Premium
Active Life Expectancy from Annual Follow–Up Data with Missing Responses
Author(s) -
Izmirlian Grant,
Brock Dwight,
Ferrucci Luigi,
Phillips Caroline
Publication year - 2000
Publication title -
biometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.298
H-Index - 130
eISSN - 1541-0420
pISSN - 0006-341X
DOI - 10.1111/j.0006-341x.2000.00244.x
Subject(s) - life expectancy , covariate , markov chain , statistics , missing data , demography , population , residence , mathematics , econometrics , gerontology , psychology , medicine , sociology
Summary. Active life expectancy (ALE) at a given age is denned as the expected remaining years free of disability. In this study, three categories of health status are defined according to the ability to perform activities of daily living independently. Several studies have used increment‐decrement life tables to estimate ALE, without error analysis, from only a baseline and one follow‐up interview. The present work conducts an individual‐level covariate analysis using a three‐state Markov chain model for multiple follow‐up data. Using a logistic link, the model estimates single‐year transition probabilities among states of health, accounting for missing interviews. This approach has the advantages of smoothing subsequent estimates and increased power by using all follow‐ups. We compute ALE and total life expectancy from these estimated single‐year transition probabilities. Variance estimates are computed using the delta method. Data from the Iowa Established Population for the Epidemiologic Study of the Elderly are used to test the effects of smoking on ALE on all 5‐year age groups past 65 years, controlling for sex and education.