Rejoinder: To Weight or Not to Weight?: On the Relation Between Inverse-probability Weighting and Principal Stratification for Truncation by Death

Tchetgen Tchetgen, Eric J.a,b; Glymour, M. Mariac; Shpitser, Ilyaa; Weuve, Jenniferd

doi: 10.1097/EDE.0b013e31823b5081
Author Information

From the Departments of aEpidemiology, bBiostatistics, and cSociety, Human Development and Health, Harvard School of Public Health, Boston, MA; and dRush Institute for Healthy Aging and Department of Internal Medicine, Rush University Medical Center, Chicago, IL.

Editors' note: Related articles appear on pages 119 and 129.

Correspondence: Eric J. Tchetgen Tchetgen, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Ave, Boston, MA 02115. E-mail:

Article Outline

We thank the editor for the opportunity to respond to the commentary by Chaix et al.1 In this rejoinder, we address the concerns voiced by Chaix et al1 that inverse-probability weighting creates an immortal population, and that instead, an application of the principal stratification framework to our data would provide a more useful causal effect of smoking on cognitive decline. We distinguish between issues pertaining to statistical inference and those related to causal inference, and we show that, under a structural equation framework, the causal effect identified by inverse-probability weighting survivors, in fact, naturally incorporates principal-strata causal effects. Therefore, we formally establish relations between these 2 seemingly unrelated analytic frameworks.

We consider a simplified study design with only 3 occasions, as in the causal directed acyclic graph in Figure 1: a baseline j = 0 at which binary smoking status S is observed, and 2 follow-up contacts, with cognitive function C(j) assessed at each j = 1, 2. All respondents participate at the first follow-up, but some die before the second follow-up, with D = 0 indicating survival. Survival is affected by S and C(1), and death is the only source of attrition. The outcome of interest is change in cognitive function ΔC = C(2) − C(1). Throughout, we assume no measurement error, and we distinguish between an exogenous time-varying common cause of D and ΔC, which we indicate by L, and an endogenous common cause of D and ΔC, for example, C(1). An exogenous common cause of D and ΔC is known not to be affected by S, whereas an endogenous common cause of the 2 variables may be an effect of S. To simplify further, we take C(1) to be the only endogenous common cause of D and ΔC. The variable U encodes possible unmeasured common causes of cognitive function measurement C(1) and change in cognitive function ΔC, the presence of which cannot be ruled out with certainty. For instance, there is evidence for genetic determinants of Alzheimer disease that suggests a genetic basis for an individual's cognitive function over time2; however, such genetic information was not available in our study. Although such a genetic component of U would not be affected by smoking behavior, U might also include an unknown epigenetic effect of smoking behavior on future cognitive function, in which case it would be an effect of smoking behavior. For simplicity and with no loss in generality, we further assume that all analyses are stratified by a set of time-constant confounders of S.

Back to Top | Article Outline

Issues of Statistical Inference

The first issue relates to statistical inference using inverse-probability weighting. Chaix et al1 claim that we are creating an immortal population by applying inverse-probability weights to the survivors, or in their words, that, inverse-probability weighting is “weighting up the dead.” Thus, according to Chaix et al, “[i]n replacing dead participants by cloning the living, inverse-probability weighting generates a sample in which participants are not allowed to die.” Although the “cloning” metaphor often used to describe inverse-probability weighting can be a helpful pedagogic tool, it should not be taken literally. The metaphor fails to convey the appropriate statistical interpretation of inverse-probability-weighted estimation, particularly in the context of attrition due to death. A more appropriate statistical interpretation is that inverse-probability weighting ensures the statistical independence between D and (measured) time-varying common causes (C(1), L) of D and ΔC. Most importantly, by virtue of weighting survivors by the inverse of their probability of surviving, inverse-probability weighting ensures such independence without altering the conditional density f1CC(1), L, U, S, D = 0) of ΔC given the past among survivors, and the joint conditional density f2(C(1), L, US) of covariates given smoking status. These 2 properties are key to our goal of making inferences about the effects of smoking on cognitive decline because, as we show later, all meaningful information about such effects is contained in f1 and f2.

Although Chaix et al argue it is “courageous” to develop an inverse-probability-weighting model to predict survival when only 20% of the sample remains at the end of follow-up, we think it is even more courageous to ignore selective survival in such a circumstance. A conventional analysis of survivors in such a situation—with regression adjustment for common causes of D and ΔC—could provide a valid test of the sharp null hypothesis of no effect of S on ΔC, as described in Figure 1 (omitting the dotted arrows), only if there were no endogenous common causes of D and ΔC, as in Figure 2. This scenario is inconsistent with prior empirical studies. If exposure and time-varying causes of ΔC are associated, as would likely be the case in an unweighted population of survivors, the time-varying covariates will bias associations between the exposure and ΔC.3 Thus, ignoring the presence of strong selective survival is almost guaranteed to bias the observed association between S and ΔC, even under the sharp null hypothesis.

In contrast, under a standard positivity assumption for survival,4 inverse-probability weighting is well known to estimate the g-formula of Robins5:

. where the second representation follows from the conditional independence of U and D given (S, L, C(1)). Thus, g(s) is a statistical object that takes as input the distribution of the observed data, and generates a number for each value of smoking status. In particular, irrespective of its causal interpretation, g(s) is always well defined for both values of s, and its definition requires no reference to counterfactuals. Later in the text, we address the causal interpretation of g(s). For now, consider again the sharp null hypothesis. It is well known that the null implies that g(s) is a constant, and therefore that g(s) is guaranteed not to find an effect of S when there is none. In other words, g(s) can be used to construct a valid test of the null of no effect of S on ΔC in the presence of dependent truncation due to death. This can be observed in Figure 1 under the sharp null (ie, omitting the dotted arrows from the diagram) in which S is independent of (d-separated from) U, and C(1) is conditionally independent of ΔC given L, U, and D = 0.

The fact that a weighted analysis is guaranteed to give the correct answer under the sharp null for the causal structure in Figure 1 in itself constitutes compelling justification for epidemiologists to routinely present analyses inverse-probability weighted for attrition due to death, even if as supplemental analyses.

Back to Top | Article Outline

Issues of Causal Inference

We now turn to issues of causal inference. Chaix et al1 question the interpretation of the inverse-probability-weighting estimator and thus the meaning of g(s). We have established earlier that g(s) recovers the sharp null when it holds, in which case the causal interpretation of g(s) should be uncontroversial. Next, we consider the alternative hypothesis. In this case, by the argument given earlier, if a test statistic based on g(s) rejects, we can safely conclude that such an effect is present at the usual type-1-error nominal level, provided no modeling error exists. The ability of such a statistical test to reject is intimately related to the causal interpretation one can assign to the contrast g(1) − g(0), corresponding to the effect of smoking measured by the g-formula. In addition, a causal interpretation is indispensable for understanding the direction and magnitude of the effects of smoking, and for providing meaningful policy recommendations.

Next, we show that g(s) can be given a meaningful causal interpretation in which principal-strata causal effects are naturally incorporated. Specifically, we establish that under a structural equation model, inverse-probability weighting for survival as implemented by Weuve et al3 delivers inferences about a causal effect of S on ΔC that is a combination of an average principal-strata direct effect of S on ΔC and indirect effects of S on ΔC. For instance, in the simple case of a linear structural model for ΔC, the direct effect is defined for the principal stratum of individuals who are alive and for whom smoking has no causal effect on (C(1), U, D). In this case, the indirect effects include some components that are defined for everyone in the population and others defined only for individuals in a principal stratum. The indirect effect combines the total direct causal effects of S on (measured and unmeasured) endogenous causes of ΔC with principal-strata direct effects of the latter variables on ΔC. Therefore, we formally establish that the g-formula (hence inverse-probability-weighted estimation) and principal stratification are in fact not disparate concepts, but are instead connected.

The exposition is framed around a structural equation theory of causal inference, described by Pearl.6 Structural equations provide a nonparametric algebraic interpretation of the diagram of Figure 1 corresponding to 6 functions, one for each variable on the causal graph:

Each of the nonparametric functions {gS, …, gΔC} represents a causal mechanism that determines the value of the left variable, known as the output, from variables on the right, known as the inputs.6 The errors (εS, εU, εC(1), εL, εD, εΔC) stand for all factors not included on the graph that could possibly affect their corresponding outputs when all other inputs are held constant. To be consistent with the causal graph presented in Figure 1, it is required that these errors be mutually independent, but we allow their distribution to remain arbitrary. Lack of a causal effect of a given variable on an output is encoded by an absence of the variable from the right-hand side. For example, the absence of U from the arguments of gD encodes the assumption that variations in U will leave D unchanged, as long as variables S, C(1), and εD remain constant, which is consistent with the assumption that there are no unmeasured common causes of death and cognitive decline.

The last equation makes explicit the fact that ΔC is observed only among survivors with (D = 0), with structural equation gΔC (S, C(1), U, L, εΔC). As stated by Pearl,6 the invariance of structural equations permits their use as a basis for modeling causal effects and counterfactuals. In fact, to emulate the intervention in which one sets {S = s} for all individuals simply amounts to replacing the equation for S with S = s, producing the following set of modified equations:

with (Us, Cs(1), Ls, Ds, ΔCs) denoting the counterfactual outcomes had smoking status been set to s (possibly contrary to fact). We emphasize that although the model specifies a structural equation for death, survival is not manipulable, and, together with ΔC, should be understood as part of the outcome produced by the system of equations. Structural equations are particularly helpful to clarify the difficulty with interpreting the effect of smoking when truncation by death is present. Specifically, we note that the individual effect of smoking is recovered by taking the contrast ΔC1 − ΔC0, which clearly is defined only for individuals in the principal stratum {D0 = D1 = 0} and is equal to the following:

Consider the submodel M1 in which equations (4), (5), and (8) are linear.

We note that βS formally encodes the principal-strata causal effect of S among all individuals who survived irrespective of smoking status under an intervention in which we hold (C(1), U) fixed, ie, among individuals with {D0,c,u = D1,c,u = 0}, where Ds,c,u is the counterfactual outcome obtained by replacing equations (3)(5) with the following:




In other words, βS is a principal-strata controlled direct effect. Similarly, βC(1) is the principal-strata causal effect of C(1) on ΔC if one could intervene to set C(1) to c + 1 versus c, while holding (S, U) fixed at (s, u), among individuals with {Ds,c,u = Ds,c+1,u = 0}; similar interpretations hold for βU and βL.

Under the structural linear model, it is easy to verify that the causal effect of S on ΔC for individuals in the principal stratum {D0 = D1 = 0} is constant and equal to the following:

As stated above, βS encodes a principal-strata controlled direct effect of smoking, whereas βC(1)αS is the product of the principal-strata controlled direct effect of C(1) on ΔC (setting S = s, L = l, U = u) and the controlled direct effect of S on C(1) (setting U = u), to produce an indirect effect of smoking through the pathway SC(1) → ΔC. Similarly, βC(1)αUθS encodes an indirect effect of S through the pathway SUC(1) → ΔC, and finally βUθS is an indirect effect of S through the pathway SU → ΔC. The equation in the above display states that the total causal effect of smoking is the sum of the effects through these various pathways. Although the parameters βS, βC(1), βU, and βL have only a causal interpretation within principal strata, the parameters αS, θS, αS, and αU are well defined as causal contrasts for all individuals. Model M1 assumes that S has a constant effect on U and C(1) in all individuals, in which case the decomposition in the above display is the average principal strata effect

The homogeneity assumption may not be realistic, as the SC(1) pathway may be activated only in a subset of the population, and similarly the pathway SUC(1) may be activated only for a subset of individuals. A generalization of M1 accommodates such heterogeneity. Consider the model M2 corresponding to the system of equations (3)(7) together with equation (14). In this model, the causal effect of S on ΔC for individuals in the principal stratum {D0 = D1 = 0} is as follows:

so that βS may be interpreted as the principal-strata direct effect of smoking among individuals with

In the Appendix, we prove the following result.

Result 1: Under the structural equation model M2, we have that the causal contrast identified by the g-formula is

The result establishes that g(s) correctly identifies a combination of causal effects of S on ΔC under the linear structural equation: a principal-strata direct effect of smoking, with indirect effects, which themselves combine principal-strata direct effects of endogenous variables, with the controlled direct effect of smoking on the latter variables. In the Appendix, we relax the linearity assumption of equation (14), and we provide a general nonparametric decomposition of the causal effect of S on ΔC identified by g(s). The result also implies that, under the submodel M1 of M2, the g-formula, and therefore inverse-probability-weighted estimation, correctly identifies the average principal-strata effect of smoking to be

which is a combination of a direct principal-strata effect of smoking βS, and indirect effects of smoking mediated through (U, C(1)).

In conclusion, instead of “weighting up the dead,” inverse-probability-weighted analyses are an accessible and flexible approach for estimating the causal effects of smoking on cognitive change in the context of dependent truncation by death. Under an assumption that there are no unmeasured common causes of mortality and cognitive decline, we have characterized the causal effect that is identified by inverse-probability weighting survivors using a structural equation framework, and we have shown that this causal effect naturally incorporates principal-strata effects. Although the identifying assumption of “no unmeasured common cause of mortality and cognitive decline” cannot be confirmed with certainty empirically, methodology is currently available for conducting sensitivity analyses to assess the impact on inferences of a violation of this assumption in inverse-probability-weighted analyses.7,8

Back to Top | Article Outline



Consider the system of nonparametric structural equations (3)(8). Then, we have that the causal contrast identified by the g-formula is as follows:

where W = (C(1), L, U), Ws = (Cs(1), Ls, Us), with distribution FWs (w) evaluated at w, s = 0, 1; DE is an average of principal-strata average direct effect and IE is the integral of an indirect effect of S on ΔC through W.

Back to Top | Article Outline
Proof of Lemma

Let ΔC (ε) denote the distribution function of εΔC evaluated at ε. Below, we use w* to indicate a fixed baseline value of W. Evaluating the g-formula under the model gives

by the independence assumptions encoded in the structural equations, which gives the result.

Back to Top | Article Outline
Proof of Result 1

The result is obtained by applying the Lemma with gΔC given by equation (14).

Back to Top | Article Outline


We thank Jamie Robins, Andrea Rotnitzky, Tyler VanderWeele, and Miguel Hernan for very helpful discussions.

Back to Top | Article Outline


1. Chaix B, Evans D, Merlo J, Suzuki E. Weighing up the dead and missing: reflections on inverse-probability weighting and principal stratification to address truncation by death. Epidemiology. 2012;23:129–131.
2. Bertram L, McQueen MB, Mullin K, Blacker D, Tanzi RE. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat Genet. 2007;39:17–23.
3. Weuve J, Tchetgen Tchetgen EJ, Glymour MM. Accounting for bias due to selective attrition: the example of smoking and cognitive decline. Epidemiology. 2012;23.
4. Robins JM, Rotnitzky A. Recovery of information and adjustment for dependent censoring using surrogate markers. In: Jewell N, Dietz K, Farewell V eds. AIDS Epidemiology—Methodological Issues. Boston: Birkhäuser; 1992:297–331.
5. Robins JM. Causal inference from complex longitudinal data. Latent variable modeling and applications to causality. In: Berkane M ed. Lecture Notes in Statistics (120). New York: Springer Verlag; 1997:69–117.
6. Pearl J. Causality: Models, Reasoning, and Inference. 2nd ed. New York: Cambridge University Press; 2009.
7. Robins JM, Rotnitzky A, Scharfstein D. Sensitivity analysis for selection bias and unmeasured confounding in missing data and causal inference models. In: Halloran ME, Berry D eds. Statistical Models in Epidemiology: The Environment and Clinical Trials. IMA. Vol 116. New York: Springer-Verlag; 1999:1–92.
8. Scharfstein DO, Rotnitzky A, Robins JM. Adjusting for non-ignorable drop-out using semi-parametric non-response models [Comments and Rejoinder]. J Am Statist Assoc. 1999;94:1096–1120.
© 2012 Lippincott Williams & Wilkins, Inc.