Cited by 933Advanced Causal Inference TechniquesStatistical Methods in EpidemiologyPhilosophy of Biology and Scientific Explanation

The DOI matches its Crossref record; 99 of the 108 references it cites were found in a registry or an index.

Confounding and Collapsibility in Causal Inference

Open access

Greenland, Sander, Pearl, Judea, Robins, James M.·1999·Statistical Science

DOI 10.1214/ss/1009211805

Open this paper in CoChat — free account Open PDF

Abstract

Consideration of confounding is fundamental to the design and analysis of studies of causal effects. Yet, apart from confounding in experimental designs, the topic is given little or no discussion in most statistics texts. We here provide an overview of confounding and related concepts based on a counterfactual model for causation. Special attention is given to definitions of confounding, problems in control of confounding, the relation of confounding to exchangeability and collapsibility, and the importance of distinguishing confounding from noncollapsibility.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Integrity signals

Registry

Crossref · 10.1214/ss/1009211805

Cites

99 of 99 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

The central role of the propensity score in observational studies for causal effects
ROSENBAUM, P. R., & RUBIN, D. B. (1983) · Biometrika · doi:10.1093/biomet/70.1.41 · Cited by 31,241
Specification Tests in Econometrics
Hausman, J. A. (1978) · Econometrica · doi:10.2307/1913827 · Cited by 18,820
A Treatise of Human Nature
Hume, D. A. (1739) · Cambridge University Press eBooks · doi:10.1017/cbo9780511620409.040 · Cited by 9,910
The Statistical Analysis of Failure Time Data.
Aitkin, M. A., Kalbfleisch, J. D., & Prentice, R. L. (1981) · Biometrics · doi:10.2307/2530581 · Cited by 9,875
Estimating causal effects of treatments in randomized and nonrandomized studies
Rubin, D. B. (1974) · Journal of Educational Psychology · doi:10.1037/h0037350 · Cited by 9,436
Statistics and Causal Inference
Holland, P. W. (1986) · Journal of the American Statistical Association · doi:10.1080/01621459.1986.10478354 · Cited by 5,293
Discrete Multivariate Analysis: Theory and Practice.
Beniger, J. R., Bishop, Y. M. M., Feinberg, S. E., Holland, P. W., Light, R. J., & Mosteller, F. C. (1975) · Contemporary Sociology A Journal of Reviews · doi:10.2307/2063625 · Cited by 5,119
Design of Experiments
Fisher, R. A. (1936) · BMJ · doi:10.1136/bmj.1.3923.554-a · Cited by 4,980
Causal Diagrams for Epidemiologic Research
Greenland, S., Pearl, J., & Robins, J. M. (1999) · Epidemiology · doi:10.1097/00001648-199901000-00008 · Cited by 3,728
Self-citation
Air Pollution and Hospital Admissions for Heart Disease in Eight U.S. Counties
Schwartz, J. D. (1999) · Epidemiology · doi:10.1097/00001648-199901000-00005 · Cited by 3,045

Flagged first, then by citations

Show all 108 references

9 of the 108 references couldn’t be matched to a known paper.

Research works printed but not linked: 9

Clayton, D. and Hills, M. (1993). Statistical Models in Epidemiology. Oxford Univ. Press.
Feynman, R. P. (1963). Lectures on Physics. Addison-Wesley, Reading, MA.
Gail, M. H. (1986). Adjusting for covariates that have the same distribution in exposed and unexposed cohorts. In Modern Statistical Methods in Chronic Disease Epidemiology (S. H. Moolgavkar, and R. L. Prentice, eds.) 3-18. Wiley, New York.
MacMahon, B. and Pugh, T. F. (1967). Causes and entities of disease. In Preventive Medicine (D. W. Clark and B. MacMahon, eds.) 11-18. Little Brown, Boston.
McKechnie, J. L. (ed.) (1979). Webster's New Twentieth Century Dictionary. Simon and Schuster, New York.
Neyman, J. (1935). Statistical problems in agricultural experimentation (with discussion). J. Roy. Statist. Soc. Suppl. 2 107-180. · DOI
Pearl, J. and Robins, J. M. (1995). Probabilitic evaluation of sequential plans from causal model with hidden variables. In Uncertainty in Artificial Intelligence (P. Besnard and S. Hanks, eds.) 11 444-453. Morgan-Kaufman, San Francisco.
Robins, J. M. (1998). Correction for non-compliance in equivalence trials. Statistics in Medicine 17 269-302. · DOI
Rothman, K. J. and Greenland, S. (1998). Modern Epidemiology, 2nd ed. Lippincott-Raven, Philadelphia.

Sorted by citations

Items 401–500 of 933

A comparison of methods to estimate the survivor average causal effect in the presence of missing data: a simulation study
McGuinness, M. B., Kasza, J., Karahalios, A., Guymer, R. H., Finger, R. P., & Simpson, J. A. (2019) · BMC Medical Research Methodology · doi:10.1186/s12874-019-0874-x · Cited by 13
A Bayesian Approach for Estimating Causal Effects from Observational Data
Pensar, J., Talvitie, T., Hyttinen, A., & Koivisto, M. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i04.5988 · Cited by 13
“Moreover, under the assumption of causal sufﬁciency (i.e., absence of latent confounders), all causal effects can be identiﬁed by back-door adjustment according to the known causal structure (Spirtes, Glymour, and Scheines 1993; Pearl 2009; Greenland, Robins, and Pearl 1999).”
Method
Selection, Confounding, and Attrition Biases in Randomized Controlled Trials of Rehabilitation Interventions
Armijo-Olivo, S., de Oliveira-Souza, A. I. S., Mohamad, N., Carletti, E. M. D. C., Fuentes, J., & Ballenberger, N. (2022) · American Journal of Physical Medicine & Rehabilitation · doi:10.1097/phm.0000000000001947 · Cited by 13
“A confounder is a variable/factor that influences the outcome (dependent variable) and is associated with the exposure (independent variable), which leads to an incorrect association.(67,68) Hence, confounding is a type of bias that may occur when confounders are not considered when attempting to investigate the influ…”
Background
Cardiovascular and mortality risks in older Medicare patients treated with varenicline or bupropion for smoking cessation: an observational cohort study
Graham, D. J., By, K., McKean, S. J., Mosholder, A. D., Kornegay, C. J., Racoosin, J. A., Young, J. C., Levenson, M. S., MaCurdy, T. E., Worrall, C. M., et al. (2014) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.3678 · Cited by 13
Exhaustion, Automation, Theory, and Confounding
Joffe, M. M. (2009) · Epidemiology · doi:10.1097/ede.0b013e3181a82501 · Cited by 13
Response: Defining and estimating causal effects
Maldonado, G., & Greenland, S. (2002) · International Journal of Epidemiology · doi:10.1093/ije/31.2.435 · Cited by 8
Self-citation
Assessing the Celerity of Arrest on 3-Year Recidivism Patterns in a Sample of Criminal Defendants
Zettler, H. R., Morris, R. G., Piquero, A. R., & Cardwell, S. M. (2015) · Journal of Criminal Justice · doi:10.1016/j.jcrimjus.2015.05.003 · Cited by 13
Funding and productivity: Does winning grants affect the scientific productivity of recipients? Evidence from the social sciences and economics
Ding, Y., & Moreira, F. (2025) · Scientometrics · doi:10.1007/s11192-025-05277-3 · Cited by 20
“Generally, exchangeability or ignorability (Greenland et al., 1999; Robins and Green-land, 1992) means that the treatment effect should not be significantly different from the control group when individuals in the treated group do not receive the intervention.”
Best practices for moving from correlation to causation in ecological research
Correia, H., Dee, L. E., Byrnes, J. E. K., Fieberg, J., Fortin, M., Glymour, C., Runge, J., Shipley, B., Shpitser, I., Siegel, K., et al. (2026) · Nature Communications · doi:10.1038/s41467-026-69878-z · Cited by 19
Heterogeneous treatment effect estimation for observational data using model-based forests
Dandl, S., Bender, A., & Hothorn, T. (2024) · Statistical Methods in Medical Research · doi:10.1177/09622802231224628 · Cited by 12
A Potential Outcomes Approach to Selection Bias
Kenah, E. (2023) · Epidemiology · doi:10.1097/ede.0000000000001660 · Cited by 12
“7 The structural approach of Hernán et al. 8 showed that selection bias occurs when conditioning on selection into the study opens a noncausal path from an exposure X to a disease D in a causal directed acyclic graph (DAG) 9,10 for the eligible population.”
Stratified Restricted Mean Survival Time Model for Marginal Causal Effect in Observational Survival Data
Ni, A., Lin, Z., & Lü, B. (2021) · Annals of Epidemiology · doi:10.1016/j.annepidem.2021.09.016 · Cited by 12
Treatment Learning Causal Transformer for Noisy Image Classification
Yang, C.-H. H., Hung, D. I.-T., Liu, Y.-C., & Chen, P. (2023) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv56688.2023.00608 · Cited by 12
“For a binary outcome, it is defined as: given by: The above mentioned metric cannot be properly estimated if there are confounding variables in the system, which will introduce bias (Greenland et al., 1999).”
BackgroundInfluential
Use of auxiliary covariates in estimating a biomarker-adjusted treatment effect model with clinical trial data
Zhang, Z., Qu, Y., Zhang, B., Nie, L., & Soon, G. G. (2013) · Statistical Methods in Medical Research · doi:10.1177/0962280213515572 · Cited by 12
“For the ratio-based model (5), the estimating function (10) can be further augmented as”
BackgroundMethod
Comparing external and internal validation methods in correcting outcome misclassification bias in logistic regression: A simulation study and application to the case of postsurgical venous thromboembolism following total hip and knee arthroplasty
Ni, J., Dasgupta, K., Kahn, S. R., Talbot, D. C., Lefebvre, G., Lix, L. M., Berry, G. K., Burman, M. L., Dimentberg, R., Laflamme, Y. G., et al. (2018) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.4693 · Cited by 12
“Noncollapsibility is a phenomenon of discrepancy between the marginal OR and the conditional OR even in the absence of confounding.(59) To eliminate issues due to noncollapsibility, collapsible measures such as the risk ratio and risk difference are recommended to detect confounding bias in cohort studies.”
Background
Confounding and Interaction
Pearce, N. E., & Greenland, S. (2005) · Springer eBooks · doi:10.1007/978-3-540-26577-1_10 · Cited by 16
Self-citation
CONFOUNDING, HOMOGENEITY AND COLLAPSIBILITY FOR CAUSAL EFFECTS IN EPIDEMIOLOGIC STUDIES
Geng, Z., Guo, J., Lau, T. S., & Fung, W. (2001) · Cited by 12
Mixing of Confounding and Non-Collapsibility: A Notable Deficiency of the Odds Ratio
Pang, M., Kaufman, J. S., & Platt, R. W. (2013) · The American Journal of Cardiology · doi:10.1016/j.amjcard.2012.09.002 · Cited by 12
Further Refinements to the Organizational Schema for Causal Effects
Suzuki, E., & Yamamoto, E. (2014) · Epidemiology · doi:10.1097/ede.0000000000000114 · Cited by 12
Principles of Systematic Reviewing
Higgins, J. P. T., Smith, G. D., Altman, D. G., & Egger, M. (2022) · Wiley eBooks · doi:10.1002/9781119099369.ch2 · Cited by 11
Is the Product Method More Efficient Than the Difference Method for Assessing Mediation?
Cheng, C., Spiegelman, D. L., & Li, F. (2022) · American Journal of Epidemiology · doi:10.1093/aje/kwac144 · Cited by 11
Robust double machine learning model with application to omics data
Wang, X., Liu, Y., Qin, G., & Yu, Y. (2024) · BMC Bioinformatics · doi:10.1186/s12859-024-05975-4 · Cited by 11
Causal Mediation Analysis: A Summary‐Data Mendelian Randomization Approach
Lin, S., Lin, S., Ge, T., Chen, C., & Lin, K. (2025) · Statistics in Medicine · doi:10.1002/sim.10317 · Cited by 11
Commentary
Hoggatt, K. J., & Greenland, S. (2013) · Epidemiology · doi:10.1097/ede.0000000000000023 · Cited by 11
“the same noncollapsibility problem arises when comparing a total population odds ratio to average subgroup odds ratios or when using person-time rate ratios or differences instead of odds to measure effects.(19,20) It does not arise, however, with risk differences or risk ratios.”
BackgroundMethodInfluentialSelf-citation
A Theorem at the Core of Colliding Bias
Shahar, D. J., & Shahar, E. (2017) · The International Journal of Biostatistics · doi:10.1515/ijb-2016-0055 · Cited by 11
Advanced Modelling Strategies: Challenges and pitfalls in robust causal inference with observational data
Tennant, P. W. G., Arnold, K. F., Berrie, L., Ellison, G. T. H., & Gilthorpe, M. S. (2017) · White Rose Research Online (University of Leeds, The University of Sheffield, University of York) · Cited by 11
Affirmative Action and Women in Uganda’s Public University Education
Odaga, G. (2020) · Higher Education Policy · doi:10.1057/s41307-020-00189-8 · Cited by 11
An underappreciated misclassification mechanism: implications of nondifferential dependent misclassification of covariate and exposure
Brennan, A. T., Getz, K., Brooks, D. R., & Fox, M. P. (2021) · Annals of Epidemiology · doi:10.1016/j.annepidem.2021.02.007 · Cited by 11
Marginal and Conditional Confounding Using Logits
Karlson, K. B., Popham, F., & Holm, A. (2021) · Sociological Methods & Research · doi:10.1177/0049124121995548 · Cited by 13
Marginal Structural Models to Estimate Causal Effects of Right-to-Carry Laws on Crime
van der Wal, W. M. (2022) · Statistics and Public Policy · doi:10.1080/2330443x.2022.2120136 · Cited by 11
Inflammation contributes to cardiovascular risk in patients receiving statin therapy
Tardif, J., & Samuel, M. (2023) · The Lancet · doi:10.1016/s0140-6736(23)00454-3 · Cited by 11
Adjusting for Confounding in Early Postlaunch Settings
Schmidt, A. F., Klungel, O. H., & Groenwold, R. H. H. (2015) · Epidemiology · doi:10.1097/ede.0000000000000388 · Cited by 11
“…Or in the test data to 0.30, 0.70, 1.00, 1.50, and 3.00. note that we focus exclusively on power because the conditional Or will differ from the marginal Or due to noncollapsibility 24 ; other performance metrics (see below) are presented in the etables 1–6 (http://links.lww.com/eDe/a972).”
Influential
Pathways between Socioeconomic Disadvantage and Childhood Growth in the Scottish Longitudinal Study, 1991–2001
Silverwood, R. J., Williamson, L., Grundy, E. M. D., & Stavola, B. L. D. (2016) · PLoS ONE · doi:10.1371/journal.pone.0164853 · Cited by 11
“Even if the above assumptions hold and there is no intermediate confounding then the ‘direct effect’ and ‘indirect effect’ estimated under the traditional mediation analysis are not consistent estimates of the NDE and NIE due to the non-collapsibility of the OR [62].”
Background
Towards non-conventional methods of designing register-based epidemiological studies: An application to pediatric research
Gong, T., Brew, B. K., Sjölander, A., & Almqvist, C. (2017) · Scandinavian Journal of Public Health · doi:10.1177/1403494817702339 · Cited by 11
“[2] Greenland S, Robins JM and Pearl J.”
Background
Retrospectively reported childhood adversity is associated with asthma and chronic bronchitis, independent of mental health
Sheikh, M. A. (2018) · Journal of Psychosomatic Research · doi:10.1016/j.jpsychores.2018.09.007 · Cited by 11
“Previous literature has repeatedly argued that neither confounding 250 nor mediation should not be assessed using odds ratios [60, 72-76], because an odds ratio is 251 not a collapsible measure [77] and attenuations in odds ratios do not necessarily correspond 252”
Background
Improved inference for fixed‐effects meta‐analysis of 2 × 2 tables
Li, K. Q., & Rice, K. M. (2020) · Research Synthesis Methods · doi:10.1002/jrsm.1401 · Cited by 11
“But homogeneity is unlikely to hold in practice; differences in study population characteristics and research protocols can induce unequal effect sizes across studies, particularly as odds ratios are non-collapsible across studies.(5) Small-study effects(6) can also induce heterogeneity.”
Background
Innovation on Machine Learning in Healthcare Services—An Introduction
Pattnayak, P., & Panda, A. R. (2021) · Studies in computational intelligence · doi:10.1007/978-981-33-4698-7_1 · Cited by 10
Contextualized Fairness for Recommender Systems in Premium Scenarios
Li, Y., Hedia, M.-L., Ma, W., Lu, H., Zhang, M., Liu, Y., & Ma, S. (2021) · Big Data Research · doi:10.1016/j.bdr.2021.100300 · Cited by 13
A Review of the Techniques Used to Control Confounding Bias and How Spatiotemporal Variation Can Be Controlled in Environmental Impact Studies
Hatami, R. (2019) · Water, Air and Soil Pollution · doi:10.1007/s11270-019-4150-9 · Cited by 10
Claims and confounds in economic experiments
Zizzo, D. J. (2013) · Journal of Economic Behavior & Organization · doi:10.1016/j.jebo.2013.05.006 · Cited by 11
“See Davidson and MacKinnon (1993) for a standard econometric analysis and Greenland et al. (1999) for an example of analysis from a statistician’s perspective on confounding.”
Method
Leukemia clusters in upstate New York: how adding covariates changes the story
Ahrens, C., Altman, N., Casella, G., Eaton, M. M., Hwang, J. T. G., STAUDENMAYER, J. W., & Ştefănescu, C. (2001) · Environmetrics · doi:10.1002/env.490 · Cited by 11
On adjustment for auxiliary covariates in additive hazard models for the analysis of randomized experiments
Vansteelandt, S., Martinussen, T., & Tchetgen, E. J. T. (2013) · Biometrika · doi:10.1093/biomet/ast045 · Cited by 11
“While justified for linear regression (Cox & McCullagh, 1982), the situation is more subtle for nonlinear regression models, where covariate adjustment may change the meaning and thus the magnitude of the treatment effect, even when the treatment is randomly assigned (Greenland et al., 1999).”
Confounding and Interaction
Pearce, N. E., & Greenland, S. (2013) · Springer eBooks · doi:10.1007/978-0-387-09834-0_10 · Cited by 12
Self-citation
Is There an Absence of Theory in Social Epidemiology? The Authors Respond to Muntaner
Cooper, R., & Kaufman, J. S. (1999) · American Journal of Epidemiology · doi:10.1093/oxfordjournals.aje.a009971 · Cited by 11
MARGINAL STRUCTURAL QUANTILE MODELS FOR LONGITUDINAL OBSERVATIONAL STUDIES WITH TIME-VARYING TREATMENT
Hogan, J. W., & Lee, J. Y. (2004) · Cited by 11
On the Composition and Limitations of Publicly Available COVID-19 X-Ray Imaging DatasetsPreprint
Cruz, B. G. S., Sölter, J., Bossa, M. N., & Husch, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.11572 · Cited by 11
Editorial for Sports Performance section of Journal of Sports Sciences : Moving match analysis forward – Guidelines and perspectives
Memmert, D. (2025) · Journal of Sports Sciences · doi:10.1080/02640414.2025.2535850 · Cited by 15
“From a statistical perspective, if a covariate is capable of affecting both the independent variable and the dependent variable of a study design, it has the potential to confound the effect between the two (Greenland et al., 1999).”
Causal inference for quantile treatment effects
Sun, S., Moodie, E., & Nešlehová, J. G. (2021) · Environmetrics · doi:10.1002/env.2668 · Cited by 21
“As noted by Greenland et al. (1999), counterfactual (or potential outcomes) approaches to causal inference emphasize the importance of randomization (Greenland, 1990; Greenland & Robins, 1986; Robins, 1986; Rubin, 1978; Rubin, 1990; Rubin, 1991).”
Background
Racial/ethnic differences in post-acute sequelae of SARS-CoV-2 in children and adolescents in the United States
Zhang, D., Zhang, B., Wu, Q., Zhou, T., Tong, J., Lu, Y., Chen, J., Wang, H., Chisolm, D. J., Jhaveri, R. R., et al. (2025) · Nature Communications · doi:10.1038/s41467-024-55273-z · Cited by 10
A Novel Criterion for Selecting Covariates
Berger, V. W. (2005) · Drug Information Journal · doi:10.1177/009286150503900303 · Cited by 10
A two-stage strategy to accommodate general patterns of confounding in the design of observational studies
Haneuse, S., Schildcrout, J. S., & Gillen, D. L. (2011) · Biostatistics · doi:10.1093/biostatistics/kxr044 · Cited by 10
Covariate balance for no confounding in the sufficient-cause model
Suzuki, E., Tsuda, T., & Yamamoto, E. (2017) · Annals of Epidemiology · doi:10.1016/j.annepidem.2017.11.005 · Cited by 10
Is the effect of Mediterranean diet on hip fracture mediated through type 2 diabetes mellitus and body mass index?
Mitchell, A. J., Fall, T., Melhus, H., Wolk, A., Michaëlsson, K., & Byberg, L. (2020) · International Journal of Epidemiology · doi:10.1093/ije/dyaa239 · Cited by 12
“Although the effect measures presented within this paper are conditional on confounders, comparison of the total effect odds ratios with the controlled direct effect estimates are still hampered due to the non-collapsible nature of the odds ratio.(33,34) Adding a true mediator to a regression model will lead to attenu…”
Another Look at the Method of Y-Standardization in Logit and Probit Models
Karlson, K. B. (2014) · Journal of Mathematical Sociology · doi:10.1080/0022250x.2014.897950 · Cited by 10
Distinguishing between primary and secondary racial identification in analyses of health disparities of a multiracial population in Hawaii
Wey, A. R., Davis, J. W., Taira, D. A., & Sentell, T. L. (2016) · Ethnicity and Health · doi:10.1080/13557858.2016.1263284 · Cited by 10
“The odds ratio for a covariate of interest can change by including a covariate associated with the outcome but unrelated to the covariate of interest, which is due to the non-collapsibility of odds ratios (Greenland, Robins, and Pearl 1999).”
Background
Propensity scores – A brief introduction for resuscitation researchers
Andersen, L. W., & Kurth, T. (2018) · Resuscitation · doi:10.1016/j.resuscitation.2018.02.003 · Cited by 10
Evaluation of Causal Effects and Local Structure Learning of Causal Networks
Geng, Z., Liu, Y., Liu, C., & Miao, W. (2018) · Annual Review of Statistics and Its Application · doi:10.1146/annurev-statistics-030718-105312 · Cited by 10
Slicing: A sustainable approach to structuring samples for analysis in long‐term studies
van Lieshout, S. H. J., Froy, H., Schroeder, J., Burke, T. A., Simons, M. J. P., & Dugdale, H. L. (2020) · Methods in Ecology and Evolution · doi:10.1111/2041-210x.13352 · Cited by 10
“This approach can overcome the experimental design and statistical issues with cross-classification and confounding variables in long-term studies (Gelman & Hill, 2006; Greenland et al., 1999; Schielzeth & Nakagawa, 2013), by bridging batches and clusters together.”
Background
Impact of time-varying cumulative bevacizumab exposures on survival: re-analysis of data from randomized clinical trial in patients with metastatic colo-rectal cancer
Guilloteau, A., Abrahamowicz, M., Boussari, O., Jooste, V., Aparicio, T., Quantin, C., Malicot, K. L., & Binquet, C. (2021) · BMC Medical Research Methodology · doi:10.1186/s12874-020-01202-9 · Cited by 10
The Uses of Propensity Scores in Randomized Controlled Trials
Loux, T. M., & Huang, Y. (2023) · Observational Studies · doi:10.1353/obs.2023.0007 · Cited by 10
“In addition to concerns about over-fitting and p-hacking, covariate-adjusted effect estimates target a different parameter - the conditional average treatment effect - than crude comparisons (Greenland, Robins, and Pearl, 1999).”
BackgroundInfluential
Conditions for Non‐confounding and Collapsibility without Knowledge of Completely Constructed Causal Diagrams
Geng, Z., & LI, G. (2001) · Scandinavian Journal of Statistics · doi:10.1111/1467-9469.00087 · Cited by 10
“Recently Greenland et al. (1999) provided an overview of concepts of confounding, and they emphasize that more thorough discussion of confounding should be included in basic statistics education. Miettinen & Cook (1981) obtained inductively a criterion for detecting confounding from various examples in epidemiologic s…”
BackgroundMethodInfluential
Exchangeability and regression models
McCullagh, P. J. (2005) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780198566540.003.0005 · Cited by 10
Trends of child undernutrition in rural Ecuadorian communities with differential access to roads, 2004–2013
Lopez, V. K., Dombecki, C., Trostle, J., Mogrovejo, P., Castro, N. P., Cevallos, W., Goldstick, J. E., Jones, A. D., & Eisenberg, J. N. S. (2018) · Maternal and Child Nutrition · doi:10.1111/mcn.12588 · Cited by 11
Methodological Issues in Population-Based Studies of Multigenerational Associations
McGee, G., Perkins, N. J., Mumford, S. L., Kioumourtzoglou, M., Weisskopf, M. G., Schildcrout, J. S., Coull, B. A., Schisterman, E. F., & Haneuse, S. (2020) · American Journal of Epidemiology · doi:10.1093/aje/kwaa125 · Cited by 11
AMDANet: Attention-Driven Multi-Perspective Discrepancy Alignment for RGB-Infrared Image Fusion and Segmentation
Zhong, H., Tang, F., Chen, Z., Chang, H. J., & Gao, Y. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00991 · Cited by 9
“Under the influence of nonlinear factors such as regularization and activation functions, the encoder in the model often exhibits empirical feature bias toward specific modalities [9, 35, 43, 47].”
Nonparametric Adjustment Techniques for Binary Covariates
Berger, V. W. (2005) · Biometrical Journal · doi:10.1002/bimj.200410100 · Cited by 9
A rank-based approach to design and analysis of pretest-posttest randomized trials, with application to COVID-19 ordinal scale data
Zou, G., Smith, E. D., Zou, L., Qiu, S., & Shu, D. (2023) · Contemporary Clinical Trials · doi:10.1016/j.cct.2023.107085 · Cited by 10
Dependence of Confounding on the Target Population: A Modification of Causal Graphs to Account for Co-Action
Flanders, W. D., Johnson, C. Y., Howards, P. P., & Greenland, S. (2011) · Annals of Epidemiology · doi:10.1016/j.annepidem.2011.05.002 · Cited by 9
“Understanding of confounding and biologic interaction has evolved along with advances in causal analysis and reasoning (1, 2).”
BackgroundSelf-citation
Conditions for confounding of interactions
Liu, A., Abrahamowicz, M., & Siemiatycki, J. (2015) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.3924 · Cited by 9
Making Causal Inferences About Macrosocial Factors as a Basis for Public Health Policies
Kaufman, J. S. (2007) · Springer eBooks · doi:10.1007/978-0-387-70812-6_17 · Cited by 9
“Therefore, it is possible, for example, that a true summary causal effect that is null (e.g., a causal risk difference equal to 0) could arise because half of the population is harmed and half the population is benefited ( Greenland, Robins, & Pearl, 1999 ).”
Background
Estimating Counterfactual Risk Under Hypothetical Interventions in the Presence of Competing Events: Crystalline Silica Exposure and Mortality From 2 Causes of Death
Neophytou, A. M., Picciotto, S., Brown, D., Gallagher, L. G., Checkoway, H., Eisen, E. A., & Costello, S. (2018) · American Journal of Epidemiology · doi:10.1093/aje/kwy077 · Cited by 9
Combining Effect Estimates Across Cohorts and Sufficient Adjustment Sets for Collaborative Research
Hamra, G. B., Lesko, C. R., Buckley, J. P., Jensen, E. T., Tancredi, D. J., Lau, B. M., Hertz‐Picciotto, I., & Outcomes, O. B. O. P. C. F. E. I. O. C. H. (2021) · Epidemiology · doi:10.1097/ede.0000000000001336 · Cited by 9
Statistical Modelling in the Age of Data Science
Vansteelandt, S. (2021) · Observational Studies · doi:10.1353/obs.2021.0013 · Cited by 9
“…question at stake; note, for instance, the generally poor understanding that the meaning of many regression coeﬃcients changes due to non-collapsibility as one adjusts for more and more covariates (even in the absence of confounding, selection or mediation eﬀects) (Greenland et al., 1999).”
Background
Do Private Providers Initiate Anti-Tuberculosis Therapy on the Basis of Chest Radiographs? A Standardised Patient Study in Urban India
Svadzian, A., Daniels, B., Sulis, G., Das, J., Daftary, A., Kwan, A. T., Das, V., Das, R., & Pai, M. (2022) · The Lancet Regional Health - Southeast Asia · doi:10.1016/j.lansea.2023.100152 · Cited by 10
Testing concordance of instrumental variable effects in generalized linear models with application to Mendelian randomization
Dai, J. Y., Chan, K. C. G., & Hsu, L. (2014) · Statistics in Medicine · doi:10.1002/sim.6217 · Cited by 9
“One criticism of the structural equation approach is that the effect measure is defined within the strata of U that are unobserved, thus harder to interpret, particularly for odds ratios due to the issue of noncollapsibility [30].”
Background
The impact of confounder selection in propensity scores when applied to prospective cohort studies in pregnancy
Xu, R., Hou, J., & Chambers, C. (2018) · Reproductive Toxicology · doi:10.1016/j.reprotox.2018.04.003 · Cited by 9
Assessing the impact of case-mix heterogeneity in individual participant data meta-analysis: Novel use of I 2 statistic and prediction interval
Vo, T., Porcher, R., & Vansteelandt, S. (2020) · Research Methods in Medicine & Health Sciences · doi:10.1177/2632084320957207 · Cited by 9
Teaching: the role of active manipulation of three-dimensional scatter plots in understanding the concept of confounding
Busstra, C. M., Hårtog, R. J. M., & Veer, P. V. ’. (2005) · Epidemiologic Perspectives & Innovations · doi:10.1186/1742-5573-2-6 · Cited by 9
“(See Newman or Greenland for more fundamental definitions of confounding [2,3].”
Re: "Estimating the Relative Risk in Cohort Studies and Clinical Trials of Common Outcomes"
Karp, I. (2014) · American Journal of Epidemiology · doi:10.1093/aje/kwt435 · Cited by 9
Commentary: Causal Effects in Mediation Modeling: An Introduction with Applications to Latent Variables
Coman, E., Thoemmes, F. J., & Fifield, J. (2017) · Frontiers in Psychology · doi:10.3389/fpsyg.2017.00151 · Cited by 9
Propensity score matching mitigates risk of faulty inferences in observational studies of effectiveness of restoration trials
Kluender, C. R., Germino, M. J., & Anthony, C. R. (2024) · Journal of Applied Ecology · doi:10.1111/1365-2664.14638 · Cited by 8
Transportability of model‐based estimands in evidence synthesis
Remiro‐Azócar, A. (2024) · Statistics in Medicine · doi:10.1002/sim.10111 · Cited by 9
“On the aforementioned additive scale, the conditional average treatment effect estimand for the trial is defined as: 44”
Estimating Time‐Varying Exposure Effects Through Continuous‐Time Modelling in Mendelian Randomization
Tian, H., Patel, A., & Burgess, S. (2024) · Statistics in Medicine · doi:10.1002/sim.10222 · Cited by 10
Effect modification and non-collapsibility together may lead to conflicting treatment decisions: A review of marginal and conditional estimands and recommendations for decision-making
Phillippo, D., Remiro‐Azócar, A., Heath, A., Baio, G., Dias, S., Ades, A. E., & Welton, N. J. (2025) · Research Synthesis Methods · doi:10.1017/rsm.2025.2 · Cited by 8
Inverse Probability of Treatment Weighting and Confounder Missingness in Electronic Health Record-based Analyses: A Comparison of Approaches Using Plasmode Simulation
Vader, D. T., Mamtani, R., Li, Y., Griffith, S. D., Calip, G. S., & Hubbard, R. A. (2023) · Epidemiology · doi:10.1097/ede.0000000000001618 · Cited by 8
“13,14 However, there is limited guidance on which missing data approaches perform adequately (i.e., with minimal bias) and under what conditions when conducting comparative effectiveness research using IPTWs.”
Background
Synthesis estimators for transportability with positivity violations by a continuous covariate
Zivich, P. N., Edwards, J. K., Shook‐Sa, B. E., Lofgren, E., Lessler, J., & Cole, S. R. (2024) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1093/jrsssa/qnae084 · Cited by 8
Marginal versus conditional causal effects
Mohammad, K., Nazari, S. S. H., Mansournia, N., & Mansournia, M. A. (2015) · DOAJ (DOAJ: Directory of Open Access Journals) · Cited by 8
The Charlson Index Is Insufficient to Control for Comorbidities in a National Trauma Registry
Renson, A., & Bjurlin, M. A. (2018) · Journal of Surgical Research · doi:10.1016/j.jss.2018.07.072 · Cited by 8
Understanding Etiologic Pathways Through Multiple Sequential Mediators: An Application in Perinatal Epidemiology
Ananth, C. V., & Loh, W. W. (2022) · Epidemiology · doi:10.1097/ede.0000000000001518 · Cited by 8
A Vegetable, Oil, and Fruit Dietary Pattern in Late Pregnancy is Linked to Reduced Risks of Adverse Birth Outcomes in a Predominantly Low-Income Hispanic and Latina Pregnancy Cohort
Maldonado, L. E., Farzan, S. F., Toledo‐Corral, C. M., Dunton, G. F., Habre, R., Eckel, S. P., Johnson, M. R., Yang, T., Grubbs, B. H., Lerner, D. L., et al. (2022) · Journal of Nutrition · doi:10.1093/jn/nxac209 · Cited by 12
The impact of unmeasured baseline effect modification on estimates from an inverse probability of treatment weighted logistic model
Delaney, J. A., Platt, R. W., & Suissa, S. (2009) · European Journal of Epidemiology · doi:10.1007/s10654-009-9341-z · Cited by 8
“This feature of the OR and hazard ratio, sometimes referred to as non-collapsibility, is often considered a disadvantage of the OR [17, 18].”
Background
Relative contributions of biomarkers in Alzheimer’s disease
Haight, T. J., & Jagust, W. (2012) · Annals of Epidemiology · doi:10.1016/j.annepidem.2012.09.004 · Cited by 8
Misspecified poisson regression models for large‐scale registry data: inference for ‘largenand smallp’
Grøn, R., Gerds, T. A., & Andersen, P. K. (2015) · Statistics in Medicine · doi:10.1002/sim.6755 · Cited by 8
“This can be seen as an example of Simpson’s paradox and is also related to the non-collapsibility of a hazard ratio [23, 24].”
Background
A Causal Inference Framework for Cancer Cluster Investigations using Publicly Available Data
Nethery, R. C., Yang, Y., Brown, A. J., & Dominici, F. (2020) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/rssa.12567 · Cited by 8
Increasing Power in Phase III Oncology Trials With Multivariable Regression: An Empirical Assessment of 535 Primary End Point Analyses
Sherry, A. D., Passy, A. H., McCaw, Z. R., Jaoude, J. A., Lin, T. A., Kouzy, R., Miller, A. M., Kupferman, G. S., Beck, E. J., Msaouel, P., et al. (2024) · JCO Clinical Cancer Informatics · doi:10.1200/cci.24.00102 · Cited by 7
Graph Fairness via Authentic Counterfactuals: Tackling Structural and Causal Challenges
Wang, Z., Yin, Z., Liu, F., Liu, Z., Lisetti, C., Yu, R., Wang, S., Liu, J., Ganapati, S., Zhou, S., et al. (2025) · ACM SIGKDD Explorations Newsletter · doi:10.1145/3715073.3715081 · Cited by 7
“Despite these advancements, these methods often produce potentially unauthentic coun-terfactuals [9, 5].”
Impact of postprogression therapies on overall survival: Recommendations from the 2023 kidney cancer association think tank meeting
Berg, S. A., Rosa, S. L., Zhang, T., Pierorazio, P. M., Albiges, L. K., Beckermann, K. E., Campbell, M. T., Carlo, M. I., Coleman, K., George, D. J., et al. (2024) · Urologic Oncology Seminars and Original Investigations · doi:10.1016/j.urolonc.2024.10.022 · Cited by 8
Understanding the effectiveness of the Comirnaty monovalent and bivalent vaccines during the Winter Coronavirus (COVID-19) Infection Study
Ward, T., Paton, R. S., Overton, C. E., Mellor, J., Aziz, N. I. B. A., Charlett, A., & Fyles, M. (2025) · Journal of Infection · doi:10.1016/j.jinf.2025.106461 · Cited by 7
Mathematical expression and sampling issues of treatment contrasts: Beyond significance testing and meta-analysis to clinically useful research synthesis
Krause, M. S. (2016) · Psychotherapy Research · doi:10.1080/10503307.2016.1222459 · Cited by 7
Accounting for Latent Covariates in Average Effects from Count Regressions
Kiefer, C., & Mayer, A. (2020) · Multivariate Behavioral Research · doi:10.1080/00273171.2020.1751027 · Cited by 7
“For an illustration of non-collapsibility for logistic regression, see Robinson and Jewell (1991) for a general overview, see Greenland et al. (1999). Thus, in”

Authors: Greenland, S., Pearl, J., & Robins, J. M.
ORCID: Greenland, Sander
Robins, James M.
Venue: Statistical Science
Publisher: Institute of Mathematical Statistics
Published: 1999-02-01
Volume: 14
Issue: 1
Pages: 29-46
ISSN: 0883-4237 (ISSN-L)
2168-8745
DOI: 10.1214/ss/1009211805
OpenAlex: W1996016456
Cited by: 933
References: 108
Citations by year: 2026: 43 · 2025: 47 · 2024: 61 · 2023: 53 · 2022: 65 · 2021: 75 · 2020: 48 · 2019: 40 · 2018: 40 · 2017: 36 · 2016: 35 · 2015: 44 · 2014: 52 · 2013: 47 · 2012: 47
FWCI: 13.67
Type: journal article
Language: en
Open access: bronze
doi.org
Locations: Statistical Science (publishedVersion)
Project Euclid (Cornell University) (submittedVersion)
Topics: Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
Social Sciences › Arts and Humanities › History and Philosophy of Science › Philosophy of Biology and Scientific Explanation
Keywords: confounding, collapsibility, counterfactual modeling, causal inference, exchangeability, causal effects, observational studies

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Confounding and Collapsibility in Causal Inference

Abstract

Overview

Integrity signals

Verification

Claims the paper makes

Claims other papers make about this one

Self-check

Citations

Cites

Cited by

Related

Metadata

Record

About these checks