New Results

Extending causality tests with genetic instruments: an integration of Mendelian Randomization and the Classical Twin Design

Camelia C. Minică, Conor V. Dolan, Dorret I. Boomsma, Eco de Geus, Michael C. Neale
doi: https://doi.org/10.1101/134585

ABSTRACT

Mendelian Randomization (MR) is an important approach to modelling causality in non-experimental settings. MR uses genetic instruments to test causal relationships between exposures and outcomes of interest. Individual genetic variants have small effects, and so, when used as instruments, render MR liable to weak instrument bias. Polygenic scores have the advantage of larger effects, but may be characterized by direct pleiotropy, which violates a central assumption of MR.

We developed the MR-DoC twin model by integrating MR with the Direction of Causation twin model. This model allows us to test pleiotropy directly. We considered the issue of parameter identification, and given identification, we conducted extensive power calculations. MR-DoC allows one to test causal hypotheses and to obtain unbiased estimates of the causal effect given pleiotropic instruments (polygenic scores), while controlling for genetic and environmental influences common to the outcome and exposure. Furthermore, MR-DoC in twins has appreciably greater statistical power than a standard MR analysis applied to singletons, if the unshared environmental effects on the exposure and the outcome are uncorrelated. Generally, power increases with: 1) decreasing residual exposure-outcome correlation, and 2) decreasing heritability of the exposure variable.

MR-DoC allows one to employ strong instrumental variables (polygenic scores, possibly pleiotropic), guarding against weak instrument bias and increasing the power to detect causal effects. Our approach will enhance and extend MR’s range of applications, and increase the value of the large cohorts collected at twin registries as they correctly detect causation and estimate effect sizes even in the presence of pleiotropy.

INTRODUCTION

Establishing causality in observational studies is important as knowledge of the relationship between a putative causal factor (exposure) and a potential outcome may inform rational treatment and prevention policies. While randomized controlled trials (RCTs) are the acid test of causality, they are expensive, time consuming, and may be practically or ethically unfeasible. For example, one cannot assign randomly individuals to a ‘low serum cholesterol condition’ in studying the causal effects of low serum cholesterol levels on cancer. An important alternative to the RCT is Mendelian Randomization (Katan 1, 2004).

Mendelian Randomization (MR) offers some traction in addressing causality by using genetic variants as instrumental variables to detect the causal effect of a modifiable risk factor (exposure) on a disease outcome in non-experimental settings 2,3. MR is quickly becoming the dominant approach to establishing causality; many recent applications have been published 2,4-10. The ascendency of MR is due to: dramatic drop in DNA genotyping costs, which has given rise to large genotyped samples, robust genetic associations established in genome-wide association studies (GWASs) 11, and the inherent advantages of MR, which include ecologic validity, robustness to reverse causation (from exposure to instrument) 12 and confounding 13.

MR requires instruments with a relatively strong direct relationship with the exposure. A disadvantage of many genetic variants is that they have weak effects 14. Weak instruments confer insufficient statistical power, and render MR liable to weak instrument bias 13,15,16. Combining the weak genetic effects into a polygenic risk score (PGS) is a possible route to increase the strength of the genetic instrument 13,17-19. However, the MR assumption that the instrument is not pleiotropic (has no direct effect on the outcome) is stronger in the case of a PGS instrument 13,16,20,21. A PGS comprises many genetic variants, any of which may directly affect both the exposure and the outcome, or may include variants in linkage disequilibrium with variants affecting the outcome. As demonstrated by twin studies 22-27 and by polygenic risk score analyses 28-30, many genetic variants associate with multiple phenotypes, suggesting pervasive pleiotropy 31-35.

Several methods are currently in use as means to tackle the ‘no pleiotropy’ assumption. Some approaches apply prior selection criteria to increase the probability that the instruments are valid. For instance, the stepwise procedure implemented in the R-package gtx 36 employs iteratively a heterogeneity test to discard from a polygenic score genetic instruments yielding significant heterogeneity in the estimates of the causal effect. Possible heterogeneity is assumed to be indicative of pleiotropy. The efficiency of this method depends on its power 36 to detect heterogeneity arising from pleiotropy. However, heterogeneity may arise from sources other than pleiotropy, so that one may needlessly weaken the instrument by removing valid genetic variants. Johnson 36 noted that the performance of this procedure in terms of bias and type I error rates within the MR context has yet to be established. Other approaches like those based on the median estimator 37 can handle up to 50% invalid instruments. However, the strong MR assumption still applies to the variant(s) yielding the median causal effect. MR-Egger regression 38 is another approach that, with weaker assumptions, gives consistent estimates even when all instruments are pleiotropic. The estimator is (asymptotically) consistent under the weaker assumption that the effect of the instrument on the exposure is uncorrelated with the effect of the instrument on the outcome (i.e., the Instrument Strength Independent of Direct Effect assumption). However, the InSIDE assumption is unlikely to hold universally 38, as shared molecular mechanisms (i.e., direct/biological pleiotropy) are expected to yield correlated effects on the associated traits (see, for examples, 34). Furthermore, Bowder et al. noted 38 that there are other plausible paths from the instrument to the outcome (except direct paths, or indirect, via the exposure), for example, via confounders affecting both traits, or due to linkage disequilibrium between the instrument and a genetic variant affecting the outcome. In this case the estimate of the estimates of the causal effect in Egger regression will be biased 39. Finally, although MR-Egger uses multiple genetic variants to estimate the causal effect, these instruments are employed individually (i.e., not combined in a polygenic score), and so the approach (like the median-based approaches) remains liable to weak instrument bias 40.

The utility of the classical twin design (CTD) in the study of direction of causality is well established 41-45. The present aim is to combine MR and CTD into a single model. A similar approach was proposed by Kohler et al. who integrated CTD with the instrumental variable method 44. Our focus is on combining MR with CTD to render testable the ‘no pleiotropy’ assumption. Particularly, we address issues of identification and statistical power associated specifically with a (poly-)genic instrument, which may be related directly to the outcome (i.e., violating the no-pleiotropy assumption). Integrating MR with CTD has three advantages: 1) it allows one to relax the strong MR assumption concerning the instrument’s conditional independence of the outcome (conditional on the exposure and confounders, i.e., the no pleiotropy assumption; 2) by accounting for pleiotropy, the approach facilitates the use of PGS as instruments; and 3) in specific circumstances, the approach confers substantial gains in power relative to the standard MR approaches.

METHODS

Direction of Causation (DoC) twin model was advanced as an exploratory approach to establish direction of causation between two correlated traits 43,46-49. In contrast, MR 3,13 is used to test unidirectional causation (from designated causal exposure to outcome). Here we propose the MR-DoC twin model, developed by imposing restrictions on the DoC parameters to represent unidirectional causality hypotheses, and by extending the model to include measured genetic variants as instrumental variables. Integrating MR with DoC allows us to test certain MR assumptions.

The Direction of Causation twin model

The Direction of causation (DoC) twin model 43,46-49 uses cross-sectional data observed in monozygotic (MZ) and DZ twins to test causal hypotheses regarding the association between two traits. In contrast to MR, DoC does not necessarily involve a prior hypothesis concerning the causal direction. The path diagram of such a model is shown in Figure 1, given an exposure variable X and an outcome variable Y, observed in DZ twins.

Figure 1:
Figure 1:

Path diagram representing the direction of causation twin model given two traits: variable X and variable Y measured in dizygotic (DZ) twins (t1 and t2). Squares represent observed variables, while circles represent latent variables. A, C and E stand for additive genetic effects, shared and unique environmental effects, respectively. The double headed arrows represent within/between twins covariances of additive genetic effects (ra), shared environmental effects (rc) and unique environmental effects (re). The cross-twin covariance between additive genetic effects is fixed to .5 (1) for DZ (MZ) twins. DZ (MZ) twins are expected to share on average 50% (100%) of the genetic effects underlying both traits, hence the cross-twin cross-trait covariance is fixed to .5(1)ra for DZ (MZ) twins.. Single headed arrows represent causal effects. Note, the model as depicted, is not identified. Typically ra, rc, and re are assumed to be zero in the application of the DoC twin model.

In Figure 1, X and Y are mutually causally related (parameters g1 and g2). The traits are subjected to the influence of latent additive genetic (AX and AY), shared (CX and CY) and unique (EX and EY) environmental effects, influences which can be direct or indirect, i.e., via the causal paths. As an instance of CTD, this model has the usual assumptions concerning random mating and the absence of genotype-environment interplay (GxE interaction, GxE covariance). The cross-twin correlation of the shared environmental variables is assumed to equal 1 within-trait, and rc across traits, regardless of zygosity. By definition, the cross-twin correlation of unique environmental effects is fixed to zero both within and across traits.

The model as depicted in Figure 1 is not identified; it requires additional restrictions to identify the parameters. By imposing restrictions on the parameters, one can model several alternative hypotheses concerning the observational association between X and Y. The ‘tertium quid’ hypothesis, that a third variable causes both traits, can be tested by constraining the parameters g1 and g2 to equal zero (i.e., the saturated bivariate model). Uni- and bidirectional causal hypotheses can be tested by fixing to zero the within- and cross-twin cross-trait genetic and environmental correlations (ra, rc, re), and estimating the causal parameters g1 and/or g2 (i.e., the uni- and the bidirectional causality models). These competing and nested alternative hypotheses can be tested by likelihood ratio 42,43, provided: 1) the two traits differ in their sources of variance 43; and 2) there are at least three sources of variance influencing the traits (i.e., A, E, and either C or D (dominance)) 50. Given the assumptions mentioned above, we have 43,49: Embedded Image Embedded Image where subscript i stands for twin pair, and j for twin (j=1,2).

Standard Mendelian Randomization

The MR model is an instrumental variable regression model, which employs genetic variants as instrumental variables to test causal hypotheses regarding the association between an exposure and an outcome 3,13. Here we assume that the instrument is a PGS. Three assumptions must hold for a genetic variant to be a valid instrumental variable, as shown in Figure 2. Assumption 1: The genetic instrument (PGS) is robustly associated with the exposure variable X (b1≠0 in Figure 2); Assumption 2: PGS is independent of confounders C (m=0; PGS ⊥ C); Assumption 3: PGS is independent of the outcome variable Y conditional on the exposure X and confounders C (b2=0; PGS ⊥ Y | X, C).

Figure 2:
Figure 2:

Assumptions in Mendelian Randomization. Hypothesis: X causes Y. By assumption, the regression coefficients (m and b2) associated with the dashed lines are zero. Abbreviations: PGS – polygenic score; X – exposure variable; Y – outcome variable; C – confounders; Abbreviations: PGS – polygenic score; X – exposure variable; Y – outcome variable; C – confounders;

In MR, the third assumption pertains to possible pleiotropic effects of the instrument (PGS), or to the likelihood of including variant(s) in linkage disequilibrium with variants affecting the outcome. In practice, the ‘no pleiotropy’ assumption may be violated, particularly when the instrument is a PGS combining the effects of multiple genetic variants (note that a single variant with pleiotropic effects in principle renders the polygenic score invalid as an instrumental variable). This core MR assumption is illustrated in Figure 3 where we consider several MR models with and without pleiotropic instruments, and pinpoint the definition of the no pleiotropy assumption.

Figure 3:
Figure 3:

A) X directly influences Y (parameter g1). In addition, the additive genetic variables AX and AY are correlated (parameter ra). B) The set of loci M underlies the variance of AX, but does not underlie the variance of AY, i.e., in set theory notation, M=AX /AY. Likewise, N=AY/AX, i.e., the set of loci N contributing to the variance of AY but not to the variance of AX. Z represents the intersection of AX and AY, that is, the set of loci Z underlies both AX and AY, i.e., Z=AY ⋂ AX. Note that the set Z may contain pleiotropic loci, where the pleiotropy is due to direct effects or due to linkage disequilibrium; C) The MR model with a polygenic instrument (PGS) and ‘no pleiotropy’. PGS is associated with X (parameter b1), but is assumed to have no direct influence on the outcome Y. This model holds only if the instrument PGS is constructed on the basis of a subset of variants from set M. In the presence of PGS, A*x is a residual (in the regression of X on PGS). D) MR with pleiotropic genetic instrument. In this model, the PGS is constructed on the basis of a sample of genetic variants taken from set Z. The parameter b2 accommodates the fact that the set of variants used to construct PGS underlies the variance of both AX and AY. The no pleiotropy assumption implies b2=0.

Among the methods of causal effect estimation in the standard MR are the two-stage least squares (2SLS) and the ratio of coefficients. In 2SLS, first, the instrumental variable (e.g., the polygenic score) is used to predict the exposure X, and second, the outcome is regressed on the predicted values of X. In the ratio of coefficients method, the causal effect is computed as a ratio of regression coefficients, with the numerator obtained in the regression of the outcome on the instrument, and the denominator obtained in the regression of the exposure on the instrument 13. Both methods are based on least squares estimation, and are expected to yield equivalent results in MR studies involving a single instrumental variable 13.

The standard MR model can also be fitted in a single step as a structural equation model (as depicted in Figure 3 (panel C), with maximum likelihood (ML) estimation. The causal parameter Embedded Image in Figure 3C can be tested by the means of a likelihood ratio or a Wald test.

Integrating Mendelian Randomization with the Direction of Causation twin model (the MR-DoC model)

In observational studies, MR and DoC twin model offer some traction in testing a hypothesized direction of causality. As demonstrated below, the combined MR-DoC model has definite advantages over the individual approaches, in terms of power and assumptions. Figure 4 displays a path diagram of the MR-DoC model. Note that, the model as depicted is not identified. We consider the issue of identification below.

Figure 4:
Figure 4:

Path diagrammatic representation of the MR-DoC model in DZ twins. The parameter x equals the standard deviation of the observed instrument, i.e., PGS in the circle is standardized. The model as depicted is not identified (see Table 1).

View this table:
Table 1:

Parameter constraints that render identified the Figure 3. Note ‘fr’ indicates that the parameter is estimated, ‘0’ – that the parameter is constrained to equal 0.

The MR-DoC model is based on the following regression model: Embedded Image Embedded Image where i stands for twin pairs, and j for twin (j=1,2). The vector of parameters is θ = [ra, rc, re, ax, cx, ex, ay, cy, ey, g1, b1, b2, x], where x is the standard deviation of the PGS. Here, and in Figure 3, the parameter of interest is g1, as it concerns the causal effect of exposure X on outcome Y. Note that this model now includes the parameter b2, i.e., the pleiotropic effects of PGS on Y, which are usually assumed to be absent. Where we refer to a pleiotropic instrument below, we mean that the parameter b2 is not zero. Using ML estimation, we can test hypotheses concerning the parameters by means of a likelihood ratio or Wald test.

Model identification

Model identification concerns the question whether the observed data provide sufficient information to yield unique estimates of the unknown parameters collected in the vector θ 51. In the present case, the observed information is summarized in the MZ and DZ (6x6) covariance matrices. We assume that the means are equal in the MZ and DZ and are equal in the twin 1 and twin 2 members (this is obviously testable). As the parameterization of the means has no bearing on the identification of the covariance structure model, we do not consider them in addressing identification.

Local identification is evaluated at a given set of parameters, say θa, and implies there are no points in the vicinity of point θa in the parameter space leading to the same expected covariance matrices ΣMZ(θa) and ΣDZ(θa) 51,52. We evaluated local identification using symbolic algebra in Maple 53. Derks et al. 54 previously used this method in the context of twin modeling. Using Maple, we checked whether the rank of the Jacobian matrix is full column rank. The Jacobian matrix contains the first order derivatives of the elements in the matrices ΣMZ(θa) and ΣDZ(θa) with respect to the free parameters. If the Jacobian is not full column rank, we require additional parameter constraints (on the elements in the parameter vector θa). Having established local identification in this manner, we proceeded to address the question of resolution by considering the statistical power to estimate the parameters of interest.

Power calculations

We used exact data simulation 55 to create data that fit a given identified model exactly (i.e., the observed covariance matrices equaled the expected covariance matrices exactly). We then dropped parameters of interest and assessed the power in the standard way using the non-central χ2 distribution, with noncentrality parameter (NCP) λ. We adopted an alpha of .05 throughout. Data were simulated in R using the MASS library function mvrnorm() with the empirical=TRUE option 56. The MR-DoC model was fitted to the population covariance matrices in OpenMx 57. We used the R-package AER 58 to conduct 2SLS estimation in the standard MR using the sample to twin 1 data, i.e., effectively a sample of unrelated individuals. We used the mRnd power calculator to calculate the power of the 2SLS procedure 59. We chose effect sizes by considering the decomposition of the variance of the outcome Y, as illustrated in Figure 4. That is, given that the outcome Y is standardized Embedded Image), we considered the components making up the explained variance, i.e., 1-σ2ξY, as a function of the regression parameters b1, b2, g1, the variances of the residuals ξX and ξY (parameters Embedded Image and Embedded Image), and the covariance between the residuals ξX and ξY (parameter σξXξY).

We varied (a) the strength of the instrumental variable, defined as the percentage of variance explained in the exposure X by instrument (PGS); (b) the variance of ξX (residual X), i.e., parameter σ2ξX, representing the percentage of variance in the exposure, not explained by the instrumental variable); (c) the variance in ξY (residual Y), i.e., parameter σ2ξY, representing the percent of variance in Y not explained by the MR model; and (d) the covariance between ξX and ξY (parameter σξXξY). Using the path tracing rules, we can distinguish five components of variance (C1 to C5, Fig.5) that involve the parameters of interest g1 (the causal effect) and b2 (the direct effect of the instrument on the outcome). In all scenarios, the predictors explained 10% of the phenotypic variance of the outcome Y. The parameter values used in simulations are included in Supplementary Tables S1 and S2. To provide an indication on the potential gains in power conferred by our approach relative to a standard MR analysis of data obtained in unrelated individuals, we report the number of unrelated persons required to attain equivalent power as MR-DoC based on 2000 twin pairs.

Figure 5:
Figure 5:

Effect size calculation for the power analyses. Abbreviations: PGS – polygenic score; X – exposure variable; Y – outcome variable; ξX - residual X; σ2ξX - variance in the exposure, not explained by the instrumental variable PGS; ξY - residual Y; σ2ξY - variance in Y not explained by the MR model; σξXξY - the covariance between ξX and ξY ; b1 – regression coefficient in the regression of the exposure X on the instrument; b2 – regression coefficient in the regression of the outcome on the PGS; g1 – the causal effect of X on Y.

RESULTS

Parameter recovery in standard MR and MR-DoC with valid or invalid (i.e., pleiotropic) instrumental variables (given re = 0)

Table 2 displays results obtained using non-pleiotropic (i.e., Figure 3, b2=0), or pleiotropic (i.e., b2 ≠ 0, see Fig.4) instrumental variables in the causal effect estimation. Given b2=0, results indicate that all estimation methods recover the true parameter value (scenario S1, Table 2). As is to be expected, b2≠0 leads to biased estimates of the causal effect when employing standard MR methods (e.g., 2SLS or ratio of coefficients). MR-DoC recovers correctly the true parameter values (scenario S2, Table 2). Finally, we checked parameter recovery when the instrument has pleiotropic effects, but there is no causal effect of the exposure X on outcome Y (i.e., b2≠0 and g1=0; scenario S3, Table 2). Results showed that standard MR detects a causal effect (when in truth there is none), while MR-DoC does not. We remind the reader that we have set re = 0 to obtain these results.

View this table:
Table 2:

Results of the comparison of the standard MR in analysis of unrelated individuals and MR-DoC twin model. Additive genetic (A), shared environmental (C) and unique environmental (E) effects contributed to the variance of both the exposure Embedded Image and the outcome Embedded Image variable, and genetic and shared environmental factors contributed to the correlations between the traits (rξXξY=0.2). Incorrectly estimated parameter values are in italics.

Model identification

We considered identification in seven models given in Table 1. Model 1, in which all parameters are estimated freely, is not identified. However, constraining to zero any of the parameters re (Model 2), b2 (Model 3) or rc (Model 4) renders the model identified (i.e., Models 2, 3, 4). Conversely, all parameters are identified if the two traits differ with respect to their ACE model. This is the case if, e.g., the exposure is an AE trait and the outcome is an ACE trait (implying the parameters cx and rc are zero, Model 7), or the exposure is an ACE trait and the outcome (conditional on the exposure) is an AE trait (implying the parameters cy and rc to zero, Model 6). Note that in the latter case Y (unconditional) is characterized by shared environmental effects on Y (transmitted from CX, through the g1 path) 41. Furthermore, we found that MR-DoC is not identified when the traits’ variances are limited to two sources (e.g., X and Y are both AE traits, Model 5). Fixing the parameter re to 0 is a constraint commonly employed in the fixed economic within MZ twin intra-pair differences model 44 and in the discordant twin design. The following results are based on the model with this identifying constraint in place, i.e., re = 0.

Power

Figure 6 (and Supplementary Tables S3-S5) displays the results pertaining to the power to detect the causal effect in standard MR and in MR-DoC.

Figure 6:
Figure 6:

Results given a non-pleiotropic instrument (parameter b2=0; left and middle panels), and given a pleiotropic instrument (parameter b2 ≠ 0; right panel). Fitting the model with the parameter g1 freely estimated, and with the parameter g1 constrained to equal zero provided the Non-Centrality Parameter. The standard MR is based on two-stage least squares in a sample of 4000 unrelateds. The MR-DoC twin model used maximum likelihood and a sample of 2000 twin pairs. Abbreviations: rξXξY - residual exposure-outcome correlation; PGS – polygenic score.

With a valid instrumental variable (no pleiotropy) and all parameters freely estimated (including the parameter re, Table S3), the main factors that influence MR-DoC’s power are: (a) instrument’s strength; (b) the genetic covariance structure of X (exposure) and Y (outcome); and (c) the magnitude of the residual X-Y correlation. As expected, and consistent with the standard MR literature, increasing instrument’s strength increases power. For instance, with an ACE trait as the exposure variable Embedded Image, an outcome variable having roughly the same mode of inheritance Embedded Image, and a residual correlation of rξXξY=.2, the power of the MR-DoC equals .627 and .905 when the instrument explains 5% and 10% of the variance in the exposure, respectively. However, contrary to the standard DoC literature, having traits with similar genetic covariance structure has no bearing on MR-DoC’s power to detect the causal effect. This is the case for instance, if both the outcome and exposure are ACE traits. Power is the highest when the outcome variable has a lower heritability than the exposure variable. For example, power increases from .622 in Scenario S1G (with a 50% heritable outcome and a 20% heritable exposure) to .658 in Scenario S1I (with a 20% heritable outcome variable and a 50% heritable exposure; Table S3). Finally, increasing X-Y residual correlation reduces MR-DoC’s power. For instance, with an outcome and an exposure having roughly the same mode of inheritance, and an instrument explaining 5% of the variance in the exposure, MR-DoC’s power drops from .627 (Scenario S1H, Table S3) to .312 (Scenario S1B) when the residual correlation increases from rξXξY=0.2 to rξXξY=0.4.

Given equal N, standard MR (4000 unrelateds) has larger power than MR-DoC (4000 twins) only when b2=0 and re ≠0 (Fig.6, left). Yet, assuming data collected in twins are readily available, greater statistical power is available in the MR-DoC model (than from reducing the twin pairs to singletons and resorting to standard MR, Table S4).

We also calculated the required N of unrelated individuals to achieve the same power as 2000 twin pairs (Table 3). We found the yield of MR-DoC substantial if the parameter re was fixed to zero (as simulated, Fig.6, middle). For example, with a sample of 2000 twin pairs, MR-DoC yields a NCP λ of 38.68 given an instrument explaining 10% of the variance in the exposure, a residual correlation of rξXξY=0.2, an exposure variable with low heritability Embedded Image, and a moderately heritable outcome Embedded Image. Standard MR needs about 56 737 unrelated individuals to achieve equivalent power (Table 3).

View this table:
Table 3:

The number of unrelated individuals (N) required by MR-standard to achieve the same NCP as MR-DoC in 2000 twin pairs. The number of unrelateds was estimated based on the NCP obtained by fitting the standard MR as a structural equation model, with estimation of the causal effect based on maximum likelihood (similar to MR-DoC).

Given b2≠0 and re=0, MR-DoC’s power increases with: (a) decreasing X-Y residual correlation, and (b) decreasing heritability of the exposure (Fig.6 right; see for details, Tables S2 and S5). Regarding the former, power is always larger when the association between the exposure and the outcome is largely causal in nature (i.e., when the residual correlation drops from rξXξY = 0.4 to rξXξY = 0.2). Regarding the latter, power is always greater in scenarios where the exposure has low heritability. For instance, the NCP λ increases from 7.29 to 39.1 when the heritability of the exposure decreases from 50% (S2I) to 20% (S2J, Table S5). The same pattern of results was observed when b2=0 and re=0 (see Fig.6 middle, and Table S4). Note that the instrument’s strength has no longer a bearing on power when b2≠0, i.e., when it affects the outcome both directly (parameter b2) and indirectly, via the exposure (parameter b1). For instance, 2000 twin pairs yield an NCP λ of 5.98 given two traits having identical variance components Embedded Image, and a large residual correlation (rξXξY = 0.4), regardless of whether the instrument explains 5% or 10% of the variance in the exposure (Scenarios S2B and S2E, Table S5). As mentioned above, contrary to the standard DoC literature 43, MR-DoC does not require the exposure and the outcome variable to have radically different covariance structures to ensure sufficient power to test unidirectional causal hypotheses.

DISCUSSION

Our aim was to integrate MR and the classical twin model to render testable the MR’s strong assumption that the instrumental variable has no direct effect on the outcome, conditional on the exposure and confounders. We showed that, with standard MR methods, violations of this assumption readily lead to biased causal effect estimates, or may even yield spurious false positives. MR-DoC correctly detects causal effects and provides accurate parameter estimates even if the instrument is pleiotropic. With traits that have the same covariance structure (e.g., when both the outcome and exposure are ACE traits), the weaker assumption (also used in the discordant twin design) that the unique environmental component influencing the exposure, influences the outcome only via its effect on the exposure, but not directly (i.e., re=0), is needed to identify model parameters. Note, however, that this assumption is not required with traits that have different covariance structures (e.g., with an ACE and an AE trait).

Aside from providing the means to relax the ‘no pleiotropy’ assumption, the MR-DoC twin model confers several other advantages in understanding causal relationships between exposures and outcomes. First, MR-DoC provides full statistical description of the observed exposure-outcome relationship, allowing one to disentangle the causal effect of the exposure on the outcome, from potential pleiotropic effects of the instrumental variable, as well as from the contribution of other genetic and environmental factors to the variances and the covariance of the two traits. Second, the twin data provide sufficient information to estimate the direct path from the instrument to the outcome (i.e., the parameter b2 in Fig.3). Importantly, this path captures not only pleiotropic effects, but also possible effects of other variants affecting the outcome which are in linkage disequilibrium with the instrument. Third, our approach opens up the possibility of using strong genetic instrumental variables in the form of polygenic scores. This is generally desirable in the standard MR design, as the strength of the valid instrumental variable has a bearing on the precision of the causal estimate (i.e., with weak instruments the estimate tends to approach the observed OLS exposure-outcome association 60), and on the distribution of the causal effect estimate (the weaker the instrument, the more skewed the distribution; see Fig.7.1 in 13). As a consequence, significance tests and the construction of the confidence intervals, which rely on asymptotic normality, are no longer accurate. Consequently, tests may suffer inflated type I error rates 61. In addition, strong instruments are desirable from the perspective of power, as our results showed (consistent with the literature 61).

Interestingly, the strength of the instrument (i.e., defined in terms of percentage of explained variance in the exposure), has no bearing on the power when the instrument has pleiotropic effects, given that these effects are correctly modeled (in MR-DoC). The reason for this is that the exposure no longer features as a full mediator variable in the presence of a direct path between the instrument and the outcome. Correspondingly, the misfit due to dropping the causal parameter from the analysis is attenuated by the presence of the parameter b2 in the model. That is, fixing the parameter g1 to zero will largely bias the parameter b2, but will not affect the between twin covariance matrix (as is the case in the standard MR design where the sole path from the instrument to the outcome is via the exposure). Stated otherwise, the bias in b2 will be greatest, leaving ra (Ax,Ay) and rc (Cx,Cy) relatively unaffected, regardless of how strong the instrument is. In this circumstance, the power to detect the causal effect will primarily depend on the magnitude of the residual correlation between the two traits and, to a lesser extent, on their modes of inheritance.

MR-DoC is tailored for the readily available datasets collected worldwide on more than 1.5 million twins at the Twin and Family Registries (see 62 for details on these rich resources of genotypic and phenotypic data). We showed that twin data correctly detect causation and estimate effect sizes even in the presence of pleiotropy. Although primarily developed to address the ‘no pleiotropy’ assumption, our results demonstrate that MR-DoC has greater statistical power than standard MR analysis in singletons if the parameter re is zero (as fixed to zero in the model). With re fixed to zero, dropping the causal parameter g1 from the model greatly impacts all paths connecting the exposure and the outcome, both within and across twins (parameters ra and rc), creating a large discrepancy between the observed and the expected covariance matrix. This misfit is evident throughout the covariance matrix. Consequently, in some specific circumstances, testing causal hypotheses requires tens of thousands of unrelated individuals for one to achieve the same power as that conferred by several hundreds of twin pairs. It should be noted that unlike MR methods 37,38 or asymmetry tests 34 based on summary statistics, the approach proposed here only requires GWAS results for the exposure variable. With MR-DoC and these rich phenotype resources - ranging from personality, diet and lifestyle to disease and psychiatric traits 62 - collected at Twin and Family registries, the availability of genetic instruments robustly associated with the exposure remains the main limiting factor in addressing causal questions in non-experimental settings.

As presented here, MR-DoC is limited to twin data, but note that twin registries often have available information on additional family members 62-66. Conversely, not all cohort studies necessarily include related individuals. To expand the model, we aim to accommodate additional family members (siblings and parents), and distantly related individuals by using Genetic Relationship Matrices (GRMs) based on average allelic correlations (where the alleles are defined at the measured single nucleotide polymorphisms). We anticipate that these extensions will further increase statistical power and robustness to assumption violation 67,68. Second, throughout the paper we assumed that the mating is random, there is no genotype-environment correlation, and no genotype by environment interaction. Indeed, assumption violation may also arise because the mating is assortative 13, or because there are other plausible paths from the instrument to the outcome (except direct paths, or indirect, via the exposure), for example, via confounders affecting both traits, i.e., implying genotype by environment interaction or genotype by environment correlation. However, we note that these effects may be captured by the MR-DoC twin model with appropriate experimental designs 67,69. On a cautionary note, although valid strong instruments are desirable in MR from the perspective of power, making up the polygenic score based on variants of unknown function should limit the testable hypotheses to whether the model is consistent or not with a direct causal effect from the exposure to the outcome (as pointed out in 21). While we considered the use of MR-DoC with polygenic scores, our conclusions also hold in scenarios where a genetic variant with known function is used as the instrument, which would improve the biological interpretation of the causal effect.

In conclusion, by integrating Mendelian Randomization with the Direction of Causation twin model, we developed a model that allows one to test and relax the strong ‘no pleiotropy’ assumption employed by standard MR. This approach therefore allows one to employ strong instrumental variables in the form of polygenic scores, guarding against the weak instrument bias and increasing the power to detect causal effects of exposures on potential outcomes. We anticipate that MR-DoC will enhance and extend MR’s range of applications, and increase the value of the large cohorts collected at twin registries as they correctly detect causation and estimate effect sizes even in the presence of pleiotropy.

FUNDING

This work was supported by the National Institute on Drug Abuse [grant number DA-018673].

REFERENCES

  1. 1.↵
    Katan, M. B. Apolipoprotein E isoforms, serum cholesterol, and cancer. Int J Epidemiol 33, 9–9 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  2. 2.↵
    Evans, D. M. & Davey Smith, G. Mendelian randomization: new applications in the coming age of hypothesis-free causality. Annual review of genomics and human genetics 16, 327–350 (2015).
    OpenUrlCrossRefPubMed
  3. 3.↵
    Smith, G. D. & Hemani, G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum. Mol. Genet. 23, R89–R98 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    Ference, B. A., Majeed, F., Penumetcha, R., Flack, J. M. & Brook, R. D. Effect of naturally random allocation to lower low-density lipoprotein cholesterol on the risk of coronary heart disease mediated by polymorphisms in NPC1L1, HMGCR, or both: a 2× 2 factorial Mendelian randomization study. Journal of the American College of Cardiology 65, 1552–1561 (2015).
    OpenUrlFREE Full Text
  5. 5.
    Vimaleswaran, K. S. et al. Causal relationship between obesity and vitamin D status: bi-directional Mendelian randomization analysis of multiple cohorts. PLoS Med 10, e1001383 (2013).
    OpenUrlCrossRefPubMed
  6. 6.
    Holmes, M. V. et al. Causal effects of body mass index on cardiometabolic traits and events: a Mendelian randomization analysis. The American Journal of Human Genetics 94, 198–208 (2014).
    OpenUrlCrossRefPubMed
  7. 7.
    Holmes, M. V. et al. Mendelian randomization of blood lipids for coronary heart disease. European heart journal, eht571 (2014).
  8. 8.
    Proitsi, P. et al. Genetic predisposition to increased blood cholesterol and triglyceride lipid levels and risk of Alzheimer disease: a Mendelian randomization analysis. PLoS Med 11, e1001713 (2014).
    OpenUrlCrossRefPubMed
  9. 9.
    Ference, B. A. et al. Effect of long-term exposure to lower low-density lipoprotein cholesterol beginning early in life on the risk of coronary heart disease: a Mendelian randomization analysis. Journal of the American College of Cardiology 60, 2631–2639 (2012).
    OpenUrlFREE Full Text
  10. 10.↵
    Holmes, M. V. et al. Association between alcohol and cardiovascular disease: Mendelian randomisation analysis based on individual participant data. Bmj 349, g4164 (2014).
    OpenUrlAbstract/FREE Full Text
  11. 11.↵
    Welter, D. et al. The NHGRI GWAS Catalog, a curated resource of SNP-trait associations. Nucleic Acids Res. 42, D1001–D1006 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.↵
    Smith, G. D. Use of genetic markers and gene-diet interactions for interrogating population-level causal influences of diet on health. Genes Nutr. 6, 27–43 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    Burgess, S. & Thompson, S. G. Mendelian randomization: methods for using genetic variants in causal estimation. (CRC Press, 2015).
  14. 14.↵
    Visscher, P. M., Brown, M. A., McCarthy, M. I. & Yang, J. Five years of GWAS discovery. The American Journal of Human Genetics 90, 7–24 (2012).
    OpenUrlCrossRefPubMed
  15. 15.↵
    Davies, N. M. et al. The many weak instruments problem and Mendelian randomization. Stat. Med. 34, 454–468 (2015).
    OpenUrlCrossRefPubMed
  16. 16.↵
    Burgess, S., Thompson, S. G. & Collaboration, C. C. G. Avoiding bias from weak instruments in Mendelian randomization studies. Int. J. Epidemiol. 40, 755–764 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    Burgess, S. & Thompson, S. G. Use of allele scores as instrumental variables for Mendelian randomization. Int. J. Epidemiol. 42, 1134–1144 (2013).
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.
    Palmer, T. M. et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat. Methods Med. Res. 21, 223–242 (2012).
    OpenUrlCrossRefPubMed
  19. 19.↵
    Pierce, B. L., Ahsan, H. & VanderWeele, T. J. Power and instrument strength requirements for Mendelian randomization studies using multiple genetic variants. Int. J. Epidemiol., dyq151 (2010).
  20. 20.↵
    Sheehan, N. A., Didelez, V., Burton, P. R. & Tobin, M. D. Mendelian randomisation and causal inference in observational epidemiology. PLoS Med 5, e177 (2008).
    OpenUrlCrossRefPubMed
  21. 21.↵
    Burgess, S., Freitag, D. F., Khan, H., Gorman, D. N. & Thompson, S. G. Using multivariable Mendelian randomization to disentangle the causal effects of lipid fractions. PloS one 9, e108891 (2014).
    OpenUrlCrossRefPubMed
  22. 22.↵
    Middeldorp, C., Cath, D., Van Dyck, R. & Boomsma, D. The co-morbidity of anxiety and depression in the perspective of genetic epidemiology. A review of twin and family studies. Psychological medicine 35, 611–624 (2005).
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.
    Kendler, K. S., Neale, M. C., Kessler, R. C., Heath, A. C. & Eaves, L. J. Major depression and generalized anxiety disorder: same genes,(partly) different environments? Archives of general psychiatry 49, 716–722 (1992).
    OpenUrlCrossRefPubMedWeb of Science
  24. 24.
    Neale, M. C. & Kendler, K. S. Models of comorbidity for multifactorial disorders. Am J Hum Genet 57, 935 (1995).
    OpenUrlPubMedWeb of Science
  25. 25.
    de Geus, E. The association between depression and heart disease: The role of genetic factors. Depression and heart disease, 57–98 (2011).
  26. 26.
    de Geus, E. J. Genetic pleiotropy in depression and coronary artery disease. Psychosomatic medicine 68, 185–186 (2006).
    OpenUrlFREE Full Text
  27. 27.↵
    Ligthart, L. & Boomsma, D. I. Causes of comorbidity: pleiotropy or causality? Shared genetic and environmental influences on migraine and neuroticism. Twin Research and Human Genetics 15, 158–165 (2012).
    OpenUrlCrossRefWeb of Science
  28. 28.↵
    Evans, D. M., Visscher, P. M. & Wray, N. R. Harnessing the information contained within genome-wide association studies to improve individual prediction of complex disease risk. Hum Mol Genet 18, 3525–3531 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  29. 29.
    Purcell, S. M. et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460, 748–752 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  30. 30.↵
    Ligthart, L. et al. Genetic risk score analysis indicates migraine with and without comorbid depression are genetically different disorders. Human genetics 133, 173–186 (2014).
    OpenUrlCrossRefPubMed
  31. 31.↵
    Sivakumaran, S. et al. Abundant pleiotropy in human complex diseases and traits. The American Journal of Human Genetics 89, 607–618 (2011).
    OpenUrlCrossRefPubMed
  32. 32.
    Bulik-Sullivan, B. et al. An atlas of genetic correlations across human diseases and traits. Nat. Genet. (2015).
  33. 33.
    Solovieff, N., Cotsapas, C., Lee, P. H., Purcell, S. M. & Smoller, J. W. Pleiotropy in complex traits: challenges and strategies. Nature Reviews Genetics 14, 483–495 (2013).
    OpenUrlCrossRefPubMed
  34. 34.↵
    Pickrell, J. K. et al. Detection and interpretation of shared genetic influences on 42 human traits. Nature genetics (2016).
  35. 35.↵
    Visscher, P. M. & Yang, J. A plethora of pleiotropy across complex traits. Nature genetics 48, 707 (2016).
    OpenUrlCrossRef
  36. 36.↵
    Johnson, T. & Johnson, M. T. Package ’gtx’. (2012).
  37. 37.↵
    Bowden, J., Davey Smith, G., Haycock, P. C. & Burgess, S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet. Epidemiol. 40, 304–314 (2016).
    OpenUrlCrossRefPubMed
  38. 38.↵
    Bowden, J., Smith, G. D. & Burgess, S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 44, 512–525 (2015).
    OpenUrlCrossRefPubMed
  39. 39.↵
    Bowden, J. & Jackson, C. On the physical interpretation of a meta-analysis in the presence of heterogeneity and bias: from clinical trials to Mendelian randomization. arXiv preprint arXiv:1508.03768 (2015).
  40. 40.↵
    Bowden, F. J., Burgess, S. & Smith, G. D. Response to Hartwig and Davies. Int. J. Epidemiol., dyw252 (2016).
  41. 41.↵
    De Moor, M. H., Boomsma, D. I., Stubbe, J. H., Willemsen, G. & de Geus, E. J. Testing causality in the association between regular exercise and symptoms of anxiety and depression. Arch. Gen. Psychiatry 65, 897–905 (2008).
    OpenUrlCrossRefPubMedWeb of Science
  42. 42.↵
    Duffy, D. L. & Martin, N. G. Inferring the direction of causation in cross-sectional twin data: Theoretical and empirical considerations. Genet. Epidemiol. 11, 483–502 (1994).
    OpenUrlCrossRefPubMedWeb of Science
  43. 43.↵
    Heath, A. C. et al. Testing hypotheses about direction of causation using cross-sectional family data. Behav. Genet. 23, 29–50 (1993).
    OpenUrlCrossRefPubMedWeb of Science
  44. 44.↵
    Kohler, H.-P., Behrman, J. R. & Schnittker, J. Social science methods for twins data: Integrating causality, endowments, and heritability. Biodemography Soc. Biol. 57, 88–141 (2011).
    OpenUrlPubMed
  45. 45.↵
    1. Harry T Reis &
    2. Charles M Judd
    Turkheimer, E. & Harden, K. P. in Handbook of research methods in social and personality psychology (eds Harry T Reis & Charles M Judd) (Cambridge University Press, 2000).
  46. 46.↵
    Gillespie, N. A., Zhu, G., Neale, M. C., Heath, A. C. & Martin, N. G. Direction of causation modeling between cross-sectional measures of parenting and psychological distress in female twins. Behavior Genetics 33, 383–396 (2003).
    OpenUrlCrossRefPubMed
  47. 47.
    Verhulst, B., Eaves, L. J. & Hatemi, P. K. Correlation not causation: The relationship between personality traits and political ideologies. American Journal of Political Science 56, 34–51 (2012).
    OpenUrlCrossRefPubMed
  48. 48.
    Verhulst, B. & Estabrook, R. Using genetic information to test causal relationships in cross-sectional data. Journal of theoretical politics 24, 328–344 (2012).
    OpenUrlCrossRefWeb of Science
  49. 49.↵
    Neale, M. & Cardon, L. Methodology for genetic studies of twins and families. Vol. 67 (Springer Science & Business Media, 1992).
  50. 50.↵
    Mather, K. & Jinks, J. L. Introduction to biometrical genetics. (Springer Science & Business Media, 2012).
  51. 51.↵
    Bollen, K. A. & Bauldry, S. Model identification and computer algebra. Sociological methods & research 39, 127–156 (2010).
    OpenUrlCrossRefPubMedWeb of Science
  52. 52.↵
    Bekker, P. A., Merckens, A. & Wansbeek, T. J. Identification, Equivalent Models, and Computer Algebra: Statistical Modeling and Decision Science. (Academic Press, 2014).
  53. 53.↵
    Morgan, M. B. et al. Maple 10 Programming Guide. (Maplesoft, 2005).
  54. 54.↵
    Derks, E. M., Dolan, C. V. & Boomsma, D. I. A test of the equal environment assumption (EEA) in multivariate twin studies. Twin Research and Human Genetics 9, 403–411 (2006).
    OpenUrlCrossRef
  55. 55.↵
    Van Der Sluis, S., Dolan, C. V., Neale, M. C. & Posthuma, D. Power calculations using exact data simulation: a useful tool for genetic study designs. Behav. Genet. 38, 202–211 (2008).
    OpenUrlPubMed
  56. 56.↵
    R-Core-Team. R: A language and environment for statistical computing.. R Foundation for Statistical Computing, Vienna, Austria. (2015).
  57. 57.↵
    Neale, M. C. et al. OpenMx 2.0: Extended structural equation and statistical modeling. Psychometrika 81, 535–549 (2016).
    OpenUrl
  58. 58.↵
    Kleiber, C. & Zeileis, A. Applied Econometrics with R. (Springer-Verlag, 2008).
  59. 59.↵
    Brion, M.-J. A., Shakhbazov, K. & Visscher, P. M. Calculating statistical power in Mendelian randomization studies. Int. J. Epidemiol. 42, 1497–1501 (2013).
    OpenUrlCrossRefPubMedWeb of Science
  60. 60.↵
    Bound, J., Jaeger, D. A. & Baker, R. M. Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak. Journal of the American statistical association 90, 443–450 (1995).
    OpenUrlCrossRefWeb of Science
  61. 61.↵
    Burgess, S., Small, D. S. & Thompson, S. G. A review of instrumental variable estimators for Mendelian randomization. Statistical methods in medical research, 0962280215597579 (2015).
  62. 62.↵
    Twin, Registries & Worldwide. Twin registries worldwide: An important resource for scientific research. Twin Research and Human Genetics 16 (2013).
  63. 63.
    Kaprio, J. The Finnish Twin Cohort Study: An Update. Twin Research and Human Genetics 16, 157–162, doi:10.1017/thg.2012.142 (2013).
    OpenUrlCrossRefPubMed
  64. 64.
    Skytthe, A. et al. The Danish Twin Registry: Linking Surveys, National Registers, and Biological Information. Twin Research and Human Genetics 16, 104–111, doi:10.1017/thg.2012.77 (2012).
    OpenUrlCrossRef
  65. 65.
    van Beijsterveldt, C. E. et al. The Young Netherlands Twin Register (YNTR): longitudinal twin and family studies in over 70,000 children. Twin Research and Human Genetics 16, 252–267 (2013).
    OpenUrlCrossRefWeb of Science
  66. 66.↵
    Willemsen, G. et al. The Adult Netherlands Twin Register: twenty-five years of survey and biological data collection. Twin Research and Human Genetics 16, 271–281 (2013).
    OpenUrlCrossRefWeb of Science
  67. 67.↵
    Keller, M. C., Medland, S. E. & Duncan, L. E. Are extended twin family designs worth the trouble? A comparison of the bias, precision, and accuracy of parameters estimated in four twin family models. Behav. Genet. 40, 377–393 (2010).
    OpenUrlCrossRefPubMed
  68. 68.↵
    Posthuma, D. & Boomsma, D. I. A note on the statistical power in extended twin designs. Behav. Genet. 30, 147–158 (2000).
    OpenUrlCrossRefPubMedWeb of Science
  69. 69.↵
    Plomin, R., DeFries, J. C., Knopik, V. S. & Neiderheiser, J. Behavioral genetics. (Palgrave Macmillan, 2013).
Back to top
Posted May 05, 2017.
Download PDF
Email
Extending causality tests with genetic instruments: an integration of Mendelian Randomization and the Classical Twin Design
Camelia C. Minică, Conor V. Dolan, Dorret I. Boomsma, Eco de Geus, Michael C. Neale
bioRxiv 134585; doi: https://doi.org/10.1101/134585
Digg logo Reddit logo Twitter logo Facebook logo Google logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area