Time-Varying Latent Effect Model for Longitudinal Data with Informative Observation Times

Na Cai, Wenbin Lu, Hao Zhang

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

In analysis of longitudinal data, it is not uncommon that observation times of repeated measurements are subject-specific and correlated with underlying longitudinal outcomes. Taking account of the dependence between observation times and longitudinal outcomes is critical under these situations to assure the validity of statistical inference. In this article, we propose a flexible joint model for longitudinal data analysis in the presence of informative observation times. In particular, the new procedure considers the shared random-effect model and assumes a time-varying coefficient for the latent variable, allowing a flexible way of modeling longitudinal outcomes while adjusting their association with observation times. Estimating equations are developed for parameter estimation. We show that the resulting estimators are consistent and asymptotically normal, with variance-covariance matrix that has a closed form and can be consistently estimated by the usual plug-in method. One additional advantage of the procedure is that it provides a unified framework to test whether the effect of the latent variable is zero, constant, or time-varying. Simulation studies show that the proposed approach is appropriate for practical use. An application to a bladder cancer data is also given to illustrate the methodology.

Original languageEnglish (US)
Pages (from-to)1093-1102
Number of pages10
JournalBiometrics
Volume68
Issue number4
DOIs
StatePublished - Dec 2012
Externally publishedYes

Fingerprint

Longitudinal Data
Time-varying
Observation
Latent Variables
Covariance matrix
Parameter estimation
Plug-in Method
Longitudinal Data Analysis
Bladder Cancer
Time-varying Coefficients
Variance-covariance Matrix
Joint Model
Repeated Measurements
Estimating Equation
Random Effects Model
data analysis
Statistical Inference
Model
Parameter Estimation
Closed-form

Keywords

  • Estimating equation method
  • Informative observation times
  • Longitudinal data analysis
  • Time-varying effect

ASJC Scopus subject areas

  • Applied Mathematics
  • Statistics and Probability
  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Medicine(all)

Cite this

Time-Varying Latent Effect Model for Longitudinal Data with Informative Observation Times. / Cai, Na; Lu, Wenbin; Zhang, Hao.

In: Biometrics, Vol. 68, No. 4, 12.2012, p. 1093-1102.

Research output: Contribution to journalArticle

@article{7b3ba098b1bb4cd4bcf4f6e2923566d5,
title = "Time-Varying Latent Effect Model for Longitudinal Data with Informative Observation Times",
abstract = "In analysis of longitudinal data, it is not uncommon that observation times of repeated measurements are subject-specific and correlated with underlying longitudinal outcomes. Taking account of the dependence between observation times and longitudinal outcomes is critical under these situations to assure the validity of statistical inference. In this article, we propose a flexible joint model for longitudinal data analysis in the presence of informative observation times. In particular, the new procedure considers the shared random-effect model and assumes a time-varying coefficient for the latent variable, allowing a flexible way of modeling longitudinal outcomes while adjusting their association with observation times. Estimating equations are developed for parameter estimation. We show that the resulting estimators are consistent and asymptotically normal, with variance-covariance matrix that has a closed form and can be consistently estimated by the usual plug-in method. One additional advantage of the procedure is that it provides a unified framework to test whether the effect of the latent variable is zero, constant, or time-varying. Simulation studies show that the proposed approach is appropriate for practical use. An application to a bladder cancer data is also given to illustrate the methodology.",
keywords = "Estimating equation method, Informative observation times, Longitudinal data analysis, Time-varying effect",
author = "Na Cai and Wenbin Lu and Hao Zhang",
year = "2012",
month = "12",
doi = "10.1111/j.1541-0420.2012.01794.x",
language = "English (US)",
volume = "68",
pages = "1093--1102",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Time-Varying Latent Effect Model for Longitudinal Data with Informative Observation Times

AU - Cai, Na

AU - Lu, Wenbin

AU - Zhang, Hao

PY - 2012/12

Y1 - 2012/12

N2 - In analysis of longitudinal data, it is not uncommon that observation times of repeated measurements are subject-specific and correlated with underlying longitudinal outcomes. Taking account of the dependence between observation times and longitudinal outcomes is critical under these situations to assure the validity of statistical inference. In this article, we propose a flexible joint model for longitudinal data analysis in the presence of informative observation times. In particular, the new procedure considers the shared random-effect model and assumes a time-varying coefficient for the latent variable, allowing a flexible way of modeling longitudinal outcomes while adjusting their association with observation times. Estimating equations are developed for parameter estimation. We show that the resulting estimators are consistent and asymptotically normal, with variance-covariance matrix that has a closed form and can be consistently estimated by the usual plug-in method. One additional advantage of the procedure is that it provides a unified framework to test whether the effect of the latent variable is zero, constant, or time-varying. Simulation studies show that the proposed approach is appropriate for practical use. An application to a bladder cancer data is also given to illustrate the methodology.

AB - In analysis of longitudinal data, it is not uncommon that observation times of repeated measurements are subject-specific and correlated with underlying longitudinal outcomes. Taking account of the dependence between observation times and longitudinal outcomes is critical under these situations to assure the validity of statistical inference. In this article, we propose a flexible joint model for longitudinal data analysis in the presence of informative observation times. In particular, the new procedure considers the shared random-effect model and assumes a time-varying coefficient for the latent variable, allowing a flexible way of modeling longitudinal outcomes while adjusting their association with observation times. Estimating equations are developed for parameter estimation. We show that the resulting estimators are consistent and asymptotically normal, with variance-covariance matrix that has a closed form and can be consistently estimated by the usual plug-in method. One additional advantage of the procedure is that it provides a unified framework to test whether the effect of the latent variable is zero, constant, or time-varying. Simulation studies show that the proposed approach is appropriate for practical use. An application to a bladder cancer data is also given to illustrate the methodology.

KW - Estimating equation method

KW - Informative observation times

KW - Longitudinal data analysis

KW - Time-varying effect

UR - http://www.scopus.com/inward/record.url?scp=84871665014&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871665014&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2012.01794.x

DO - 10.1111/j.1541-0420.2012.01794.x

M3 - Article

C2 - 23025338

AN - SCOPUS:84871665014

VL - 68

SP - 1093

EP - 1102

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -