Abstract
BACKGROUND: Insomnia is a highly prevalent condition, with heterogeneous clinical presentations and underlying mechanisms. Traditional assessment methods often fail to capture this complexity, thereby hindering the development of personalized treatments. This paper details a protocol for a study that employs a "deep phenotyping" approach to comprehensively characterize insomnia. METHODS: This single-center prospective observational study recruited adults with insomnia and a parallel cohort of normal sleepers as the controls. Participants undergo a 4-week multimodal assessment. The assessment framework integrates four key data domains: (1) clinical assessment, involving self-reported data from a comprehensive battery of clinical and psychological questionnaires; (2) digital phenotyping, capturing real-world behavioral and physiological data through a wrist-worn wearable device and a smartphone application; (3) functional neuroimaging, using a baseline functional near-infrared spectroscopy (fNIRS) scan to measure prefrontal cortex activity; and (4) genomic and biomarker collection from blood samples for genomic and exploratory biomarker analyses. The study was conducted between March 2023 and October 2024, and all recruitment and data collection have been completed. The core analysis will employ advanced computational methods, including clustering and machine learning, to identify the distinct subtypes of insomnia. DISCUSSION: By applying multivariate pattern analysis and machine learning techniques to this rich, integrated dataset, we aimed to identify distinct biopsychosocial phenotypes of insomnia. This deep phenotyping approach is expected to elucidate the heterogeneity of insomnia, paving the way for the development of targeted and personalized management strategies for individuals with sleep disorders. CLINICAL TRIAL REGISTRATION: Clinical Research Information Service KCT0009175; https://cris.nih.go.kr/cris/search/detailSearch.do?seq=26133.