and are available in Supplement 1.38 Briefly, inclusion criteria comprised the following: a positive screening results based on the DSM-5 criteria for insomnia disorder; a score of 16 or less on the 8-item Sleep Condition Indicator (SCI; scale, 0-4; range: 0-32, with higher scores indicating better sleep)39; aged 18 years or older; reliable internet access; and ability to read and understand English. We screened for comorbid conditions and medication use at baseline but excluded only those people whose health was expected to necessitate hospital admission or who had a life expectancy of less than 6 months, who currently received psychological treatment for insomnia or were expecting treatment within 6 months, and who reported suicidal thoughts. We did not exclude participants taking medication for sleep problems or for any other physical or mental health problem. Several methods were used to direct people to the online recruitment page. Persons who had completed previous sleep surveys (ie, Great British Sleep Survey40; World Sleep Survey41) were contacted by email, a recruitment button was placed on an insomnia intervention website,42 advertisements were placed on Facebook and announcements on Twitter, and information about the study was presented via broadcast media.
Digital cognitive behavioral therapy was delivered using the Sleepio program (Big Health Ltd)43 and an associated iOS app (Big Health Ltd). The program is fully automated, and its underlying algorithms feed the delivery of information, support, and advice in a personally tailored manner. Delivery is structured into 6 sessions typically lasting 20 minutes each, and participants had access to the intervention for up to 12 weeks. Treatment content is based on CBT manuals44-46 and includes behavioral, cognitive, and educational components. A more extensive description can be found in the study protocol (Supplement 1).38 The program has been evaluated in multiple randomized clinical trials.31,33,43,47-50
SHE was selected for the control arm because this is the behavioral comparator that people with insomnia are most typically offered in routine care. SHE therefore was based on recognized sleep hygiene advice, for example, recommendations about bed routines and use of alcohol and caffeine.46,51 To ensure consistency of approach and content, SHE was delivered on a dedicated study webpage where materials could be viewed and downloaded in a single session.
Assessments took place at weeks 0 (baseline), 4 (midtreatment), 8 (posttreatment), and 24 (follow-up). At week 25, all participants in the control group (SHE) were offered dCBT, which finished the controlled element of the trial. Uncontrolled follow-up data were collected at week 36 and week 48; these data are not presented herein.
The 3 primary measures used to index physical health, psychological well-being, and sleep-related quality of life were the Patient-Reported Outcomes Measurement Information System: Global Health scale52 for physical health (PROMIS-10; 10 items scored 1 to 5; range: 10-50, with higher scores indicating better health), the Warwick-Edinburgh Mental Well-being Scale53 for psychological well-being (WEMWBS, 14 items scored 1 to 5; range: 14-70, with higher scores indicating better well-being), and the Glasgow Sleep Impact Index54 (GSII), a patient-generated outcome rating in which participants rate self-defined sleep-related impairments (range: 0-100, with higher scores indicating greater impairment).
Secondary outcomes related to specific measurement of 6 areas of daytime consequences typically associated with the clinical diagnosis of insomnia disorder.1,10 These outcomes were mood (9-item Patient Health Questionnaire55 [PHQ-9]; scale, 0-3; range: 0-27, and the 7-item Generalized Anxiety Disorder56 [GAD-7]; scale, 0-3; range: 0-21), energy (7-item Flinders Fatigue Scale57 [FFS]; scale, 0-4; range: 0-28), relationship satisfaction (7-item Relationship Assessment Scale58 [RAS]; scale, 1-5; range: 7-35), cognitive functioning (25-item Cognitive Failures Questionnaire59 [CFQ]; scale, 0-4; range: 0-100), work performance and satisfaction (Work Productivity and Activity Impairment questionnaire: Specific Health Problem60 [WPAI:SHP] and 1 item on job satisfaction61), and sleepiness (8-item Epworth Sleepiness Scale62 [ESS]; scale, 0-3; range: 0-24]). As an exploratory measure, participants also completed 1 item about their general life satisfaction.63 To appraise the mediating effects of improvement of insomnia symptoms per se, we used the SCI.34,55 At week 8, potential adverse effects were assessed by asking participants to rate a number of potential adverse effects for frequency and severity.64
All analyses were performed as intention to treat and blinded using Stata version 14 (StataCorp). In accordance with the Consolidated Standards of Reporting Trials guidelines, all participant flow is reported, including descriptive statistics of recruitment, dropout, and completeness of interventions.
The main efficacy analysis was based on randomized allocation including all participants with an outcome recorded at the relevant time point. No interim analyses for efficacy or futility were conducted.
Each primary outcome was analyzed using a linear mixed-effects model to account for the repeated measures at 4, 8, and 24 weeks.65 This method implicitly accounts for missing-at-random outcome data. The baseline outcome measure, treatment assignment, and categorical time point were included as fixed effects, with a random participant-level effect. Adjusted, absolute between-group changes in outcome at each time point were obtained by including an interaction term between time point and treatment assignment in each model. Results are presented with 95% CIs and 2-sided P values. Cohen d standardized effect sizes were estimated by dividing the adjusted between-group difference by the baseline pooled SD of the corresponding outcome. Assumptions of normality were assessed graphically using histograms. The robustness of the assumptions regarding missing outcome data were examined in a series of sensitivity analyses (pattern mixture models, inclusion of baseline characteristics associated with having a missing outcome, and last observation carried forward). Secondary outcomes were analyzed using similar mixed-effects models as the primary outcomes.
A series of linear mixed-effects models were fitted to assess the extent to which the effect of the intervention on each outcome at 8 weeks was mediated by changes in insomnia at 4 weeks. The total effect of the intervention on outcome at 8 weeks was estimated from a model adjusting for baseline SCI but not the baseline outcome. The direct effect of the intervention at 8 weeks was estimated from a model adjusting for baseline SCI and SCI at 4 weeks. The effect of SCI at 4 weeks was also extracted from this model and multiplied by the estimated effect of the intervention on SCI at 4 weeks to obtain the indirect effect. This approach is similar to that described by Baron and Kenny66 but uses linear mixed-effects models.67 The percentage mediated, estimated as the indirect effect divided by the total effect, was obtained. The extent of mediation of the outcome effects at 24 weeks by insomnia at 8 weeks was evaluated in the same way.
The final sample comprised 1711 adults, of whom 1329 (77.7%) were female and the mean (SD) age was 48.0 (13.8) years; 1558 (91.1%) were white, 45 (0.3%) were Asian, 19 (0.1%) were black, 36 (0.2%) were of mixed race/ethnicity, 35 (0.2%) were of another race/ethnicity, and 17 (0.1%) did not wish to state race/ethnicity. Participants were recruited between December 1, 2015, and December 1, 2016, and allocated to either dCBT plus TAU (n = 853) or SHE plus TAU (n = 858). An overview of sample descriptive statistics and baseline scores for the primary and secondary outcomes can be found in Table 1. Full details of sample characteristics and missing data can be found in eTable 1 in Supplement 2. Dropout from study assessments was greater in the treatment group than in the control group (Figure).
In the dCBT group, 689 of the 853 participants (80.8%) logged on for at least 1 session, 491 participants (57.6%) completed at least 4 sessions, and 413 participants (48.4%) completed all 6 sessions (Figure). Sleep hygiene education was accessed at least once by 759 of 858 participants (88.5%).
Quiz Ref IDAt weeks 4, 8, and 24, dCBT was associated with significant improvement in global health (Cohen d for week 4, 0.16; week 8, 0.31; and week 24, 0.31) and mental well-being (Cohen d for week 4, 0.13; week 8, 0.35; and week 24, 0.38), and a significant reduction in sleep-related impairment to quality of life (Cohen d for week 4, –0.69; week 8, –1.38; and week 24, –1.46; RQ 1 [Table 2]). Effects were robust across sensitivity analyses investigating assumptions regarding missingness of the outcome.
The results of the mediation analyses can be found in Table 3 (indirect effects, RQ 2). The mediator (SCI) demonstrated an improvement compared with SHE (Cohen d for week 4, 0.89; week 8, 1.51; and week 24, 1.51; all P < .01 Table 4). Quiz Ref IDWhen we considered functional health on PROMIS-10 as the outcome, the proportion of the effect of dCBT on PROMIS-10 score at 8 weeks that was mediated by changes in insomnia symptoms at 4 weeks was 50.5%, and the proportion of the effect of the intervention on PROMIS-10 scores at 24 weeks that was mediated by changes in insomnia symptoms at 8 weeks was 83.8%. For psychological well-being, these values were 47.0% at 8 weeks and 74.9% at 24 weeks. When we considered sleep-related quality of life, with GSII (rank 1: the most important impairment) as the outcome, the proportion of the effect of the intervention on GSII at 8 weeks that was mediated by changes in insomnia symptoms was 45.5% and increased to 65.9% at 24 weeks. eTable 2 in Supplement 2 provides mediation analyses on PROMIS subscales and the GSII total score and second- and third-ranked reported impairments.
Symptoms of depression (PHQ-9), anxiety (GAD-7), sleepiness (ESS), and cognitive failures (CFQ) all demonstrated significant differences in favor of dCBT at weeks 4, 8, and 24, reflecting small effect sizes (Table 4, RQ 3). There were moderate to large effects observed at weeks 4, 8, and 24 for fatigue (FFS). On the WPAI, productivity at work (presenteeism) that was attributed to sleep problems showed a small to moderate improvement after dCBT relative to control. A significant but small effect in terms of reduced absenteeism attributed to poor sleep and increased job satisfaction was observed at week 24. There were no significant effects at any time point on relationship functioning (RAS). Mediation analyses indicated that changes in insomnia symptoms also accounted for significant and sizeable proportions of effects on these secondary outcomes at both week 8 and week 24 (eTable 3 in Supplement 2, RQ 4).
There was 1 serious adverse event reported, which was unrelated to the intervention. The event was reported to the University of Oxford Medical Sciences Inter-Divisional Ethics Committee. Participants who received dCBT reported a higher occurrence of difficulty remembering things, headache and/or migraine, fatigue and/or exhaustion, extreme sleepiness, difficulty with concentration and attention, reduced motivation and/or energy, and irritability attributed to the insomnia improvement program at week 8 (eTable 4 in Supplement 2).
It is well established that CBT is the first-line treatment for people with chronic insomnia14,17-19,21,22 and that sleep-related outcomes, whether on index measures of insomnia or on derivations from sleep diaries, show sustained improvement.17 The findings from this study confirm large effects on insomnia symptoms. However, we primarily addressed whether CBT for insomnia affects functional health, psychological well-being, and sleep-related quality of life. We used dCBT instead of face-to-face CBT to facilitate an adequately powered scientific inquiry of explanatory processes.
Our findings provide strong support for the hypothesis that dCBT improved participants’ functional health relative to sleep hygiene education. At 24 weeks, effect sizes were small for functional health and psychological well-being (Cohen d, 0.31 and 0.38) and large for sleep-related quality of life (Cohen d, −1.46). The greater effect on sleep-related quality of life, measured with the GSII, likely occurs because the GSII specifically asks participants to define areas of impairment attributed to poor sleep, whereas functional health and well-being were assessed with global measures (PROMIS-10 and WEMWBS). Improvements in all 3 primary outcomes were mediated by insomnia improvements associated with dCBT. These mediation effects were substantial: 51% mediation of functional health, 47% of well-being, and 46% of sleep-related quality of life at 8 weeks, and 84% of functional health, 75% of well-being, and 66% of sleep-related quality of life at 24 weeks’ follow-up. To our knowledge, this is the first large-scale study demonstrating a causal relationship between CBT-mediated reduction in insomnia symptoms and perceived health status and quality of life.
A report of a parallel study (Oxford Access for Students Improving Sleep [OASIS]) exploring mental health symptoms as the outcome variable of interest and likewise demonstrated a causal relationship between insomnia improvement and reductions in mental health symptoms.33 Together, the mediation analyses in these 2 studies (with a total of 5466 participants) provide novel and convincing evidence that insomnia may be a legitimate and important target for mental health and well-being. Consideration of the secondary variables in the present study yields further data on the generalized effects of insomnia improvement on symptoms of depression, anxiety, fatigue, sleepiness, and cognitive failures and productivity, with mediation in substantial part through changes in insomnia.
Quiz Ref IDWe are mindful of limitations of this research. Although participant demographic characteristics were typical of clinical populations (eg, typically female, middle-aged) and had some similar characteristics (eg, two-thirds reported diagnosed health comorbidities), participants were not drawn from patient groups or health care services and therefore reflect individuals who experience insomnia in the general population. This difference may reduce the generalizability to the insomnia patient group but increases the generalizability to those experiencing insomnia symptoms in the general population. As is typical of digital programs, and indeed of CBT in general, there was a substantial dropout from treatment (58% of participants completed ≥4 dCBT sessions); however, intention-to-treat analyses still identified significant improvements. Sensitivity and missing data analyses did not change the study conclusions. In addition, the instrument used to measure insomnia symptoms includes items around daytime functioning because these items are part of the DSM-5 insomnia disorder diagnosis; these items may have inflated the mediation effects. Finally, we note that attributed adverse effects were more common in the group receiving dCBT than in the control participants. Despite widespread generalized benefits, dCBT can have adverse effects; we hypothesize that these effects might result from the sleep restriction component that is introduced in week 3. Potentially, these adverse effects were short lived; at weeks 8 and 24, sleepiness, fatigue, cognitive failures, and mood symptoms were more common in participants receiving SHE than in those receiving dCBT.
The results of this definitive trial suggest that dCBT not only is effective in improving insomnia symptoms but also demonstrates positive effects around the clock by improving the functional health, psychological well-being, and sleep-related quality of life of people with positive screening results for insomnia disorder. In addition, improved insomnia is a mediator of these benefits. These findings indicate that dCBT improves both daytime and nighttime aspects of insomnia, lending further weight to the clinical guideline recommendation of CBT as the treatment of choice for insomnia.
Accepted for Publication: August 4, 2018.
Corresponding Author: Colin A. Espie, PhD, Sleep & Circadian Neuroscience Institute, Nuffield Department of Clinical Neurosciences, University of Oxford, Sir William Dunn School of Pathology, South Parks Road, Oxford OX1 3RE, United Kingdom (email@example.com).
Published Online: September 25, 2018. doi:10.1001/jamapsychiatry.2018.2745
Author Contributions: Dr Espie had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Espie, Kyle, Gordon, Drake, Siriwardena, Cape, Ong, Sheaves, Foster, Freeman, Costa-Font, Luik.
Acquisition, analysis, or interpretation of data: Emsley, Kyle, Gordon, Drake, Costa-Font, Marsden, Luik.
Drafting of the manuscript: Espie, Emsley, Kyle, Marsden, Luik.
Critical revision of the manuscript for important intellectual content: Espie, Emsley, Kyle, Gordon, Drake, Siriwardena, Cape, Ong, Sheaves, Foster, Freeman, Costa-Font, Luik.
Statistical analysis: Emsley, Marsden.
Administrative, technical, or material support: Kyle, Gordon, Siriwardena, Sheaves, Freeman, Luik.
Supervision: Espie, Gordon, Costa-Font.
Conflict of Interest Disclosures: Dr Espie reports being a cofounder, chief medical officer, and shareholder of and receiving salary from Big Health Ltd and being a developer of Sleepio. Drs Kyle and Drake report receiving nonfinancial support from Big Health Ltd (provision of Sleepio for use in clinical trials). Dr Cape reports providing clinical advice and support to Sleepio and receiving payment from Big Health Ltd. Dr Ong reports receiving nonfinancial support from Big Health Ltd (provision of Sleepio for use in clinical trials), providing consultancy support for Sleepio, and receiving payment from Big Health Ltd. Dr Sheaves reports providing monthly support for an online discussion forum run by Sleepio and receiving payment from Big Health Ltd. Dr Freeman reports being a cofounder of the University of Oxford spinout company, Oxford VR; receiving nonfinancial support from Big Health Ltd (provision of Sleepio for use in clinical trials); and being supported by an NIHR Research Professorship. Dr Luik held a research position at the University of Oxford during the conduct of the study that was funded by Big Health Ltd. No other conflicts were reported.
Funding/Support: The study was funded by Big Health Ltd. The work was supported in part by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre, NIHR Oxford Health Biomedical Research Centre, NIHR Biomedical Research Centre at South London, Maudsley National Health Service (NHS) Foundation Trust, King’s College London, and the Dr Mortimer & Theresa Sackler Foundation.
Role of the Funder/Sponsor: Big Health Ltd was involved in the design and conduct of the study; collection, management, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The funder was not involved in the analysis of the data. Other funders had no role in the design and conduct of the study, collection of the data, data analysis, management, interpretation, or review or approval of the manuscript.
Disclaimer: The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. Dr Freeman is supported by an NIHR Research Professorship.
Meeting Presentation: This paper was presented at the 24th Congress of the European Sleep Research Society; September 25, 2018; Basel, Switzerland.
Additional Contribution: Alasdair Henry, PhD, Big Health Ltd, helped with formatting the manuscript, which was performed as part of his regular duties; he was not additionally compensated. Sleepio was provided to participants at no cost. The study was conducted at the University of Oxford, Sleep & Circadian Neuroscience Institute. The University of Oxford has a memorandum of understanding with Big Health for the conduct of joint research.
Create a personal account or sign in to: