Causal Mediation Analysis for the Cox Proportional Hazards Model with a Smooth Baseline Hazard Estimator

Religion and survival among European older adults

There are several pathways through which religion can affect longevity. Previous research, predominately from North America, has shown decreased mortality risk for participants that attended religious services. This study aims to examine the association between religion and all-cause mortality in a large sample of older European adults, comparing religious affiliations, and using prayer frequency as well as frequency of participation in a religious organisation as measures of religiousness. To this end, a total of 16,062 participants from the Survey of Health Ageing and Retirement in Europe were employed for a survival analysis (median follow-up 11.3 years; 3790 recorded deaths). Following a religion was negatively associated with mortality regardless of demographic and socioeconomic factors (HR = 0.81; 95% CI 0.74-0.89). Large differences in the median survival of participants from different religious affiliations can be mostly attributed to demographic and socioeconomic factors. Both frequency of prayer and religious participation exhibited a significant positive dose-response relationship with survival despite adjustments, although the results for religious participation were more profound. Changes on the religiosity levels of the European population will require additional research on the subject in the future.

Risk-aware survival time prediction from whole slide pathological images

Deep-learning-based survival prediction can assist doctors by providing additional information for diagnosis by estimating the risk or time of death. The former focuses on ranking deaths among patients based on the Cox model, whereas the latter directly predicts the survival time of each patient. However, it is observed that survival time prediction for the patients, particularly with close observation times, possibly has incorrect orders, leading to low prediction accuracy. Therefore, in this paper, we present a whole slide image (WSI)-based survival time prediction method that takes advantage of both the risk as well as time prediction. Specifically, we propose to combine these two approaches by extracting the risk prediction features and using them as guides for the survival time prediction. Considering the high resolution of WSIs, we extract tumor patches from WSIs using a pre-trained tumor classifier and apply the graph convolutional network to aggregate information across these patches effectively. Extensive experiments demonstrate that the proposed method significantly improves the time prediction accuracy when compared with direct prediction of the survival times without guidance and outperforms existing methods.

Pet ownership and survival of European older adults

With pet ownership on the rise, millions of individuals are exposed to this environmental exposure. Although the subject has been largely studied, more evidence is needed to clarify the potential association of pet ownership with human health. The aim of this research is to study the potential association of pet exposure (any pet, cat, dog, bird, fish) with all-cause, cardiovascular and cancer mortality of older ([Formula: see text] 50 years) European residents. To this end, a total of 23,274 participants from the Survey of Health Ageing and Retirement in Europe (SHARE) were employed (median follow-up 119 months). All-cause mortality (5163 events), as well as cardiovascular (CVD) (1832 events), and cancer mortality (1346 events) were examined using Cox Proportional Hazards models for their relation with pet exposure at baseline. Stratified analyses were also performed by gender and for single or multi-person households. No significant association was observed for any of the pets with all-cause mortality on the whole sample and the fully adjusted models. In stratified analyses, bird exposure significantly increased the risk of all-cause mortality in women [Hazard Ratio [Formula: see text]; 95% CI 1.04-1.44] as well as women living alone [Formula: see text]; 95% CI 1.02-1.85). Cause-specific models revealed an increased risk of death for women bird owners for causes other than cancer and CVD [Formula: see text]; 95% CI 1.05-1.99). In conclusion, bird ownership may be negatively associated with survival of older women in Europe.

Exposure assessment for Cox proportional hazards cure models with interval-censored survival data

Mixture cure models have been developed as an effective tool to analyze failure time data with a cure fraction. Used in conjunction with the logistic regression model, this model allows covariate-adjusted inference of an exposure effect on the cured probability and the hazard of failure for the uncured subjects. However, the covariate-adjusted inference for the overall exposure effect is not directly provided. In this paper, we describe a Cox proportional hazards cure model to analyze interval-censored survival data in the presence of a cured fraction and then apply a post-estimation approach by using model-predicted estimates difference to assess the overall exposure effect on the restricted mean survival time scale. For baseline hazard/survival function estimation, simple parametric models as fractional polynomials or restricted cubic splines are utilized to approximate the baseline logarithm cumulative hazard function, or, alternatively, the full likelihood is specified through a piecewise linear approximation for the cumulative baseline hazard function. Simulation studies were conducted to demonstrate the unbiasedness of both estimation methods for the overall exposure effect estimates over various baseline hazard distribution shapes. The methods are applied to analyze the interval-censored relapse time data from a smoking cessation study.

Causal inference with a mediated proportional hazards regression model

The natural direct and indirect effects in causal mediation analysis with survival data having one mediator is addressed by VanderWeele (2011) [1]. He derived an approach for (1) an accelerated failure time regression model in general cases and (2) a proportional hazards regression model when the time-to-event outcome is rare. If the outcome is not rare, then VanderWeele (2011) [1] did not derive a simple closed-form expression for the log-natural direct and log-natural indirect effects for the proportional hazards regression model because the baseline cumulative hazard function does not approach zero. We develop two approaches to extend VanderWeele's approach, in which the assumption of a rare outcome is not required. We obtain the natural direct and indirect effects for specific time points through numerical integration after we calculate the cumulative baseline hazard by (1) applying the Breslow method in the Cox proportional hazards regression model to estimate the unspecified cumulative baseline hazard; (2) assuming a piecewise constant baseline hazard model, yielding a parametric model, to estimate the baseline hazard and cumulative baseline hazard. We conduct simulation studies to compare our two approaches with other methods and illustrate our two approaches by applying them to data from the ASsessment, Serial Evaluation, and Subsequent Sequelae in Acute Kidney Injury (ASSESS-AKI) Consortium.

Association of life satisfaction with disability-free survival: role of chronic diseases and healthy lifestyle

BACKGROUND

this article investigates the association between life satisfaction and disability-free survival, and explores the roles of chronic diseases and healthy lifestyle in this association.

METHOD

a cohort of 2,116 functionally independent adults aged ≥60 was followed up to 12 years. At baseline, life satisfaction was assessed with the Life Satisfaction Index A (LSI-A). Disability-free survival was defined as the survival till the first occurrence of either death, dementia or physical disability. Information on lifestyle factors was collected via questionnaire. Chronic diseases were ascertained through clinical examinations at baseline and each follow-up. Data were analysed using Cox proportional hazard regression models and Laplace regression.

RESULTS

over follow-up, 1,121 participants died, developed dementia, or became disabled. High LSI-A versus Low LSI-A had a lower risk of death, dementia and physical disability (hazard ratio [HR] 0.79, 95% confidence intervals [CI] 0.67-0.94), and had a longer disability-free period by 1.73 (95% CI 0.18-3.32) years. In mediation analysis, accumulation of chronic diseases mediated 17.8% of the association between LSI-A and disability-free survival. In joint effect analysis, participants with high LSI-A and a favourable lifestyle profile had a HR of 0.53 (95% CI 0.41-0.69) for the composite endpoint, and lived 3.2 (95% CI 1.35-5.11) disability-free years longer than those with low life satisfaction and an unfavourable lifestyle profile.

DISCUSSION

high life satisfaction is independently associated with longer disability-free survival. This association is partially mediated by a lower burden of chronic diseases and is reinforced by healthy lifestyle.

Mediation analysis for survival data with High-Dimensional mediators

MOTIVATION

Mediation analysis has become a prevalent method to identify causal pathway(s) between an independent variable and a dependent variable through intermediate variable(s). However, little work has been done when the intermediate variables (mediators) are high-dimensional and the outcome is a survival endpoint. In this paper, we introduce a novel method to identify potential mediators in a causal framework of high-dimensional Cox regression.

RESULTS

We first reduce the data dimension through a mediation-based sure independence screening (SIS) method. A de-biased Lasso inference procedure is used for Cox's regression parameters. We adopt a multiple-testing procedure to accurately control the false discovery rate (FDR) when testing high-dimensional mediation hypotheses. Simulation studies are conducted to demonstrate the performance of our method. We apply this approach to explore the mediation mechanisms of 379,330 DNA methylation markers between smoking and overall survival among lung cancer patients in the TCGA lung cancer cohort. Two methylation sites (cg08108679 and cg26478297) are identified as potential mediating epigenetic markers.

AVAILABILITY

Our proposed method is available with the R package HIMA at https://cran.r-project.org/web/packages/HIMA/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

ProTICS reveals prognostic impact of tumor infiltrating immune cells in different molecular subtypes

Different subtypes of the same cancer often show distinct genomic signatures and require targeted treatments. The differences at the cellular and molecular levels of tumor microenvironment in different cancer subtypes have significant effects on tumor pathogenesis and prognostic outcomes. Although there have been significant researches on the prognostic association of tumor infiltrating lymphocytes in selected histological subtypes, few investigations have systemically reported the prognostic impacts of immune cells in molecular subtypes, as quantified by machine learning approaches on multi-omics datasets. This paper describes a new computational framework, ProTICS, to quantify the differences in the proportion of immune cells in tumor microenvironment and estimate their prognostic effects in different subtypes. First, we stratified patients into molecular subtypes based on gene expression and methylation profiles by applying nonnegative tensor factorization technique. Then we quantified the proportion of cell types in each specimen using an mRNA-based deconvolution method. For tumors in each subtype, we estimated the prognostic effects of immune cell types by applying Cox proportional hazard regression. At the molecular level, we also predicted the prognosis of signature genes for each subtype. Finally, we benchmarked the performance of ProTICS on three TCGA datasets and another independent METABRIC dataset. ProTICS successfully stratified tumors into different molecular subtypes manifested by distinct overall survival. Furthermore, the different immune cell types showed distinct prognostic patterns with respect to molecular subtypes. This study provides new insights into the prognostic association between immune cells and molecular subtypes, showing the utility of immune cells as potential prognostic markers. Availability: R code is available at https://github.com/liu-shuhui/ProTICS.

Mediation analysis for mixture Cox proportional hazards cure models

Mediation analysis aims to decompose a total effect into specific pathways and investigate the underlying causal mechanism. Although existing methods have been developed to conduct mediation analysis in the context of survival models, none of these methods accommodates the existence of a substantial proportion of subjects who never experience the event of interest, even if the follow-up is sufficiently long. In this study, we consider mediation analysis for the mixture of Cox proportional hazards cure models that cope with the cure fraction problem. Path-specific effects on restricted mean survival time and survival probability are assessed by introducing a partially latent group indicator and applying the mediation formula approach in a three-stage mediation framework. A Bayesian approach with P-splines for approximating the baseline hazard function is developed to conduct analysis. The satisfactory performance of the proposed method is verified through simulation studies. An application of the Alzheimer's disease (AD) neuroimaging initiative dataset investigates the causal effects of APOE-ϵ4 allele on AD progression.

Loneliness, Social Isolation, and Chronic Disease Outcomes

Background

Research suggests that loneliness and social isolation (SI) are serious public health concerns. However, our knowledge of the associations of loneliness and SI with specific chronic diseases is limited.

Purpose

The present prospective cohort study investigated (a) the longitudinal associations of loneliness and SI with four chronic diseases (cardiovascular disease [CVD], chronic obstructive pulmonary disease [COPD], diabetes mellitus Type 2 [T2D], and cancer), (b) the synergistic association of loneliness and SI with chronic disease, and (c) baseline psychological and behavioral explanatory factors.

Methods

Self-reported data from the 2013 Danish “How are you?” survey (N = 24,687) were combined with individual-level data from the National Danish Patient Registry on diagnoses in a 5 year follow-up period (2013–2018).

Results

Cox proportional hazard regression analyses showed that loneliness and SI were independently associated with CVD (loneliness: adjusted hazard ratio (AHR) = 1.20, 95% confidence interval [CI; 1.03, 1.40]; SI: AHR = 1.23, 95% CI [1.04, 146]) and T2D (loneliness: AHR =1.90, 95% CI [1.42, 2.55]; SI: AHR = 1.59, 95% CI [1.15, 2.21]). No significant associations were found between loneliness or SI and COPD and cancer, respectively. Likewise, loneliness and SI did not demonstrate a synergistic effect on chronic disease. Multiple mediation analysis indicated that loneliness and SI had an indirect effect on CVD and T2D through both baseline psychological and behavioral factors.

Conclusion

Loneliness and SI were independently associated with a diagnosis of CVD and T2D within a 5 year follow-up period. The associations of loneliness and SI with CVD and T2D were fully explained by baseline psychological and behavioral factors.

Receipt of opioid agonist treatment halves the risk of HIV-1 RNA viral load rebound through improved ART adherence for HIV-infected women who use illicit drugs

BACKGROUND

Women living with HIV who use illicit drugs may be particularly vulnerable to HIV-1 RNA viral load (VL) rebound.

METHODS

We used longitudinal data from 2006 to 2017 to evaluate the impact of sociodemographic, behavioral, social-structural, and clinical factors on the hazard of viral rebound for women enrolled in the ACCESS study, a prospective cohort with systematic VL monitoring. Women were included if they achieved VL suppression (<50 copies/mL) following antiretroviral therapy (ART) initiation and had more than one study interview. Sociodemographic as well as substance use, social-structural, addiction treatment, and HIV clinical factors were evaluated as predictors of viral rebound (VL > 1000 copies/mL). Cox regressions using a recurrent events framework, time-varying covariates, robust standard errors, and a frailty component were used.

RESULTS

Of the 185 women included, 62 (34%) experienced at least one viral rebound event over an 11-year period, accumulating a total of 87 viral rebound events. In adjusted analysis, stimulant use more than doubled the hazard of viral rebound (adjusted hazard ratio [AHR]: 2.35, 95% confidence interval [CI]: 1.07-5.14) while the only factor protective against viral rebound was receipt of opioid agonist treatment (OAT) in the past six months (AHR: 0.46, 95% CI: 0.26-0.81). After adjusting for ART adherence in the past six months, the effect of OAT was attenuated (AHR: 0.57, 95% CI: 0.32-1.02).

CONCLUSIONS

Efforts to improve access to and retention within OAT programs and decrease stimulant use may improve rates of viral suppression for HIV-positive women who use illicit drugs.

Effect decomposition through multiple causally nonordered mediators in the presence of exposure-induced mediator-outcome confounding

Avin et al (2005) showed that, in the presence of exposure-induced mediator-outcome confounding, decomposing the total causal effect (TCE) using standard conditional exchangeability assumptions is not possible even under a nonparametric structural equation model with all confounders observed. Subsequent research has investigated the assumptions required for such a decomposition to be identifiable and estimable from observed data. One approach was proposed by VanderWeele et al (2014). They decomposed the TCE under three different scenarios: (1) treating the mediator and the exposure-induced confounder as joint mediators; (2) generating path-specific effects albeit without distinguishing between multiple distinct paths through the exposure-induced confounder; and (3) using so-called randomised interventional analogues where sampling values from the distribution of the mediator within the levels of the exposure effectively marginalises over the exposure-induced confounder. In this paper, we extend their approach to the case where there are multiple mediators that do not influence each other directly but which are all influenced by an exposure-induced mediator-outcome confounder. We provide a motivating example and results from a simulation study based on from our work in dental epidemiology featuring the 1982 Pelotas Birth Cohort in Brazil.

The Impact of Antidepressants on the Risk of Developing Obstructive Sleep Apnea in Posttraumatic Stress Disorder: A Nationwide Cohort Study in Taiwan

STUDY OBJECTIVES

The association between posttraumatic stress disorder (PTSD) and obstructive sleep apnea (OSA) has been reported inconsistently, and the association between antidepressant use and the risk of developing OSA in patients with PTSD has not been previously studied. Therefore, we used the Longitudinal National Health Insurance Database (LHID) to investigate the impact of PTSD and antidepressant use on the risk of OSA development.

METHODS

Identified from the LHID, 2,316 individuals aged ≥ 18 years with PTSD, but with no history of OSA, and 23,160 control individuals matched for age, sex, obesity and index date were enrolled between 2000 and 2015 and followed up until the end of 2015 to identify the development of OSA. A two-tailed Bonferroni-corrected P < .00038 (.05/13) was considered statistically significant as we examined 13 antidepressants.

RESULTS

Individuals with PTSD had increased risk of developing OSA (adjusted hazard ratio 4.672, 95% confidence interval 2.246-9.787, P < .001) after adjusting for demographic data, medical comorbidities, and medication. Treatment with antidepressants was not significantly associated with an increased risk of developing OSA compared to no antidepressant treatment.

CONCLUSIONS

Asian patients with PTSD had increased risk of developing OSA, and treatment with antidepressants did not play a key role in increasing the risk of OSA development. Further studies are required to investigate the underlying mechanisms of PTSD and the roles of antidepressants on the risk of developing OSA.

CITATION

Lin C-E, Chung C-H, Chen L-F, Chien W-C, Chou P-H. The impact of antidepressants on the risk of developing obstructive sleep apnea in posttraumatic stress disorder: a nationwide cohort study in taiwan. J Clin Sleep Med. 2019;15(9):1233-1241.

Marginal Time-Dependent Causal Effects in Mediation Analysis With Survival Data

The main aim of mediation analysis is to study the direct and indirect effects of an exposure on an outcome. To date, the literature on mediation analysis with multiple mediators has mainly focused on continuous and dichotomous outcomes. However, the development of methods for multiple mediation analysis of survival outcomes is still limited. Here we extend to survival outcomes a method for multiple mediation analysis based on the computation of appropriate weights. The approach considered has the advantages of not requiring specific models for mediators, allowing nonindependent mediators of any nature, and not relying on the assumption of rare outcomes. Simulation studies show good performance of the proposed estimator in terms of bias and coverage probability. The method is further applied to an example from a published study on prostate cancer mortality aimed at understanding the extent to which the effect of DNA methyltransferase 3b (DNMT3b) genotype on mortality was explained by DNA methylation and tumor aggressiveness. This approach can be used to quantify the marginal time-dependent direct and indirect effects carried by multiple indirect pathways, and software code is provided to facilitate its application.