Variable selection for joint models with time-varying coefficients

Yujing Xie, Zangdong He, Wanzhu Tu, Zhangsheng Yu

Research output: Contribution to journalArticle

Abstract

Many clinical studies collect longitudinal and survival data concurrently. Joint models combining these two types of outcomes through shared random effects are frequently used in practical data analysis. The standard joint models assume that the coefficients for the longitudinal and survival components are time-invariant. In many applications, the assumption is overly restrictive. In this research, we extend the standard joint model to include time-varying coefficients, in both longitudinal and survival components, and we present a data-driven method for variable selection. Specifically, we use a B-spline decomposition and penalized likelihood with adaptive group LASSO to select the relevant independent variables and to distinguish the time-varying and time-invariant effects for the two model components. We use Gaussian-Legendre and Gaussian-Hermite quadratures to approximate the integrals in the absence of closed-form solutions. Simulation studies show good selection and estimation performance. Finally, we use the proposed procedure to analyze data generated by a study of primary biliary cirrhosis.

Original languageEnglish (US)
JournalStatistical Methods in Medical Research
DOIs
StateAccepted/In press - Jan 1 2019

Fingerprint

Time-varying Coefficients
Joint Model
Variable Selection
Joints
Standard Model
Penalized Likelihood
Invariant
Survival Data
Longitudinal Data
Legendre
Component Model
B-spline
Random Effects
Hermite
Closed-form Solution
Data-driven
Quadrature
Biliary Liver Cirrhosis
Time-varying
Data analysis

Keywords

  • adaptive LASSO
  • B-spline
  • Gaussian quadrature

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability
  • Health Information Management

Cite this

Variable selection for joint models with time-varying coefficients. / Xie, Yujing; He, Zangdong; Tu, Wanzhu; Yu, Zhangsheng.

In: Statistical Methods in Medical Research, 01.01.2019.

Research output: Contribution to journalArticle

@article{cfea5c849b924f8f858186be00af5389,
title = "Variable selection for joint models with time-varying coefficients",
abstract = "Many clinical studies collect longitudinal and survival data concurrently. Joint models combining these two types of outcomes through shared random effects are frequently used in practical data analysis. The standard joint models assume that the coefficients for the longitudinal and survival components are time-invariant. In many applications, the assumption is overly restrictive. In this research, we extend the standard joint model to include time-varying coefficients, in both longitudinal and survival components, and we present a data-driven method for variable selection. Specifically, we use a B-spline decomposition and penalized likelihood with adaptive group LASSO to select the relevant independent variables and to distinguish the time-varying and time-invariant effects for the two model components. We use Gaussian-Legendre and Gaussian-Hermite quadratures to approximate the integrals in the absence of closed-form solutions. Simulation studies show good selection and estimation performance. Finally, we use the proposed procedure to analyze data generated by a study of primary biliary cirrhosis.",
keywords = "adaptive LASSO, B-spline, Gaussian quadrature",
author = "Yujing Xie and Zangdong He and Wanzhu Tu and Zhangsheng Yu",
year = "2019",
month = "1",
day = "1",
doi = "10.1177/0962280219873125",
language = "English (US)",
journal = "Statistical Methods in Medical Research",
issn = "0962-2802",
publisher = "SAGE Publications Ltd",

}

TY - JOUR

T1 - Variable selection for joint models with time-varying coefficients

AU - Xie, Yujing

AU - He, Zangdong

AU - Tu, Wanzhu

AU - Yu, Zhangsheng

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Many clinical studies collect longitudinal and survival data concurrently. Joint models combining these two types of outcomes through shared random effects are frequently used in practical data analysis. The standard joint models assume that the coefficients for the longitudinal and survival components are time-invariant. In many applications, the assumption is overly restrictive. In this research, we extend the standard joint model to include time-varying coefficients, in both longitudinal and survival components, and we present a data-driven method for variable selection. Specifically, we use a B-spline decomposition and penalized likelihood with adaptive group LASSO to select the relevant independent variables and to distinguish the time-varying and time-invariant effects for the two model components. We use Gaussian-Legendre and Gaussian-Hermite quadratures to approximate the integrals in the absence of closed-form solutions. Simulation studies show good selection and estimation performance. Finally, we use the proposed procedure to analyze data generated by a study of primary biliary cirrhosis.

AB - Many clinical studies collect longitudinal and survival data concurrently. Joint models combining these two types of outcomes through shared random effects are frequently used in practical data analysis. The standard joint models assume that the coefficients for the longitudinal and survival components are time-invariant. In many applications, the assumption is overly restrictive. In this research, we extend the standard joint model to include time-varying coefficients, in both longitudinal and survival components, and we present a data-driven method for variable selection. Specifically, we use a B-spline decomposition and penalized likelihood with adaptive group LASSO to select the relevant independent variables and to distinguish the time-varying and time-invariant effects for the two model components. We use Gaussian-Legendre and Gaussian-Hermite quadratures to approximate the integrals in the absence of closed-form solutions. Simulation studies show good selection and estimation performance. Finally, we use the proposed procedure to analyze data generated by a study of primary biliary cirrhosis.

KW - adaptive LASSO

KW - B-spline

KW - Gaussian quadrature

UR - http://www.scopus.com/inward/record.url?scp=85073834283&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073834283&partnerID=8YFLogxK

U2 - 10.1177/0962280219873125

DO - 10.1177/0962280219873125

M3 - Article

C2 - 31512571

AN - SCOPUS:85073834283

JO - Statistical Methods in Medical Research

JF - Statistical Methods in Medical Research

SN - 0962-2802

ER -