DAGs intro with exercises 3h DirectedAcyclicGraph Hein Stigum http://folk.uio.no/heins/ courses May-16 H.S. 1 Agenda • Background • DAG concepts – – – – Association, Cause Confounder, Collider, Mediator Causal thinking Paths Exercises • Analyzing DAGs – Examples • More on DAGs May-16 H.S. 2 Why causal graphs? • Estimate effect of exposure on disease • Problem – Association measures are biased • Causal graphs help: – Understanding • Confounding, mediation, selection bias – Analysis • Adjust or not – Discussion • Precise statement of prior assumptions May-16 H.S. 3 Causal versus casual CONCEPTS May-16 H.S. 4 Precision and validity • Measures of populations – precision - random error – validity - systematic error Precision Bias True value Estimate DAGs only consider bias 5/29/2016 H.S. 5 god-DAG Causal Graph: Node = variable Arrow = cause E=exposure, D=disease DAG=Directed Acyclic Graph Read of the DAG: Estimations: Causality = arrows Associations = paths Independencies = no paths E-D association has two parts: ED causal effect keep open ECUD bias try to close E[C]UD Condition (adjust) to close Time May-16 H.S. 6 Association and Cause Association 3 possible causal Association 3 possible causal structures structures 3 possible causal structure Association 1 1 Yellow Yellow fingers fingers Lung Lung cancer cancer Cause Cause (Reversed cause) E Yellow Yellow fingers fingers Smoke Smoke D Lung Lung cancer cancer 2 2 Yellow Yellow fingers fingers Confounder Confounder Lung Lung cancer cancer U U 3 3 Yellow Yellow fingers fingers Collider Collider Lung Lung cancer cancer + more complicated structures May-16 H.S. 7 Confounder idea A common cause Smoking + Adjust for smoking Smoking + Yellow fingers Lung cancer + Yellow fingers + Lung cancer + • A confounder induces an association between its effects • Conditioning on a confounder removes the association • Condition = (restrict, stratify, adjust) • Paths • Simplest form • Causal confounding, may have non-causal confounding May-16 H.S. 8 Collider idea Two causes for selection to study Selected + Yellow fingers Selected subjects Selected + Lung cancer + + Yellow fingers Lung cancer - or + and • Conditioning on a collider induces an association between its causes • “And” and “or” selection leads to different bias • Paths • Simplest form May-16 H.S. 9 Mediator M • Have found a cause (E) • How does it work? – Mediator (M) E direct effect D – Paths 𝑇𝑜𝑡𝑎𝑙 𝑒𝑓𝑓𝑒𝑐𝑡 = 𝑖𝑛𝑑𝑖𝑟𝑒𝑐𝑡 + 𝑑𝑖𝑟𝑒𝑐𝑡 𝑖𝑛𝑑𝑖𝑟𝑒𝑐𝑡 𝑀𝑒𝑑𝑖𝑎𝑡𝑒𝑑 𝑝𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 = 𝑡𝑜𝑡𝑎𝑙 Use ordinary regression methods if: linear model and no E-M interaction. Otherwise, need new methods May-16 H.S. 10 Causal thinking in analyses May-16 H.S. 11 Regression before DAGs Use statistical criteria for variable selection Risk factors for D: Variable OR E 2.0 C 1.2 Comments Surprisingly low association Association Both can be misleading! C E May-16 Report all variables in the model as equals D H.S. 12 Statistical criteria for variable selection C E D - Want the effect of E on D (E precedes D) - Observe the two associations C-E and C-D - Assume statistical criteria dictates adjusting for C (likelihood ratio, Akaike (赤池 弘次) or 10% change in estimate) The undirected graph above is compatible with three DAGs: C C E D Confounder 1. Adjust Conclusion: E C D Mediator 2. Direct: adjust 3. Total: not adjust E D Collider 4. Not adjust The data driven method is correct in 2 out of 4 situations Need information from outside the data to do a proper analysis DAGs variable selection: close all non-causal paths May-16 H.S. 13 Association versus causation Use statistical criteria for variable selection Risk factors for D: Variable OR E 2.0 C 1.2 Comments Surprisingly low association Association Causation C E Symmetrical C is a confounder for E-D E is a confounder for C-D May-16 Base adjustments on a DAG C D Report all variables in the model as equals E D Directional Report only the E-effect C is a confounder for ED E is a mediator for CD H.S. 14 The Path of the Righteous Paths May-16 H.S. 15 Path definitions Path: any trail from E to D (without repeating itself) Type: causal, non-causal State: open, closed 1 2 3 4 Four paths: Path ED EMD ECD EKD Goal: Keep causal paths of interest open Close all non-causal paths May-16 H.S. 16 Four rules 1. Causal path: ED (all arrows in the same direction) otherwise non-causal Before conditioning: 2. Closed path: K (closed at a collider, otherwise open) Conditioning on: 3. a non-collider closes: [M] or [C] 4. a collider opens: [K] (or a descendant of a collider) May-16 H.S. 17 ANALYZING DAGs May-16 H.S. 18 Confounding examples May-16 H.S. 19 Vitamin and birth defects 1. Is there a bias in the crude E-D effect? 2. Should we adjust for C? 3. What happens if age also has a direct effect on D? Unconditional Path 1 ED 2 ECUD Type Status Causal Open Non-causal Open Conditioning on C Path 1 ED 2 EC]UD 3 EC] D Type Causal Non-causal Non-causal May-16 May-16 Status Open Closed Closed Bias No bias H.S. This is an example of confounding Question: Is U a confounder? 20 Exercise: Physical activity and Coronary Heart Disease (CHD) We want the total effect of Physical Activity on CHD. 1. Write down the paths. 2. Are they causal/non-causal, open/closed? 3. What should we adjust for? 5 minutes May-16 H.S. 21 Intermediate variables May-16 H.S. 22 Exercise: Tea and depression 1. Write down the paths. Show type and status. O coffee E tea 2. You want the total effect of tea on depression. What would you adjust for? C caffeine D depression 3. You want the direct effect of tea on depression. What would you adjust for? 4. Is caffeine an intermediate variable or a confounder? 10 minutes May-16 H.S. 23 Exercise: Statin and CHD C cholesterol E U lifestyle 1. Write down the paths. Show type and status. D 2. You want the total effect of statin on CHD. What would you adjust for? CHD statin 3. If lifestyle is unmeasured, can we estimate the direct effect of statin on CHD (not mediated through cholesterol)? 4. Is cholesterol an intermediate variable or a collider? 10 minutes May-16 H.S. 24 Direct and indirect effects So far: Causal interpretation: Controlled (in)direct effect linear model and no E-M interaction New concept: Natural (in)direct effect Causal interpretation also for: non-linear model , E-M interaction Controlled effect = Natural effect if linear model and no E-M interaction Hafeman and Schwartz 2009; Lange and Hansen 2011; Pearl 2012; Robins and Greenland 1992; VanderWeele 2009, 2014 May-16 H.S. 25 Confounder, collider and mediator Mixed May-16 H.S. 26 Diabetes and Fractures We want the total effect of Diabetes (type 2) on fractures Conditional Unconditional Path Path 11 E→D E→D 22 E→F→D E→F→D 33 E→B→D E→B→D 44 E←[V]→B→D E←V→B→D 55 E←[P]→B→D E←P→B→D May-16 Type Type Causal Causal Causal Causal Causal Causal Non-causal Non-causal Non-causal Non-causal Status Status Open Open Open Open Open Open Closed Open Closed Open H.S. Questions: Mediators More paths? B a collider? V and P ind? Confounders 27 Selection bias May-16 H.S. 28 Convenience sample, homogenous sample H 1. Convenience sample: Conduct the study among hospital patients? hospital E D 2. Homogeneous sample: Population data, exclude hospital patients? fractures diabetes Conditional Unconditional Path 1 E→D 2 E→H←D E→[H]←D Type Causal Non-causal Non-Causal Status Open Closed Open Collider, selection bias Collider stratification bias: at least on stratum is biased This type of selection bias can be adjusted for! May-16 H.S. 29 Adjusting for Selection bias S sex age smoke CHD Paths Type Status smokeCHD Causal Open smokesexSageCHD Non-causal Open Adjusting for sex or age or both removes the selection bias Selection bias types: Berkson’s, loss to follow up, nonresponse, self-selection, healthy worker Hernan et al, A structural approach to selection bias, Epidemiology 2004 May-16 H.S. 30 Drawing DAGs with DAGitty May-16 H.S. 31 Drawing a causal DAG Start: E and D add: [S] add: C-s 1 exposure, 1 disease variables conditioned by design all common causes of 2 or more variables in the DAG C V E D C must be included V may be excluded M may be excluded K may be excluded common cause exogenous mediator unless we condition M K May-16 H.S. 32 DAGitty drawing tool • DAGitty – Draw DAGs – Find adjustment sets – Test DAGs • Web page – http://www.dagitty.net/ – Run or download May-16 HS 33 Interface May-16 HS 34 Draw model • Draw new model – Model>New model, Exposure, Outcome • New variables, connect – – – – n c r d new variable (or double click) connect (hit c over V1 and over V2 to connect) rename delete • Status (toggle on/off) –u –a May-16 unobserved adjusted HS 35 Export DAG • Export to Word or PowerPoint – “Zoom” the DAGitty drawing first (Ctrl-roll) – Use “Snipping tool” or – use Model>Export as PDF Without zooming May-16 With zooming HS 36 Randomized experiments May-16 H.S. 37 Strength of arrow E Not deterministic C1 D C2 C1, C2, C3 exogenous Adjust for C1, C2, C3? C3 C R full compliance deterministic E Full compliance no E-D confounding D U R not full compliance E D Sub analysis conditioning on E may lead to bias May-16 Not full compliance weak E-D confounding but R-D is unconfounded Path 1 RED 2 REUD H.S. Type Status Causal open non-causal closed 38 Randomized experiments C E Observational study D C R E D U R c E IVe ITTe D Randomized experiment with full compliance R= randomized treatment E= actual treatment. R=E Randomized experiment with less than full compliance (c) If linear model: ITTe=c*IVe, c<1 IntentionToTreat effect: effect of R on D (unconfounded) population PerProtocol: crude effect of E on D (confounded by U) InstrumentalVariable effect: adjusted effect of E on D (if c is known, 2SLS) individual May-16 H.S. 39 Exercise: R E R + - + 85 0 85 15 100 115 D R + - + 43 63 106 57 37 94 D E + - + 38 68 106 47 47 94 C andomized T ontrolled rial N Risk 1. 100 100 0.85 0.00 Calculate the compliance (c) as a risk difference from the table. 2. Calculate the intention to treat effect (ITTe) as a risk difference. N Risk 3. 100 100 0.43 0.63 Calculate the per-protocol effect (PP) as a risk difference. 4. Calculate the instrumental variable effect (IVe). N Risk 5. Explain the results in words. 85 115 0.45 0.59 U R c E IVe D 10 minutes ITTe May-16 H.S. 40 MORE ON DAGs May-16 H.S. 41 Confounding versus selection bias Path: Any trail from E to D (without repeating itself) Open non-causal path = biasing path Confounding and selection bias not always distinct May use DAG to give distinct definitions: C A E A B D Causal C B A B E D Confounding: E D Selection bias: Non-causal path without colliders Non-causal path open due to conditioning on a collider Note: interaction based selection bias not included Hernan et al, A structural approach to selection bias, Epidemiology 2004 May-16 H.S. 42 Summing up • Data driven analyses do not work. Need (causal) information from outside the data. • DAGs are intuitive and accurate tools to display that information. • Paths show the flow of causality and of bias and guide the analysis. • DAGs clarify concepts like confounding and selection bias, and show that we can adjust for both. Better discussion based on DAGs Draw your assumptions before your conclusions May-16 H.S. 43 Recommended reading • Books – – – – – Hernan, M. A. and J. Robins. Causal Inference. Web: Rothman, K. J., S. Greenland, and T. L. Lash. Modern Epidemiology, 2008. Morgan and Winship, Counterfactuals and Causal Inference, 2009 Pearl J, Causality – Models, Reasoning and Inference, 2009 Veierød, M.B., Lydersen, S. Laake,P. Medical Statistics. 2012 • Papers – Greenland, S., J. Pearl, and J. M. Robins. Causal diagrams for epidemiologic research, Epidemiology 1999 – Hernandez-Diaz, S., E. F. Schisterman, and M. A. Hernan. The birth weight "paradox" uncovered? Am J Epidemiol 2006 – Hernan, M. A., S. Hernandez-Diaz, and J. M. Robins. A structural approach to selection bias, Epidemiology 2004 – Berk, R.A. An introduction to selection bias in sociological data, Am Soc R 1983 – Greenland, S. and B. Brumback. An overview of relations among causal modeling methods, Int J Epidemiol 2002 – Weinberg, C. R. Can DAGs clarify effect modification? Epidemiology 2007 May-16 H.S. 44