EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Heterogeneous causal mediation analysis using Bayesian additive regression trees

Chen Liu, Xu Qin, Victor B Talisa, Jiebiao Wang iD

Biometrics · 09.04.2026 · 82 (2)

DOI: 10.1093/biomtc/ujag079 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biometrics, online erschienen am 25.05.2026. Verfasst von Chen Liu und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biostatistics and Health Data Science, University of Pittsburgh, Pittsburgh, PA 15261, United States. Hauptthemen laut MEDLINE: Mediation Analysis. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NIH HHS, NIA NIH HHS, U.S. Department of Education Institute of Education Sciences Grant.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Causal mediation analysis provides insights into the mechanisms through which treatments affect outcomes. While mediation effects often vary across individuals, most existing methods focus solely on population-average effects, overlooking individual-level heterogeneity. To address this limitation, we propose a Bayesian regression tree ensemble method that flexibly models nonlinear relationships and captures treatment-by-mediator interactions in the mediation process. Using hierarchical posterior sampling, our approach provides credible intervals with nominal coverage rates for inferring heterogeneous mediation effects. Additionally, we leverage regression tree summaries to identify subgroups with distinct mediation effects and employ SHapley Additive exPlanation values to highlight key moderators and their influence on the mediation process. Comprehensive simulations demonstrate the method’s accuracy in estimating and inferring heterogeneous mediation effects. Finally, we apply our method to investigate the heterogeneous mediation role of Alzheimer’s disease pathology burden in the effect of apolipoprotein E genotype on late-life cognition.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Mediation AnalysisBayes TheoremComputer SimulationAlzheimer DiseaseHumansModels, StatisticalRegression AnalysisCausalityApolipoproteins ETreatment Effect Heterogeneity

Stichwörter der Autor:innen

Bayesian tree ensemblescausal mediation analysisheterogeneous effectsmoderation mechanismsnonlinear interactions

Substanzen

Apolipoproteins E

Autor:innen und Institutionen

  • Department of Biostatistics and Health Data Science, University of Pittsburgh, Pittsburgh, PA 15261, United States.
  • Department of Health and Human Development, University of Pittsburgh, Pittsburgh, PA 15260, United States.
  • Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA 15213, United States.
  • Clinical and Translational Science Institute, University of Pittsburgh, Pittsburgh, PA 15261, United States.

ORCID-Profile: Jiebiao Wang

Förderung

  • NIH HHS
  • NIA NIH HHS
  • U.S. Department of Education Institute of Education Sciences Grant
  • National Science Foundation Faculty Early Career Development Program (CAREER) Award
  • National Institutes of Health

Zitieren

APA

Liu, C., Qin, X., Talisa, V. B., & Wang, J. (2026). Heterogeneous causal mediation analysis using Bayesian additive regression trees. Biometrics, 82(2). https://doi.org/10.1093/biomtc/ujag079

Vancouver

Liu C, Qin X, Talisa VB, Wang J. Heterogeneous causal mediation analysis using Bayesian additive regression trees. Biometrics. 2026;82(2). doi:10.1093/biomtc/ujag079

BibTeX
@article{Liu2026,
  author = {Liu, Chen and Qin, Xu and Talisa, Victor B and Wang, Jiebiao},
  title = {Heterogeneous causal mediation analysis using Bayesian additive regression trees},
  journal = {Biometrics},
  year = {2026},
  volume = {82},
  number = {2},
  doi = {10.1093/biomtc/ujag079}
}
Kostenlos starten →