| Abstract | The delivery and effectiveness of Cognitive Behavioral Therapy for insomnia (CBT-I) could be enhanced using digital phenotyping (DP). DP involves collecting and analysing data typically from smartphones and wearables, to gain insights into an individual's health and behaviour. Real-time data gathered on patients' sleep patterns and co-factors (e.g. mental health), could be used to personalise and adjust CBT-I treatment plans, ensuring they are aligned with patient-specific needs, potentially leading to improved outcomes. By providing personalised feedback, DP could enhance patient engagement and motivation, increasing adherence. Objective assessments provided by DP could allow for evidence-based decision-making. In this context, we built a DP platform and performed a proof-of-concept study with adults (65+), living at home (20 with and 5 without insomnia), to investigate the feasibility of using DP in informing CBT-I within this population. At onboarding, neuropsychological tests and standard sleep questionnaires were administered. For 14 days, participants completed daily ecological momentary assessments and cognitive games through a mobile application. Data was passively collected using wearables. Our platform was used to compute digital phenotypes pertaining to mood, cognitive state, and daily activities. Results show >80% of expected data was recorded, indicating experimental protocol adherence. Statistically significant correlations were found between self-reported sleep efficiency and morning irritability (r=-0.179, p>0.05), and sadness (r=-0.266, p>0.001). Baseline insomnia severity index correlated with mean sleep efficiency (r=-0.475, p>0.05). In terms of digital phenotypes, daily step number correlated with self-reported anxiety (r=-0.17, p>0.01) and stress (r=-0.19, p>0.01); similarly for daily call duration (ranxiety=-0.33, panxiety>0.001, rstress=-0.23, pstress>0.001). Thus, our digital phenotyping framework could be used to inform CBT-I interventions; a pilot study using this approach is underway. |
|---|
