Mediation misgivings: ambiguous clinical and public health interpretations of natural direct and indirect effects

Causal Inference Challenges in the Relationship Between Social Determinants and Cardiovascular Outcomes

The effects of social determinants on cardiovascular outcomes are frequently estimated in epidemiologic analyses, but the profound causal and statistical challenges of this research program are not widely discussed. Here, we carefully review definitions and measures for social determinants of cardiovascular health and then examine the various assumptions required for valid causal inference in multivariable analyses of observational data, such as what one would typically encounter in cohorts, population surveys, health care databases, and vital statistics databases. We explain the necessity of the "well-defined exposure" and show how this goal relates to the "consistency assumption" that is necessary for valid causal inference. Well-defined exposure is especially challenging for social determinants of health because they are seldom simple atomistic interventions that are easily conceptualized and measured. We then review threats to valid inference that arise from confounding, selection bias, information bias, and positivity violations. Other causal considerations are reviewed and explained, such as correct model specification, absence of immortal time, and avoidance of the "Table 2 Fallacy," and their application to social determinants of cardiovascular outcomes are discussed. Fruitful approaches, including focusing on policy interventions and the "target trial" frameworks are proposed and provide a pathway for a more efficacious research program that can more reliably improve population health. Valid causal inference in this setting is quite challenging, but-with clever design and thoughtful analysis-the important role of social factors in patterning cardiovascular outcomes can be quantified and reported.

Does the effect of adolescent health behaviours on adult cardiometabolic health differ by socioeconomic background? Protocol for a population-based cohort study

INTRODUCTION

Adolescence is a sensitive period for cardiometabolic health. Yet, it remains unknown if adolescent health behaviours, such as alcohol use, smoking, diet and physical activity, have differential effects across socioeconomic strata. Adopting a life-course perspective and a causal inference framework, we aim to assess whether the effects of adolescent health behaviours on adult cardiometabolic health differ by levels of neighbourhood deprivation, parental education and occupational class. Gaining a better understanding of these social disparities in susceptibility to health behaviours can inform policy initiatives that aim to improve population health and reduce socioeconomic inequalities in cardiometabolic health.

METHODS AND ANALYSIS

We will conduct a secondary analysis of the Young Finns Study, which is a longitudinal population-based cohort study. We will use measures of health behaviours-smoking, alcohol use, fruit and vegetable consumption, and physical activity-as exposure and parental education, occupational class and neighbourhood deprivation as effect modifiers during adolescence (ages 12-18 years). Eight biomarkers of cardiometabolic health (outcomes)-waist circumference, body mass index, blood pressure, low-density lipoprotein cholesterol, apolipoprotein B, plasma glucose and insulin resistance-will be measured when participants were aged 33-40. A descriptive analysis will investigate the clustering of health behaviours. Informed by this, we will conduct a causal analysis to estimate effects of single or clustered adolescent health behaviours on cardiometabolic health conditional on socioeconomic background. This analysis will be based on a causal model implemented via a directed acyclic graph and inverse probability-weighted marginal structural models to estimate effect modification.

ETHICS AND DISSEMINATION

The Young Finns study was conducted according to the guidelines of the Declaration of Helsinki, and the protocol was approved by ethics committees of University of Helsinki, Kuopio, Oulu, Tampere and Turku. We will disseminate findings at international conferences and a manuscript in an open-access peer-reviewed journal.

The role of differential exposure and susceptibility to heavy drinking linking income inequalities and tooth loss: An investigation of the alcohol harm paradox using a four-way decomposition analysis

OBJECTIVES

To examine the extent to which the effect of income inequalities on tooth loss is attributable to differential exposure and susceptibility to heavy drinking in older Brazilian adults.

METHODS

We conducted a secondary analysis using data from The Brazilian Longitudinal Study of Ageing (ELSI-Brazil 2015-2016), a nationally representative sample of community-dwelling people aged 50 years and over. Causal mediation analysis based on the counterfactual outcome framework decomposed the effect of income on tooth loss mediated by heavy drinking into four components (four-way decomposition): controlled direct effect (neither mediation nor interaction), reference interaction (interaction only), mediated interaction (both mediation and interaction) and pure indirect effect (mediation only). Proportions of effect attributable to each component were calculated to estimate the differential exposure (the sum of the third and fourth components) and differential susceptibility (the sum of the second and third components) to heavy drinking.

RESULTS

The analytical sample comprised 8114 participants. After adjusting for covariates, 7.3% (95% CI: 3.8%; 10.9%) and -39.5% (95% CI: -75.8%; -3.3%) of the effects of income on tooth loss were attributable to differential exposure and susceptibility to heavy drinking, respectively, consistent with the alcohol harm paradox. When setting non-functional dentition as outcome, only the effect of differential susceptibility remained (-81.7% [95% CI: -128.2%; -35.2%]).

CONCLUSION

Our findings suggest that individuals of low-income groups appear to be more susceptible to the effects of heavy drinking on tooth loss.

Examining the Causal Mediating Role of Cardiovascular Disease on the Effect of Subclinical Cardiovascular Disease on Cognitive Impairment via Separable Effects

BACKGROUND

An important epidemiological question is understanding how vascular risk factors contribute to cognitive impairment. Using data from the Cardiovascular Health Cognition Study, we investigated how subclinical cardiovascular disease (sCVD) relates to cognitive impairment risk and the extent to which the hypothesized risk is mediated by the incidence of clinically manifested cardiovascular disease (CVD), both overall and within apolipoprotein E-4 (APOE-4) subgroups.

METHODS

We adopted a novel "separable effects" causal mediation framework that assumes that sCVD has separably intervenable atherosclerosis-related components. We then ran several mediation models, adjusting for key covariates.

RESULTS

We found that sCVD increased overall risk of cognitive impairment (risk ratio [RR] = 1.21, 95% confidence interval [CI]: 1.03, 1.44); however, there was little or no mediation by incident clinically manifested CVD (indirect effect RR = 1.02, 95% CI: 1.00, 1.03). We also found attenuated effects among APOE-4 carriers (total effect RR = 1.09, 95% CI: 0.81, 1.47; indirect effect RR = 0.99, 95% CI: 0.96, 1.01) and stronger findings among noncarriers (total effect RR = 1.29, 95% CI: 1.05, 1.60; indirect effect RR = 1.02, 95% CI: 1.00, 1.05). In secondary analyses restricting cognitive impairment to only incident dementia cases, we found similar effect patterns.

CONCLUSIONS

We found that the effect of sCVD on cognitive impairment does not seem to be mediated by CVD, both overall and within APOE-4 subgroups. Our results were critically assessed via sensitivity analyses, and they were found to be robust. Future work is needed to fully understand the relationship between sCVD, CVD, and cognitive impairment.

Disparities in job characteristics by race and sex in a Southern aluminum smelting facility

BACKGROUND

Former workers at a Southern aluminum smelting facility raised concerns that the most hazardous jobs were assigned to Black workers, but the role of workplace segregation had not been quantified or examined in the company town. Prior studies discuss race and gender disparities in working conditions, but few have documented them in the aluminum industry.

METHODS

We obtained workers' company records for 1985-2007 and characterized four job metrics: prestige (sociologic rankings), worker-defined danger (worker assessments), annual wage (1985 dollars), and estimated total particulate matter (TPM) exposure (job exposure matrix). Characteristics of job at hire and trajectories were compared by race and sex using linear binomial models.

RESULTS

Non-White males had the highest percentage of workers in low prestige and high danger jobs at hire and up to 20 years after. After 20 years tenure, 100% of White workers were in higher prestige and lower danger jobs. Most female workers, regardless of race, entered and remained in low-wage jobs, while 50% of all male workers maintained their initial higher-wage jobs. Non-White females had the highest prevalence of workers in low-wage jobs at hire and after 20 years-increasing from 63% (95% CI: 59-67) to 100% (95% CI: 78-100). All female workers were less likely to be in high TPM exposure jobs. Non-White males were most likely to be hired into high TPM exposure jobs, and this exposure prevalence increased as time accrued, while staying constant for other race-sex groups.

CONCLUSIONS

There is evidence of job segregation by race and sex in this cohort of aluminum smelting workers. Documentation of disparities in occupational hazards is important for informing health interventions and research.

An exact regression-based approach for the estimation of natural direct and indirect effects with a binary outcome and a continuous mediator

In the causal mediation framework, a number of parametric regression-based approaches have been introduced in recent years for estimating natural direct and indirect effects for a binary outcome in an exact manner, without invoking simplifying assumptions based on the rareness or commonness of the outcome. However, most of these works have focused on a binary mediator. In this article, we aim at a continuous mediator and introduce an exact approach for the estimation of natural effects on the odds ratio, risk ratio, and risk difference scales. Our approach relies on logistic and linear models for the outcome and mediator, respectively, and uses numerical integration to calculate the nested counterfactual probabilities underlying the definition of natural effects. Formulas for the delta method standard errors for all effects estimators are provided. The performance of our proposed exact estimators was evaluated in simulation studies that featured scenarios with different levels of outcome rareness/commonness, including a marginally but not conditionally rare outcome scenario. Furthermore, we evaluated the merit of Firth's penalization to mitigate the bias in the logistic regression coefficients estimators for the smallest outcome prevalences and sample sizes investigated. Using a SAS macro provided, we implemented our approach to assess the effect of placental abruption on low birth weight mediated by gestational age. We found that our exact natural effects estimators worked properly in both simulated and real data applications.

Evaluating the Mediating Role of Recall of Intervention Knowledge in the Relationship Between a Peer-Driven Intervention and HIV Risk Behaviors Among People Who Inject Drugs

Peer-driven interventions can be effective in reducing HIV injection risk behaviors among people who inject drugs (PWID). We employed a causal mediation framework to examine the mediating role of recall of intervention knowledge in the relationship between a peer-driven intervention and subsequent self-reported HIV injection-related risk behavior among PWID in the HIV Prevention Trials Network (HPTN) 037 study. For each intervention network, the index participant received training at baseline to become a peer educator, while non-index participants and all participants in the control networks received only HIV testing and counseling; recall of intervention knowledge was measured at the 6-month visit for each participant, and each participant was followed to ascertain HIV injection-related risk behaviors at the 12-month visit. We used inverse probability weighting to fit marginal structural models to estimate the total effect (TE) and controlled direct effect (CDE) of the intervention on the outcome. The proportion eliminated (PE) by intervening to remove mediation by the recall of intervention knowledge was computed. There were 385 participants (47% in intervention networks) included in the analysis. The TE and CDE risk ratios for the intervention were 0.47 [95% confidence interval (CI): 0.28, 0.78] and 0.73 (95% CI: 0.26, 2.06) and the PE was 49%. Compared to participants in the control networks, the peer-driven intervention reduced the risk of HIV injection-related risk behavior by 53%. The mediating role of recall of intervention knowledge accounted for less than 50% of the total effect of the intervention, suggesting that other potential causal pathways between the intervention and the outcome, such as motivation and skill, self-efficacy, social norms and behavior modeling, should be considered in future studies.

Child maltreatment, anxiety and depression, and asthma among British adults in the UK Biobank

BACKGROUND

Child maltreatment is associated with asthma in adults. We examined whether lifetime major depressive disorder (MDD) or lifetime generalised anxiety disorder (GAD) mediate an association between child maltreatment and current asthma among 81 105 British adults in the UK Biobank who completed a mental health survey and had complete data on child maltreatment, GAD, MDD, asthma and relevant covariates but no diagnosis of chronic obstructive pulmonary disease.

METHODS

Child maltreatment was ascertained based on answers to the five questions in the Childhood Trauma Screener. Two mediators, lifetime MDD and GAD, were assessed based on the Composite International Diagnostic Interview Short Form. Current asthma was defined as physician-diagnosed asthma and wheeze or whistling in the chest in the previous year. Logistic regression was used for the multivariable analysis of child maltreatment and current asthma, and a mediation analysis was conducted to estimate the contributions of lifetime MDD and lifetime GAD to the child maltreatment-current asthma association.

RESULTS

In a multivariable analysis, any child maltreatment was associated with asthma (adjusted OR 1.22, 95% CI 1.15-1.28; p<0.01). In a mediation analysis adjusted for household income, educational attainment, smoking status, pack-years of smoking and other covariates, lifetime GAD and lifetime MDD explained 21.8% and 32.5%, respectively, of the child maltreatment-current asthma association. Similar results were obtained after excluding current smokers and former smokers with ≥10 pack-years of smoking from the mediation analysis.

CONCLUSION

Our findings suggest that GAD and MDD mediate an association between child maltreatment and asthma in adults, independently of smoking.

Causal inference and effect estimation using observational data

Observational studies aiming to estimate causal effects often rely on conceptual frameworks that are unfamiliar to many researchers and practitioners. We provide a clear, structured overview of key concepts and terms, intended as a starting point for readers unfamiliar with the causal inference literature. First, we introduce theoretical frameworks underlying causal effect estimation methods: the counterfactual theory of causation, the potential outcomes framework, structural equations and directed acyclic graphs. Second, we define the most common causal effect estimands, and the issues of effect measure modification, interaction and mediation (direct and indirect effects). Third, we define the assumptions required to estimate causal effects: exchangeability, positivity, consistency and non-interference. Fourth, we define and explain biases that arise when attempting to estimate causal effects, including confounding, collider bias, selection bias and measurement bias. Finally, we describe common methods and study designs for causal effect estimation, including covariate adjustment, G-methods and natural experiment methods.

Remission and Relapse of Dyslipidemia After Vertical Sleeve Gastrectomy vs Roux-en-Y Gastric Bypass in a Racially and Ethnically Diverse Population

IMPORTANCE

The comparative effectiveness of the most common operations in the long-term management of dyslipidemia is not clear.

OBJECTIVE

To compare 4-year outcomes associated with vertical sleeve gastrectomy (VSG) vs Roux-en-Y gastric bypass (RYGB) for remission and relapse of dyslipidemia.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective comparative effectiveness study was conducted from January 1, 2009, to December 31, 2016, with follow-up until December 31, 2018. Participants included patients with dyslipidemia at the time of surgery who underwent VSG (4142 patients) or RYGB (2853 patients). Patients were part of a large integrated health care system in Southern California. Analysis was conducted from January 1, 2018, to December 31, 2021.

EXPOSURES

RYGB and VSG.

MAIN OUTCOMES AND MEASURES

Dyslipidemia remission and relapse were assessed in each year of follow-up for as long as 4 years after surgery.

RESULTS

A total of 8265 patients were included, with a mean (SD) age of 46 (11) years; 6591 (79.8%) were women, 3545 (42.9%) were Hispanic, 1468 (17.8%) were non-Hispanic Black, 2985 (36.1%) were non-Hispanic White, 267 (3.2%) were of other non-Hispanic race, and the mean (SD) body mass index (calculated as weight in kilograms divided by height in meters squared) was 44 (7) at the time of surgery. Dyslipidemia outcomes at 4 years were ascertained for 2168 patients (75.9%) undergoing RYGB and 3999 (73.9%) undergoing VSG. Remission was significantly higher for those who underwent RYGB (824 [38.0%]) compared with VSG (1120 [28.0%]) (difference in the probability of remission, 0.10; 95% CI, 0.01-0.19), with no differences in relapse (455 [21.0%] vs 960 [24.0%]). Without accounting for relapse, remission of dyslipidemia after 4 years was 58.9% (1279) for those who underwent RYGB and 51.9% (2079) for those who underwent VSG. Four-year differences between operations were most pronounced for patients 65 years or older (0.39; 95% CI, 0.27-0.51), those with cardiovascular disease (0.43; 95% CI, 0.24-0.62), or non-Hispanic Black patients (0.13; 95% CI, 0.01-0.25) and White patients (0.13; 95% CI, 0.03-0.22).

CONCLUSIONS AND RELEVANCE

In this large, racially and ethnically diverse cohort of patients who underwent bariatric and metabolic surgery in clinical practices, RYGB was associated with higher rates of dyslipidemia remission after 4 years compared with VSG. However, almost one-quarter of all patients experienced relapse, suggesting that patients should be monitored closely throughout their postoperative course to maximize the benefits of these operations for treatment of dyslipidemia.

Roles of participation in social activities in the association between adverse childhood experiences and health among older Japanese adults

Adverse childhood experiences (ACEs) have shown strong associations with later-life health such as depression and subjective health. Social participation is also associated with later-life health but it is unclear to what extent this could contribute to alleviating harmful impacts of ACEs, nor is it clear whether ACEs are themselves associated with later-life social participation. Thus, this study aims to understand: (1) the influence of ACEs on social participation in later life and (2) whether social participation can alleviate the harmful influences of ACEs on depression and subjective health among Japanese older adults. Data were from 5,671 Japanese older adults (aged 65+) in surveys in 2013 and 2016 as part of the Japan Gerontological Evaluation Study (JAGES). Logistic regression analyses were conducted to estimate the relations between ACEs and later-life social participation, adjusting for potential confounders and mediators. Inverse probability weighting was used to estimate average effects of ACEs on later-depression and subjective health, adjusting for potential confounders, and these were compared against controlled direct effect (CDE) estimates from marginal structural models based on all respondents experiencing weekly social participation. We found that ACEs were associated with reduced later-life social participation (OR for >1 ACEs = 0.88, 95% CI = 0.79, 0.99). The estimated effect of ACEs on depression ( adjusted total effect estimates: OR = 1.23, 95% CI = 1.05, 1.45) was marginally alleviated in estimates assuming weekly social participation for everyone (CDE = 1.18, 95% CI = 0.98, 1.43). A similar tendency was seen for poor subjective heath. Negative impacts of ACEs on depression may be marginally mitigated through social participation, but mitigating effects were moderate. Further investigation on other potential later-life mitigating factors is needed.

A principled approach to mediation analysis in perinatal epidemiology

For many research questions in perinatal epidemiology, gestational age is a mediator that features the causal pathway between exposure and outcome. A mediator is an intermediate variable between an exposure and outcome, which is influenced by the exposure on the causal pathway to the outcome. Therefore, conventional analyses that adjust, stratify, or match for gestational age or its proxy (eg, preterm vs term deliveries) are problematic. This practice, which is entrenched in perinatal research, induces an overadjustment bias. Depending on the causal question, it may be inappropriate to adjust (or condition) for a mediator, such as gestational age, by either design or statistical analysis, but its effect can be quantified through causal mediation analysis. In an exposition of such methods, we demonstrated the relationship between the exposure and outcome and provided a formal analytical framework to quantify the extent to which a causal effect is influenced by a mediator. We reviewed concepts of confounding and causal inference, introduced the concept of a mediator and illustrated the perils of adjusting for a mediator in an exposure-outcome paradigm for a given causal question, adopted causal methods that call for an evaluation of a mediator in a causal exposure effect on the outcome, and discussed unmeasured confounding assumptions in mediation analysis. Furthermore, we reviewed other developments in the causal mediation analysis literature, including decomposition of a total effect when the mediator interacts with the exposure (4-way decomposition), methods for multiple mediators, mediation methods for case-control studies, mediation methods for time-to-event outcomes, sample size and power analysis for mediation analysis, and available software to apply these methods. To illustrate these methods, we provided a clinical example to estimate the risk of perinatal mortality (outcome) concerning placental abruption (exposure) and to determine the extent to which preterm delivery (mediator; a proxy for gestational age) plays a role in this causal effect. We hoped that the adoption of mediation methods described in this review will move research in perinatal epidemiology away from biased adjustments of mediators toward a more nuanced quantification of effects that pose unique challenges and provide unique insights in our field.

Comparative Effectiveness of Gastric Bypass and Vertical Sleeve Gastrectomy for Hypertension Remission and Relapse: The ENGAGE CVD Study

Weight loss is an effective strategy for the management of hypertension, and bariatric surgery is the most effective weight loss and maintenance strategy for obesity. The importance of bariatric surgery in the long-term management of hypertension and which operation is most effective is less clear. We compared the effectiveness of vertical sleeve gastrectomy (VSG) and Roux-en-Y gastric bypass (RYGB) for remission and relapse of hypertension after surgery in the ENGAGE CVD cohort study (Effectiveness of Gastric Bypass Versus Gastric Sleeve for Cardiovascular Disease). Operations were done by 23 surgeons across 9 surgical practices. Hypertension remission and relapse were assessed in each year of follow-up beginning 30 days and up to 5 years postsurgery. We used a local instrumental variable approach to account for selection bias in the choice of VSG or RYGB. The study population included 4964 patients with hypertension at the time of surgery (n=3186 VSG and n=1778 RYGB). At 1 year, 27% of patients with RYGB and 28% of patients with VSG achieved remission. After 5 years, without accounting for relapse, 42% of RYGB and 43% of VSG patients had experienced hypertension remission. After accounting for relapse, only 17% of RYGB and 18% of VSG patients remained in remission 5 years after surgery. There were no statistically significant differences between VSG and RYGB for hypertension remission, relapse, or mean systolic and diastolic blood pressure at any time during follow-up.

Gender-specific effects of adverse childhood experiences on incidence of activities of daily life disability in middle-age and elderly Chinese population

BACKGROUND

Adverse Childhood Experiences (ACEs) may have long-lasting effects on late life health, probably through life-course mediators. However, whether such effects still exist when these mediators have been appropriately controlled is unclear.

OBJECTIVES

To estimate the controlled direct effect of ACEs on Activities of Daily Life (ADL) disability in middle-aged people and examine the gender-difference of this effect.

PARTICIPANTS AND SETTING

We used data from the China Health and Retirement Longitudinal Study, a nationally representative longitudinal survey of persons aged 45+ years.

METHODS

ACEs were measured by the Adverse Childhood Experiences International Questionnaire and number of ACEs was classified as 0, 1, 2 and 3+, while ADL disability was measured using the Katz Index. Gender-specific controlled direct effects of ACEs on the incidence of ADL disability were estimated by marginal structural model (MSM) with stabilized inverse-probability-of-treatment weights of mediators (unhealthy behaviors, chronic diseases and depression).

RESULTS

4,544 males and 4,767 females were included. Gender differences existed in most categories of ACEs, and about 10 % participants had 3+ ACEs. Participants who had 3+ ACEs had 39 % and 59 % higher risk of ADL disability than those with 0 ACEs among males and females, respectively. After controlling for the mediators, the direct effect was slightly increased in males (risk ratio (RR) = 1.45, p < 0.001) but decreased in females (RR=1.28, p < 0.05).

CONCLUSIONS

Precautions targeted in reducing ACEs may be beneficial in preventing ADL disability, but gender-specific prevention should be considered.

Pre-school childcare and inequalities in child development

Centre-based childcare may benefit pre-school children and alleviate inequalities in early childhood development, but evidence on socio-emotional and physical health outcomes is limited. Data were from the UK Millennium Cohort Study (n = 14,376). Inverse-probability weighting was used to estimate confounder-adjusted population-average effects of centre and non-centre-based childcare (compared to parental care only) between ages 26-31 months on (age 3): internalising and externalising symptoms, pro-social behaviour, independence, emotional dysregulation, vocabulary, school readiness, and body mass index. To assess impacts on inequalities, controlled direct effects of low parental education and lone parenthood on all outcomes were estimated under two hypothetical scenarios: 1) universal take-up of centre-based childcare; and 2) parental care only. On average, non-centre based childcare improved vocabulary and centre-based care improved school readiness, with little evidence of other benefits. However, socio-economic inequalities were observed for all outcomes and were attenuated in scenario 1 (universal take-up). For example, inequalities in externalising symptoms (according to low parental education) were reduced from a confounder-adjusted standard deviation difference of 7.8 (95% confidence intervals: 6.7-8.8), to 1.7 (0.6-2.7). Inequalities by parental education in scenario 2 (parental care only) were wider than in scenario 1 for externalising symptoms (at 3.4; 2.4-4.4), and for emotional dysregulation and school readiness. Inequalities by lone parenthood, which were smaller, fell in scenario 1, and fell further in scenario 2. Universal access to centre-based pre-school care may alleviate inequalities, while restricted access (e.g. during lockdown for a pandemic such as Covid-19) may widen some inequalities in socioemotional and cognitive development.

Insights into the Cross-world Independence Assumption of Causal Mediation Analysis

Causal mediation analysis is a useful tool for epidemiologic research, but it has been criticized for relying on a "cross-world" independence assumption that counterfactual outcome and mediator values are independent even in causal worlds where the exposure assignments for the outcome and mediator differ. This assumption is empirically difficult to verify and problematic to justify based on background knowledge. In the present article, we aim to assist the applied researcher in understanding this assumption. Synthesizing what is known about the cross-world independence assumption, we discuss the relationship between assumptions for causal mediation analyses, causal models, and nonparametric identification of natural direct and indirect effects. In particular, we give a practical example of an applied setting where the cross-world independence assumption is violated even without any post-treatment confounding. Further, we review possible alternatives to the cross-world independence assumption, including the use of bounds that avoid the assumption altogether. Finally, we carry out a numeric study in which the cross-world independence assumption is violated to assess the ensuing bias in estimating natural direct and indirect effects. We conclude with recommendations for carrying out causal mediation analyses.

Mediation effects that emulate a target randomised trial: Simulation-based evaluation of ill-defined interventions on multiple mediators

Many epidemiological questions concern potential interventions to alter the pathways presumed to mediate an association. For example, we consider a study that investigates the benefit of interventions in young adulthood for ameliorating the poorer mid-life psychosocial outcomes of adolescent self-harmers relative to their healthy peers. Two methodological challenges arise. First, mediation methods have hitherto mostly focused on the elusive task of discovering pathways, rather than on the evaluation of mediator interventions. Second, the complexity of such questions is invariably such that there are no well-defined mediator interventions (i.e. actual treatments, programs, etc.) for which data exist on the relevant populations, outcomes and time-spans of interest. Instead, researchers must rely on exposure (non-intervention) data, that is, on mediator measures such as depression symptoms for which the actual interventions that one might implement to alter them are not well defined. We propose a novel framework that addresses these challenges by defining mediation effects that map to a target trial of hypothetical interventions targeting multiple mediators for which we simulate the effects. Specifically, we specify a target trial addressing three policy-relevant questions, regarding the impacts of hypothetical interventions that would shift the mediators' distributions (separately under various interdependence assumptions, jointly or sequentially) to user-specified distributions that can be emulated with the observed data. We then define novel interventional effects that map to this trial, simulating shifts by setting mediators to random draws from those distributions. We show that estimation using a g-computation method is possible under an expanded set of causal assumptions relative to inference with well-defined interventions, which reflects the lower level of evidence that is expected with ill-defined interventions. Application to the self-harm example in the Victorian Adolescent Health Cohort Study illustrates the value of our proposal for informing the design and evaluation of actual interventions in the future.

The impact of adjusting for pure predictors of exposure, mediator, and outcome on the variance of natural direct and indirect effect estimators

It is now well established that adjusting for pure predictors of the outcome, in addition to confounders, allows unbiased estimation of the total exposure effect on an outcome with generally reduced standard errors (SEs). However, no analogous results have been derived for mediation analysis. Considering the simplest linear regression setting and the ordinary least square estimator, we obtained theoretical results showing that adjusting for pure predictors of the outcome, in addition to confounders, allows unbiased estimation of the natural indirect effect (NIE) and the natural direct effect (NDE) on the difference scale with reduced SEs. Adjusting for pure predictors of the mediator increases the SE of the NDE's estimator, but may increase or decrease the variance of the NIE's estimator. Adjusting for pure predictors of the exposure increases the variance of estimators of the NIE and NDE. Simulation studies were used to confirm and extend these results to the case where the mediator or the outcome is binary. Additional simulations were conducted to explore scenarios featuring an exposure-mediator interaction as well as the relative risk and odds ratio scales for the case of binary mediator and outcome. Both a regression approach and an inverse probability weighting approach were considered in the simulation study. A real-data illustration employing data from the Canadian Study of Health and Aging is provided. This analysis is concerned with the mediating effect of vitamin D in the effect of physical activity on dementia and its results are overall consistent with the theoretical and empirical findings.

Multivariable Mendelian Randomization and Mediation

Mendelian randomization (MR) is the use of genetic variants associated with an exposure to estimate the causal effect of that exposure on an outcome. Mediation analysis is the method of decomposing the effects of an exposure on an outcome, which act directly, and those that act via mediating variables. These effects are decomposed through the use of multivariable analysis to estimate the causal effects between three types of variables: exposures, mediators, and an outcome. Multivariable MR (MVMR) is a recent extension to MR that uses genetic variants associated with multiple, potentially related exposures to estimate the effect of each exposure on a single outcome. MVMR allows for equivalent analysis to mediation within the MR framework and therefore can also be used to estimate mediation effects. This approach retains the benefits of using genetic instruments for causal inference, such as avoiding bias due to confounding, while allowing for estimation of the different effects required for mediation analysis. This review explains MVMR, what is estimated when one exposure is a mediator of another in an MVMR estimation, and how MR and MVMR can therefore be used to estimate mediated effects. This review then goes on to consider the advantages and limitations of using MR and MVMR to conduct mediation analysis.

Examining chronic inflammatory markers on blood pressure measures in the presence of vitamin D insufficiency among indigenous cree adults: results from the cross-sectional Multi-Community Environment-and-Health Study in Eeyou Istchee, Quebec, Canada

OBJECTIVE

High blood pressure (BP) is a risk factor for cardiovascular disease. Examining the role of inflammatory mediators on BP is important since vitamin D (VD) is a modifiable risk factor, which possibly modulates inflammatory cytokines. This study simulated what are known as average 'controlled direct effects (CDE)' of inflammatory markers, C reactive protein (CRP), tumour necrosis factor-α (TNF-α), and interlukin-6 (IL-6) on continuous BP measures, while fixing VD, an intermediate variable to specific level.

DESIGN

Cross-sectional study.

SETTING

We analysed data from the Multi-Community Environment-and-Health Study, 2005-2009, conducted in Eeyou Istchee, Quebec, Canada.

PARTICIPANTS

This study recruited 1425 study Indigenous Cree participants from seven Cree communities. Only adults with serum VD levels, inflammatory markers and BP measures were included in this data analysis.

PRIMARY AND SECONDARY OUTCOMES MEASURES

Inflammatory markers examined the top 25th exposure percentiles. VD 'insufficiency' (ie, 25-hydroxyvitamin-D levels<50 nmol/L) defined by the Institute of Medicine. CDE for each inflammatory marker in the presence and absence of population VD insufficiency simulated the average direct effect change for systolic and diastolic BP (SBP and DBP) measures. All models were adjusted for exposure-and-mediator outcome relationship.

RESULTS

Among 161 participants, 97 (60 %) were female. The prevalence of VD insufficiency was 32%. CDE estimates show in the presence and absence of population vitamin D insufficiency, inflammatory markers have a slightly different association on BP. TNF-α significantly and inversely associated with SBP in the presence of vitamin D insufficiency, fully adjusted model β = -13.61 (95% CI -24.42 to -2.80); however, TNF-α was not associated with SBP in the absence of vitamin D insufficiency. CRP, IL-6 were also not significantly associated with BP measures, although the magnitude of association was greater for those with elevated inflammation and VD insufficiency.

CONCLUSION

This novel analysis shows in the presence of VD insufficiency, inflammation (particularly TNF-α) may affect SBP. Additional research is needed to elucidate these findings, and the temporal relationship between these variables.

Social contact and inequalities in depressive symptoms and loneliness among older adults: A mediation analysis of the English Longitudinal Study of Ageing

Background

Social contact, including remote contact (by telephone, email, letter or text), could help reduce social inequalities in depressive symptoms and loneliness among older adults.

Methods

Data were from the 8th wave of the English Longitudinal Study of Aging (2016/17), stratified by age (n = 1578 aged <65; n = 4026 aged 65+). Inverse probability weighting was used to estimate average effects of weekly in-person and remote social contact on depressive symptoms (score of 3+ on 8-item CES-D scale) and two measures of loneliness (sometimes/often feels lonely vs hardly ever/never; and top quintile of UCLA loneliness scale vs all others). We also estimated controlled direct effects of education, partner status, and wealth on loneliness and depressive symptoms under two scenarios: 1) universal infrequent (<weekly) in-person social contact; and 2) universal weekly remote social contact.

Results

Weekly in-person social contact was associated on average with reduced odds of loneliness, but associations with remote social contact were weak. Lower education raised odds of depressive symptoms and loneliness, but differences were attenuated with infrequent in-person contact. Respondents living alone experienced more depressive symptoms and loneliness than those living with a partner, and less wealth was associated with more depressive symptoms. With universal infrequent in-person contact, these differences narrowed among those aged under 65 but widened among those aged 65+. Universal weekly remote contact had relatively little impact on inequalities.

Conclusions

Reduced in-person social contact may increase depressive symptoms and loneliness among older adults, especially for those aged 65+ who live alone. Reliance on remote social contact seems unlikely to compensate for social inequalities.

•

In-person but not remote contact associated with less depression/loneliness.

•

In-person contact < weekly reduces educational inequalities in depression/loneliness.

•

In those aged 65+ partner/wealth inequalities widen with <weekly in-person contact.

•

Weekly remote contact has little impact on inequalities in depression/loneliness.

Inferential challenges when assessing racial/ethnic health disparities in environmental research

Numerous epidemiologic studies have documented environmental health disparities according to race/ethnicity (R/E) to inform targeted interventions aimed at reducing these disparities. Yet, the use of R/E under the potential outcomes framework implies numerous underlying assumptions for epidemiologic studies that are often not carefully considered in environmental health research. In this commentary, we describe the current state of thinking about the interpretation of R/E variables in etiologic studies. We then discuss how such variables are commonly used in environmental epidemiology. We observed three main uses for R/E: i) as a confounder, ii) as an effect measure modifier and iii) as the main exposure of interest either through descriptive analysis or under a causal framework. We identified some common methodological concerns in each case and provided some practical solutions. The use of R/E in observational studies requires particular cautions in terms of formal interpretation and this commentary aims at providing a practical resource for future studies assessing racial/ethnic health disparities in environmental research.

Direct and indirect pathways between low income status and becoming a high-cost health care user in Ontario, Canada: a mediation analysis of health risk behaviors

PURPOSE

We investigated natural direct and indirect pathways between low-income status and high-cost user (HCU) transitions considering health risk behaviors as potential mediators.

METHODS

We analyzed data from respondents (aged 18+) from four pooled cycles of the Canadian Community Health Survey (2005-2010/2011) linked to administrative data in Ontario, Canada. HCUs were defined as the top 5% of the population, ranked by cost consumption in any of the five years after survey interview. Low-income status was defined from the provincial distribution of self-reported household income, with missing values imputed from neighborhood-level data. In mediation analyses based on marginal structural models, we quantified the contributions of smoking, physical inactivity, alcohol consumption, and body mass index to income-HCU associations.

RESULTS

115,091 respondents (representative of 9,661,764 Ontarians) were included in the study, of which 7.2% became HCUs. The odds of becoming HCUs were 1.36 times (95% CI: 1.25-1.48) greater for low (vs. high) income status respondents. Smoking, physical activity, alcohol consumption, and body mass index contributed 9.4%, 6.5%, 10.6%, and 4.4% to this association, respectively. Tests for exposure-mediator interactions were not statistically significant.

CONCLUSIONS

Health risk behaviors only partially explain income inequalities in future HCU transitions.

A Regression Framework for Causal Mediation Analysis with Applications to Behavioral Science

We introduce and extend the classical regression framework for conducting mediation analysis from the fit of only one model. Using the essential mediation components (EMCs) allows us to estimate causal mediation effects and their analytical variance. This single-equation approach reduces computation time and permits the use of a rich suite of regression tools that are not easily implemented on a system of three equations. Additionally, we extend this framework to non-nested mediation systems, provide a joint measure of mediation for complex mediation hypotheses, propose new visualizations for mediation effects, and explain why estimates of the total effect may differ depending on the approach used. Using data from social science studies, we also provide extensive illustrations of the usefulness of this framework and its advantages over traditional approaches to mediation analysis. The example data are freely available for download online and we include the R code necessary to reproduce our results.

The future of activity space and health research

We discuss the future of activity space and health research in the context of a recently published systematic review. Our discussion outlines a number of elements for reflection among the research community. We need to think beyond activity space and reconceptualize exposure in era of high volume, high precision location data. We need to develop standardized methods for understanding global positioning system data. We must adopt replicable scientific computing processes and machine learning models. Finally, we must embrace modern notions of causality in order to contend with the conceptual challenges faced by our research field.

Commentary: Causal Inference for Social Exposures

Social epidemiology seeks to describe and quantify the causal effects of social institutions, interactions, and structures on human health. To accomplish this task, we define exposures as treatments and posit populations exposed or unexposed to these well-defined regimens. This inferential structure allows us to unambiguously estimate and interpret quantitative causal parameters and to investigate how these may be affected by biases such as confounding. This paradigm has been challenged recently by some critics who favor broadening the exposures that may be studied beyond treatments to also consider states. Defining the exposure protocol of an observational study is a continuum of specificity, and one may choose to loosen this definition, incurring the cost of causal parameters that become commensurately more vague. The advantages and disadvantages of broader versus narrower definitions of exposure are matters of continuing debate in social epidemiology as in other branches of epidemiology.

The Role of Access to a Regular Primary Care Physician in Mediating Immigration-Based Disparities in Colorectal Screening: Application of Multiple Mediation Methods

BACKGROUND

Colorectal cancer screening participation is lower among recent immigrants than among Canadian-born individuals. We assessed whether this screening disparity is mediated by access to regular primary care physicians (PCP).

METHODS

Pooling years 2003 to 2014 of the Canadian Community Health Survey, lifetime screening in respondents aged 50 to 75 years of age who immigrated in the previous 10 years (n = 1,067) was compared with Canadian-born respondents (N = 102,366). Regression- and inverse probability weighting-based methods were used to estimate the total effect (TE) and controlled direct effect (CDE) of recent immigration on never having received either a stool- or endoscopic-based screening test. The proportion of the TE that would be eliminated if all had a PCP was computed using these estimates [proportion eliminated (PE) = (TE - CDE)/(TE - 1)]. Analyses were stratified by visible minority status and adjusted for income, rurality, age, sex, marital status, education, and exposure to a provincially organized colorectal screening program.

RESULTS

The prevalence of never having been screened was 71% and 57% in visible minority and white recent immigrants, respectively, and 46% in white Canadian-born respondents. If all had regular PCPs, there would be no reduction in the screening inequality between white recent immigrants and Canadian-born (null PE), and the inequality between visible minority immigrants and white Canadian-born may increase by 6% to 13%.

CONCLUSIONS

Ensuring all have regular PCPs may lead to greater screening gains among Canadian-born than recent immigrants.

IMPACT

Improving access to PCPs may increase colorectal screening overall, but not reduce immigration-based disparities screening. Alternative interventions to reduce this disparity should be explored.

Child maltreatment, psychopathological symptoms, and onset of diabetes mellitus, hypothyroidism and COPD in adulthood

BACKGROUND

The aim of this study was to assess the associations between child maltreatment (CM), psychopathological symptoms, and onset of diabetes mellitus, hypothyroidism (i.e., low metabolism), and chronic bronchitis/emphysema/COPD in adulthood.

METHODS

The present analysis used cross-sectional data collected in 2007-2008 within the Tromsø Study, Norway (N = 12,981). CM was measured with a single item, and self-reported information on psychopathological symptoms and physical health outcomes was used. The associations between CM, psychopathological symptoms, and physical health outcomes were assessed with linear and Poisson regression models. Mediation was assessed with difference-in-coefficients method.

RESULTS

In the fully-adjusted models, CM was associated with higher levels of anxiety and depression, psychological distress, difficulty in sleeping, insomnia, and use of sleeping pills and antidepressants in adulthood (p < 0.05). Moreover, CM was associated with a more than two-folds increased risk of consultation with psychiatrist (p < 0.001), a 26% increased risk of forgetfulness (p < 0.001), a 15% increased risk of decline in memory (p < 0.001), and a 96% increased risk of psychiatric problems (p < 0.001) over the course of life. In the fully-adjusted models, CM was associated with a 27-82% increased risk of physical health outcomes in adulthood (p < 0.05). Indicators of psychopathological symptoms significantly (p < 0.05) mediate the associations between CM and physical health outcomes.

LIMITATIONS

The design of this study is cross-sectional, and all measures are self-reported.

CONCLUSION

The associations between retrospectively-reported CM and physical health outcomes in adulthood are partially driven by psychopathological symptoms in adulthood.

A classical regression framework for mediation analysis: fitting one model to estimate mediation effects

Mediation analysis explores the degree to which an exposure's effect on an outcome is diverted through a mediating variable. We describe a classical regression framework for conducting mediation analyses in which estimates of causal mediation effects and their variance are obtained from the fit of a single regression model. The vector of changes in exposure pathway coefficients, which we named the essential mediation components (EMCs), is used to estimate standard causal mediation effects. Because these effects are often simple functions of the EMCs, an analytical expression for their model-based variance follows directly. Given this formula, it is instructive to revisit the performance of routinely used variance approximations (e.g., delta method and resampling methods). Requiring the fit of only one model reduces the computation time required for complex mediation analyses and permits the use of a rich suite of regression tools that are not easily implemented on a system of three equations, as would be required in the Baron-Kenny framework. Using data from the BRAIN-ICU study, we provide examples to illustrate the advantages of this framework and compare it with the existing approaches.

The mediating role of the venules between smoking and ischemic stroke

A potential mechanism by which smoking affects ischemic stroke is through wider venules, but this mediating role of wider venules has never been quantified. Here, we aimed to estimate to what extent the effect of smoking on ischemic stroke is possibly mediated by the venules via the recently developed four-way effect decomposition. This study was part of a population-based study including 9109 stroke-free persons participated in the study in 1990, 2004, or 2006 (mean age: 63.7 years; 58% women). Smoking behavior (smoking versus non-smoking) was identified by interview. Retinal venular calibers were measured semi-automatically on retinal photographs. Incident strokes were assessed until January 2016. A regression-based approach was used with venular calibers as mediator to decompose the total effect of smoking compared to non-smoking into four components: controlled direct effect (neither mediation nor interaction), pure indirect effect (mediation only), reference interaction effect (interaction only) and mediated interaction effect (both mediation and interaction). During a mean follow-up of 12.5 years, 665 persons suffered an ischemic stroke. Smoking increased the risk of developing ischemic stroke compared to non-smoking with an excess risk of 0.41 (95% confidence interval 0.10; 0.67). With retinal venules as a potential mediator, the excess relative risk could be decomposed into 77% controlled direct effect, 4% mediation only, 4% interaction only, and 15% mediated interaction. To conclude, in the pathophysiology of ischemic stroke, the effect of smoking on ischemic stroke may partly explained by changes in the venules, where there is both pure mediation and mediated interaction.

Does Partner Selection Mediate the Relationship Between School Attendance and HIV/Herpes Simplex Virus-2 Among Adolescent Girls and Young Women in South Africa: An Analysis of HIV Prevention Trials Network 068 Data

OBJECTIVE

School attendance prevents HIV and herpes simplex virus-2 (HSV-2) in adolescent girls and young women, but the mechanisms to explain this relationship remain unclear. Our study assesses the extent to which characteristics of sex partners, partner age, and number mediate the relationship between attendance and risk of infection in adolescent girls and young women in South Africa.

DESIGN

We use longitudinal data from the HIV Prevention Trials Network 068 randomized controlled trial in rural South Africa, where girls were enrolled in early adolescence and followed in the main trial for more than 3 years. We examined older partners and the number of partners as possible mediators.

METHODS

We used the parametric g-formula to estimate 4-year risk differences for the effect of school attendance on the cumulative incidence of HIV/HSV-2 overall and the controlled direct effect (CDE) for mediation. We examined mediation separately and jointly for the mediators of interest.

RESULTS

We found that young women with high attendance in school had a lower cumulative incidence of HIV compared with those with low attendance (risk difference = -1.6%). Partner age difference (CDE = -1.2%) and the number of partners (CDE = -0.4%) mediated a large portion of this effect. In fact, when we accounted for the mediators jointly, the effect of schooling on HIV was almost removed, showing full mediation (CDE = -0.3%). The same patterns were observed for the relationship between school attendance and cumulative incidence of HSV-2 infection.

CONCLUSION

Increasing school attendance reduces the risk of acquiring HIV and HSV-2. Our results indicate the importance of school attendance in reducing partner number and partner age difference in this relationship.

Racial Disparities in Cardiovascular Health Behaviors: The Coronary Artery Risk Development in Young Adults Study

INTRODUCTION

There are known racial differences in cardiovascular health behaviors, including smoking, physical activity, and diet quality. A better understanding of these differences may help identify intervention targets for reducing cardiovascular disease disparities. This study examined whether socioeconomic, psychosocial, and neighborhood environmental factors, in isolation or together, mediate racial differences in health behaviors.

METHODS

Participants were 3,081 men and women from the Coronary Artery Risk Development in Young Adults study who were enrolled in 1985-1986 (Year 0) and completed a follow-up examination in 2015-2016 (Year 30). A health behavior score was created at Years 0, 7, 20, and 30 using smoking, physical activity, and diet assessed that year. The race difference in health behavior score was estimated using linear regression in serial cross-sectional analyses. Mediation analyses computed the proportion of the race and health behavior score association attributable to socioeconomic, psychosocial, and neighborhood factors.

RESULTS

Data analysis conducted in 2016-2017 found that blacks had significantly lower health behavior scores than whites across 30 years of follow-up. Individual socioeconomic factors mediated 48.9%-70.1% of the association between race and health behavior score, psychosocial factors 20.3%-30.0%, and neighborhood factors 22.1%-41.4% (p<0.01 for all).

CONCLUSIONS

Racial differences in health behavior scores appear to be mediated predominately by correspondingly large differences in socioeconomic factors. This study highlights the profound impact of socioeconomic factors, which are mostly not under an individual's control, on health behaviors. Policy action targeting socioeconomic factors may help reduce disparities in health behaviors.

Unequal Exposure or Unequal Vulnerability? Contributions of Neighborhood Conditions and Cardiovascular Risk Factors to Socioeconomic Inequality in Incident Cardiovascular Disease in the Multi-Ethnic Study of Atherosclerosis

Risk factors can drive socioeconomic inequalities in cardiovascular disease (CVD) through differential exposure and differential vulnerability. In this paper, we show how econometric decomposition directly enables simultaneous, policy-oriented assessment of these 2 mechanisms. We specifically estimate contributions of neighborhood environment and proximal risk factors to socioeconomic inequality in CVD incidence via these mechanisms. We followed 5,608 participants in the Multi-Ethnic Study of Atherosclerosis (2000-2012) to their first CVD event (median length of follow-up, 12.2 years). We used a summary measure of baseline socioeconomic position (SEP). Covariates included baseline demographics, neighborhood characteristics, and psychosocial, behavioral, and biomedical risk factors. Using Poisson models, we decomposed the difference (inequality) in incidence rates between low- and high-SEP groups into contributions of 1) differences in covariate means (differential exposure) and 2) differences in CVD risk associated with covariates (differential vulnerability). Notwithstanding large uncertainty in neighborhood estimates, our analysis suggested that differential exposure to poorer neighborhood socioeconomic conditions, adverse social environment, diabetes, and hypertension accounted for most of the inequality. Psychosocial and behavioral contributions were negligible. Further, neighborhood SEP, female sex, and white race were more strongly associated with CVD among low-SEP (vs. high-SEP) participants. These differentials in vulnerability also accounted for nontrivial portions of the inequality and could have important implications for intervention.

Associations Between Maternal Prepregnancy Body Mass Index and Gestational Weight Gain and Daughter's Age at Menarche: The Avon Longitudinal Study of Parents and Children

Earlier puberty and menarche are associated with adverse health outcomes. Reported associations of maternal adiposity with daughter’s age at menarche are inconsistent. We examined associations between maternal prepregnancy body mass index (BMI; weight (kg)/height (m)²) and gestational weight gain (GWG) and daughter’s ages at menarche (n = 3,935 mother-offspring pairs), pubarche (Tanner stage 2 for pubic hair) (n = 2,942 pairs), and thelarche (Tanner stage 2 for breast development) (n = 2,942 pairs) in the Avon Longitudinal Study of Parents and Children, a prospective United Kingdom pregnancy cohort study (baseline 1991–1992). During a follow-up period of up to 17 years (1991–2008), mean menarcheal age was 12.6 (standard deviation, 1.2) years. Both maternal prepregnancy BMI and GWG were inversely associated with daughter’s age at menarche after adjustment for maternal age, parity, socioeconomic status, smoking, maternal menarcheal age, and ethnicity (mean differences were −0.34 months (95% confidence interval: −0.45, −0.22) per BMI unit and −0.17 months (95% confidence interval: −0.26, −0.07) per kg, respectively). Associations remained unchanged after adjustment for birth weight and gestational age but were attenuated to the null when results were adjusted for daughter’s prepubertal BMI. Similar results were found for ages at pubarche and thelarche. These findings indicate that greater prepregnancy BMI and GWG are associated with earlier puberty in daughters and that these associations are mediated by daughters’ prepubertal BMIs.

Interventional Effects for Mediation Analysis with Multiple Mediators

The mediation formula for the identification of natural (in)direct effects has facilitated mediation analyses that better respect the nature of the data, with greater consideration of the need for confounding control. The default assumptions on which it relies are strong, however. In particular, they are known to be violated when confounders of the mediator-outcome association are affected by the exposure. This complicates extensions of counterfactual-based mediation analysis to settings that involve repeatedly measured mediators, or multiple correlated mediators. VanderWeele, Vansteelandt, and Robins introduced so-called interventional (in)direct effects. These can be identified under much weaker conditions than natural (in)direct effects, but have the drawback of not adding up to the total effect. In this article, we adapt their proposal to achieve an exact decomposition of the total effect, and extend it to the multiple mediator setting. Interestingly, the proposed effects capture the path-specific effects of an exposure on an outcome that are mediated by distinct mediators, even when-as often-the structural dependence between the multiple mediators is unknown, for instance, when the direction of the causal effects between the mediators is unknown, or there may be unmeasured common causes of the mediators.

Conditional cash transfers and the reduction in partner violence for young women: an investigation of causal pathways using evidence from a randomized experiment in South Africa (HPTN 068)

INTRODUCTION

Evidence has shown that the experience of violence by a partner has important influences on women's risk of HIV acquisition. Using a randomized experiment in northeast South Africa, we found that a conditional cash transfer (CCT) targeted to poor girls in high school reduced the risk of physical intimate partner violence (IPV) in the past 12 months by 34%. The purpose of this analysis is to understand the pathways through which the CCT affects IPV.

METHODS

HPTN 068 was a phase 3, randomized controlled trial in rural Mpumalanga province, South Africa. Eligible young women (aged 13-20) and their parents or guardians were randomly assigned (1:1) to either receive a monthly cash transfer conditional on monthly high school attendance or no cash transfer. Between 2011 and 2015, participants (N = 2,448) were interviewed at baseline, then at annual follow-up visits at 12, 24 and 36 months. The total effect of the CCT on IPV was estimated using a GEE log-binomial regression model. We then estimated controlled direct effects to examine mediation of direct effects through intermediate pathways. Mediators include sexual partnership measures, the sexual relationship power scale, and household consumption measures.

RESULTS

We found evidence that the CCT works in part through delaying sexual debut or reducing the number of sexual partners. The intervention interacts with these mediators leading to larger reductions in IPV risk compared to the total effect of the CCT on any physical IPV [RR 0.66, CI(95%):0.59-0.74]. The largest reductions are seen when we estimate the controlled direct effect under no sexual debut [RR 0.57, CI(95%):0.48-0.65] or under no sexual partner in the last 12 months [RR 0.53, CI(95%):0.46-0.60].

CONCLUSIONS

Results indicate that a CCT for high school girls has protective effects on their experience of IPV and that the effect is due in part to girls choosing not to engage in sexual partnerships, thereby reducing the opportunity for IPV. As a lower exposure to IPV and safer sexual behaviours also protect against HIV acquisition, this study adds to the growing body of evidence on how cash transfers may reduce young women's HIV risk.

Mediators of the relationship between social activities and cognitive function among older Irish adults: results from the Irish longitudinal study on ageing

OBJECTIVE

Evidence points to an association between social and leisure activity (SLA) engagement and cognitive outcomes, but the mechanisms underlying this link remain unknown. We aimed to investigate three potential mechanisms: Vascular function, Perceived Stress, and Cognitive Reserve.

METHODS

With data from 8163 adults aged over 50 in the Irish Longitudinal Study of Ageing, we used a structural equation model to evaluate Vascular Function and Perceived Stress as potential mediators, and Cognitive Reserve as a potential antecedent in the relationship between SLA at baseline (2009), and cognitive outcomes collected at a two-year follow-up point (2011).

RESULTS

Cognitive Reserve was strongly associated both with cognitive outcomes (β = 0.306; p < 0.001) and with SLA (β = 0.694; p < 0.001). Perceived stress (β = 0.018) acted as a significant mediator in the relationships between SLA and cognitive outcomes (p < 0.001), although Vascular Function did not (β = 0.000).

CONCLUSION

These results indicate that SLA may protect cognitive function partly because of its association with cognitive reserve, and partly through its impact on perceived stress. Results have policy implications for those interested in facilitating SLA to protect cognitive outcomes among older adults.

Associations of Early- and Later-Childhood Poverty With Child Cognitive Function in Indonesia: Effect Decomposition in the Presence of Exposure-Induced Mediator-Outcome Confounding

The amount of family financial resources available in early life influences child health and development. Using data from the 2000 and 2007 waves of the Indonesian Family Life Survey, we estimated the associations of early-life poverty (at age <7 years) and poverty in later childhood (at age 7-14 years) with cognitive function at age 7-14 years. Our analysis provided little support for the idea that an early intervention to support household income has a larger effect than intervention later in childhood; both seemed equally important. We also decomposed the effect of poverty at age <7 years into direct and indirect effects mediated through poverty and schooling/home environment at age 7-14 years. For decomposing the effects, we used 3 approaches: 1) joint mediators, 2) path-specific, and 3) intervention analog. Being exposed to poverty before age 7 years had a larger direct effect (difference in cognitive function z score) on child cognitive function at age 7-14 years (i.e., joint mediators β = -0.07, 95% confidence interval: -0.12, -0.02) than the indirect effects mediated through later poverty at age 7-14 years (β = -0.01, 95% confidence interval: -0.04, 0.01) and school attendance/home environment at age 7-14 years. The effect of poverty on cognitive function was small; nevertheless, financial intervention may still benefit children's cognitive function.