Inference for causal interactions for continuous exposures under dichotomization

Tyler J. Vanderweele, Yu Chen, Habibul Ahsan

Research output: Contribution to journalArticle

Abstract

Dichotomization of continuous exposure variables is a common practice in medical and epidemiological research. The practice has been cautioned against on the grounds of efficiency and bias. Here we consider the consequences of dichotomization of a continuous covariate for the study of interactions. We show that when a continuous exposure has been dichotomized certain inferences concerning causal interactions can be drawn with regard to the original continuous exposure scale. Within the context of interaction analyses, dichotomization and the use of the results in this article can furthermore help prevent incorrect conclusions about the presence of interactions that result simply from erroneous modeling of the exposure variables. By considering different dichotomization points one can gain considerable insight concerning the presence of causal interaction between exposures at different levels. The results in this article are applied to a study of the interactive effects between smoking and arsenic exposure from well water in producing skin lesions.

Original languageEnglish (US)
Pages (from-to)1414-1421
Number of pages8
JournalBiometrics
Volume67
Issue number4
DOIs
StatePublished - Dec 2011

Fingerprint

exposure models
Arsenic
skin lesions
arsenic
Biomedical Research
Skin
Smoking
Water
Interaction
water
Causal Inference
Covariates
Modeling

Keywords

  • Causal inference
  • Counterfactuals
  • Dichotomization
  • Interaction
  • Synergism

ASJC Scopus subject areas

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

Cite this

Inference for causal interactions for continuous exposures under dichotomization. / Vanderweele, Tyler J.; Chen, Yu; Ahsan, Habibul.

In: Biometrics, Vol. 67, No. 4, 12.2011, p. 1414-1421.

Research output: Contribution to journalArticle

Vanderweele, Tyler J. ; Chen, Yu ; Ahsan, Habibul. / Inference for causal interactions for continuous exposures under dichotomization. In: Biometrics. 2011 ; Vol. 67, No. 4. pp. 1414-1421.
@article{ad52a1388f1b4ae7aa5b275d8f8ae761,
title = "Inference for causal interactions for continuous exposures under dichotomization",
abstract = "Dichotomization of continuous exposure variables is a common practice in medical and epidemiological research. The practice has been cautioned against on the grounds of efficiency and bias. Here we consider the consequences of dichotomization of a continuous covariate for the study of interactions. We show that when a continuous exposure has been dichotomized certain inferences concerning causal interactions can be drawn with regard to the original continuous exposure scale. Within the context of interaction analyses, dichotomization and the use of the results in this article can furthermore help prevent incorrect conclusions about the presence of interactions that result simply from erroneous modeling of the exposure variables. By considering different dichotomization points one can gain considerable insight concerning the presence of causal interaction between exposures at different levels. The results in this article are applied to a study of the interactive effects between smoking and arsenic exposure from well water in producing skin lesions.",
keywords = "Causal inference, Counterfactuals, Dichotomization, Interaction, Synergism",
author = "Vanderweele, {Tyler J.} and Yu Chen and Habibul Ahsan",
year = "2011",
month = "12",
doi = "10.1111/j.1541-0420.2011.01629.x",
language = "English (US)",
volume = "67",
pages = "1414--1421",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Inference for causal interactions for continuous exposures under dichotomization

AU - Vanderweele, Tyler J.

AU - Chen, Yu

AU - Ahsan, Habibul

PY - 2011/12

Y1 - 2011/12

N2 - Dichotomization of continuous exposure variables is a common practice in medical and epidemiological research. The practice has been cautioned against on the grounds of efficiency and bias. Here we consider the consequences of dichotomization of a continuous covariate for the study of interactions. We show that when a continuous exposure has been dichotomized certain inferences concerning causal interactions can be drawn with regard to the original continuous exposure scale. Within the context of interaction analyses, dichotomization and the use of the results in this article can furthermore help prevent incorrect conclusions about the presence of interactions that result simply from erroneous modeling of the exposure variables. By considering different dichotomization points one can gain considerable insight concerning the presence of causal interaction between exposures at different levels. The results in this article are applied to a study of the interactive effects between smoking and arsenic exposure from well water in producing skin lesions.

AB - Dichotomization of continuous exposure variables is a common practice in medical and epidemiological research. The practice has been cautioned against on the grounds of efficiency and bias. Here we consider the consequences of dichotomization of a continuous covariate for the study of interactions. We show that when a continuous exposure has been dichotomized certain inferences concerning causal interactions can be drawn with regard to the original continuous exposure scale. Within the context of interaction analyses, dichotomization and the use of the results in this article can furthermore help prevent incorrect conclusions about the presence of interactions that result simply from erroneous modeling of the exposure variables. By considering different dichotomization points one can gain considerable insight concerning the presence of causal interaction between exposures at different levels. The results in this article are applied to a study of the interactive effects between smoking and arsenic exposure from well water in producing skin lesions.

KW - Causal inference

KW - Counterfactuals

KW - Dichotomization

KW - Interaction

KW - Synergism

UR - http://www.scopus.com/inward/record.url?scp=83655163913&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=83655163913&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2011.01629.x

DO - 10.1111/j.1541-0420.2011.01629.x

M3 - Article

VL - 67

SP - 1414

EP - 1421

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -