A Novel Strategy to Develop a Rapid Screen, Simple Subgrouping and Precise Intervention Model for Older Adults With Poor Sleep Quality: Integrating Latent Class Analysis With Causal- Pie Analysis to Implement Population-based Precision Medicine

Poor sleep quality is a major issue of public health in older adults. However, the underlying cause of poor sleep quality in older adults is multifactorial and complex. A highly accessible and effective instrument for screening and subgrouping older adults with poor sleep quality is warranted. In particular, the instrument is expected to suggest respective interventions. The present project will use two groups, including non-disabled and disabled older adults, to develop the short screen and subgrouping scale along with the respective intervention. During the first and second year of the project, we will apply the established non-disabled version of' the 'Short Screen and Subgrouping Scale' to screen and to subgroup older adults with poor sleep quality in non-disabled older population. Consecutively, the screened subgroups will be performed the causal-pie analysis to elucidate the complex cause of poor sleep quality. The number of participants is estimated as 500. Meanwhile, additional 500 disabled older adults, who will be invited in the hospital, will be used in the latent class analysis (LCA) to subgroup reasonable classes of poor sleep quality in disabled older adults. According to the result of LCA, the disabled version of the 'Short Screen and Subgrouping Scale' will be produced and relevant causal-pie analysis will be conducted accordingly. In the second and third year of this project, two randomized control trials (RCT) with two arms will be conducted, respectively. One arm is non-pharmacological intervention group (n=30) and the other is the control group (n=30). In the non-disabled older adults, the screened mild insomnia group, which has been defined in our pilot study, will be invited to attend the RCT. In the disabled older adults, after the LCA, the subgroup with mild symptoms will be considered to enter the RCT study phase. The intervention for both RCTs will be designed according to the result of causal-pie analysis. The non-pharmacological intervention will comprise sleep hygiene education, exercise training and relevant modules of cognitive behavioral therapy. This 3-year project is expected to frame a population-level model of precision medicine, which targets poor sleep quality in older adults.