Pavlos Topalidis

The rapid expansion of consumer sleep trackers (CST) in clinical and basic research has intensified the need for standardised performance evaluation frameworks capable of examining their limitations and capabilities against polysomnography (PSG). In this study, we systematically compared eight CST: sleep² with Polar Verity Sense (VS) and Polar H10, Oura Ring 3, Apple Watch Series 9, Fitbit Charge 6, Garmin Vivoactive 6 & Venu 3, WHOOP 4, and Circul+, against home-based ambulatory PSG. Eighteen participants completed five consecutive nights (Monday to Friday) of home PSG while wearing two paired devices of each CST. Controlled sleep manipulations, such as restricted and extended sleep, were included in the protocol to challenge the devices’ algorithms under varied levels of difficulty. Performance evaluation followed a standardised framework incorporating both epoch-by-epoch and discrepancy analyses to assess multi-class sleep-staging performance and agreement across key sleep parameters. Epoch-by-epoch accuracy and Cohen’s κ were as follows: sleep² -Polar VS = 83.7% (κ = 0.76), sleep²-Polar H10 = 84.04% (κ = 0.76), Oura Ring 3 = 72.5% (κ = 0.59), Apple Watch Series 9 = 72.3% (κ = 0.56), Fitbit Charge 6 = 66.16% (κ = 0.47), Garmin Vivoactive 6 & Venu 3 = 63.4% (κ = 0.41), WHOOP 4 = 65.15% (κ = 0.48), and Circul+ = 55.59% (κ = 0.33). Discrepancy analysis revealed systematic and proportional biases across CST devices: most wrist-worn devices overestimated sleep time and underestimated wake after sleep onset, with error magnitude increasing drastically in atypical nights characterised by fragmented, restricted, or extended sleep. Cardiac-derived arm- and chest-band sleep² derived sleep metrics showed minor deviation from PSG, whereas Oura devices exhibited moderate but variable levels of bias. Overall, these findings reveal substantial variability across CST and highlight the importance of standardised bench marking to support the scientific and clinical use of wearable-derived sleep macrostructure.

Purpose: Given rising numbers in sleep disorders like insomnia and insufficient availability of treatment options, the need for well-validated digital interventions rises. This pilot study aims at assessing the feasibility of an app-program which combines sleep-training based on core elements of Cognitive Behavioural Therapy for Insomnia (CBT-I) with reliable sleep-monitoring based on heart rate variability via an ECG-sensor. Patients and Methods: About 48 participants (26 females) aged 30– 73 (M = 50.33 ± 11.88) were included in the study. At the beginning of the baseline (T0), at start (T1) and end (T2) of the 6-week training phase as well as 4 weeks after the end of the program (T3; follow-up) several questionnaires assessing sleep quality, insomnia severity, general psychological symptom severity, depression, anxiety as well as quality of life were completed. Furthermore, ambulatory polysomnography (PSG) was conducted three times at T0, T1 and T2. General feasibility was assessed by conducting interviews. Results: Overall, the app-program as well as the study protocol was deemed as feasible according to the participants, besides some difficulties regarding app-instructions and certain technical issues, as well as some expected complaints about worse sleep quality during PSG-recordings. For statistical results, insomnia severity (p < 0.001, r = 0.67), sleep quality (p < 0.001, r = 0.56), general psychological symptom severity (p < 0.001, r = 0.68), depression (p = 0.002, r = 0.50) and anxiety (p < 0.001, r = 0.60) improved significantly during the training phase, while quality of life [physical (p = 0.014, r = 0.41) and psychological health (p = 0.049, r = 0.35)] improved significantly during the follow-up-period. PSG data revealed a significant decrease in Wake After Sleep Onset over the course of the study (p = 0.025, r = 0.36), yet no significant changes were found for other sleep parameters. Conclusion: The app-program was largely feasible and potentially effective in improving sleep and well-being. PSG-derived WASO changes highlight the value of objective sleep measures. Future studies should refine protocols and include control conditions for greater generalizability.

Imagine you are listening to a familiar song on the radio. As the melody and rhythm unfold, you can often anticipate the next note or beat, even before it plays. This ability demonstrates the brain’s capacity to extract statistical regularities from sensory input and to generate predictions about future sensory events. It is considered automatic, requiring no conscious effort or attentional resources (1–4). But to what extent does this predictive ability operate when our attention is greatly reduced, such as during sleep? Experimental findings from animal and human studies reveal a complex picture of how the brain engages in predictive processing during sleep (5–13). Although evidence suggests that the brain differentially reacts to unexpected stimuli and rhythmic music (5,7,13), there is a notable disruption in feedback processing, which is essential for generating accurate predictions of upcoming stimuli (10). Here, for the first time, we examine the brain’s ability during sleep to predict or pre-activate low-level features of expected stimuli before presentation. We use sequences of predictable or unpredictable/random tones in a passive-listening paradigm while recording simultaneous electroencephalography (EEG) and magnetoencephalography (MEG) during wakefulness and sleep. We found that during wakefulness, N1 sleep and N2 sleep, subtle changes in tone frequencies elicit unique/distinct neural activations. However, these activations are less distinct and less sustained during sleep than during wakefulness. Critically, replicating previous work in wakefulness (4), we find evidence that neural activations specific to the anticipated tone occur before its presentation. Extending previous findings, we show that such predictive neural patterns fade as individuals fall into sleep. In Brief The extent to which predictive processing takes place in sleep is yet to be determined. Using a passive-listening EEG/MEG paradigm, Topalidis et al. show that auditory representations in sleep are brief and unstable, easily overwritten by subsequent inputs, which possibly hinders the tracking and extraction of sensory associations. Highlights Participants passively listened to random and predictable sequences of tones during both wakefulness and sleep, without being made aware of the underlying pattern. The brain reta ins the ability to process basic low-level features during sleep. While these feature-specific responses are preserved during sleep, they are less distinct and sustained than in wakefulness. Unlike in wakefulness, during sleep, the brain does not predict or anticipate upcoming sounds, despite continuing to process basic auditory information.

Summary: Due to insufficient treatment options for insomnia, effective solutions are urgently needed. We evaluated the effects of a CBT‐I‐based app combining sleep training with subjective and objective sleep monitoring on (i) sleep and (ii) subjective‐objective sleep discrepancies (SOSD). Fifty‐seven volunteers (20–76 years; M_Age = 45.67 ± 16.38; 39 female) suffering from sleep problems were randomly assigned to an experimental group (EG, n = 28) or a waitlist control group (CG, n = 29). During the 6‐week app phase, the EG used the CBT‐I‐based programme and a heart rate sensor for daily sleep monitoring and ‐feedback, while the CG used sleep monitoring only. Sleep was measured (i) subjectively via questionnaires (Insomnia Severity Index, ISI; Pittsburgh Sleep Quality Index, PSQI), (ii) objectively via ambulatory polysomnography (PSG), and (iii) continuously via heart‐rate sensor and sleep diaries. Data revealed interactions for ISI (p = 0.003, ƞ²_part = 0.11) and PSQI (p = 0.050, ƞ²_part = 0.05), indicating training‐specific improvements in EG, yet not in CG. While PSG‐derived outcomes appear to be less training‐specific, a tendential reduction in wake after sleep onset (WASO) was found in EG (p = 0.061, d = 0.55). Regarding changes in SOSD, the results indicate improvements during the app phase (EG) for sleep efficiency, sleep onset latency, and WASO (p ≤ 0.022, d ≥ 0.46); for total sleep time both groups showed a SOSD reduction. The findings indicate beneficial effects of a novel smartphone app on sleep and SOSD. More scientific evaluation of such digital programmes is needed to ultimately help in reducing the gap in non‐pharmacological insomnia treatment.

More and more people quantify their sleep using wearables and are getting obsessed in their pursuit of optimal sleep (“orthosomnia”). However, it is criticized that many of these wearables are giving inaccurate feedback and even can lead to negative daytime consequences. Acknowledging these facts, we here aim to extend previous findings [] in a new sample of 136 self-reported poor sleepers by implementing optimization procedures to minimize erroneous classification when ambulatory sensing sleep. Firstly, here, we introduce an advanced interbeat-interval (IBI) quality control using a random forest method to account for wearable recordings in naturalistic and more noisy settings. We further aim to improve sleep classification by opting for a loss function model instead of overall epoch-by-epoch accuracy to avoid model biases towards the majority class (i.e., “light sleep”). Using these implementations, we compare the classification performance between the optimized (loss function model) and the accuracy model. We use signals derived from PSG, one-channel ECG, and two consumer wearables: the ECG breast belt Polar® H10 (H10) and the Polar® Verity Sense (VS), an optical Photoplethysmography (PPG) heart-rate sensor. Results reveal high overall accuracy for the new model for ECG (86.3 %, κ=.79), H10 (84.4%, κ=.76), and VS (84.2%, κ=.75) with intended improvements in deep sleep and wake. In addition, the new optimized model displays moderate to high correlations and agreement with PSG on primary sleep parameters, while measures of reliability, expressed in intra-class correlations, suggest excellent reliability for most sleep parameters. Finally, it is demonstrated that the new model is still classifying sleep accurately in 4-classes in users taking heart-affecting and/or psychoactive medication which can be considered a prerequisite for use also in high age groups and/or with common disorders. Further improving and validating sleep stage classification algorithms with affordable wearables may resolve existing scepticism and open the door for such approaches in clinical practice.

There is an urgent need for easily accessible treatment options for sleep problems to reduce the current treatment gap in receiving cognitive behavioral therapy for insomnia (CBT-I). Using a randomized controlled trial, we evaluated the efficacy of a CBT-I-based online program on sleep. Fifty-three volunteers (21–71 years; M_Age = 44.6 ± 12.5; 27 female) suffering from impaired sleep were randomly allocated either to the experimental group (EG, n = 27) or to an active control group (CG, n = 26). The EG participated in a 6-week CBT-I-based online program, while the CG received psychoeducation and sleep hygiene instructions. Sleep was assessed both objectively via ambulatory polysomnography (PSG) as well as subjectively via questionnaires at three time points (baseline, pre- and post-intervention). A one-month follow-up assessment was performed using questionnaires. The EG showed small but reliable improvements from pre- to post-intervention in PSG-derived wake after sleep onset (from 58.6 min to 42.5 min; p < 0.05) and sleep efficiency (from 86.0% to 89.2%; p < 0.05). Furthermore, subjective sleep quality (assessed via Pittsburgh Sleep Quality Index) improved significantly during intervention (p = 0.011) and follow-up (p = 0.015) in the EG alone. The Insomnia Severity Index decreased from pre- to post-intervention in both groups (EG: p = 0.003, CG: p = 0.008), while it further improved during follow-up (p = 0.035) in the EG alone. We show that a CBT-I-based online program can improve sleep not only subjectively but also objectively and can be a viable alternative when face-to-face interventions are not available.

Due to insufficient treatment options for sleep disorders, effective solutions are urgently needed. We evaluate the effects of a CBT-I-based app combining sleep training with subjective and objective sleep monitoring on i) sleep and ii) subjective-objective sleep discrepancies (SOSD) as known for insomnia. Fifty-seven volunteers (20-76 years; MAge=45.67±16.38; 39 female) suffering from sleep problems were randomly assigned to an experimental group (EG, n=28) or a waitlist control group (CG, n=29). During the 6-week app phase, the EG used the CBT-I-based program and a heart-rate sensor for daily sleep monitoring and -feedback, while the CG used the sleep monitoring only. Sleep was measured i) subjectively via questionnaires (Insomnia Severity Index, ISI; Pittsburgh Sleep Quality Index, PSQI), ii) objectively via ambulatory polysomnography (PSG), and iii) continuously via heart-rate sensor and sleep diaries during the app phase. Data revealed interactions for ISI (p=.003) and PSQI (p=.050), indicating training-specific improvements in EG, yet not in CG. While PSG-derived outcomes generally appear to be less training-specific, a statistical trend towards a reduction in WASO was found in EG (p=.061). Regarding changes in SOSD, results indicate improvements during app phase (EG) for SE, SOL and WASO (p’s ≤.022); for TST both groups show a SOSD reduction. The findings indicate beneficial effects of a novel smartphone app on sleep and SOSD. More scientific evaluation of such digital programs is needed and should ultimately help to reduce the gap in non-pharmacological treatment care for insomnia.

Objectives: There is an urgent need for easily accessible treatment options for sleep problems to reduce the current treatment gap in receiving cognitive behavioral therapy for insomnia (CBT-I). Using a randomized controlled trial, we evaluate the efficacy of an CBT-I-based online-program on sleep. Methods Fifty-three volunteers (21-71 years; MAge = 44.6±12.5; 27 female) suffering from impaired sleep were randomly allocated either to the experimental group (EG, n = 27) or to an active control group (CG, n = 26). The EG participated in a 6-week CBT-I-based online-program, while the CG received psychoeducation and sleep hygiene instructions. Sleep was assessed both objectively, via ambulatory polysomnography (PSG), as well as subjectively, via questionnaires at three time points (baseline, pre- and post-intervention). A one-month follow-up assessment was done using questionnaires. Results: The EG showed small but reliable improvements from pre- to post-intervention in PSG-derived wake after sleep onset (from 58.6min to 42.5min.; p <.05) and sleep efficiency (from 86.0% to 89.2%; p <.05). Furthermore, subjective sleep quality (assessed via Pittsburgh Sleep Quality Index) improved significantly during intervention (p = .011) and follow-up (p = .015) in the EG only (T0: 8.2, T3: 4.5). The Insomnia Severity Index decreased from pre- to post-intervention in both groups (EG: p = .003, CG: p = .008), while it further improved during follow-up (p = .035) in the EG only. Conclusions: We show that a CBT-I-based online program can improve sleep not only subjectively but also objectively and can be a viable alternative when face-to-face interventions are not available (T0: 13.2, T3: 6.7).