Cognitive behavioural therapy for insomnia : a comparison between sleep compression and sleep restriction

Background: Insomnia is a common and debilitating disorder with significant personal and societal costs. Cognitive behavioural therapy for insomnia (CBT-I) is an effective treatment, with sleep restriction therapy being a key component. However, sleep restriction therapy can be difficult for some patients to tolerate due to the initial sleep deprivation often associated with the treatment. Sleep compression therapy has been proposed as a gentler alternative. This thesis investigates the efficacy, tolerability, and impact of sleep compression therapy in comparison to sleep restriction therapy, as well as the correlation between change in subjective and objective sleep measures. Additionally, it explores the variations in time-in-bed manipulation therapies across the literature.

Methods: The thesis includes three studies and preliminary results from a fourth study. Studies 1-3 were based on the CompRest trial (NCT02743338), a randomised controlled trial of 234 participants with chronic insomnia, recruited via the Internet Psychiatry Clinic in Stockholm, Sweden. Participants were randomised to either sleep compression therapy (COMP) or sleep restriction therapy (REST), both delivered online with therapist support for five weeks followed by five weeks of self-administered treatment. Sleep was measured subjectively using the Insomnia Severity Index (ISI), and sleep diaries, while objective data were collected via actigraphy, and also polysomnography (PSG) for a subgroup (n=36). Study 1 focused on insomnia severity and tolerability of treatment, Study 2 on objective sleep, and Study 3 explored the correlation between subjective and objective sleep changes. Study 4 was a scoping review of time-in-bed manipulation therapies in the literature, registered with the Open Science Framework.

Results: In Study 1, both COMP and REST significantly reduced insomnia severity after ten weeks. However, COMP was not found to be non-inferior to REST, as the confidence interval of the ISI difference crossed the non-inferiority threshold of 1.6 points. In terms of tolerability, COMP had fewer reported side effects in the early weeks and demonstrated better adherence in two out of three measures, though there were no differences in client satisfaction or daytime functioning between the treatments.

In Study 2, objective sleep measures (PSG) showed that both treatments resulted in similar sleep outcomes after 10 weeks, though REST was associated with a larger and more rapid decrease of total sleep time and time in bed. No significant differences were observed for sleep continuity variables or sleep stages between the treatments.

Study 3 investigated correlations between subjective and objective sleep changes. Significant correlations were found between improvements in subjective sleep quality (Karolinska Sleep Quality Index) and reductions in time in bed, sleep onset latency, and increases in sleep efficiency, as measured by PSG. Additionally, improvements in the Restorative Sleep Index were correlated with sleep efficiency increases, fewer awakenings, and a reduction in N3 sleep. However, no significant correlations were found between changes in the ISI and objective sleep measures.

In Study 4 (preliminary results), a scoping review of time-in-bed manipulation therapies, 52 studies, covering 60 interventions, were included. There was considerable variation in the methods for calculating the initial sleep window, with nine different approaches identified. No clear consistency emerged between the naming of the intervention and the calculation of the sleep window. Other instructions, such as nap allowances and the safe lower limits for time in bed, also varied widely. Half of the included studies were randomised controlled trials, and eight directly compared different brief time-in-bed manipulation interventions.

Conclusions: This thesis provides new insights into the efficacy and tolerability of sleep compression therapy as an alternative to sleep restriction therapy for treating insomnia. While sleep compression therapy was not non-inferior to sleep restriction therapy in reducing insomnia severity, it was better tolerated, suggesting it may be a suitable alternative for patients who struggle with the initial sleep deprivation associated with sleep restriction therapy. Both treatments had similar impacts on subjective and objective sleep after ten weeks, though with different trajectories of change. Additionally, subjective improvements in sleep quality were linked to objective changes in sleep parameters, highlighting the potential of addressing change when investigating objective correlates of subjective experiences of sleep. The scoping review revealed substantial variation in time-in- bed manipulation therapies, underscoring the need for further research to compare different approaches.

List of scientific papers

The following manuscripts and publications were included in the thesis:

I. Jernelöv, S .* , Rosén, A .* , Forsell, E., Blom, K., Ivanova, E., Maurex, L., Jansson-Fröjmark, M., Åkerstedt, T., Kaldo, V. Is sleep compression therapy non-inferior to sleep restriction therapy? A single blind randomized controlled non-inferiority trial comparing sleep compression therapy to sleep restriction therapy as treatment for insomnia. [Manuscript]

II. Rosén, A., D'Onofrio, P., Åkerstedt, T. & Jernelöv, S. (2023). A comparison of sleep restriction and sleep compression on objective measures of sleep: a sub-sample from a large randomized controlled trial, Journal of sleep research, 32(4), e13826. https://doi.org/10.1111/jsr.13826

III. d'Onofrio, P., Jernelöv, S., Rosén, A., Blom, K., Kaldo, V., Schwarz, J., & Åkerstedt, T. (2023). The Polysomnographical Meaning of Changed Sleep Quality-A Study of Treatment with Reduced Time in Bed. Brain Sciences, 13(10), 1426. https://doi.org/10.3390/brainsci13101426

*Authors contributed equally