Digital cognitive behaviour therapy for insomnia (dCBT-I): Chronotype moderation on intervention outcomes

Abstract

Using data from 1721 participants in a community-based randomized control trial of digital cognitive behavioural therapy for insomnia compared with patient education, we employed linear mixed modelling analyses to examine whether chronotype moderated the benefits of digital cognitive behavioural therapy for insomnia on self-reported levels of insomnia severity, fatigue and psychological distress. Baseline self-ratings on the reduced version of the Horne-Östberg Morningness-Eveningness Questionnaire were used to categorize the sample into three chronotypes: morning type (n = 345; 20%); intermediate type (n = 843; 49%); and evening type (n = 524; 30%). Insomnia Severity Index, Chalder Fatigue Questionnaire, and Hospital Anxiety and Depression Scale were assessed pre- and post-intervention (9 weeks). For individuals with self-reported morning or intermediate chronotypes, digital cognitive behavioural therapy for insomnia was superior to patient education on all ratings (Insomnia Severity Index, Chalder Fatigue Questionnaire, and Hospital Anxiety and Depression Scale) at follow-up (p-values ≤ 0.05). For individuals with self-reported evening chronotype, digital cognitive behavioural therapy for insomnia was superior to patient education for Insomnia Severity Index and Chalder Fatigue Questionnaire, but not on the Hospital Anxiety and Depression Scale (p = 0.139). There were significant differences in the treatment effects between the three chronotypes on the Insomnia Severity Index (p = 0.023) estimated difference between evening and morning type of -1.70, 95% confidence interval: -2.96 to -0.45, p = 0.008, and estimated difference between evening and intermediate type -1.53, 95% confidence interval: -3.04 to -0.03, p = 0.046. There were no significant differences in the treatment effects between the three chronotypes on the Chalder Fatigue Questionnaire (p = 0.488) or the Hospital Anxiety and Depression Scale (p = 0.536). We conclude that self-reported chronotype moderates the effects of digital cognitive behavioural therapy for insomnia on insomnia severity, but not on psychological distress or fatigue.

Keywords: chronotype; circadian preferences; digital cognitive therapy for insomnia; insomnia; moderator; randomized-controlled trial.

FIGURE 1

Observed means and standard error (SE) for each group at baseline and at 9‐week follow‐up for chronotypes who were allocated to either digital cognitive behaviour therapy for insomnia (dCBT‐I; n = 867) or patient education about sleep (PE; n = 853). The p‐values represent differences between the groups in change from baseline to 9 weeks (group × time interaction term in the linear mixed model), with Insomnia Severity Index (ISI), Hospital Anxiety and Depression Scale (HADS) and Chalder Fatigue Questionnaire (CFQ) as dependent variables