Clocks & Sleep

Digital Cognitive Behavioral Therapy for Insomnia (dCBT-I) is an effective alternative to therapist-delivered CBT-I. However, there is a lack of meta-analyses assessing its effects on other sleep-related outcomes. We aimed to conduct a meta-analysis of randomized controlled trials (RCTs) evaluating dCBT-I in adults with insomnia through polysomnography (PSG) and sleep diary. Systematic searches were performed in PubMed and Web of Science. The outcomes considered were total sleep time (TST), sleep onset latency (SOL), sleep efficiency (SE), wake after sleep onset (WASO), and number of awakenings (NWAK). Meta-analyses were performed using random-effects models to compare dCBT-I with active (in-person or telehealth CBT-I) or inactive (waiting list, no treatment, or minimal intervention) control groups. Of the fourteen RCTs included, only three employed an active control. As no trials used PSG, the analyses relied solely on sleep diary data. DCBT-I showed no statistically significant differences from active controls, indicating comparable effects with therapist-delivered CBT-I. In contrast, it demonstrated statistically significant effects against inactive controls; TST increased by 0.20 h, SOL decreased by 15.53 min, SE improved by 7.91%, WASO reduced by 15.61 min, and NWAK decreased by 0.53. Future research should prioritize comparisons with therapist-delivered CBT-I and incorporate PSG for measuring these parameters.

Anticipation of stressful events can impair sleep quality. In a recent study, we reported that anticipating a stressful task before a nap led to negative changes in sleep parameters, particularly at the end of the nap. In our previous study, we compared stress anticipation with the anticipation of relaxation; thus, the observed effects may have been amplified by sleep quality improvements in the relaxation condition. In the current study, we aimed to replicate these findings using an alternative neutral control condition. The data from a newly collected sample (n = 31) were compared with the data from our previous study (n = 33) using identical analyses. The results reveal an opposite pattern from our previous study: participants in the neutral control condition showed poorer sleep (longer sleep onset latency, reduced slow-wave sleep, and lower SWA/beta ratio) compared to those anticipating stress. In a direct comparison of both studies, sleep parameters in the stress conditions were highly similar across the two studies, suggesting that the divergent outcomes are driven by differences in the control conditions. The temporal dynamic changes observed in our previous study could not be replicated. These findings highlight the importance of carefully considering control conditions in experimental sleep research and suggest that even “neutral” instructions can evoke anticipatory effects. Moreover, the observed benefits of anticipating post-sleep relaxation highlight opportunities for relaxation-based interventions to improve sleep quality.

The circadian phase difference between morning and evening types is a fundamental aspect of chronotype. However, results of categorizations into chronotypes based on reported sleep times show low concordance with those based on measurements of the hormonal or physiological or molecular rhythm–markers of the circadian phase. This might be partially explained by the profound individual differences in the phase angle between the sleep–wake cycle and these rhythms that depends on chronotype, age, sex, and other factors. Here, we examined the possibility of using self-reported sleep times in the condition of 5-days-on/2-days-off school/work schedule to estimate circadian phase differences between various chronotypes. In an in silico study, we determined that, for such an estimation, similarities of the compared chronotypes in weekend sleep duration and weekend–weekday gap and in risetime are required. In the following empirical and simulation studies of sleep times reported by 4940 survey participants, we provided examples of the estimation of circadian differences between chronotypes, and the model-based simulations of sleep times in morning and evening types exemplified a way to confirm such estimations. The results of in silico, empirical, and simulation studies underscore the possibility of using bedtimes and risetimes for direct estimation of the circadian phase differences between individuals in real-life situations, such as a 5-days-on/2-days-off school/work schedule. Additionally, the results of these studies on different chronotypes provided further mathematical modeling and empirical evidence for our failure to sleep more on weekends to recover/compensate/pay back/ catch up on lost sleep.