Mental Health
Conditions
Keywords
Cognitive Behavioral Therapy for Insomnia, insomnia, large language model, chatbot, university student
Brief summary
This study aims to evaluate the feasibility, acceptability, and potential effectiveness of a nurse-led digital cognitive behavioral therapy for insomnia (dCBT-I) via large language model based-chatbot among university students with mental health problems.
Detailed description
This is a two-arm, parallel pilot randomized controlled trial recruiting 100 university students in Hong Kong with mild to moderate symptoms of depression or anxiety who also screen positive for insomnia symptoms. All participants will receive brief sleep health advice and a booklet providing behavioral guidance on lifestyle and environmental factors related to sleep and insomnia. Participants in the intervention group will additionally receive 12 weeks of chatbot-delivered insomnia management intervention, including 2 weeks of daily sleep diary monitoring and 10 weeks of nurse-led instant messaging support grounded in cognitive behavioral therapy for insomnia (CBT-I). The primary outcome is self-reported insomnia severity at 6-month follow-up. Secondary outcomes include health-related quality of life, sleep related quality of life, sleep effort, pre-sleep arousal, anxiety and depression assessed at 3 and 6 months. Analyses will be conducted on intention-to-treat basis.
Interventions
BEHAVIORALBrief sleep health advice
At baseline, both groups will receive a 1-on-1 brief sleep health advice developed using guidelines by U.S. Centers for Diseases Control and Prevention and a booklet providing behavioral guidance on lifestyle and environmental factors related to sleep and insomnia.
BEHAVIORAL2 weeks of chatbot-based daily sleep diary
Following randomization, participants in the intervention group will complete a 2-week chatbot-based daily sleep diary. Each morning (e.g., 8:00 am), participants will be prompted via Whatsapp message delivered by chatbot to report key sleep parameters from the previous night, including bedtime, wake time, sleep onset latency, number and duration of nocturnal awakenings, total sleep time, and perceived sleep quality. The collected sleep diary data will be processed by the chatbot to generate tailored and personalized feedback, supporting individualized sleep recommendations and behavioral adjustments. At the initiation of the intervention, participants will also complete a brief baseline questionnaire to establish personalized treatment goals related to sleep improvement. Throughout the intervention period, participants will have continuous access to chatbot-delivered support, including adaptive guidance based on diary inputs, reinforcement of recommended sleep strategies, and progress
BEHAVIORAL10 weeks chatbot-based IM support
Participants in the intervention group will also receive cognitive behavioral therapy for insomnia (CBT-I) delivered via an LLM-based chatbot through WhatsApp over a 10-week period. The intervention is fully automated and available 24/7, with underlying algorithms enabling the delivery of information, therapeutic support, and advice in a personalized and adaptive manner, informed by participants' ongoing inputs and interaction patterns. The intervention will be structured into 20 chatbot-delivered instant messaging (2 pieces per week). The therapeutic content is adapted from established CBT-I manuals and led by research nurse, incorporating evidence-based behavioral, cognitive, and educational components. Behavioral strategies include sleep restriction, stimulus control, and relaxation techniques. Cognitive components comprise paradoxical intention, cognitive restructuring, mindfulness-based techniques, positive imagery, and strategies for "putting the day to rest." Educational compone
Sponsors
The University of Hong Kong
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
SINGLE (Outcomes Assessor)
Masking description
Blinding of the participants and treatment providers will not be possible because the behavior nature of the intervention, but the outcome assessors and statistical analysts will be blinded to group allocation.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Students aged ≥ 18 years from 22 local higher education institutions (University or Colleges) listed in Education Bureau. * Screening score indicating mild to moderate anxiety or depression, defined as a GAD-7 score of 5-14 and/or a PHQ-9 score of 5-14 at baseline. * Screened positive for insomnia symptoms using the Sleep Condition Indicator (SCI) and meeting DSM-5 diagnostic criteria for insomnia disorder. * Reporting a self-reported sleep efficiency of less than 85% over the past month. * Able to communicate in Chinese.
Exclusion criteria
* Participants who self-reported diagnosis of epilepsy, schizophrenia, or bipolar disorder * Participants who self-reported current suicidal ideation with intent or a suicide attempt within the past 2 months * Participants who self-reported current receipt of psychological or behavioral treatment for insomnia from a health professional * Participants who self-reported participation in an online or digital insomnia treatment programme at the time of enrolment.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Self-reported insomnia severity | 6-month follow-up | Assessed using the Insomnia Severity Index (ISI). The ISI is a 7-item self-report instrument that evaluates both nocturnal and daytime symptoms of insomnia, with total scores ranging from 0 to 28, higher scores indicate greater insomnia severity. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Health-related quality of life | 3- and 6-month follow-ups | Assessed using 36-Item Short Form Health Survey \[SF-36\]. The SF-36 is a 36-item self-reported instrument that measures both Physical Component Scores and Mental Component Scores, with scores ranging from 0 to 100; higher scores indicate better health-related quality of life. |
| Sleep-related quality of life | 3- and 6-month follow-ups | Assessed using Glasgow Sleep Impact Index \[GSII\]. The GSII is a 7-item self-report instrument that evaluates the sleep-related quality of life, with total scores ranging from 0 to 28, higher scores indicate lower level of sleep-related quality of life. |
| Depressive symptoms | 3- and 6-month follow-ups | Assessed using Patient Health Questionnaire-9 \[PHQ-9\]. The PHQ-9 is a 9-item self-report instrument that evaluates frequency and severity of depressive symptoms, with total scores ranging from 0 to 27, higher scores indicate higher level of depression. |
| Anxiety symptoms | 3- and 6-month follow-ups | Assessed using Generalized Anxiety Disorder-7 \[GAD-7\]. The GAD-7 is a 7-item self-report instrument that evaluates the frequency and severity of anxious symptoms, with total scores ranging from 0 to 21, higher scores indicate higher level of anxiety. |
| Sleep effort | 3- and 6-month follow-ups | Assessed using Glasgow Sleep Effort Scale \[GSES\]. The GSES is a 7-item self-reported instrument, with scores ranging from 0 to 14, higher scores indicate greater sleep effort. |
| Pre-sleep arousal | 3- and 6-month follow-ups | Assessed using Pre-Sleep Arousal Scale \[PSAS\]. The PSAS is a 16-item self-reported instrument consists of two subscales (cognitive arousal and somatic arousal), with scores ranging from 8 to 40 on each subscale, higher scores indicate greater pre-sleep arousal. |
Countries
Hong Kong
Contacts
CONTACTShengzhi Zhao
CONTACTTsz Kiu Chan
PRINCIPAL_INVESTIGATORShengzhi Zhao
The University of Hong Kong
Outcome results
None listed