Mediation Analysis

Lee, Hopin; Herbert, Robert D.; McAuley, James H.

doi:10.1001/jama.2018.21973

Cited by 129 publications

(109 citation statements)

References 6 publications

Supporting

Mentioning

109

Contrasting

Order By: Relevance

“…There are several analytical approaches to mediation analysis [8]. The fundamental goal is to decompose the 'total effect' of an intervention (or exposure) on an outcome into an…”

Section: Introductionmentioning

confidence: 99%

An overview of systematic reviews found suboptimal reporting and methodological limitations of mediation studies investigating causal mechanisms

Cashin

Lee

Lamb

et al. 2019

Journal of Clinical Epidemiology

Self Cite

View full text Add to dashboard Cite

Objective: To investigate whether systematic reviews of mediation studies identify limitations in reporting quality and methodological conduct. Study Design and Setting: An overview of systematic reviews. We searched four databases (MEDLINE, PsycINFO, Cochrane Database of Systematic Reviews and PubMed) to identify systematic reviews of studies that used mediation analysis to investigate mechanisms of healthcare interventions or exposures in clinical populations between 2007 and 2017. Two reviewers independently screened titles and abstracts. Summary data on the characteristics, reporting quality and methodological conduct of the studies included in the systematic reviews were extracted independently by two reviewers. The protocol was prospectively registered on PROSPERO (CRD42017059834). Results: 54 systematic reviews were included, representing 11 healthcare fields, 26 health conditions and 2008 mediation studies. 18/54 systematic reviews (33%) explicitly stated that the reporting of primary studies was suboptimal. Of these, 14/18 (78%) reviews noted incomplete reporting of effect sizes and precision estimates from mediation analyses. 29/54 systematic reviews (54%) identified limitations in the methodological conduct of primary studies. Conclusion: The reporting and methodological conduct of studies investigating mechanisms in healthcare seems to be suboptimal. Guidance is needed to improve the quality, completeness and transparency of mediation studies.

show abstract

“…There are several analytical approaches to mediation analysis [8]. The fundamental goal is to decompose the 'total effect' of an intervention (or exposure) on an outcome into an…”

Section: Introductionmentioning

confidence: 99%

An overview of systematic reviews found suboptimal reporting and methodological limitations of mediation studies investigating causal mechanisms

Cashin

Lee

Lamb

et al. 2019

Journal of Clinical Epidemiology

Self Cite

View full text Add to dashboard Cite

show abstract

“…Methodology for causal mediation analysis in epidemiological research is a rapidly advancing research field (197). The purpose of mediation analysis is to quantify possible mechanisms, i.e., pathways, through which an exposures executes or mediates effect on the outcome under study (198). The total outcome effect is separated into indirect and direct effects, where the indirect effect works through the mediator(s) under study.…”

Section: Motivationmentioning

confidence: 99%

“…A mediator is on the causal pathway between the exposure and the outcome. The ratio of the indirect effect and the total effect is the proportion mediated and a relative assessment of the pathway of interest (198).…”

Section: Motivationmentioning

confidence: 99%

“…A causal interpretation of indirect and direct effects, irrespective of methodology, requires fairly strong assumptions regarding confounding (181,197,198). Exposure-outcome (applies to all observational studies), mediator-outcome, and exposure-mediator confounders can all be controlled for by including the covariates of interest in the regression models (197).…”

Section: Motivationmentioning

confidence: 99%

See 1 more Smart Citation

Sex differences in cancer risk and survival.

Radkiewicz

Johansson

Dickman

et al. 2017

JCO

View full text Add to dashboard Cite

e13074 Background: The ongoing global cancer increase poses immense challenges for cancer care. Few comprehensive analyses of sex differences in cancer risk and survival based on robust data have been conducted. Methods: We performed a large cohort study using nationwide Swedish Cancer Register data for years 1970-2014, including all adult incident, non-breast, non-genital cancer cases (504,379 male, 368,018 female). To assess the strength of association between sex and cancer we estimated male-to-female incidence rate ratios†. Relative survival models were used to quantify differences in survival and were expressed as excess mortality ratios*. We applied Poisson regression and adjusted for age and calendar year in both. Results: Men were at significantly increased risk for 34 of 39 sites, and had significantly poorer prognosis for 27 of 39, compared to increased risk for 5 of 39 sites and poorer survival for 2 of 39, in women. Incidence rate ratios in male predominant sites ranged from 1.05; 95% CI, 1.03-1.08 (lung adenocarcinoma) to 8.00; 95% CI, 7.50-8.53 (larynx), corresponding to 0.38; 95% CI, 0.36-0.39 (thyroid well-differentiated) to 0.69; 95% CI, 0.63-0.75 (thyroid anaplastic) in female predominant sites. In sites with poorer survival in men, excess mortality ratios ranged from 1.07; 95% CI, 1.03-1.10 (colon) to 2.08; 95% CI, 1.53-2.84 (well-differentiated thyroid). Conclusions: Male sex is an independent risk factor and associated with poorer survival for a large majority of cancer sites. The excess risk in men remains elevated over time for most sites. Identifying and eliminating factors driving sex differences may reduce global cancer burden considerably. [Table: see text]

show abstract

“…Three types of effects of the intervention on the outcomes were quantified: indirect effects, direct effects and their sum -total effects. Indirect effects refers to the portions of total effects of the intervention that function through the mediator of interest, whereas direct effects account for the remaining part of the total effect(36).…”

mentioning

confidence: 99%

Activity in Nature Mediates a Park Prescription Intervention’s Effects on Physical Activity, Park Use and Quality of Life: A Mixed-Methods Process Evaluation

Petrunoff¹,

Yao²,

Sia³

et al. 2019

Preprint

View full text Add to dashboard Cite

Background: This process evaluation explored the implementation and mechanisms of impact of a Park Prescription intervention trial (PPI), including the effects of hypothesised mediators (motivation, social support, recreational physical activity [PA], park use and park PA ) on trial outcomes. Methods: Participants from the community were randomly allocated to intervention (n=80) or control (n=80) group. The intervention included baseline counselling, a prescription of exercise in parks, materials, three-month follow-up counselling and 26 weekly group exercise sessions in parks. Process evaluation indicators were assessed at three- and six-months. Implementation indicators included participation rates in intervention components and survey questions plus focus group discussions (FGDs) to understand which components participants valued. FGDs further assessed barriers and facilitators to intervention participation. To explore mechanisms of impact, linear regression was used to compare objectively measured PA between quantiles of group exercise participation. Structural equation modelling (SEM) explored hypothesised mediation of the significant intervention effects. Framework analysis was conducted for FGDs. Results: Participants were middle-aged (mean 51, SD ± 6.3 years), predominantly female (79%) and of Chinese ethnicity (81%). All intervention participants received baseline counselling, the park prescription and materials, whilst 94% received the follow-up counselling. Mean minutes of moderate-to-vigorous PA/week (95% CI) differed by group exercise participation (p=0.018): 0% participation (n=18) 128.3 (69.3, 187.2) minutes, >0-35.9% participation (n=18) 100.3 (36.9, 163.6) minutes, >35.9-67.9% participation (n=17) 50.5 (-4.9, 105.9) minutes and >67.9% participation (n=18) 177.4 (122.0, 232.8) minutes. Park PA at three-months had significant mediating effects (95% CI) on recreational PA 26.50 (6.65, 49.37) minutes/week, park use 185.38 (45.40, 353.74) minutes/month, park PA/month 165.48 (33.14, 334.16) minutes and psychological quality of life score 1.25 (0.19, 2.69) at six-months. Prioritising time with family and preferences for unstructured activities were barriers to intervention participation. Human interaction via follow-up or group exercise were facilitators. Conclusion: This process evaluation showed park PA consistently mediated effects of the PPI, suggesting activity in parks was a mechanism of its effects. To optimise effectiveness, participants’ preference for prioritising time with family through family involvement and tailoring the intervention to participants’ preferences for structured or unstructured PA could be considered in future studies.

show abstract

Mediation Analysis

Cited by 129 publications

References 6 publications

An overview of systematic reviews found suboptimal reporting and methodological limitations of mediation studies investigating causal mechanisms

An overview of systematic reviews found suboptimal reporting and methodological limitations of mediation studies investigating causal mechanisms

Sex differences in cancer risk and survival.

Activity in Nature Mediates a Park Prescription Intervention’s Effects on Physical Activity, Park Use and Quality of Life: A Mixed-Methods Process Evaluation

Contact Info

Product

Resources

About