Applied causal inference methods for sequential mediators

BACKGROUND

Mediation analysis aims at estimating to what extent the effect of an exposure on an outcome is explained by a set of mediators on the causal pathway between the exposure and the outcome. The total effect of the exposure on the outcome can be decomposed into an indirect effect, i.e. the effect explained by the mediators jointly, and a direct effect, i.e. the effect unexplained by the mediators. However finer decompositions are possible in presence of independent or sequential mediators.

METHODS

We review four statistical methods to analyse multiple sequential mediators, the inverse odds ratio weighting approach, the inverse probability weighting approach, the imputation approach and the extended imputation approach. These approaches are compared and implemented using a case-study with the aim to investigate the mediating role of adverse reproductive outcomes and infant respiratory infections in the effect of maternal pregnancy mental health on infant wheezing in the Ninfea birth cohort.

RESULTS

Using the inverse odds ratio weighting approach, the direct effect of maternal depression or anxiety in pregnancy is equal to a 59% (95% CI: 27%,94%) increased prevalence of infant wheezing and the mediated effect through adverse reproductive outcomes is equal to a 3% (95% CI: -6%,12%) increased prevalence of infant wheezing. When including infant lower respiratory infections in the mediation pathway, the direct effect decreases to 57% (95% CI: 25%,92%) and the indirect effect increases to 5% (95% CI: -5%,15%). The estimates of the effects obtained using the weighting and the imputation approaches are similar. The extended imputation approach suggests that the small joint indirect effect through adverse reproductive outcomes and lower respiratory infections is due entirely to the contribution of infant lower respiratory infections, and not to an increased prevalence of adverse reproductive outcomes.

CONCLUSIONS

The four methods revealed similar results of small mediating role of adverse reproductive outcomes and early respiratory tract infections in the effect of maternal pregnancy mental health on infant wheezing. The choice of the method depends on what is the effect of main interest, the type of the variables involved in the analysis (binary, categorical, count or continuous) and the confidence in specifying the models for the exposure, the mediators and the outcome.

Association of leukocyte DNA methylation changes with dietary folate and alcohol intake in the EPIC study

Background

There is increasing evidence that folate, an important component of one-carbon metabolism, modulates the epigenome. Alcohol, which can disrupt folate absorption, is also known to affect the epigenome. We investigated the association of dietary folate and alcohol intake on leukocyte DNA methylation levels in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Leukocyte genome-wide DNA methylation profiles on approximately 450,000 CpG sites were acquired with Illumina HumanMethylation 450K BeadChip measured among 450 women control participants of a case-control study on breast cancer nested within the EPIC cohort. After data preprocessing using surrogate variable analysis to reduce systematic variation, associations of DNA methylation with dietary folate and alcohol intake, assessed with dietary questionnaires, were investigated using CpG site-specific linear models. Specific regions of the methylome were explored using differentially methylated region (DMR) analysis and fused lasso (FL) regressions. The DMR analysis combined results from the feature-specific analysis for a specific chromosome and using distances between features as weights whereas FL regression combined two penalties to encourage sparsity of single features and the difference between two consecutive features.

Results

After correction for multiple testing, intake of dietary folate was not associated with methylation level at any DNA methylation site, while weak associations were observed between alcohol intake and methylation level at CpG sites cg03199996 and cg07382687, with q_val = 0.029 and q_val = 0.048, respectively. Interestingly, the DMR analysis revealed a total of 24 and 90 regions associated with dietary folate and alcohol, respectively. For alcohol intake, 6 of the 15 most significant DMRs were identified through FL.

Conclusions

Alcohol intake was associated with methylation levels at two CpG sites. Evidence from DMR and FL analyses indicated that dietary folate and alcohol intake may be associated with genomic regions with tumor suppressor activity such as the GSDMD and HOXA5 genes. These results were in line with the hypothesis that epigenetic mechanisms play a role in the association between folate and alcohol, although further studies are warranted to clarify the importance of these mechanisms in cancer.

Electronic supplementary material

The online version of this article (10.1186/s13148-019-0637-x) contains supplementary material, which is available to authorized users.

Dietary glycemic index, glycemic load, and cancer risk: results from the EPIC-Italy study

Factors linked to glucose metabolism are involved in the etiology of several cancers. High glycemic index (GI) or high glycemic load (GL) diets, which chronically raise postprandial blood glucose, may increase cancer risk by affecting insulin-like growth factor. We prospectively investigated cancer risk and dietary GI/GL in the EPIC-Italy cohort. After a median 14.9 years, 5112 incident cancers and 2460 deaths were identified among 45,148 recruited adults. High GI was associated with increased risk of colon and bladder cancer. High GL was associated with: increased risk of colon cancer; increased risk of diabetes-related cancers; and decreased risk of rectal cancer. High intake of carbohydrate from high GI foods was significantly associated with increased risk of colon and diabetes-related cancers, but decreased risk of stomach cancer; whereas high intake of carbohydrates from low GI foods was associated with reduced colon cancer risk. In a Mediterranean population with high and varied carbohydrate intake, carbohydrates that strongly raise postprandial blood glucose may increase colon and bladder cancer risk, while the quantity of carbohydrate consumed may be involved in diabetes-related cancers. Further studies are needed to confirm the opposing effects of high dietary GL on risks of colon and rectal cancers.