Tutorial on directed acyclic graphs
- PMID: 34371103
- PMCID: PMC8821727
- DOI: 10.1016/j.jclinepi.2021.08.001
Tutorial on directed acyclic graphs
Abstract
Directed acyclic graphs (DAGs) are an intuitive yet rigorous tool to communicate about causal questions in clinical and epidemiologic research and inform study design and statistical analysis. DAGs are constructed to depict prior knowledge about biological and behavioral systems related to specific causal research questions. DAG components portray who receives treatment or experiences exposures; mechanisms by which treatments and exposures operate; and other factors that influence the outcome of interest or which persons are included in an analysis. Once assembled, DAGs - via a few simple rules - guide the researcher in identifying whether the causal effect of interest can be identified without bias and, if so, what must be done either in study design or data analysis to achieve this. Specifically, DAGs can identify variables that, if controlled for in the design or analysis phase, are sufficient to eliminate confounding and some forms of selection bias. DAGs also help recognize variables that, if controlled for, bias the analysis (e.g., mediators or factors influenced by both exposure and outcome). Finally, DAGs help researchers recognize insidious sources of bias introduced by selection of individuals into studies or failure to completely observe all individuals until study outcomes are reached. DAGs, however, are not infallible, largely owing to limitations in prior knowledge about the system in question. In such instances, several alternative DAGs are plausible, and researchers should assess whether results differ meaningfully across analyses guided by different DAGs and be forthright about uncertainty. DAGs are powerful tools to guide the conduct of clinical research.
Copyright © 2021 Elsevier Inc. All rights reserved.
Figures
Comment in
-
A cause should not be automatically taken as an effect modifier of other causes: author's reply.J Clin Epidemiol. 2022 Jun;146:127-128. doi: 10.1016/j.jclinepi.2022.02.014. Epub 2022 Feb 25. J Clin Epidemiol. 2022. PMID: 35219802 No abstract available.
-
A cause should not be automatically taken as an effect modifier of other causes: response to Digitale et al.J Clin Epidemiol. 2022 Jun;146:128-130. doi: 10.1016/j.jclinepi.2022.02.013. Epub 2022 Feb 25. J Clin Epidemiol. 2022. PMID: 35219804 No abstract available.
Dataset use reported in
-
Harm of early dexamethasone for COVID-19 and bias in randomized trials.Eur J Intern Med. 2023 Jan;107:100-101. doi: 10.1016/j.ejim.2022.09.014. Epub 2022 Sep 19. Eur J Intern Med. 2023. PMID: 36150979 Free PMC article. No abstract available.
Similar articles
-
Software Application Profile: The daggle app-a tool to support learning and teaching the graphical rules of selecting adjustment variables using directed acyclic graphs.Int J Epidemiol. 2023 Oct 5;52(5):1659-1664. doi: 10.1093/ije/dyad038. Int J Epidemiol. 2023. PMID: 36952629 Free PMC article.
-
Use of directed acyclic graphs (DAGs) to identify confounders in applied health research: review and recommendations.Int J Epidemiol. 2021 May 17;50(2):620-632. doi: 10.1093/ije/dyaa213. Int J Epidemiol. 2021. PMID: 33330936 Free PMC article. Review.
-
Evidence synthesis for constructing directed acyclic graphs (ESC-DAGs): a novel and systematic method for building directed acyclic graphs.Int J Epidemiol. 2020 Feb 1;49(1):322-329. doi: 10.1093/ije/dyz150. Int J Epidemiol. 2020. PMID: 31325312 Free PMC article.
-
Combining directed acyclic graphs and the change-in-estimate procedure as a novel approach to adjustment-variable selection in epidemiology.BMC Med Res Methodol. 2012 Oct 11;12:156. doi: 10.1186/1471-2288-12-156. BMC Med Res Methodol. 2012. PMID: 23058038 Free PMC article.
-
Using directed acyclic graphs to guide analyses of neighbourhood health effects: an introduction.J Epidemiol Community Health. 2008 Sep;62(9):842-6. doi: 10.1136/jech.2007.067371. J Epidemiol Community Health. 2008. PMID: 18701738 Review.
Cited by
-
Occupational, academic, and personal determinants of wellbeing and psychological distress in residents: results of a survey in Lyon, France.Front Psychol. 2024 May 6;15:1347513. doi: 10.3389/fpsyg.2024.1347513. eCollection 2024. Front Psychol. 2024. PMID: 38770261 Free PMC article.
-
Risk of Early Childhood Caries Estimated by Maternal Dental Caries during Pregnancy: A Retrospective Cohort Study.Eur J Dent. 2024 Feb;18(1):329-340. doi: 10.1055/s-0043-1769896. Epub 2024 Mar 22. Eur J Dent. 2024. PMID: 38518798 Free PMC article.
-
Environmental predictors of Escherichia coli concentration at marine beaches in Vancouver, Canada: a Bayesian mixed-effects modelling analysis.Epidemiol Infect. 2024 Feb 26;152:e38. doi: 10.1017/S0950268824000311. Epidemiol Infect. 2024. PMID: 38403890 Free PMC article.
-
Exploring the Influence of Social Class and Sex on Self-Reported Health: Insights from a Representative Population-Based Study.Life (Basel). 2024 Jan 26;14(2):184. doi: 10.3390/life14020184. Life (Basel). 2024. PMID: 38398693 Free PMC article.
-
Risk of long COVID and associated symptoms after acute SARS-COV-2 infection in ethnic minorities: A nationwide register-linked cohort study in Denmark.PLoS Med. 2024 Feb 20;21(2):e1004280. doi: 10.1371/journal.pmed.1004280. eCollection 2024 Feb. PLoS Med. 2024. PMID: 38377114 Free PMC article.
References
-
- Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiol Camb Mass. 1999. Jan;10(1):37–48. - PubMed
-
[Seminal introduction of DAGs in epidemiologic research.]
-
- Glymour MM. Using causal diagrams to understand common problems in social epidemiology. In: Oakes JM, Kaufman JS, editors. Methods in Social Epidemiology. 2nd ed. San Francisco: Jossey-Bass; 2017. p. 458–92.
-
[Comprehensive and approachable explanation of the practical use of DAGs.]
-
- Hernán MA, Hernández-Díaz S, Werler MM, Mitchell AA. Causal knowledge as a prerequisite for confounding evaluation: An application to birth defects epidemiology. Am J Epidemiol. 2002. Jan 15;155(2):176–84. - PubMed
-
[Landmark convincing argument as to why prior causal knowledge about the ambient biologic or behavioral system, encoded in DAGs, should be used to decide whether a variable is contributing to confounding and therefore must be managed in order to avoid bias. Overturns decades of varying and ill-defined approaches to the evaluation of confounding.]
-
- Didelez V, Sheehan N. Mendelian randomization as an instrumental variable approach to causal inference. Stat Methods Med Res. 2007. Aug 1;16(4):309–30. - PubMed
Publication types
MeSH terms
Grants and funding
LinkOut - more resources
Full Text Sources