Assumptions of IV methods for observational epidemiology

Instrumental variable (IV) methods are becoming increas- ingly popular as they seem to offer the only viable way to overcome the problem of unobserved confounding in observational studies. However, some attention has to be paid to the details, as not all such methods target the same causal parameters and some rely on more restrictive parametric assumptions than others. We therefore discuss and contrast the most common IV approaches with relevance to typical applications in observational epidemiology. Further, we illustrate and compare the asymptotic bias of these IV estimators when underlying assumptions are violated in a numerical study. One of our conclusions is that all IV methods encounter problems in the presence of effect modification by unobserved confounders. Since this can never be ruled out for sure, we recommend that practical applications of IV estimators be accompa- nied routinely by a sensitivity analysis.

[1]  Blai Bonet,et al.  Instrumentality Tests Revisited , 2001, UAI.

[2]  A. Dawid Causal Inference without Counterfactuals , 2000 .

[3]  Roger M Harbord,et al.  Meta-analyses of observational and genetic association studies of folate intakes or levels and breast cancer risk. , 2006, Journal of the National Cancer Institute.

[4]  J. S. Silva,et al.  Endogeneity in Count Data Models: An Application to Demand for Health Care , 1997 .

[5]  A Yoshida,et al.  Molecular abnormality of an inactive aldehyde dehydrogenase variant commonly found in Orientals. , 1984, Proceedings of the National Academy of Sciences of the United States of America.

[6]  J. Pearl Causal diagrams for empirical research , 1995 .

[7]  George Davey Smith,et al.  Capitalizing on Mendelian randomization to assess the effects of treatments. , 2007, Journal of the Royal Society of Medicine.

[8]  J. Robins Correcting for non-compliance in randomized trials using structural nested mean models , 1994 .

[9]  M Alan Brookhart,et al.  American Journal of Epidemiology Practice of Epidemiology Instrumental Variable Analysis for Estimation of Treatment Effects with Dichotomous Outcomes , 2022 .

[10]  Charles Kooperberg,et al.  Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trial. , 2002, JAMA.

[11]  Roland R. Ramsahai,et al.  Causal inference with instruments and other supplementary variables , 2008 .

[12]  J. Paul Leigh,et al.  Instrumental variables technique: cigarette price provided better estimate of effects of smoking on SF-12. , 2004, Journal of clinical epidemiology.

[13]  Antonio Forcina Causal effects in the presence of non compliance: a latent variable interpretation , 2006 .

[14]  L. Smeeth,et al.  Limits to causal inference based on Mendelian randomization: a comparison with randomized controlled trials. , 2006, American journal of epidemiology.

[15]  D A Savitz,et al.  Econometric approaches to epidemiologic data: relating endogeneity and unobserved heterogeneity to confounding. , 1997, Annals of epidemiology.

[16]  Keith R Abrams,et al.  An integrated approach to the meta-analysis of genetic association studies using Mendelian randomization. , 2004, American journal of epidemiology.

[17]  Judea Pearl,et al.  Causal inference from indirect experiments , 1995, Artif. Intell. Medicine.

[18]  David J. Spiegelhalter,et al.  Probabilistic Networks and Expert Systems , 1999, Information Science and Statistics.

[19]  J. Angrist,et al.  Two-Stage Least Squares Estimation of Average Causal Effects in Models with Variable Treatment Intensity , 1995 .

[20]  N. Sheehan,et al.  Mendelian randomisation: Why epidemiology needs a formal language for causality , 2007 .

[21]  J. Pearl,et al.  Causal diagrams for epidemiologic research. , 1999, Epidemiology.

[22]  Debbie A Lawlor,et al.  Cardiovascular risk and hormone replacement therapy , 2006, Current opinion in obstetrics & gynecology.

[23]  Steffen L. Lauritzen,et al.  Causal Inference from Graphical Models , 2001 .

[24]  J. Angrist,et al.  Identification and Estimation of Local Average Treatment Effects , 1995 .

[25]  Ap Dawid,et al.  Causal inference using influence diagrams: The problem of partial compliance (with Discussion) , 2003 .

[26]  L. Cardon,et al.  Population stratification and spurious allelic association , 2003, The Lancet.

[27]  Shah Ebrahim,et al.  Association of C-Reactive Protein With Blood Pressure and Hypertension: Life Course Confounding and Mendelian Randomization Tests of Causality , 2005, Arteriosclerosis, thrombosis, and vascular biology.

[28]  M. Hernán A definition of causal effect for epidemiological research , 2004, Journal of Epidemiology and Community Health.

[29]  Glenn Shafer,et al.  Comments on "Causal Inference without Counterfactuals" by A.P. Dawid , 1999 .

[30]  L. Smeeth,et al.  For Personal Use. Only Reproduce with Permission from Elsevier Ltd Homocysteine and Stroke: Evidence on a Causal Link from Mendelian Randomisation , 2022 .

[31]  Joshua D. Angrist,et al.  Identification of Causal Effects Using Instrumental Variables , 1993 .

[32]  A. Dawid Influence Diagrams for Causal Modelling and Inference , 2002 .

[33]  S. Lewis,et al.  Alcohol, ALDH2, and Esophageal Cancer: A Meta-analysis Which Illustrates the Potentials and Limitations of a Mendelian Randomization Approach , 2005, Cancer Epidemiology Biomarkers & Prevention.

[34]  J. Mullahy Instrumental-Variable Estimation of Count Data Models: Applications to Models of Cigarette Smoking Behavior , 1997, Review of Economics and Statistics.

[35]  J. Robins,et al.  Instruments for Causal Inference: An Epidemiologist's Dream? , 2006, Epidemiology.

[36]  David A. Jaeger,et al.  Problems with Instrumental Variables Estimation when the Correlation between the Instruments and the Endogenous Explanatory Variable is Weak , 1995 .

[37]  D. Rubin Estimating causal effects of treatments in randomized and nonrandomized studies. , 1974 .

[38]  D. Rubin Comment on "Causal inference without counterfactuals," by Dawid AP , 2000 .

[39]  N Enomoto,et al.  Acetaldehyde metabolism in different aldehyde dehydrogenase-2 genotypes. , 1991, Alcoholism, clinical and experimental research.

[40]  M. B. Katan,et al.  Apolipoprotein E isoforms, serum cholesterol, and cancer , 2004 .

[41]  Debbie A Lawlor,et al.  Exploring the Developmental Overnutrition Hypothesis Using Parental–Offspring Associations and FTO as an Instrumental Variable , 2008, PLoS medicine.

[42]  David V Conti,et al.  Commentary: the concept of 'Mendelian Randomization'. , 2004, International journal of epidemiology.

[43]  Jeffrey M. Woodbridge Econometric Analysis of Cross Section and Panel Data , 2002 .

[44]  C. Manski Nonparametric Bounds on Treatment Effects , 1989 .

[45]  D. Lawlor,et al.  Clustered Environments and Randomized Genes: A Fundamental Distinction between Conventional and Genetic Epidemiology , 2007, PLoS medicine.

[46]  Ting-kai Li,et al.  Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism , 1986, Hepatology.

[47]  Donald B. Rubin,et al.  Bayesian Inference for Causal Effects: The Role of Randomization , 1978 .

[48]  Richard D Riley,et al.  Meta‐analysis of genetic studies using Mendelian randomization—a multivariate approach , 2005, Statistics in medicine.

[49]  Sebastian Schneeweiss,et al.  Preference-Based Instrumental Variable Methods for the Estimation of Treatment Effects: Assessing Validity and Interpreting Results , 2007, The international journal of biostatistics.

[50]  Stijn Vansteelandt,et al.  Causal inference with generalized structural mean models , 2003 .

[51]  N. Sheehan,et al.  Mendelian randomization as an instrumental variable approach to causal inference , 2007, Statistical methods in medical research.

[52]  Judea Pearl,et al.  Counterfactual Probabilities: Computational Methods, Bounds and Applications , 1994, UAI.

[53]  J. Robins,et al.  Estimation of treatment effects in randomised trials with non-compliance and a dichotomous outcome using structural mean models , 2004 .

[54]  Sarah Parish,et al.  Fibrinogen and coronary heart disease: test of causality by 'Mendelian randomization'. , 2006, International journal of epidemiology.

[55]  S. Ebrahim,et al.  'Mendelian randomization': can genetic epidemiology contribute to understanding environmental determinants of disease? , 2003, International journal of epidemiology.

[56]  Frank Windmeijer,et al.  Instrumental Variable Estimators for Binary Outcomes , 2009 .

[57]  Miguel A. Martínez-González,et al.  Critical Appraisal of Epidemiological Studies and Clinical Trials , 2007 .

[58]  M Alan Brookhart,et al.  Instrumental variables II: instrumental variable application-in 25 variations, the physician prescribing preference generally was strong and reduced covariate imbalance. , 2009, Journal of clinical epidemiology.

[59]  Shah Ebrahim,et al.  Does Elevated Plasma Fibrinogen Increase the Risk of Coronary Heart Disease?: Evidence from a Meta-Analysis of Genetic Association Studies , 2005, Arteriosclerosis, thrombosis, and vascular biology.

[60]  Els Goetghebeur,et al.  Correcting Instrumental Variables Estimators for Systematic Measurement Error. , 2009, Statistica Sinica.

[61]  M. Tobin,et al.  Mendelian Randomisation and Causal Inference in Observational Epidemiology , 2008, PLoS medicine.

[62]  Sara Geneletti,et al.  The effect of treatment on the treated: a decision theoretic perspective , 2006 .

[63]  M Alan Brookhart,et al.  Instrumental variables I: instrumental variables exploit natural variation in nonexperimental data to estimate causal relationships. , 2009, Journal of clinical epidemiology.

[64]  Debbie A Lawlor,et al.  Mendelian randomization suggests no causal association between C-reactive protein and carotid intima-media thickness in the young Finns study. , 2007, Arteriosclerosis, thrombosis, and vascular biology.

[65]  Wiebe R. Pestman,et al.  Instrumental Variables: Application and Limitations , 2006, Epidemiology.

[66]  S. Lewis,et al.  Alcohol Intake and Blood Pressure: A Systematic Review Implementing a Mendelian Randomization Approach , 2008, PLoS medicine.

[67]  A. Wald The Fitting of Straight Lines if Both Variables are Subject to Error , 1940 .

[68]  A. Dawid,et al.  Defining and Identifying the Effect of Treatment on the Treated , 2011 .

[69]  Juan P Casas,et al.  Insight into the nature of the CRP-coronary event association using Mendelian randomization. , 2006, International journal of epidemiology.

[70]  Roland Ramsahai,et al.  Causal Bounds and Instruments , 2007, UAI.

[71]  George Davey Smith,et al.  Mendelian randomization: Using genes as instruments for making causal inferences in epidemiology , 2008, Statistics in medicine.

[72]  Ailsa E Gebbie,et al.  Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women's Health Initiative Randomized Controlled Trial. Writing Group for the Women's Health Initiative Investigators. JAMA 2002; 288(3): 321-333 , 2002, Journal of Family Planning and Reproductive Health Care.