An overview of longitudinal data analyses in sleep research

Area of Science:

• Sleep research
• Biostatistics
• Data science

Background:

• Longitudinal data is crucial in sleep apnea research for tracking changes in weight, symptoms, and treatment adherence.
• Analyzing this data requires careful statistical method selection to avoid bias in estimates and predictions.

Purpose of the Study:

• To provide a structured overview of longitudinal data analysis methods for sleep research.
• To illustrate the application of these methods using a simulated dataset.

Main Methods:

• A simulated dataset of 300 patients with 90 repeated measurements was used.
• A stepwise analytical framework was applied, covering data description, visualization, research question formulation, and method selection.
• Ten common analytical approaches were presented: descriptive, clustering, and modeling methods.

Main Results:

• The study outlines ten statistical methods for longitudinal data analysis in sleep research.
• It emphasizes the assumptions, scope, and interpretation of each method.
• Implementation examples are provided for clarity.

Conclusions:

• This overview provides a clear analytical pathway for longitudinal data in sleep research.
• It aids researchers and clinicians in navigating complex analyses.
• The goal is to improve the accuracy and reliability of findings in sleep studies.