Sleep-PFI CHARLS

DescriptionThis repository contains the raw analytical dataset, R code, and supplementary documentation supporting the study titled: "Nonlinear associations between sleep duration and psychological frailty in an aging Chinese population: Insights from a national cohort."Data DescriptionThe raw analytical dataset is derived from the China Health and Retirement Longitudinal Study (CHARLS) 2011 baseline and 2018 fourth wave. This processed dataset includes key variables for 9,476 participants (aged $\ge 45$ years), including:Sleep metrics: Nighttime sleep duration, nap duration, and calculated total sleep time.Outcome measure: The calculated 15-item Psychological Frailty Index (PFI).Covariates: Essential sociodemographic factors, lifestyle behaviors (smoking, alcohol consumption), chronic health conditions, and laboratory biomarkers.Note: This file represents the processed analytical dataset used in our specific study. Direct access to the full raw CHARLS microdata requires application via the official CHARLS website.Code DescriptionThe accompanying R scripts provide the complete workflow for data processing and statistical modeling as presented in our manuscript:Data Preprocessing: Cleaning and variable derivation (e.g., PFI computation, sleep categorization).Main Statistical Analyses: Multivariable logistic regression and restricted cubic spline (RCS) modeling to assess nonlinear dose-response relationships.Robustness & Sensitivity Checks: Code for stratified analyses, longitudinal trajectory modeling, cluster-robust standard error estimation, multiple imputation (MICE), and inverse-probability-weighting (IPW).Visualization: Scripts for generating all tables and figures included in the publication.Supplementary MaterialThe repository also includes the comprehensive Supplementary Material file (PDF), which provides detailed documentation of our sensitivity analyses and methodological robustness checks, including:Multicollinearity assessments (VIF).Longitudinal sensitivity analyses using intermediate waves (2013, 2015) and sleep trajectory modeling.Little's MCAR test results and multiple imputation outputs.Comparison of baseline characteristics and inverse-probability-weighting (IPW) diagnostics.Purpose of SharingWe share these resources to enhance the transparency, reproducibility, and verifiability of our research. Researchers are encouraged to use these files to replicate our analyses or to conduct further investigations into aging and mental health.