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 sleep and subjective-objective sleep discrepancies (SOSD). Fifty-seven volunteers suffering from sleep problems were randomly assigned to an experimental group or a waitlist control group. 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 subjectively via questionnaires (InsomniaSeverity Index, ISI; Pittsburgh Sleep Quality Index, PSQI), objectively via ambulatory polysomnography (PSG), and continuously via heart-rate sensor and sleep diaries.
F I G U R E 4 Changes in objective, PSG-derived sleep parameters. (a) TST did not significantly change for the EG ( p > 0.961). However, TST significantly decreased from T0T2 ( p = 0.004) in the CG. (b) TIB did not significantly change for the EG ( p > 0.123). However, TIB significantly decreased from T0T2 ( p < 0.001) in the CG. (c) WASO tendentially decreased from T0T2 ( p = 0.061) in the EG. The CG showed no significant changes (p > 0.219). Horizontal lines represent the medians, boxes the interquartile range, with whiskers depicting the 1.5 interquartile range. The grey cross corresponds to the mean. CG, control group; EG, experimental group; TIB, time in bed; TST, total sleep time; T0, start of the study; T1, beginning of app phase; T2, end of app phase; WASO, wake after sleep onset. Asterisks indicate significance: ***p < 0.001; **p < 0.010; +p < 0.10. 3.3.3
id: 81475730982b50dd562d452f600283c3 - page: 9
Sleep onset latency | Analysis revealed a significant effect of TIME (F (1, 52) = 4.90, p = 0.031, 2 part = 0.09), yet no significant effect of GROUP (F (1, 52) = 0.95, p = 0.335, 2 part = 0.02), and no significant interaction effect of TIME*GROUP (F (1, 52) = 1.54, p = 0.220, 2 part = 0.03), for changes in SOSD of SOL. In contrast to the CG (n = 28), the EG (n = 26) exhibited significant improvements towards less overestimation of SOL from week 1 to 6. (cf. Figure 5 c). 3.3.4 | Wake after sleep onset part = 0.04), and no significant interaction effect of for p = 0.221, 2 TIME*GROUP (F (2, 80) = 0.20, p = 0.822, 2 part = 0.01) changes in PSG-derived WASO (see Table 3 for results and
id: cf4bf261d3d94607d3c02ff20871a029 - page: 9
Figure 4 c for visualisation). Analysis revealed a significant effect of TIME (F (1, 51) = 9.37, part = 0.16), yet no significant effect of GROUP (F (1, p = 0.004, 2 13652869, 0, Downloaded from by Turkey Cochrane Evidence Aid, Wiley Online Library on [02/01/2024]. See the Terms and Conditions ( on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 10 of 15 HINTERBERGER ET AL. T A B L E 4 Descriptive statistics and pairwise comparisons for changes in in discrepancy between subjective (sleep diary-derived) and objective (HR sensor-derived) sleep parameters during the app phase. Week 6 Week 1 Weeks 16 M SD M SD TST (cid:3)6.80 27.07 MDiff = 20.43, CI95% [3.75, 37.10], p = 0.017, d = 0.47* MDiff = 18.26, CI95% [1.58, 34.94], p = 0.032, d = 0.42* 13.63 56.93 EG (cid:3)8.22 30.16 CG 10.04 32.81
id: 3cdc63494ddbfa461b44f84520b9a136 - page: 9
SE MDiff = 5.57, CI95% [0.85, 10.29], p = 0.022, d = 0.46* MDiff = 1.63, CI95% [(cid:3)2.92, 6.18], p = 0.476, d = 0.14 (cid:3)1.49 5.46 4.08 12.86 EG 0.58 9.25 CG 2.21 8.28 SOL (cid:3)3.88 7.19 MDiff = 8.05, CI95% [1.32, 14.79], p = 0.020, d = 0.47* MDiff = 2.26, CI95% [(cid:3)4.23, 8.75], p = 0.487, d = 0.13 EG 4.18 17.30 CG 2.71 16.32 4.97 20.90 WASO MDiff = 21.98, CI95% [8.07, 35.89], p = 0.003, d = 0.61* MDiff = 8.31, CI95% [(cid:3)5.87, 22.48], p = 0.245, d = 0.22 (cid:3)29.75 28.27 (cid:3)7.78 40.97 EG (cid:3)20.38 17.23 (cid:3)12.08 19.11
id: c104ca5c94c8fe22b475e961a1f64d07 - page: 10