Adherence and Effects of a Short-term Digital Cognitive Behavioral Therapy for Insomnia Based on the Health Belief Model: A Randomized Controlled Trial

In this study, we aim to evaluate the intervention effects of a short-term digital cognitive behavioral therapy for insomnia (dCBTI) and to examine whether enhancing health beliefs can improve intervention adherence, which may, in turn, affect intervention outcomes. A randomized controlled trial was conducted in this study, a total of 153 participants randomly assigned to three groups: an enhanced health belief group, a regular group, and a treatment-waiting group. The enhanced and regular groups received a two-week dCBTI intervention, while the enhanced group also received additional intervention to enhance health belief. The Insomnia Severity Index (ISI), the dCBTI adherence assessment questionnaire, and the Health Belief Model (HBM) questionnaire were administered to all participants for evaluation at baseline and post-intervention. Results showed that: Firstly, the ISI change for the enhanced group was 11.028 (d = 35, p < 0.001), with Cohen’s d effect size of 4.469. As for the regular group, the ISI change was 9.829 (d = 34, p < 0.001), with Cohen’s d effect size of 4.706. As for the waiting group, the ISI change was -0.356 (p = 0.160 > 0.05). Secondly, the difference in HBM at the mid-test between the enhanced and regular groups was 5.521 (p = 0.017 < 0.05); at the post-test, the difference was 5.245 (p = 0.022 < 0.05). Thirdly, the overall adherence for the enhanced group was 5.623, compared to 5.133 for the regular group, with a difference of 0.490 (p = 0.004 < 0.01). Lastly, mediation analysis revealed an indirect effect estimate of 0.957, with a regular error of 0.06 and a 95% confidence interval of [0.022, 0.252], excluding 0. Thus, it can be concluded that: Firstly, both the enhanced and regular groups showed significant improvements in ISI, while the waiting group showed no significant change. The enhanced group demonstrated a greater improvement range in ISI compared to the regular group, although the effect size for the regular group was slightly higher. Secondly, the enhanced group showed significantly better HBM in the mid to late stages of the intervention. Thirdly, the enhanced group had significantly better overall adherence compared to the regular group. Lastly, adherence acted as a mediator of the grouping effect on ISI.