The effects of a CBT-I based app-program on sleep quality, insomnia severity, psychological strain and quality of life: A first efficacy study

doi:10.21203/rs.3.rs-3218498/v1

Download PDF

Article

The effects of a CBT-I based app-program on sleep quality, insomnia severity, psychological strain and quality of life: A first efficacy study

https://doi.org/10.21203/rs.3.rs-3218498/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Given rising numbers in sleep disorders and insufficient availability of treatment options, the need for well-validated digital interventions rises. This study aims at assessing the efficacy 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. Data of 48 participants (26 female) aged 30-73 (M=50.33±11.88) were included in the analyses. 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. Insomnia severity (p<.001), sleep quality (p<.001), general psychological symptom severity (p<.001), depression (p=.002) and anxiety (p<.001) improved significantly during the training phase and persisted until one-month follow-up assessment, while in sleep quality a trend for further improvement during the follow-up period was found (p=.097). Furthermore, quality of life [physical (p=.014) and psychological health (p=.049)] improved significantly during the follow-up-period. This study adds evidence on the potential of digital interventions and indicates that digital CBT-I can have substantial effects on relevant sleep- and health-related outcome measures.

Health sciences/Health care/Quality of life

Biological sciences/Psychology

Health sciences/Health care

Health sciences/Health care/Disease prevention

Health sciences/Neurology/Neurological disorders/Sleep disorders

Insomnia

CBT-I

prevention

digital intervention

app-program

Restorative and sufficient sleep is needed for vital processes such as physical recovery, growth of bones and muscles [1], the formation and strengthening of the immune system [2] and it is furthermore beneficial for cognitive abilities such as memory consolidation and processing of emotional stimuli [3]. Particularly in the western world, however, more and more people are not achieving the recommended amount of sleep, as also shown in an Austrian survey: Only about half of nearly 1000 participants reported sleeping between 7h and 9h per night [4], which is the recommended amount of sleep for adults aged 18-64 years according to the National Sleep Foundation [5]. While in 2010 sleep disturbances were reported by 18% of the Austrian population [6], ten years later the number was at 46% [4] indicating a 2.5-fold increase. This sharp and rapid augmentation illustrates the steadily growing prevalence of sleep problems such as insomnia [7, 8] in modern societies and asks for further attention. Affecting currently 6-12% of the population, insomnia is considered the most common sleep disorder in western societies [9, 10]. However, up to 30% of the general population suffer from disordered sleep and symptoms of insomnia, without necessarily meeting criteria for an insomnia diagnosis [11, 12]. Furthermore, only an alarming low number of 31% of the Austrian population qualify as good sleepers according to Pittsburgh Sleep Quality Index [PSQI; 13] self-reporting [4]. Insufficient sleep and chronic insomnia are associated with multiple health constraints, including an increased risk for developing or dying from cardio-vascular diseases [14] as well as for affective disorders [15, 16], type 2 diabetes [17] and substance abuse such as increased alcohol consumption [18]. Moreover, the socio-economical costs of a person with symptoms of insomnia or an insomnia disorder are increased 3-12 times due to decreased working performance and absenteeism [19], resulting in annual costs of insomnia of more than $ 60 billion in the US [20] and € 40-50 billion in Germany [21].

The first-line treatment as suggested by the European Sleep Research Society as well as by the World Sleep Society [22] is the Cognitive Behavioural Therapy for Insomnia (CBT-I), which includes elements such as stimulation control, sleep restriction, psychoeducation, relaxation methods and cognitive training to restructure dysfunctional thoughts [23]. CBT-I, traditionally held in face-to-face settings, has been shown efficacious regarding improvements in the Insomnia Severity Index [ISI; 24] and the PSQI, as well as for subjectively (and sometimes objectively) derived sleep parameters such as total sleep time (TST), sleep onset latency (SOL), wake after sleep onset (WASO) and sleep efficiency (SE) from pre- to post-treatment [25-27].

Since the need for CBT-I offered in person cannot be met sufficiently due to a lack of trained clinicians, digital CBT-I (dCBT-I), which can be conducted without the supervision of a clinician, represents a good, additional treatment option [28]. Additionally, not every patient is willing to consult clinical support for their sleep complaints and dCBT-I might be a more accepted low-threshold alternative in such cases. For dCBT-I, the content of CBT-I is digitized and made available in the form of a website or smartphone application. Various studies and meta-analyses indicate good efficacy of dCBT-I programs in terms of improvements in ISI and sleep parameters such as TST, SOL, WASO, SE and number of awakenings (NOA) from pre-intervention to post-intervention [29-32], while these effects, except for the improvement of TST remained stable in follow-ups of 4 weeks to 6 months [31]. It should be emphasized, however, that the improvement in sleep parameters refers to sleep parameters collected subjectively with a sleep diary. Objective measures of sleep with polysomnography (PSG) or actigraphy are usually not performed in these studies.

To summarize, digital CBT-I programs are a well-performing and effective complement to face-to-face CBT-I therapies. What unites existing and established CBT-I programs is the content which consists of sleep hygiene exercises, cognitive training, as well as techniques for behavioural changes, as well as self-monitoring in terms of a subjective sleep diary. One limitation is that for evaluation of such programs most solely need to rely on subjective self-reports of improved sleep quality over time. What these programs usually lack is a way to objectively and reliably record and evaluate sleep measures over the full intervention period. Objective assessment is usually disregarded because insomnia is understood to be a disorder that manifests mainly in the subjective experience, and it is argued that subjective sleep quality cannot be adequately predicted by objective measurement methods such as polysomnography [33]. However, in addition to subjective sleep perception and its change over the course of treatment, consideration of potential changes in objective sleep parameters is worthwhile and necessary as some meta-studies do find objective sleep alterations [27] that ask for intervention and physiological normalization. Several studies also describe a significant discrepancy between subjectively and objectively assessed sleep parameters in insomnia patients, such as an overestimation of sleep onset latency [34] or an underestimation of total sleep time [35], which call into question how accurate sleep self-reports can be especially in disordered sleep patients. This discrepancy between subjective and objective sleep measures, which is typical for insomnia sufferers, can be reduced through CBT-I therapy [36], and can only be made visible by collecting and comparing both. In addition, as described before, insufficient sleep is associated with a variety of physical health problems such as an increased risk of developing cardiovascular diseases [14], type 2 diabetes [17], or an increased susceptibility to infection [37]. In order to assess these risks as well as to monitor changes in sleep across treatment, continuous and valid objective measurements of sleep are needed.

This paper aims to investigate the efficacy of a smartphone-based app-program that integrates a i) sleep training program based on CBT-I (over the course of 6 weeks) with ii) sleep evaluation done subjectively as well objectively using an affordable and easy to wear ECG chest belt. In the present paper we focus on subjective changes of sleep quality, insomnia severity, psychological strain and quality of life over the course of the 6-week training period.

Participants

A total of 66 volunteers (34 female) aged 30-74 years (M = 50.27±12.44) participated in the study. The final sample consisted of 48 participants (26 female) aged 30-73 (M = 50.33±11.88), who finished the whole study protocol including the follow-up measurement, resulting in a dropout rate of 27%, which is significantly less compared to dropout rates of other studies investigating the efficacy of digital treatment programs for insomnia in outpatient samples (34.4% dropout in Yeung et al. [38]; 40% dropout in Peter et al. [39]). The majority of the sample (58.3%; n = 28) had a university degree, further 14.6% (n = 7) had graduated from high school and 10.4% (n = 5) had gone to secondary school. The remaining 16.7% had undergone educational training (8.3%, n = 4), went to vocational school (6.3%, n = 3) or primary school (2.1%, n = 1). Although we did not explicitly define sleep problems as inclusion criterion for this first efficacy study, in more than 80% of participants we found a PSQI score of above 5 at the beginning of the study, indicating bad sleep quality. The study had been approved by the ethical committee of the University of Salzburg (GZ 46/2020) and was conducted in accordance with the Declaration of Helsinki. All participants had given informed consent prior to enrolment in the study protocol.

Study design and procedure

In the study we assessed the efficacy of the VSL-program (Nukkuaa^TM) during a training phase of 6 weeks, following upon a 2-week waitlist baseline (cf. Figure 1). At the beginning of the baseline (T0), at the start (T1) and end (T2) of the 6-week training phase, as well as at the follow-up (T3) 4 weeks after the end of the training, several online questionnaires were completed. The questionnaires consisted of the PSQI, the ISI, the Brief Symptom Inventory [BSI; 40] and the WHO Quality of Life Questionnaire [WHOQOL-BREF; 41]. Additionally, three nights with ambulatory polysomnography (PSG) were conducted (T0-T2) to investigate and validate changes in objectively assessed sleep parameters. During the training phase, participants worked on the content of the VSL-program and used an ECG sensor (Polar® H10, Polar Electro; Kempele, Finnland) daily to continuously monitor their sleep. Furthermore, a short sleep diary was filled in every morning to assess subjective changes in sleep quality and sleep parameters like TST, SOL or nocturnal awakenings. After the training phase, participants did not need to continue the VSL-program but were free to so until the end of the follow-up period (T3). The approximate further usage of the program was assessed during the follow-up questionnaire, resulting in the following groups: no usage (n = 17), rare usage (1-2 days a week; n = 11), frequent usage (3-4 days a week; n = 5), regular usage (5-7 days a week; n = 15).

Figure 1. Study design. Effects of the ‘Virtual Sleep Lab’ app-program (NukkuaaTM) on sleep were assessed subjectively via questionnaires at baseline, pre- and post-training and follow-up (T0-T3) as well as objectively via ambulatory polysomnography at baseline, pre- and post-training (T0-T2). Furthermore, a sleep diary was completed daily (T0-T2), and an ECG chest belt (Polar® H10) was used for continuous objective sleep monitoring from pre-training (T1) to post-training (T2). The survey symbol illustrates additional questionnaire evaluations during the study (PSQI, ISI, Brief Symptom Inventory (BSI) and the WHO Quality of Life Questionnaire (WHOQOL-BREF).

Sleep Training

We evaluated the app-based sleep training program VSL (Nukkuaa^TM) aiming at easing sleep problems such as insomnia. The training program is built up in consecutive levels, which contain short daily exercises (of max. 10-15 minutes each) based on the core elements of CBT-I, such as psychoeducation, relaxation exercises, sleep restriction, stimulus control and cognitive training. The main program consisted of 6 levels, which could be completed during the 6-week training phase when using the app daily. To complete a level and advance to the next one, certain conditions must be met: i) completing each exercise at least once, ii) filling in the sleep diary at least 5 times and iii) administering audio relaxation techniques at least 7 times. The program is structured in a way that an engagement with the app for around 10 to 15 minutes per day is sufficient to make progress in order to make daily usage and integration into everyday life easier for the users. Importantly, the app combines sleep training with the possibility of daily sleep-monitoring, both subjective and objective, by use of a sleep diary as well as an ECG sensor, which is worn at night during sleep. The combination of sleep training and both subjective and practically daily objective sleep measurements has to our knowledge not been done before, making the present concept interesting to be evaluated. The algorithm used in the app to enable reliable 4-class sleep-staging using the ECG sensor has been validated against PSG and published accordingly [42].

Material and measures

Subjective sleep assessment

Insomnia Severity

For the assessment of the perceived severity of insomnia symptoms the ISI was used at four different time points (T0-T3). Seven items are answered on a 5-point Likert scale. A total sum score ranging from 0 to 28 provides information about the degree of impairment through insomnia symptoms, while higher scores indicate a higher severity of insomnia symptoms. A score from 0 to 7 is considered not clinically significant, 8 to 14 as sub-threshold insomnia, 15 to 21 as moderate insomnia and a score above 21 is classified as severe clinical insomnia.

Sleep quality

To measure subjective sleep quality the PSQI was used at four different time points (T0-T3). It consists of 18 items which result in a global score between 0 and 21, with higher scores corresponding to a worse sleep quality. An empirical cut-off score at 5 divides between good and bad sleepers, while a score > 5 is considered as ‘bad’ sleep; a score >10 is rated as ‘chronic sleep disorder’.

Psychological strain and quality of life

Furthermore, we assessed psychological strain and quality of life for a measure of the participants current life situation. For assessing psychological strain including symptoms of depression and anxiety the Brief Symptom Inventory [BSI; 40] was used. We investigated the global score as a measure of general psychological symptom severity and the subscales “depression” and “anxiety” as a measure of depression and anxiety respectively.

Quality of life was measured with the short version of the WHO Quality of Life Questionnaire [WHOQOL-BREF; 41] which differentiates into four domains: physical health, psychological health, social relationships and environment. We analyzed and report transformed scores ranging from 0 to 100, higher values indicating higher quality of life.

Statistical analyses

Statistical analyses were performed using SPSS 29 (SPSS Inc., Chicago, Illinois) and R.Studio, Version 2023.03.0 (RStudio PBC, Boston, Massachusetts). Analyses were mainly based on non-parametric Friedman analyses, as of the ordinal scale of questionnaire data. The significance level was set to p <.05, while p-values <.10 are reported as statistical trends. For a measure of effect size the correlation coefficient (r) is provided. For post-hoc comparisons, Wilcoxon-rank-sum-tests were applied. Two-tailed critical p-values adjusted for multiple comparisons using the Bonferroni correction, as implemented in R’s ‘p.adjust’ function, are reported. For descriptive values of the outcome variables divided by assessment point refer to table 1.

Subjective parameters

Insomnia Severity (ISI)

A non-parametric Friedman-test with the independent variable ‘Insomnia Severity Index’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in insomnia severity depending on the time point, χ²(3) = 49.44, p <.001.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 2A). As expected, there was no significant change from baseline to pre-training (Z = -0.809, p = 1.00). There was however a significant change from pre- to post-training (Z = -4.670, p <.001, r = .67) and the results remained stable from post-training to the follow-up assessment one month later (Z = -1.552, p = .362).

Sleep quality (PSQI)

A non-parametric Friedman-test with the independent variable ‘PSQI’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in sleep quality depending on the time point, χ²(3) = 46.21, p <.001.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 2B). As expected, there was no significant change from baseline to pre-training, but a statistical trend towards an improvement (Z = -2.274, p = .069). Furthermore, there was a significant change from pre- to post-training (Z = -3.852, p <.001, r = .56) and there was a trend for a further improvement from post-training to the follow-up assessment one month later (Z = -2.141, p = .097, r = .31).

Figure 2. Improvement in (A) the severity of insomnia symptoms assessed by the Insomnia Severity Index and (B) the subjective sleep quality assessed by the Pittsburgh Sleep Quality Index. (A) Insomnia severity improved significantly from pre-training (T1) to post-training (T2) and remained stable until one-month follow-up (T3); N = 48. (B) Sleep quality improved tendentially from baseline (T0) to pre-training and significantly from pre-training (T1) to post-training (T2) and was by trend lower in the one-month follow-up (T3); N = 48. Higher values represent a stronger impairment in the respective measure. Horizontal lines represent the medians, boxes the interquartile range, with whiskers depicting the 1.5 interquartile range. The black cross corresponds to the mean. Asterisks indicate significance: ***p < .001, +p < .10.

Psychological strain (BSI - GSI values)

A non-parametric Friedman-test with the independent variable ‘GSI’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in the participant’s psychological strain depending on the time point, χ²(3) = 32.12, p <.001.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 3A). As expected, there was no significant change from baseline to pre-training (Z = -1.421, p = .466). There was, however, a significant change from pre- to post-training (Z = -4.688, p <.001, r = .68) and results remained stable from post-training to the follow-up assessment one month later (Z = -0.641, p = 1.00).

Subscale Depression (BSI)

A non-parametric Friedman-test with the independent variable ‘depression’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in the domain depression depending on the time point, χ²(3) = 12.01, p = .007.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 3B). As expected, there was no significant change from baseline to pre-training (Z = -0.172, p = 1.00). There was, however, a significant change from pre- to post-training (Z = -3.435, p = .002, r = .50) and results remained stable from post-training to the follow-up assessment one month later (Z = -0.669, p = 1.00).

Subscale Anxiety (BSI)

A non-parametric Friedman-test with the independent variable ‘anxiety’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in the domain anxiety depending on the time point, χ²(3) = 34.70, p <.001.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 3C). As expected, there was no significant change from baseline to pre-training (Z = -0.887, p = 1.00). There was, however, a significant change from pre- to post-training (Z = -4.130, p <.001, r = .60) and results remained stable from post-training to the follow-up assessment one month later (Z = -0.563, p = 1.00).

Figure 3. Improvement in (A) psychological strain measured by the Global Severity Index of the Brief Symptom Inventory (BSI), (B) the BSI-subscale depression and (C) the BSI-subscale anxiety. (A) Psychological strain, (B) symptoms of depression and (C) symptoms of anxiety improved significantly from pre-training (T1) to post-training (T2) and remained stable until one-month follow-up (T3); N = 48. Higher values represent a stronger impairment in the respective measure. Horizontal lines represent the medians, boxes the interquartile range, with whiskers depicting the 1.5 interquartile range. The black cross corresponds to the mean. Asterisks indicate significance: ***p < .001, **p < .010.

Quality of life (WHOQOL-BREF)

Domain Physical health

A non-parametric Friedman-test with the independent variable ‘physical health’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in the domain physical health depending on the time point, χ²(3) = 18.96, p <.001.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 4A). There were no significant changes from baseline to pre-training (Z = -0.657, p = 1.00), nor from pre- to post-training (Z = -0.206, p = 1.00), but a significant change from post-training to the follow-up assessment one month later (Z = -2.829, p = .014, r = 0.41).

Domain Psychological health

A non-parametric Friedman-test with the independent variable ‘psychological health’ and the repeated measures factor ‘Time’ (T0, T1, T2, T3) revealed a statistically significant change in the domain psychological health depending on the time point, χ²(3) = 14.96, p =.002.

Post hoc analyses with Wilcoxon signed-rank tests were conducted and a Bonferroni correction was applied (cf. Figure 4B). There were no significant changes from baseline to pre-training (Z = -0.374, p = 1.00), nor from pre- to post-training (Z = -0.909, p = 1.00), but a significant change from post-training to the follow-up assessment one month later (Z = -2.405, p = .049, r = .35).

Figure 4. Improvement in quality of life assessed by the components (A) physical health and (B) psychological health measured by the WHO Quality of life questionnaire (WHOQOL-BREF). (A) Physical health and (B) psychological health improved significantly from post-training (T2) to follow-up (T3); N = 48. Lower values represent a stronger impairment in the respective measure. Horizontal lines represent the medians, boxes the interquartile range, with whiskers depicting the 1.5 interquartile range. The black cross corresponds to the mean. Asterisks indicate significance: *p < .050.

[Table 1]

Usage of the program within the follow-up-period

Further we analyzed if different frequencies in the usage of the app-program within the follow-up period resulted in different outcomes in terms of changes in the various questionnaires from T0-T3, T1-T3 and T2-T3. Hereby, we used non-parametric Kruskal-Wallis-tests to investigate the differences between the 4 groups (no usage, rare usage, frequent usage, regular usage).

No significant differences between groups were found in PSQI changes T0-T3 (H(3) = 2.29, p = .515), T1-T3 ( (3) = 5.83, p = .120) and T2-T3 (H(3) = 2.97, p = .396), in ISI changes T0-T3 (H(3) = 0.73, p = .867), T1-T3 (H(3) = 1.48, p = .688) and T2-T3 (H(3) = 6.02, p = .111), in BSI global score changes T0-T3 (H(3) = 1.27, p = .736), T1-T3 (H(3) = 1.54, p = .673), T2-T3 (H(3) = 1.18, p = .758), in BSI subscale depression T0-T3 (H(3) = 1.72, p = .632), T1-T3 (H(3) = 4.67, p = .197), T2-T3 (H(3) = 3.70, p = .296), in BSI subscale anxiety T0-T3 (H(3) = 1.05, p = .788), T1-T3 (H(3) = 1.52, p = .678), T2-T3 (H(3) = 1.17, p = .759), in WHOQOL-BREF domain “physical health” T0-T3 (H(3) = 0.90, p = .824), T1-T3 (H(3) = 0.34, p = .953), T2-T3 (H(3) = 1.00, p = .803), and in WHOQOL-BREF domain “psychological health” T0-T3 (H(3) = 1.23, p = .746), T1-T3 (H(3) = 1.29, p = .733). For changes in WHOQOL-BREF domain “psychological health" a statistical trend was observed for T2-T3 (H(3) = 7.47, p = .058). For an overview on descriptive statistic values please refer to supplementary tables S1-S7.

Factors predicting improvements during the training phase

A multiple linear regression model was conducted to assess potential predictors for the improvement of insomnia symptoms during the app training phase, defined by the difference of the ISI scores from T1 to T2. The following predictors were included in a forced entry model: Sex (male vs. female), age, education (higher, i.e., graduated from high school or university degree, vs. lower), last completed level at T2, PSQI at T1 (low, i.e., 0-5 points vs. high, i.e., 6 or more points), ISI at T1 (low, i.e., 0-7 points, vs. high, i.e., 8 or more points). The overall regression was statistically significant (R² = 0.31, F (6, 41) = 3.04, p = .015), showing that the predictors accounted for 31% of the variability in the improvement in ISI scores (cf. table 2).

[Table 2]

This paper focuses on examining the efficacy of the VSL-program regarding changes in subjective sleep quality, insomnia severity, psychological strain and quality of life. Analysis of changes in the sleep-related questionnaires PSQI and ISI showed significant improvements in subjective sleep quality as well as in insomnia severity, including symptoms such as difficulties falling asleep, maintaining sleep and early awakenings. These improvements occurred between the expected measurement time points, that is from pre- to the post-training 6 weeks later, and sometimes with even further improvements to the follow-up (4 weeks later). This indicates that subjective improvements in both clinical questionnaires can be observed after 6-week sleep training using the VSL-program (Nukkuaa^TM).

Specifically, ISI-scores improved from a mean score of 14.21 (Mdn = 14.00) at baseline to a mean score of 9.90 (Mdn = 9.00) at follow-up. Both scores are assigned to the category "low-threshold insomnia" according to the manual, which is defined by a score range of 8-14 [24].Thus, despite a statistical improvement, there was on average no clinical improvement to the aspired category of "non-significant insomnia," which is categorized by a score range of 0 to 7 points in the ISI [24]. While at start of the baseline (i.e., T0) only 6.3% of the participants fell into the category of a “non-significant insomnia”, by the end of the training phase (i.e., T2) already 20.8% did, as well as 35.4% by one-month follow-up (i.e., T3). The number thus tripled by the end of the training phase and was even 6-times higher by the follow-up-assessment.

Similarly, PSQI-scores improved from a mean score of 8.98 (Mdn = 9.50) at baseline to a mean score of 6.27 (Mdn = 5.50) at follow-up. Both scores fall into the category of "poor sleep quality," which is defined by a score range of 6 to 10. Values of 5 or less would correspond to good sleep quality [43]. Thus, again, there was on average no clinical improvement in terms of changes in the sleep quality category of the PSQI. While at the beginning of the baseline (i.e., T0) only 16.7% had been defined as a “good sleeper” according to the PSQI global score, after the training phase (i.e., T2) 35.4% of the participants were scoring ≤ 5 on the PSQI global score, as well as 50% at the one-month follow-up (i.e., T3) and can therefore be defined as “good sleepers” after administration of the VSL-program. The number of "good sleepers" thus doubled by the end of the program and even tripled by the follow-up-assessment. It is relevant to mention that in this first efficacy study, no clinical sample was recruited, but we simply recruited from the general population using public announcements in Austrian TV and radio. This fact is also reflected in the mean values of the sample in the ISI described in the last paragraph, which were also only in the range of “subthreshold insomnia” at the beginning of the study (Mdn = 14.00). It is believed that stronger improvements would have been shown if the sample had had higher insomnia symptoms on average by beginning of the study.

Another point to consider is the program progress during the training phase of 6 weeks: The objective was to complete the first 6 levels of the VSL-program, which is considered the core program of the training. In fact, only 12.5% of the sample completed these 6 levels, while another 31.3% completed the program including the 5th level. That means that only 21 individuals and thus less than half of the final sample completed at least most of the core program (including level 5). In summary, it is expected that participants on average would have shown greater improvements if more individuals had completed the core VSL-program. Additional analyses support this assumption: it was shown that the further the progress in the program, the greater the improvement in terms of alleviation of insomnia symptoms during the training phase. These results suggest greater effectiveness of the training program with consistent use of the app (cf. table 2). The findings are in line with Zachariae and colleagues [32], who showed in their meta-analysis that longer treatment duration was associated with greater effect sizes.

Furthermore, there were also improvements in the other questionnaires assessed in this study. Specifically, there were improvements in general psychological symptom severity, symptoms of depression and anxiety, and an increase in quality of life in terms of perceived physical and psychological health over the course of the study.

Regarding the questionnaire data, it is also interesting to note that the different frequencies of continued usage of the app-program between the end of the training phase and the follow-up did not have an influence on the improvement. Specifically, this means that people who continued to use the program did not show greater improvements at follow-up than people who did not. It was believed that individuals who continued to use the VSL-app (frequently or regularly) would benefit more than individuals who stopped using the program after the training phase. The results suggest that the main efficacy and thus impact on subjective parameters was already achieved in the first 6 weeks or during the training phase, no matter the time spent in the app-program during the follow-up period.

With regard to the limitations of the study, some essential points should be noted: In this first efficacy study, no control group was implemented, yet a non-training baseline and follow-up period. Therefore, it cannot be investigated in more detail whether the effects that were shown in the sample regarding an improvement in subjective sleep quality as well as an alleviation of insomnia severity are actually intervention-specific effects, or whether these also could be placebo effects. Upcoming studies need to evaluate the program against a waiting list control group as well as an active control group. As the literature shows, and what we also observe as trend in our data subjective improvements sometimes can occur by minimal intervention [44] or even mere inclusion in a study protocol, even though no training or therapy has yet been started (Perlis et al., 2005). Although such improvements are good for the patient, they cannot be considered treatment-specific and accordingly need to be clearly named as such.

As this was the first efficacy study of the VSL-program (Nukkuaa^TM), some technical issues were sometimes still present. For example, the type of feedback of NOA and WASO after waking up was changed in the app during the course of the study. Also, the daily relaxation exercises for program advancements were not counted in certain circumstances, which for some subjects delayed the progression to the next sleep training “level”. These disruptive factors may have prevented subjects from fully engaging and may have led to less trust in the digital program.

With a sample size of 48 we tried to allow first reliable inferences about the efficacy of the VSL-program and over the course of the 12-week study period. Yet for replication and better generalizability of the results it is highly needed to add further efficacy studies with larger samples, and of varying degrees of sleep problems. A fact up to date completely unknown is also whether concomitant face-to-face sleep coaching would be able to boost the positive effects observed.

In summary and considering the points raised, it can be stated that the VSL-program, with its concept of combining CBT-I-oriented sleep training together with continuous objective sleep monitoring, is effective in improving subjective sleep parameters, components of quality of life and comorbid symptoms such as depression and anxiety. These effects are seen after 6 weeks of regular usage of the VSL-program and occur even without personal guidance.

Data availability statement

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

Pavlos Topalidis was supported by the Doctoral College “Imaging the Mind” (FWF, Austrian Science Fund: W 1233-B). In addition, part of the study was supported by an innovation grant of the county of Salzburg (20102-F2002176-FPR).

Author contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analyses, and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Competing interests

E-SE, AH and PT declare no potential conflict of interest. MS is Co-Founder and CSO of Nukkuaa^TM.

van Cauter, E., et al., Reciprocal interactions between the GH axis and sleep. Growth Horm IGF Res. 14 Suppl A, S10-7 (2004) https://doi.org/10.1016/j.ghir.2004.03.006.
Besedovsky, L., T. Lange, and M. Haack, The Sleep-Immune Crosstalk in Health and Disease. Physiol Rev. 99(3), 1325-1380 (2019) https://doi.org/10.1152/physrev.00010.2018.
Walker, M.P. and E. van der Helm, Overnight therapy? The role of sleep in emotional brain processing. Psychol Bull. 135(5), 731-48 (2009) https://doi.org/10.1037/a0016570.
Blume, C., et al., "How does Austria sleep?" self-reported sleep habits and complaints in an online survey. Sleep Breath. 24(2), 735-741 (2020) https://doi.org/10.1007/s11325-019-01982-5.
Hirshkowitz, M., et al., National Sleep Foundation's updated sleep duration recommendations: final report. Sleep Health. 1(4), 233-243 (2015) https://doi.org/10.1016/j.sleh.2015.10.004.
Zeitlhofer, J., et al., Sleep habits and sleep complaints in Austria: current self-reported data on sleep behaviour, sleep disturbances and their treatment. Acta Neurol Scand. 122(6), 398-403 (2010) https://doi.org/10.1111/j.1600-0404.2010.01325.x.
Kronholm, E., et al., Trends in self-reported sleep duration and insomnia-related symptoms in Finland from 1972 to 2005: a comparative review and re-analysis of Finnish population samples. J Sleep Res. 17(1), 54-62 (2008) https://doi.org/10.1111/j.1365-2869.2008.00627.x.
Marschall, J., et al., Gesundheitsreport 2017. Beiträge zur Gesundheitsökonomie und Versorgungsforschung., A. Storm, Editor. 2017: Hamburg.
Ohayon, M.M., Prevalence of DSM-IV diagnostic criteria of insomnia: distinguishing insomnia related to mental disorders from sleep disorders. J Psychiatr Res. 31(3), 333-46 (1997) https://doi.org/10.1016/s0022-3956(97)00002-2.
Wittchen, H.U., et al., The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol. 21(9), 655-79 (2011) https://doi.org/10.1016/j.euroneuro.2011.07.018.
Ohayon, M.M., Epidemiological Overview of sleep Disorders in the General Population. Sleep Med Res. 2(1), 1-9 (2011) https://doi.org/10.17241/smr.2011.2.1.1.
Roth, T., Insomnia: definition, prevalence, etiology, and consequences. J Clin Sleep Med. 3(5 Suppl), S7-10 (2007).
Buysse, D.J., et al., The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 28(2), 193-213 (1989) https://doi.org/10.1016/0165-1781(89)90047-4.
Sofi, F., et al., Insomnia and risk of cardiovascular disease: a meta-analysis. Eur J Prev Cardiol. 21(1), 57-64 (2014) https://doi.org/10.1177/2047487312460020.
Baglioni, C., et al., Insomnia as a predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J Affect Disord. 135(1-3), 10-9 (2011) https://doi.org/10.1016/j.jad.2011.01.011.
Hertenstein, E., et al., Insomnia as a predictor of mental disorders: A systematic review and meta-analysis. Sleep Med Rev. 43, 96-105 (2019) https://doi.org/10.1016/j.smrv.2018.10.006.
Cappuccio, F.P., et al., Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care. 33(2), 414-20 (2010) https://doi.org/10.2337/dc09-1124.
Hertenstein, E., et al., Cognitive behavioral therapy for insomnia in patients with mental disorders and comorbid insomnia: A systematic review and meta-analysis. Sleep Med Rev. 62, 101597 (2022) https://doi.org/10.1016/j.smrv.2022.101597.
Daley, M., et al., The economic burden of insomnia: direct and indirect costs for individuals with insomnia syndrome, insomnia symptoms, and good sleepers. Sleep. 32(1), 55-64 (2009).
Kessler, R.C., et al., Insomnia and the performance of US workers: results from the America insomnia survey. Sleep. 34(9), 1161-71 (2011) https://doi.org/10.5665/SLEEP.1230.
Thiart, H., et al., Internet-Based Cognitive Behavioral Therapy for Insomnia: A Health Economic Evaluation. Sleep. 39(10), 1769-1778 (2016) https://doi.org/10.5665/sleep.6152.
Morin, C.M., et al., Endorsement of European guideline for the diagnosis and treatment of insomnia by the World Sleep Society. Sleep Med. 81, 124-126 (2021) https://doi.org/10.1016/j.sleep.2021.01.023.
Riemann, D., et al., European guideline for the diagnosis and treatment of insomnia. J Sleep Res. 26(6), 675-700 (2017) https://doi.org/10.1111/jsr.12594.
Dieck, A., C.M. Morin, and J. Backhaus, A German version of the Insomnia Severity Index. Somnologie. 22(1), 27-35 (2018) https://doi.org/10.1007/s11818-017-0147-z.
Okajima, I., Y. Komada, and Y. Inoue, A meta‐analysis on the treatment effectiveness of cognitive behavioral therapy for primary insomnia. Sleep and Biological Rhythms. 9(1), 24-34 (2011) https://doi.org/10.1111/j.1479-8425.2010.00481.x.
van Straten, A., et al., Cognitive and behavioral therapies in the treatment of insomnia: A meta-analysis. Sleep Med Rev. 38, 3-16 (2018) https://doi.org/10.1016/j.smrv.2017.02.001.
Baglioni, C., et al., Sleep changes in the disorder of insomnia: a meta-analysis of polysomnographic studies. Sleep Med Rev. 18(3), 195-213 (2014) https://doi.org/10.1016/j.smrv.2013.04.001.
Dopheide, J.A., Insomnia overview: epidemiology, pathophysiology, diagnosis and monitoring, and nonpharmacologic therapy. Am J Manag Care. 26(4 Suppl), S76-S84 (2020) https://doi.org/10.37765/ajmc.2020.42769.
Blom, K., et al., Internet-vs. group-delivered cognitive behavior therapy for insomnia: A randomized controlled non-inferiority trial. Behav Res Ther. 70, 47-55 (2015) https://doi.org/10.1016/j.brat.2015.05.002.
Lancee, J., et al., Guided Online or Face-to-Face Cognitive Behavioral Treatment for Insomnia: A Randomized Wait-List Controlled Trial. Sleep. 39(1), 183-91 (2016) https://doi.org/10.5665/sleep.5344.
Soh, H.L., et al., Efficacy of digital cognitive behavioural therapy for insomnia: a meta-analysis of randomised controlled trials. Sleep Med. 75, 315-325 (2020) https://doi.org/10.1016/j.sleep.2020.08.020.
Zachariae, R., et al., Efficacy of internet-delivered cognitive-behavioral therapy for insomnia - A systematic review and meta-analysis of randomized controlled trials. Sleep Med Rev. 30, 1-10 (2016) https://doi.org/10.1016/j.smrv.2015.10.004.
Kaplan, K.A., et al., When a gold standard isn't so golden: Lack of prediction of subjective sleep quality from sleep polysomnography. Biol Psychol. 123, 37-46 (2017) https://doi.org/10.1016/j.biopsycho.2016.11.010.
Ma, Y., et al., Profile of subjective-objective sleep discrepancy in patients with insomnia and sleep apnea. J Clin Sleep Med. 17(11), 2155-2163 (2021) https://doi.org/10.5664/jcsm.9348.
Harvey, A.G. and N.K. Tang, (Mis)perception of sleep in insomnia: a puzzle and a resolution. Psychol Bull. 138(1), 77-101 (2012) https://doi.org/10.1037/a0025730.
Crönlein, T., et al., Changes in Subjective-Objective Sleep Discrepancy Following Inpatient Cognitive Behavior Therapy for Insomnia. Behav Ther. 50(5), 994-1001 (2019) https://doi.org/10.1016/j.beth.2019.03.002.
Irwin, M.R., Why sleep is important for health: a psychoneuroimmunology perspective. Annu Rev Psychol. 66, 143-72 (2015) https://doi.org/10.1146/annurev-psych-010213-115205.
Yeung, W.F., et al., Predictors of dropout from internet-based self-help cognitive behavioral therapy for insomnia. Behav Res Ther. 73, 19-24 (2015) https://doi.org/10.1016/j.brat.2015.07.008.
Peter, L., et al., Effectiveness of an Online CBT-I Intervention and a Face-to-Face Treatment for Shift Work Sleep Disorder: A Comparison of Sleep Diary Data. Int J Environ Res Public Health. 16(17) (2019) https://doi.org/10.3390/ijerph16173081.
Franke, G., Franke, G.H. (2000). BSI. Brief Symptom Inventory - Deutsche Version. Manual. Göttingen: Beltz. 2000.
World Health Organization, Development of the World Health Organization WHOQOL-BREF quality of life assessment. The WHOQOL Group. Psychol Med. 28(3), 551-8 (1998) https://doi.org/10.1017/s0033291798006667.
Topalidis, P., et al., The Virtual Sleep Lab-A Novel Method for Accurate Four-Class Sleep Staging Using Heart-Rate Variability from Low-Cost Wearables. Sensors (Basel). 23(5) (2023) https://doi.org/10.3390/s23052390.
Riemann, D. and J. Backhaus, Behandlung von Schlafstörungen. Ein psychologisches Gruppenprogramm. 1996: Beltz Psychologie Verlags Union.
Eigl, E.S., L.K. Urban-Ferreira, and M. Schabus, A low-threshold sleep intervention for improving sleep quality and well-being. Front Psychiatry. 14, 1117645 (2023) https://doi.org/10.3389/fpsyt.2023.1117645.

Table 1. Mean (M), median (Mdn) and standard deviation (SD) for the outcome measures: Pittsburg Sleep Quality Index (PSQI), Insomnia Severity Index (ISI), Brief Symptom Inventory (BSI) and WHO Quality of Life Questionnaire (WHOQOL-BREF) at baseline (T0), pre-training (T1), post-training (T2) and one-month follow-up (T3).

Outcome measures
Time	ISI				PSQI				BSI Global Severity Index				BSI Subscale Depression				BSI Subscale Anxiety				WHOQOL-BREF Domain Physical health				WHOQOL-BREF Domain Psychological health
	M	Mdn	SD	M		Mdn	SD	M		Mdn	SD	M		Mdn	SD	M		Mdn	SD	M		Mdn	SD	M		Mdn	SD
T0	14.21	14.00	5.10	8.98		9.50	3.46	0.51		0.39	0.40	0.58		0.33	0.59	0.52		0.50	0.46	69.48		69.00	15.17	64.56		69.00	14.34
T1	13.85	13.00	4.65	8.29		8.00	2.88	0.53		0.43	0.38	0.570		0.50	0.59	0.53		0.50	0.43	70.04		69.00	15.76	64.79		69.00	16.07
T2	10.67	10.00	4.83	6.79		6.00	2.85	0.37		0.33	0.30	0.40		0.33	0.40	0.32		0.17	0.36	71.58		75.00	16.08	66.77		69.00	15.13
T3	9.90	9.00	4.84	6.27		5.50	3.15	0.35		0.26	0.31	0.38		0.25	0.43	0.29		0.17	0.29	75.46		75.00	15.00	69.73		69.00	15.17

Notes: N = 48

Table 2. Statistical overview over the coefficients included in the multiple regression model to assess possible predictors in regard to improvements in Insomnia severity (ISI) from pre-training (T1) to post-training (T2).

Dependent variable: ISI Change T1-T2
						95% KI
Coefficients	b	SE	β	T	p	UG	OG
(Constant)	2.806	3.826		0.733	0.468	-4.921	10.533
Sex	-0.571	1.127	-0.071	-0.506	0.615	-2.848	1.706
Age	-0.096	0.046	-0.282	-2.09	0.043	-0.189	-0.003
Education (binominal)	-1.601	1.223	-0.178	-1.310	0.198	-4.070	0.868
Last complete level at T2	0.751	0.315	0.330	2.381	0.022	0.114	1.388
PSQI T1 (binominal)	0.854	1.880	0.083	0.454	0.652	-2.942	-2.942
ISI T1 (binominal)	4.189	2.567	0.289	1.632	0.110	-0.996	-0.996
Notes: N = 48, R² = 0.308, corr. R² = 0.207, F (6, 41) = 3.04, p = .015

Competing interest reported. E-SE, AH and PT declare no potential competing interests. MS is Co-Founder and CSO of Nukkuaa TM.

VSLFS2SupplementaryMaterial300723.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

The effects of a CBT-I based app-program on sleep quality, insomnia severity, psychological strain and quality of life: A first efficacy study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Participants

Study design and procedure

Sleep Training

Material and measures

Subjective sleep assessment

Insomnia Severity

Sleep quality

Results

Statistical analyses

Subjective parameters

Insomnia Severity (ISI)

Sleep quality (PSQI)

Psychological strain (BSI - GSI values)

Subscale Depression (BSI)

Subscale Anxiety (BSI)

Quality of life (WHOQOL-BREF)

Discussion

Conclusion and limitations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1