This repository serves as data bank for investigating the importance of performance-based physical function in dementia using data from the National Health and Aging Trends Study (NHATS).
Multi-state modeling was performed using the msm package (version 1.6.9), and the estimation of life expectancies were computed using the elect package (version 1.2).
Regarding the NHATS_data data object, it is the dataset including only forward transitions (as we stated in our research based on the medical evidence of irreversible-and-progressive conditions of dementia), and older adults that completed, at least, two interviews. The variables contained in this dataset are:
id: specific identifier for each participant in the NHATSstate: clinical diagnosis for dementia. State 1 refers to older adults free of dementia (dementia-free state); state 2 refers to older adults with any mild cognitive impairment (MCI state; known as the prelude of dementia); state 3 corresponds with a clinical diagnosis of dementia; and the state 4 corresponds with participant death.dep: dummy variable indicating a depression diagnosis.smk_before: smoking status; whether older adults had been smoked during their lifetimes.exercise: whether older adults were considered physically active on a regular basis.income: stratified income groups from 1 (very-low income) to 5 (high-income).sex: 1 for men, and 2 for women.wave: rounds of NHATS (from 1 in 2011 to 12 in 2023).sppb10: whether or not the person reached 10 or more score in the Short Physical Performance Battery.income2: classification in low-income and high-income; whether the person earned more than 20,000$/year.age: the centered age of the participants; this variable is centered subtracting the mean sample age (80.55) to the real age.sppb: original score in the Short Physical Performance Battery (from 0 [the worse score] to 12 [the best score]).
