Back to search

University of South Florida

Bayesian Multivariate Joint Modeling for Skewed-longitudinal and Time-to-event Data

Abstract

dc:description

In epidemiologic and clinical studies, a relatively large number of biomarkers are repeatedly measured in patients over time, often associated with data on epidemiologic and clinical interest events. So, much attention is focused on developing the specific patterns of the longitudinal measurements, and the associations between those patterns and the time to a certain event, such as heart attack, diagnose of disease, time to transplantation, or death. In the last two decades, the research into joint modeling of longitudinal and time-to-event data has received a tremendous amount of attention. Numerous researchers have proposed joint modeling approaches for a single longitudinal exposure and time-to-event data, herein referred to as univariate joint modeling. Those model-based analyses may not provide robust inference when longitudinal measurements exhibit skewness and/or heavy tails. Additionally, the collected data are often featured by multivariate longitudinal exposures, which are significantly correlated, and ignore their correlation may lead to biased estimation. To the best of our knowledge, few studies focus on the multivariate joint modeling (MVJM) with a skewed distribution for longitudinal responses to cope with correlated multiple longitudinal exposures and adjust departures from normality and tailor linkage in specifying a survival process. This dissertation research contributes to the Bayesian inference statistical methodology in the field of joint modeling analyses, which is applied to a type 1 diabetes (T1D) study and a primary biliary cirrhosis (PBC) study. In this dissertation, firstly, we develop multivariate mixed-effects joint modeling for skewed-longitudinal and time-to-event data with skew distribution; the associated Bayesian inferential approach is introduced; the proposed joint modeling is applied to analyze the diabetes study data. Simulation studies are also conducted to assess the performance of the proposed method under different scenarios. Secondly, a Bayesian trivariate linear mixed-effects models with skew-normal distribution is applied to analyze Mayo Clinical PBC study data. This biostatistical methodology can be widely used in epidemiological and clinical study. Our multivariate joint modeling can provide more precise information and evidence to physicians, benefiting their treatment evaluation and clinical decision-making. All the analyses introduced in this dissertation have been implemented under the Bayesian framework, in the R and WinBUGS public and freely available software environment. The software codes are available from the author upon request.

Degree

thesis:*
Name thesis:degree_name
Doctor of Public Health (Dr.PH.)
Level thesis:degree_level
Dissertation
Discipline thesis:degree_discipline
Epidemiology and Biostatistics
Grantor
University of South Florida
Year dc:date
2021

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Xu, Lan
Contributors dc:contributor
  • Yangxin Huang, Ph.D.
  • Henian Chen, Ph.D.
  • Alfred Mbah, Ph.D.
  • Feng Cheng, Ph.D.

Subjects

dc:subject × 5

Identifiers

dc:identifier.*
Repository record dc:identifier
https://digitalcommons.usf.edu/etd/9268

Chain of custody

source
Harvested from
University of South Florida
Base URL
digitalcommons.usf.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Xu, Lan. Bayesian Multivariate Joint Modeling for Skewed-longitudinal and Time-to-event Data. Dissertation thesis, University of South Florida, 2021. https://digitalcommons.usf.edu/etd/9268