Bayesian Variable Selection Methods for Matched Case-Control Studies

Josephine Asafu-Adjei, G. Tadesse Mahlet, Brent Coull, Raji Balasubramanian, Michael Lev, Lee Schwamm, Rebecca Betensky

Research output: Contribution to journalArticle

Abstract

Matched case-control designs are currently used in many biomedical applications. To ensure high efficiency and statistical power in identifying features that best discriminate cases from controls, it is important to account for the use of matched designs. However, in the setting of high dimensional data, few variable selection methods account for matching. Bayesian approaches to variable selection have several advantages, including the fact that such approaches visit a wider range of model subsets. In this paper, we propose a variable selection method to account for case-control matching in a Bayesian context and apply it using simulation studies, a matched brain imaging study conducted at Massachusetts General Hospital, and a matched cardiovascular biomarker study conducted by the High Risk Plaque Initiative.

Original languageEnglish (US)
Article number20160043
JournalInternational Journal of Biostatistics
Volume13
Issue number1
DOIs
StatePublished - Jan 1 2017

Fingerprint

Matched Case-control Study
Bayesian Variable Selection
Variable Selection
Case-control
Statistical Power
Biomedical Applications
Biomarkers
High-dimensional Data
Bayesian Approach
Control Design
High Efficiency
Imaging
Simulation Study
Subset
Range of data
Variable selection
Model

Keywords

  • Bayesian analysis
  • Conditional logistic regression
  • Matched case-control studies
  • Variable selection methods

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Asafu-Adjei, J., Mahlet, G. T., Coull, B., Balasubramanian, R., Lev, M., Schwamm, L., & Betensky, R. (2017). Bayesian Variable Selection Methods for Matched Case-Control Studies. International Journal of Biostatistics, 13(1), [20160043]. https://doi.org/10.1515/ijb-2016-0043

Bayesian Variable Selection Methods for Matched Case-Control Studies. / Asafu-Adjei, Josephine; Mahlet, G. Tadesse; Coull, Brent; Balasubramanian, Raji; Lev, Michael; Schwamm, Lee; Betensky, Rebecca.

In: International Journal of Biostatistics, Vol. 13, No. 1, 20160043, 01.01.2017.

Research output: Contribution to journalArticle

Asafu-Adjei, J, Mahlet, GT, Coull, B, Balasubramanian, R, Lev, M, Schwamm, L & Betensky, R 2017, 'Bayesian Variable Selection Methods for Matched Case-Control Studies', International Journal of Biostatistics, vol. 13, no. 1, 20160043. https://doi.org/10.1515/ijb-2016-0043
Asafu-Adjei J, Mahlet GT, Coull B, Balasubramanian R, Lev M, Schwamm L et al. Bayesian Variable Selection Methods for Matched Case-Control Studies. International Journal of Biostatistics. 2017 Jan 1;13(1). 20160043. https://doi.org/10.1515/ijb-2016-0043
Asafu-Adjei, Josephine ; Mahlet, G. Tadesse ; Coull, Brent ; Balasubramanian, Raji ; Lev, Michael ; Schwamm, Lee ; Betensky, Rebecca. / Bayesian Variable Selection Methods for Matched Case-Control Studies. In: International Journal of Biostatistics. 2017 ; Vol. 13, No. 1.
@article{cedc68f915104cf69a74d3e2d8ad8ec8,
title = "Bayesian Variable Selection Methods for Matched Case-Control Studies",
abstract = "Matched case-control designs are currently used in many biomedical applications. To ensure high efficiency and statistical power in identifying features that best discriminate cases from controls, it is important to account for the use of matched designs. However, in the setting of high dimensional data, few variable selection methods account for matching. Bayesian approaches to variable selection have several advantages, including the fact that such approaches visit a wider range of model subsets. In this paper, we propose a variable selection method to account for case-control matching in a Bayesian context and apply it using simulation studies, a matched brain imaging study conducted at Massachusetts General Hospital, and a matched cardiovascular biomarker study conducted by the High Risk Plaque Initiative.",
keywords = "Bayesian analysis, Conditional logistic regression, Matched case-control studies, Variable selection methods",
author = "Josephine Asafu-Adjei and Mahlet, {G. Tadesse} and Brent Coull and Raji Balasubramanian and Michael Lev and Lee Schwamm and Rebecca Betensky",
year = "2017",
month = "1",
day = "1",
doi = "10.1515/ijb-2016-0043",
language = "English (US)",
volume = "13",
journal = "International Journal of Biostatistics",
issn = "1557-4679",
publisher = "Berkeley Electronic Press",
number = "1",

}

TY - JOUR

T1 - Bayesian Variable Selection Methods for Matched Case-Control Studies

AU - Asafu-Adjei, Josephine

AU - Mahlet, G. Tadesse

AU - Coull, Brent

AU - Balasubramanian, Raji

AU - Lev, Michael

AU - Schwamm, Lee

AU - Betensky, Rebecca

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Matched case-control designs are currently used in many biomedical applications. To ensure high efficiency and statistical power in identifying features that best discriminate cases from controls, it is important to account for the use of matched designs. However, in the setting of high dimensional data, few variable selection methods account for matching. Bayesian approaches to variable selection have several advantages, including the fact that such approaches visit a wider range of model subsets. In this paper, we propose a variable selection method to account for case-control matching in a Bayesian context and apply it using simulation studies, a matched brain imaging study conducted at Massachusetts General Hospital, and a matched cardiovascular biomarker study conducted by the High Risk Plaque Initiative.

AB - Matched case-control designs are currently used in many biomedical applications. To ensure high efficiency and statistical power in identifying features that best discriminate cases from controls, it is important to account for the use of matched designs. However, in the setting of high dimensional data, few variable selection methods account for matching. Bayesian approaches to variable selection have several advantages, including the fact that such approaches visit a wider range of model subsets. In this paper, we propose a variable selection method to account for case-control matching in a Bayesian context and apply it using simulation studies, a matched brain imaging study conducted at Massachusetts General Hospital, and a matched cardiovascular biomarker study conducted by the High Risk Plaque Initiative.

KW - Bayesian analysis

KW - Conditional logistic regression

KW - Matched case-control studies

KW - Variable selection methods

UR - http://www.scopus.com/inward/record.url?scp=85020465942&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020465942&partnerID=8YFLogxK

U2 - 10.1515/ijb-2016-0043

DO - 10.1515/ijb-2016-0043

M3 - Article

VL - 13

JO - International Journal of Biostatistics

JF - International Journal of Biostatistics

SN - 1557-4679

IS - 1

M1 - 20160043

ER -