Robust mixed effects model for clustered failure time data: Application to Huntington’s disease event measures

Tanya P. Garcia, Yanyuan Ma, Karen Marder, Yuanjia Wang

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

An important goal in clinical and statistical research is properly modeling the distribution for clustered failure times which have a natural intra-class dependency and are subject to censoring. We handle these challenges with a novel approach that does not impose restrictive modeling or distribu-tional assumptions. Using a logit transformation, we relate the distribution for clustered failure times to covariates and a random, subject-specific effect. The covariates are modeled with unknown functional forms, and the random effect may depend on the covariates and have an unknown and unspecified distribution. We introduce pseudovalues to handle censoring and splines for functional covariate effects, and frame the problem into fitting an additive logistic mixed effects model. Unlike existing approaches for fitting such models, we develop semiparametric techniques that estimate the functional model parameters without specifying or estimating the random effect distribution. We show both theoretically and empirically that the resulting estimators are consistent for any choice of random effect distribution and any dependency structure between the random effect and covariates. Last, we illustrate the method’s utility in an application to a Huntington’s disease study where our method provides new insights into differences between motor and cognitive impairment event times in at-risk subjects.

Original languageEnglish (US)
Pages (from-to)1085-1116
Number of pages32
JournalAnnals of Applied Statistics
Volume11
Issue number2
DOIs
StatePublished - Jun 1 2017

Fingerprint

Failure Time Data
Mixed Effects Model
Clustered Data
Covariates
Random Effects
Failure Time
Censoring
Splines
Logistics
Unknown
Logit
Functional Model
Model Fitting
Modeling
Spline
Estimator
Random effects
Estimate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Modeling and Simulation
  • Statistics, Probability and Uncertainty

Cite this

@article{188e2fd1057643a1871e74d69f8b778c,
title = "Robust mixed effects model for clustered failure time data: Application to Huntington’s disease event measures",
abstract = "An important goal in clinical and statistical research is properly modeling the distribution for clustered failure times which have a natural intra-class dependency and are subject to censoring. We handle these challenges with a novel approach that does not impose restrictive modeling or distribu-tional assumptions. Using a logit transformation, we relate the distribution for clustered failure times to covariates and a random, subject-specific effect. The covariates are modeled with unknown functional forms, and the random effect may depend on the covariates and have an unknown and unspecified distribution. We introduce pseudovalues to handle censoring and splines for functional covariate effects, and frame the problem into fitting an additive logistic mixed effects model. Unlike existing approaches for fitting such models, we develop semiparametric techniques that estimate the functional model parameters without specifying or estimating the random effect distribution. We show both theoretically and empirically that the resulting estimators are consistent for any choice of random effect distribution and any dependency structure between the random effect and covariates. Last, we illustrate the method’s utility in an application to a Huntington’s disease study where our method provides new insights into differences between motor and cognitive impairment event times in at-risk subjects.",
author = "Garcia, {Tanya P.} and Yanyuan Ma and Karen Marder and Yuanjia Wang",
year = "2017",
month = "6",
day = "1",
doi = "10.1214/17-AOAS1038",
language = "English (US)",
volume = "11",
pages = "1085--1116",
journal = "Annals of Applied Statistics",
issn = "1932-6157",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

Robust mixed effects model for clustered failure time data : Application to Huntington’s disease event measures. / Garcia, Tanya P.; Ma, Yanyuan; Marder, Karen; Wang, Yuanjia.

In: Annals of Applied Statistics, Vol. 11, No. 2, 01.06.2017, p. 1085-1116.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Robust mixed effects model for clustered failure time data

T2 - Application to Huntington’s disease event measures

AU - Garcia, Tanya P.

AU - Ma, Yanyuan

AU - Marder, Karen

AU - Wang, Yuanjia

PY - 2017/6/1

Y1 - 2017/6/1

N2 - An important goal in clinical and statistical research is properly modeling the distribution for clustered failure times which have a natural intra-class dependency and are subject to censoring. We handle these challenges with a novel approach that does not impose restrictive modeling or distribu-tional assumptions. Using a logit transformation, we relate the distribution for clustered failure times to covariates and a random, subject-specific effect. The covariates are modeled with unknown functional forms, and the random effect may depend on the covariates and have an unknown and unspecified distribution. We introduce pseudovalues to handle censoring and splines for functional covariate effects, and frame the problem into fitting an additive logistic mixed effects model. Unlike existing approaches for fitting such models, we develop semiparametric techniques that estimate the functional model parameters without specifying or estimating the random effect distribution. We show both theoretically and empirically that the resulting estimators are consistent for any choice of random effect distribution and any dependency structure between the random effect and covariates. Last, we illustrate the method’s utility in an application to a Huntington’s disease study where our method provides new insights into differences between motor and cognitive impairment event times in at-risk subjects.

AB - An important goal in clinical and statistical research is properly modeling the distribution for clustered failure times which have a natural intra-class dependency and are subject to censoring. We handle these challenges with a novel approach that does not impose restrictive modeling or distribu-tional assumptions. Using a logit transformation, we relate the distribution for clustered failure times to covariates and a random, subject-specific effect. The covariates are modeled with unknown functional forms, and the random effect may depend on the covariates and have an unknown and unspecified distribution. We introduce pseudovalues to handle censoring and splines for functional covariate effects, and frame the problem into fitting an additive logistic mixed effects model. Unlike existing approaches for fitting such models, we develop semiparametric techniques that estimate the functional model parameters without specifying or estimating the random effect distribution. We show both theoretically and empirically that the resulting estimators are consistent for any choice of random effect distribution and any dependency structure between the random effect and covariates. Last, we illustrate the method’s utility in an application to a Huntington’s disease study where our method provides new insights into differences between motor and cognitive impairment event times in at-risk subjects.

UR - http://www.scopus.com/inward/record.url?scp=85026289641&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85026289641&partnerID=8YFLogxK

U2 - 10.1214/17-AOAS1038

DO - 10.1214/17-AOAS1038

M3 - Article

C2 - 29399240

AN - SCOPUS:85026289641

VL - 11

SP - 1085

EP - 1116

JO - Annals of Applied Statistics

JF - Annals of Applied Statistics

SN - 1932-6157

IS - 2

ER -