Advanced search
1 file | 673.71 KB

Semiparametric tests for sufficient cause interaction

Author
Organization
Project
Bioinformatics: from nucleotids to networks (N2N)
Abstract
A sufficient cause interaction between two exposures signals the presence of individuals for whom the outcome would occur only under certain values of the two exposures. When the outcome is dichotomous and all exposures are categorical, then, under certain no confounding assumptions, empirical conditions for sufficient cause interactions can be constructed on the basis of the sign of linear contrasts of conditional outcome probabilities between differently exposed subgroups, given confounders. It is argued that logistic regression models are unsatisfactory for evaluating such contrasts, and that Bernoulli regression models with linear link are prone to misspecification. We therefore develop semiparametric tests for sufficient cause interactions under models which postulate probability contrasts in terms of a finite dimensional parameter, but which are otherwise unspecified. Estimation is often not feasible in these models because it would require non-parametric estimation of auxiliary conditional expectations given high dimensional variables. We therefore develop multiply robust tests under a union model which assumes that at least one of several working submodels holds. In the special case of a randomized experiment or a family-based genetic study in which the joint exposure distribution is known by design or Mendelian inheritance, the procedure leads to asymptotically distribution-free tests of the null hypothesis of no sufficient cause interaction.
Keywords
IDENTIFICATION, ASSOCIATION, MODELS, INFERENCE, Gene-environment interaction, GENE-ENVIRONMENT INTERACTIONS, Synergism, Sufficient cause, INCOMPLETE DATA, EPISTASIS, STATISTICAL INTERACTIONS, Gene-gene interaction, Semiparametric inference, Effect modification, Double robustness, EFFICIENCY, ROBUSTNESS

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 673.71 KB

Citation

Please use this url to cite or link to this publication:

Chicago
Vansteelandt, Stijn, Tyler J VanderWeele, and James M Robins. 2012. “Semiparametric Tests for Sufficient Cause Interaction.” Journal of the Royal Statistical Society Series B-statistical Methodology 74 (2): 223–244.
APA
Vansteelandt, S., VanderWeele, T. J., & Robins, J. M. (2012). Semiparametric tests for sufficient cause interaction. JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY, 74(2), 223–244.
Vancouver
1.
Vansteelandt S, VanderWeele TJ, Robins JM. Semiparametric tests for sufficient cause interaction. JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY. 2012;74(2):223–44.
MLA
Vansteelandt, Stijn, Tyler J VanderWeele, and James M Robins. “Semiparametric Tests for Sufficient Cause Interaction.” JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY 74.2 (2012): 223–244. Print.
@article{2122740,
  abstract     = {A sufficient cause interaction between two exposures signals the presence of individuals for whom the outcome would occur only under certain values of the two exposures. When the outcome is dichotomous and all exposures are categorical, then, under certain no confounding assumptions, empirical conditions for sufficient cause interactions can be constructed on the basis of the sign of linear contrasts of conditional outcome probabilities between differently exposed subgroups, given confounders. It is argued that logistic regression models are unsatisfactory for evaluating such contrasts, and that Bernoulli regression models with linear link are prone to misspecification. We therefore develop semiparametric tests for sufficient cause interactions under models which postulate probability contrasts in terms of a finite dimensional parameter, but which are otherwise unspecified. Estimation is often not feasible in these models because it would require non-parametric estimation of auxiliary conditional expectations given high dimensional variables. We therefore develop multiply robust tests under a union model which assumes that at least one of several working submodels holds. In the special case of a randomized experiment or a family-based genetic study in which the joint exposure distribution is known by design or Mendelian inheritance, the procedure leads to asymptotically distribution-free tests of the null hypothesis of no sufficient cause interaction.},
  author       = {Vansteelandt, Stijn and VanderWeele, Tyler J and Robins, James M},
  issn         = {1369-7412},
  journal      = {JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY},
  keyword      = {IDENTIFICATION,ASSOCIATION,MODELS,INFERENCE,Gene-environment interaction,GENE-ENVIRONMENT INTERACTIONS,Synergism,Sufficient cause,INCOMPLETE DATA,EPISTASIS,STATISTICAL INTERACTIONS,Gene-gene interaction,Semiparametric inference,Effect modification,Double robustness,EFFICIENCY,ROBUSTNESS},
  language     = {eng},
  number       = {2},
  pages        = {223--244},
  title        = {Semiparametric tests for sufficient cause interaction},
  url          = {http://dx.doi.org/10.1111/j.1467-9868.2011.01011.x},
  volume       = {74},
  year         = {2012},
}

Altmetric
View in Altmetric
Web of Science
Times cited: