A robust and adaptive framework for interaction testing in quantitative traits between multiple genetic loci and exposure variables.

The identification and understanding of gene-environment interactions can provide insights into the pathways and mechanisms underlying complex diseases. However, testing for gene-environment interaction remains a challenge since a.) statistical power is often limited and b.) modeling of environmental effects is nontrivial and such model misspecifications can lead to false positive interaction findings. To address the lack of statistical power, recent methods aim to identify interactions on an aggregated level using, for example, polygenic risk scores. While this strategy can increase the power to detect interactions, identifying contributing genes and pathways is difficult based on these relatively global results. Here, we propose RITSS (Robust Interaction Testing using Sample Splitting), a gene-environment interaction testing framework for quantitative traits that is based on sample splitting and robust test statistics. RITSS can incorporate sets of genetic variants and/or multiple environmental factors. Based on the user's choice of statistical/machine learning approaches, a screening step selects and combines potential interactions into scores with improved interpretability. In the testing step, the application of robust statistics minimizes the susceptibility to main effect misspecifications. Using extensive simulation studies, we demonstrate that RITSS controls the type 1 error rate in a wide range of scenarios, and we show how the screening strategy influences statistical power. In an application to lung function phenotypes and human height in the UK Biobank, RITSS identified highly significant interactions based on subcomponents of genetic risk scores. While the contributing single variant interaction signals are weak, our results indicate interaction patterns that result in strong aggregated effects, providing potential insights into underlying gene-environment interaction mechanisms.

Dealing with time-dependent exposures and confounding when defining and estimating attributable fractions-Revisiting estimands and estimators.

The population-attributable fraction (PAF) is commonly interpreted as the proportion of events that can be ascribed to a certain exposure in a certain population. Its estimation is sensitive to common forms of time-dependent bias in the face of a time-dependent exposure. Predominant estimation approaches based on multistate modeling fail to fully eliminate such bias and, as a result, do not permit a causal interpretation, even in the absence of confounding. While recently proposed multistate modeling approaches can successfully eliminate residual time-dependent bias, and moreover succeed to adjust for time-dependent confounding by means of inverse probability of censoring weighting, inadequate application, and misinterpretation prevails in the medical literature. In this paper, we therefore revisit recent work on previously proposed PAF estimands and estimators in settings with time-dependent exposures and competing events and extend this work in several ways. First, we critically revisit the interpretation and applied terminology of these estimands. Second, we further formalize the assumptions under which a causally interpretable PAF estimand can be identified and provide analogous weighting-based representations of the identifying functionals of other proposed estimands. This representation aims to enhance the applied statistician's understanding of different sources of bias that may arise when the aim is to obtain a valid estimate of a causally interpretable PAF. To illustrate and compare these representations, we present a real-life application to observational data from the Ghent University Hospital ICUs to estimate the fraction of ICU deaths attributable to hospital-acquired infections.

Inverse probability of treatment weighting with generalized linear outcome models for doubly robust estimation.

There are now many options for doubly robust estimation; however, there is a concerning trend in the applied literature to believe that the combination of a propensity score and an adjusted outcome model automatically results in a doubly robust estimator and/or to misuse more complex established doubly robust estimators. A simple alternative, canonical link generalized linear models (GLM) fit via inverse probability of treatment (propensity score) weighted maximum likelihood estimation followed by standardization (the g $$ g $$ -formula) for the average causal effect, is a doubly robust estimation method. Our aim is for the reader not just to be able to use this method, which we refer to as IPTW GLM, for doubly robust estimation, but to fully understand why it has the doubly robust property. For this reason, we define clearly, and in multiple ways, all concepts needed to understand the method and why it is doubly robust. In addition, we want to make very clear that the mere combination of propensity score weighting and an adjusted outcome model does not generally result in a doubly robust estimator. Finally, we hope to dispel the misconception that one can adjust for residual confounding remaining after propensity score weighting by adjusting in the outcome model for what remains 'unbalanced' even when using doubly robust estimators. We provide R code for our simulations and real open-source data examples that can be followed step-by-step to use and hopefully understand the IPTW GLM method. We also compare to a much better-known but still simple doubly robust estimator.

The potential impact fraction of population weight reduction scenarios on non-communicable diseases in Belgium: application of the g-computation approach.

BACKGROUND: Overweight is a major risk factor for non-communicable diseases (NCDs) in Europe, affecting almost 60% of all adults. Tackling obesity is therefore a key long-term health challenge and is vital to reduce premature mortality from NCDs. Methodological challenges remain however, to provide actionable evidence on the potential health benefits of population weight reduction interventions. This study aims to use a g-computation approach to assess the impact of hypothetical weight reduction scenarios on NCDs in Belgium in a multi-exposure context. METHODS: Belgian health interview survey data (2008/2013/2018, n = 27 536) were linked to environmental data at the residential address. A g-computation approach was used to evaluate the potential impact fraction (PIF) of population weight reduction scenarios on four NCDs: diabetes, hypertension, cardiovascular disease (CVD), and musculoskeletal (MSK) disease. Four scenarios were considered: 1) a distribution shift where, for each individual with overweight, a counterfactual weight was drawn from the distribution of individuals with a "normal" BMI 2) a one-unit reduction of the BMI of individuals with overweight, 3) a modification of the BMI of individuals with overweight based on a weight loss of 10%, 4) a reduction of the waist circumference (WC) to half of the height among all people with a WC:height ratio greater than 0.5. Regression models were adjusted for socio-demographic, lifestyle, and environmental factors. RESULTS: The first scenario resulted in preventing a proportion of cases ranging from 32.3% for diabetes to 6% for MSK diseases. The second scenario prevented a proportion of cases ranging from 4.5% for diabetes to 0.8% for MSK diseases. The third scenario prevented a proportion of cases, ranging from 13.6% for diabetes to 2.4% for MSK diseases and the fourth scenario prevented a proportion of cases ranging from 36.4% for diabetes to 7.1% for MSK diseases. CONCLUSION: Implementing weight reduction scenarios among individuals with excess weight could lead to a substantial and statistically significant decrease in the prevalence of diabetes, hypertension, cardiovascular disease (CVD), and musculoskeletal (MSK) diseases in Belgium. The g-computation approach to assess PIF of interventions represents a straightforward approach for drawing causal inferences from observational data while providing useful information for policy makers.

Counterfactual Mediation Analysis with a Latent Class Exposure.

Latent classes are a useful tool in developmental research, however there are challenges associated with embedding them within a counterfactual mediation model. We develop and test a new method "updated pseudo class draws (uPCD)" to examine the association between a latent class exposure and distal outcome that could easily be extended to allow the use of any counterfactual mediation method. UPCD extends an existing group of methods (based on pseudo class draws) that assume that the true values of the latent class variable are missing, and need to be multiply imputed using class membership probabilities. We simulate data based on the Avon Longitudinal Study of Parents and Children, examine performance for existing techniques to relate a latent class exposure to a distal outcome ("one-step," "bias-adjusted three-step," "modal class assignment," "non-inclusive pseudo class draws," and "inclusive pseudo class draws") and compare bias in parameter estimates and their precision to uPCD when estimating counterfactual mediation effects. We found that uPCD shows minimal bias when estimating counterfactual mediation effects across all levels of entropy. UPCD performs similarly to recommended methods (one-step and bias-adjusted three-step), but provides greater flexibility and scope for incorporating the latent grouping within any commonly-used counterfactual mediation approach.

Understanding the effects of a complex psychological intervention on symptoms of depression in Goa, India: findings from a causal mediation analysis.

BACKGROUND: Understanding how and under what circumstances a highly effective psychological intervention, improved symptoms of depression is important to maximise its clinical effectiveness. AIMS: To address this complexity, we estimate the indirect effects of potentially important mediators to improve symptoms of depression (measured with the Patient Health Questionnaire (PHQ-9)) in the Healthy Activity Program trial. METHOD: Interventional in(direct) effects were used to decompose the total effect of the intervention on PHQ-9 scores into the direct and indirect effects. The following indirect effects were considered: characteristics of sessions, represented by the number of sessions and homework completed; behavioural activation, according to an adapted version of the Behavioural Activation for Depression Scale - Short Form; and extra sessions offered to participants who did not respond to the intervention. RESULTS: Of the total effect of the intervention measured through the difference in PHQ-9 scores between treatment arms (mean difference: -2.1, bias-corrected 95% CI -3.2 to -1.5), 34% was mediated through improved levels of behavioural activation (mean difference: -0.7, bias-corrected 95% CI -1.2 to -0.4). There was no evidence to support the mediating role of characteristics of the sessions nor the extra sessions offered to participants who did not respond to the treatment. CONCLUSIONS: Findings from our robust mediation analyses confirmed the importance of targeting behavioural activation. Contrary to published literature, our findings suggest that neither the number of sessions nor proportion of homework completed improved outcomes. Moreover, in this context, alternative treatments other than extra sessions should be considered for patients who do not respond to the intervention.

Ensuring valid inference for Cox hazard ratios after variable selection.

The problem of how to best select variables for confounding adjustment forms one of the key challenges in the evaluation of exposure effects in observational studies, and has been the subject of vigorous recent activity in causal inference. A major drawback of routine procedures is that there is no finite sample size at which they are guaranteed to deliver exposure effect estimators and associated confidence intervals with adequate performance. In this work, we will consider this problem when inferring conditional causal hazard ratios from observational studies under the assumption of no unmeasured confounding. The major complication that we face with survival data is that the key confounding variables may not be those that explain the censoring mechanism. In this paper, we overcome this problem using a novel and simple procedure that can be implemented using off-the-shelf software for penalized Cox regression. In particular, we will propose tests of the null hypothesis that the exposure has no effect on the considered survival endpoint, which are uniformly valid under standard sparsity conditions. Simulation results show that the proposed methods yield valid inferences even when covariates are high-dimensional.

Instrumental variable estimation of the causal hazard ratio.

Cox's proportional hazards model is one of the most popular statistical models to evaluate associations of exposure with a censored failure time outcome. When confounding factors are not fully observed, the exposure hazard ratio estimated using a Cox model is subject to unmeasured confounding bias. To address this, we propose a novel approach for the identification and estimation of the causal hazard ratio in the presence of unmeasured confounding factors. Our approach is based on a binary instrumental variable, and an additional no-interaction assumption in a first-stage regression of the treatment on the IV and unmeasured confounders. We propose, to the best of our knowledge, the first consistent estimator of the (population) causal hazard ratio within an instrumental variable framework. A version of our estimator admits a closed-form representation. We derive the asymptotic distribution of our estimator and provide a consistent estimator for its asymptotic variance. Our approach is illustrated via simulation studies and a data application.

Rejoinder to discussions on "Instrumental variable estimation of the causal hazard ratio".

In this paper, we respond to comments on our paper, "Instrumental variable estimation of the causal hazard ratio."

Causal inference in survival analysis using longitudinal observational data: Sequential trials and marginal structural models.

Longitudinal observational data on patients can be used to investigate causal effects of time-varying treatments on time-to-event outcomes. Several methods have been developed for estimating such effects by controlling for the time-dependent confounding that typically occurs. The most commonly used is marginal structural models (MSM) estimated using inverse probability of treatment weights (IPTW) (MSM-IPTW). An alternative, the sequential trials approach, is increasingly popular, and involves creating a sequence of "trials" from new time origins and comparing treatment initiators and non-initiators. Individuals are censored when they deviate from their treatment assignment at the start of each "trial" (initiator or noninitiator), which is accounted for using inverse probability of censoring weights. The analysis uses data combined across trials. We show that the sequential trials approach can estimate the parameters of a particular MSM. The causal estimand that we focus on is the marginal risk difference between the sustained treatment strategies of "always treat" vs "never treat." We compare how the sequential trials approach and MSM-IPTW estimate this estimand, and discuss their assumptions and how data are used differently. The performance of the two approaches is compared in a simulation study. The sequential trials approach, which tends to involve less extreme weights than MSM-IPTW, results in greater efficiency for estimating the marginal risk difference at most follow-up times, but this can, in certain scenarios, be reversed at later time points and relies on modelling assumptions. We apply the methods to longitudinal observational data from the UK Cystic Fibrosis Registry to estimate the effect of dornase alfa on survival.

Using random-forest multiple imputation to address bias of self-reported anthropometric measures, hypertension and hypercholesterolemia in the Belgian health interview survey.

BACKGROUND: In many countries, the prevalence of non-communicable diseases risk factors is commonly assessed through self-reported information from health interview surveys. It has been shown, however, that self-reported instead of objective data lead to an underestimation of the prevalence of obesity, hypertension and hypercholesterolemia. This study aimed to assess the agreement between self-reported and measured height, weight, hypertension and hypercholesterolemia and to identify an adequate approach for valid measurement error correction. METHODS: Nine thousand four hundred thirty-nine participants of the 2018 Belgian health interview survey (BHIS) older than 18 years, of which 1184 participated in the 2018 Belgian health examination survey (BELHES), were included in the analysis. Regression calibration was compared with multiple imputation by chained equations based on parametric and non-parametric techniques. RESULTS: This study confirmed the underestimation of risk factor prevalence based on self-reported data. With both regression calibration and multiple imputation, adjusted estimation of these variables in the BHIS allowed to generate national prevalence estimates that were closer to their BELHES clinical counterparts. For overweight, obesity and hypertension, all methods provided smaller standard errors than those obtained with clinical data. However, for hypercholesterolemia, for which the regression model's accuracy was poor, multiple imputation was the only approach which provided smaller standard errors than those based on clinical data. CONCLUSIONS: The random-forest multiple imputation proves to be the method of choice to correct the bias related to self-reported data in the BHIS. This method is particularly useful to enable improved secondary analysis of self-reported data by using information included in the BELHES. Whenever feasible, combined information from HIS and objective measurements should be used in risk factor monitoring.

The health and economic burden of musculoskeletal disorders in Belgium from 2013 to 2018.

INTRODUCTION: Low back pain (LBP), neck pain (NKP), osteoarthritis (OST) and rheumatoid arthritis (RHE) are among the musculoskeletal (MSK) disorders causing the greatest disability in terms of Years Lived with Disability. The current study aims to analyze the health and economic impact of these MSK disorders in Belgium, providing a summary of morbidity and mortality outcomes from 2013 to 2018, as well as direct and indirect costs from 2013 to 2017. METHODS: The health burden of LBP, NKP, OST and RHE in Belgium from 2013 to 2018 was summarized in terms of prevalence and disability-adjusted life years (DALY) using data from the Belgian health interview surveys (BHIS), the INTEGO database (Belgian registration network for general practitioners) and the Global Burden of Diseases study 2019. The economic burden included estimates of direct medical costs and indirect costs, measured by cost of work absenteeism. For this purpose, data of the respondents to the BHIS-2013 were linked with the national health insurance data (intermutualistic agency [IMA] database) 2013-2017. RESULTS: In 2018, 2.5 million Belgians were affected by at least one MSK disorder. OST represented the disorder with the highest number of cases for both men and women, followed by LBP. In the same year, MSK disorders contributed to a total of 180,746 DALYs for female and 116,063 DALYs for men. LBP appeared to be the largest contributor to the health burden of MSK. Having at least one MSK disorder costed on average 3 billion € in medical expenses and 2 billion € in indirect costs per year, with LBP being the most costly. CONCLUSION: MSK disorders represent a major health and economic burden in Belgium. As their burden will probably continue to increase in the future, acting on the risk factors associated to these disorders is crucial to mitigate both the health and economic burden.

Caution against examining the role of reverse causality in Mendelian Randomization.

Recently, Mendelian Randomization (MR) has gained in popularity as a concept to assess the causal relationship between phenotypes in genetic association studies. An extension of standard MR methodology, the MR Steiger approach, has recently been developed to infer the causal direction between two phenotypes in prospective studies. Through simulation studies, we examined and quantified the ability of the MR Steiger approach to determine the causal direction between two phenotypes (i.e., effect direction). Through simulation studies, our results show that the MR Steiger approach may fail to correctly identify the direction of causality. This is true, especially in the presence of pleiotropy. We also applied the MR Steiger method to the COPDGene study, a case-control study of chronic obstructive pulmonary disease (COPD) in current and former smokers, to examine the role of smoking on lung function. We have created an R package on Github called reverseDirection which runs simulations for user-specified scenarios to examine when the MR Steiger approach can correctly determine the causal direction between two phenotypes in any user specified scenario. In summary, our results emphasize the importance of caution when the MR Steiger approach is used in to infer the direction of causality.

Challenges in Systematic Reviews and Meta-Analyses of Mediation Analyses.

Systematic reviews and meta-analyses of mediation studies are increasingly being implemented in practice. Nonetheless, the methodology for conducting such review and analysis is still in a development phase, with much room for improvement. In this paper, we highlight and discuss challenges that investigators face in systematic reviews and meta-analyses of mediation studies and propose ways of accommodating these in practice.

Nonlinear mediation analysis with high-dimensional mediators whose causal structure is unknown.

With multiple possible mediators on the causal pathway from a treatment to an outcome, we consider the problem of decomposing the effects along multiple possible causal path(s) through each distinct mediator. Under a path-specific effects framework, such fine-grained decompositions necessitate stringent assumptions, such as correctly specifying the causal structure among the mediators, and no unobserved confounding among the mediators. In contrast, interventional direct and indirect effects for multiple mediators can be identified under much weaker conditions, while providing scientifically relevant causal interpretations. Nonetheless, current estimation approaches require (correctly) specifying a model for the joint mediator distribution, which can be difficult when there is a high-dimensional set of possibly continuous and noncontinuous mediators. In this article, we avoid the need to model this distribution, by developing a definition of interventional effects previously suggested for longitudinal mediation. We propose a novel estimation strategy that uses nonparametric estimates of the (counterfactual) mediator distributions. Noncontinuous outcomes can be accommodated using nonlinear outcome models. Estimation proceeds via Monte Carlo integration. The procedure is illustrated using publicly available genomic data to assess the causal effect of a microRNA expression on the 3-month mortality of brain cancer patients that is potentially mediated by expression values of multiple genes.

On estimation and cross-validation of dynamic treatment regimes with competing risks.

The optimal moment to start renal replacement therapy in a patient with acute kidney injury (AKI) remains a challenging problem in intensive care nephrology. Multiple randomized controlled trials have tried to answer this question, but these contrast only a limited number of treatment initiation strategies. In view of this, we use routinely collected observational data from the Ghent University Hospital intensive care units (ICUs) to investigate different prespecified timing strategies for renal replacement therapy initiation based on time-updated levels of serum potassium, pH, and fluid balance in critically ill patients with AKI with the aim to minimize 30-day ICU mortality. For this purpose, we apply statistical techniques for evaluating the impact of specific dynamic treatment regimes in the presence of ICU discharge as a competing event. We discuss two approaches, a nonparametric one - using an inverse probability weighted Aalen-Johansen estimator - and a semiparametric one - using dynamic-regime marginal structural models. Furthermore, we suggest an easy to implement cross-validation technique to assess the out-of-sample performance of the optimal dynamic treatment regime. Our work illustrates the potential of data-driven medical decision support based on routinely collected observational data.

A statistical test to reject the structural interpretation of a latent factor model.

Factor analysis is often used to assess whether a single univariate latent variable is sufficient to explain most of the covariance among a set of indicators for some underlying construct. When evidence suggests that a single factor is adequate, research often proceeds by using a univariate summary of the indicators in subsequent research. Implicit in such practices is the assumption that it is the underlying latent, rather than the indicators, that is causally efficacious. The assumption that the indicators do not have effects on anything subsequent, and that they are themselves only affected by antecedents through the underlying latent is a strong assumption, effectively imposing a structural interpretation on the latent factor model. In this paper, we show that this structural assumption has empirically testable implications, even though the latent variable itself is unobserved. We develop a statistical test to potentially reject the structural interpretation of a latent factor model. We apply this test to data concerning associations between the Satisfaction-with-Life-Scale and subsequent all-cause mortality, which provides strong evidence against a structural interpretation for a univariate latent underlying the scale. Discussion is given to the implications of this result for the development, evaluation, and use of measures, and for the use of factor analysis itself.

Timing of dialysis in acute kidney injury using routinely collected data and dynamic treatment regimes.

BACKGROUND AND OBJECTIVES: Defining the optimal moment to start renal replacement therapy (RRT) in acute kidney injury (AKI) remains challenging. Multiple randomized controlled trials (RCTs) addressed this question whilst using absolute criteria such as pH or serum potassium. However, there is a need for identification of the most optimal cut-offs of these criteria. We conducted a causal analysis on routinely collected data (RCD) to compare the impact of different pre-specified dynamic treatment regimes (DTRs) for RRT initiation based on time-updated levels of potassium, pH, and urinary output on 30-day ICU mortality. DESIGN, SETTING, PARTICIPANTS, AND MEASUREMENTS: Patients in the ICU of Ghent University Hospital were included at the time they met KDIGO-AKI-stage ≥ 2. We applied inverse-probability-of-censoring-weighted Aalen-Johansen estimators to evaluate 30-day survival under 81 DTRs prescribing RRT initiation under different thresholds of potassium, pH, or persisting oliguria. RESULTS: Out of 13,403 eligible patients (60.8 ± 16.8 years, SOFA 7.0 ± 4.1), 5622 (63.4 ± 15.3 years, SOFA 8.2 ± 4.2) met KDIGO-AKI-stage ≥ 2. The DTR that delayed RRT until potassium ≥ 7 mmol/l, persisting oliguria for 24-36 h, and/or pH < 7.0 (non-oliguric) or < 7.2 (oliguric) despite maximal conservative treatment resulted in a reduced 30-day ICU mortality (from 12.7% [95% CI 11.9-13.6%] under current standard of care to 10.5% [95% CI 9.5-11.7%]; risk difference 2.2% [95% CI 1.3-3.8%]) with no increase in patients starting RRT (from 471 [95% CI 430-511] to 475 [95% CI 342-572]). The fivefold cross-validation benchmark for the optimal DTR resulted in 30-day ICU mortality of 10.7%. CONCLUSIONS: Our causal analysis of RCD to compare RRT initiation at different thresholds of refractory low pH, high potassium, and persisting oliguria identified a DTR that resulted in a decrease in 30-day ICU mortality without increase in number of RRTs. Our results suggest that the current criteria to start RRT as implemented in most RCTs may be suboptimal. However, as our analysis is hypothesis generating, this optimal DTR should ideally be validated in a multicentric RCT.

Health care costs and lost productivity costs related to excess weight in Belgium.

BACKGROUND: This study aimed to estimate annual health care and lost productivity costs associated with excess weight among the adult population in Belgium, using national health data. METHODS: Health care costs and costs of absenteeism were estimated using data from the Belgian national health interview survey (BHIS) 2013 linked with individual health insurance data (2013-2017). Average yearly health care costs and costs of absenteeism were assessed by body mass index (BMI) categories - i.e., underweight (BMI < 18.5 kg/m2), normal weight (18.5 ≤ BMI < 25 kg/m2), overweight (25 ≤ BMI < 30 kg/m2) and obesity (BMI ≥ 30 kg/m2). Health care costs were also analysed by type of cost (i.e. ambulatory, hospital, reimbursed medication). The cost attributable to excess weight and the contribution of various other chronic conditions to the incremental cost of excess weight were estimated using the method of recycled prediction (a.k.a. standardisation). RESULTS: According to BHIS 2013, 34.7% and 13.9% of the Belgian adult population were respectively affected by overweight or obesity. They were mostly concentrated in the age-group 35-65 years and had significantly more chronic conditions compared to the normal weight population. Average total healthcare expenses for people with overweight and obesity were significantly higher than those observed in the normal weight population. The adjusted incremental annual health care cost of excess weight in Belgium was estimated at €3,329,206,657 (€651 [95% CI: €144-€1,084] and €1,015 [95% CI: €343-€1,697] per capita for individuals with overweight and obesity respectively). The comorbidities identified to be the main drivers for these incremental health care costs were hypertension, high cholesterol, serious gloom and depression. Mean annual incremental cost of absenteeism for overweight accounted for €242 per capita but was not statistically significant, people with obesity showed a significantly higher cost (p < 0.001) compared to the normal weight population: €2,015 [95% CI: €179-€4,336] per capita. The annual total incremental costs due to absenteeism of the population affected by overweight and obesity was estimated at €1,209,552,137. Arthritis, including rheumatoid arthritis and osteoarthritis, was the most important driver of the incremental cost of absenteeism in individuals with overweight and obesity, followed by hypertension and low back pain. CONCLUSIONS: The mean annual incremental cost of excess weight in Belgium is of concern and stresses the need for policy actions aiming to reduce excess body weight. This study can be used as a baseline to evaluate the potential savings and health benefits of obesity prevention interventions.

Longitudinal mediation analysis of time-to-event endpoints in the presence of competing risks.

This proposal is motivated by an analysis of the English Longitudinal Study of Ageing (ELSA), which aims to investigate the role of loneliness in explaining the negative impact of hearing loss on dementia. The methodological challenges that complicate this mediation analysis include the use of a time-to-event endpoint subject to competing risks, as well as the presence of feedback relationships between the mediator and confounders that are both repeatedly measured over time. To account for these challenges, we introduce path-specific effect proportional (cause-specific) hazard models. These extend marginal structural proportional (cause-specific) hazard models to enable effect decomposition on either the cause-specific hazard ratio scale or the cumulative incidence function scale. We show that under certain ignorability assumptions, the path-specific direct and indirect effects indexing this model are identifiable from the observed data. We next propose an inverse probability weighting approach to estimate these effects. On the ELSA data, this approach reveals little evidence that the total effect of hearing loss on dementia is mediated through the feeling of loneliness, with a non-statistically significant indirect effect equal to 1.01 (hazard ratio (HR) scale; 95% confidence interval (CI) 0.99 to 1.05).