Average Causal Effect Estimation Via Instrumental Variables: the No Simultaneous Heterogeneity Assumption.

BACKGROUND: Instrumental variables (IVs) can be used to provide evidence as to whether a treatment has a causal effect on an outcome . Even if the instrument satisfies the three core IV assumptions of relevance, independence, and exclusion restriction, further assumptions are required to identify the average causal effect (ACE) of on . Sufficient assumptions for this include homogeneity in the causal effect of on ; homogeneity in the association of with ; and no effect modification. METHODS: We describe the no simultaneous heterogeneity assumption, which requires the heterogeneity in the - causal effect to be mean independent of (i.e., uncorrelated with) both and heterogeneity in the - association. This happens, for example, if there are no common modifiers of the - effect and the - association, and the - effect is additive linear. We illustrate the assumption of no simultaneous heterogeneity using simulations and by re-examining selected published studies. RESULTS: Under no simultaneous heterogeneity, the Wald estimand equals the ACE even if both homogeneity assumptions and no effect modification (which we demonstrate to be special cases of-and therefore stronger than-no simultaneous heterogeneity) are violated. CONCLUSIONS: The assumption of no simultaneous heterogeneity is sufficient for identifying the ACE using IVs. Since this assumption is weaker than existing assumptions for ACE identification, doing so may be more plausible than previously anticipated.

Homogeneity in the Instrument-exposure Association and Point Estimation Using Binary Instrumental Variables.

BACKGROUND: Interpreting instrumental variable results often requires further assumptions in addition to the core assumptions of relevance, independence, and the exclusion restriction. METHODS: We assess whether instrument-exposure additive homogeneity renders the Wald estimand equal to the average derivative effect (ADE) in the case of a binary instrument and a continuous exposure. RESULTS: Instrument-exposure additive homogeneity is insufficient for ADE identification when the instrument is binary, the exposure is continuous, and the effect of the exposure on the outcome is nonlinear on the additive scale. For a binary exposure, the exposure-outcome effect is necessarily additive linear, so the homogeneity condition is sufficient. CONCLUSIONS: For binary instruments, instrument-exposure additive homogeneity identifies the ADE if the exposure is also binary. Otherwise, additional assumptions (such as additive linearity of the exposure-outcome effect) are required.

Is population structure in the genetic biobank era irrelevant, a challenge, or an opportunity?

Replicable genetic association signals have consistently been found through genome-wide association studies in recent years. The recent dramatic expansion of study sizes improves power of estimation of effect sizes, genomic prediction, causal inference, and polygenic selection, but it simultaneously increases susceptibility of these methods to bias due to subtle population structure. Standard methods using genetic principal components to correct for structure might not always be appropriate and we use a simulation study to illustrate when correction might be ineffective for avoiding biases. New methods such as trans-ethnic modeling and chromosome painting allow for a richer understanding of the relationship between traits and population structure. We illustrate the arguments using real examples (stroke and educational attainment) and provide a more nuanced understanding of population structure, which is set to be revisited as a critical aspect of future analyses in genetic epidemiology. We also make simple recommendations for how problems can be avoided in the future. Our results have particular importance for the implementation of GWAS meta-analysis, for prediction of traits, and for causal inference.

Correction to: Is population structure in the genetic biobank era irrelevant, a challenge, or an opportunity?

In the original article publication, there is an incorrect impression that Fig. 1 formed a formal Directed Acyclic Graph (DAG) by describing it as a causal model. However, it was not correct if interpreted in this way.

Multivariable two-sample Mendelian randomization estimates of the effects of intelligence and education on health.

Intelligence and education are predictive of better physical and mental health, socioeconomic position (SEP), and longevity. However, these associations are insufficient to prove that intelligence and/or education cause these outcomes. Intelligence and education are phenotypically and genetically correlated, which makes it difficult to elucidate causal relationships. We used univariate and multivariable Mendelian randomization to estimate the total and direct effects of intelligence and educational attainment on mental and physical health, measures of socioeconomic position, and longevity. Both intelligence and education had beneficial total effects. Higher intelligence had positive direct effects on income and alcohol consumption, and negative direct effects on moderate and vigorous physical activity. Higher educational attainment had positive direct effects on income, alcohol consumption, and vigorous physical activity, and negative direct effects on smoking, BMI and sedentary behaviour. If the Mendelian randomization assumptions hold, these findings suggest that both intelligence and education affect health.

Effect modification of FADS2 polymorphisms on the association between breastfeeding and intelligence: results from a collaborative meta-analysis.

BACKGROUND: Accumulating evidence suggests that breastfeeding benefits children's intelligence, possibly due to long-chain polyunsaturated fatty acids (LC-PUFAs) present in breast milk. Under a nutritional adequacy hypothesis, an interaction between breastfeeding and genetic variants associated with endogenous LC-PUFAs synthesis might be expected. However, the literature on this topic is controversial. METHODS: We investigated this gene × environment interaction through a collaborative effort. The primary analysis involved >12 000 individuals and used ever breastfeeding, FADS2 polymorphisms rs174575 and rs1535 coded assuming a recessive effect of the G allele, and intelligence quotient (IQ) in Z scores. RESULTS: There was no strong evidence of interaction, with pooled covariate-adjusted interaction coefficients (i.e. difference between genetic groups of the difference in IQ Z scores comparing ever with never breastfed individuals) of 0.12[(95% confidence interval (CI): -0.19; 0.43] and 0.06 (95% CI: -0.16; 0.27) for the rs174575 and rs1535 variants, respectively. Secondary analyses corroborated these results. In studies with ≥5.85 and <5.85 months of breastfeeding duration, pooled estimates for the rs174575 variant were 0.50 (95% CI: -0.06; 1.06) and 0.14 (95% CI: -0.10; 0.38), respectively, and 0.27 (95% CI: -0.28; 0.82) and -0.01 (95% CI: -0.19; 0.16) for the rs1535 variant. CONCLUSIONS: Our findings did not support an interaction between ever breastfeeding and FADS2 polymorphisms. However, subgroup analysis suggested that breastfeeding may supply LC-PUFAs requirements for cognitive development if breastfeeding lasts for some (currently unknown) time. Future studies in large individual-level datasets would allow properly powered subgroup analyses and further improve our understanding on the breastfeeding × FADS2 interaction.

Bias in Mendelian randomization due to assortative mating.

Mendelian randomization (MR) has been increasingly used to strengthen causal inference in observational epidemiology. Methodological developments in the field allow detecting and/or adjusting for different potential sources of bias, mainly bias due to horizontal pleiotropy (or "off-target" genetic effects). Another potential source of bias is nonrandom matching between spouses (i.e., assortative mating). In this study, we performed simulations to investigate the bias caused in MR by assortative mating. We found that bias can arise due to either cross-trait assortative mating (i.e., assortment on two phenotypes, such as highly educated women selecting taller men) or single-trait assortative mating (i.e., assortment on a single phenotype), even if the exposure and outcome phenotypes are not the phenotypes under assortment. The simulations also indicate that bias due to assortative mating accumulates over generations and that MR methods robust to horizontal pleiotropy are also affected by this bias. Finally, we show that genetic data from mother-father-offspring trios can be used to detect and correct for this bias.

Correction: Breastfeeding effects on DNA methylation in the offspring: A systematic literature review.

[This corrects the article DOI: 10.1371/journal.pone.0173070.].

Breastfeeding effects on DNA methylation in the offspring: A systematic literature review.

BACKGROUND: Breastfeeding benefits both infants and mothers. Recent research shows long-term health and human capital benefits among individuals who were breastfed. Epigenetic mechanisms have been suggested as potential mediators of the effects of early-life exposures on later health outcomes. We reviewed the literature on the potential effects of breastfeeding on DNA methylation. METHODS: Studies reporting original results and evaluating DNA methylation differences according to breastfeeding/breast milk groups (e.g., ever vs. never comparisons, different categories of breastfeeding duration, etc) were eligible. Six databases were searched simultaneously using Ovid, and the resulting studies were evaluated independently by two reviewers. RESULTS: Seven eligible studies were identified. Five were conducted in humans. Studies were heterogeneous regarding sample selection, age, target methylation regions, methylation measurement and breastfeeding categorisation. Collectively, the studies suggest that breastfeeding might be negatively associated with promoter methylation of LEP (which encodes an anorexigenic hormone), CDKN2A (involved in tumour suppression) and Slc2a4 genes (which encodes an insulin-related glucose transporter) and positively with promoter methylation of the Nyp (which encodes an orexigenic neuropeptide) gene, as well as influence global methylation patterns and modulate epigenetic effects of some genetic variants. CONCLUSIONS: The findings from our systematic review are far from conclusive due to the small number of studies and their inherent limitations. Further studies are required to understand the actual potential role of epigenetics in the associations of breastfeeding with later health outcomes. Suggestions for future investigations, focusing on epigenome-wide association studies, are provided.

Effect modification of FADS2 polymorphisms on the association between breastfeeding and intelligence: protocol for a collaborative meta-analysis.

INTRODUCTION: Evidence from observational studies and randomised controlled trials suggests that breastfeeding is positively associated with IQ, possibly because breast milk is a source of long-chain polyunsaturated fatty acids. Different studies have detected gene-breastfeeding interactions involving FADS2 variants and intelligence. However, findings are inconsistent regarding the direction of such effect modification. METHODS/DESIGN: To clarify how FADS2 and breastfeeding interact in their association with IQ, we are conducting a consortium-based meta-analysis of independent studies. Results produced by each individual study using standardised analysis scripts and harmonised data will be used. INCLUSION CRITERIA: breastfeeding, IQ and either rs174575 or rs1535 polymorphisms available; and being of European ancestry. EXCLUSION CRITERIA: twin studies; only poorly imputed genetic data available; or unavailability of proper ethics approval. Studies will be invited based on being known to have at least some of the required data, or suggested by participating studies as potentially eligible. This inclusive approach will favour achieving a larger sample size and be less prone to publication bias. DISCUSSION: Improving current understanding of FADS2-breastfeeding interaction may provide important biological insights regarding the importance of long-chain polyunsaturated fatty acids for the breastfeeding-IQ association. This meta-analysis will help to improve such knowledge by replicating earlier studies, conducting additional analysis and evaluating different sources of heterogeneity. Publishing this protocol will minimise the possibility of bias due to post hoc changes to the analysis protocol.

Associations of angiotensin targeting antihypertensive drugs with mortality and hospitalization in primary care patients with dementia.

We investigated whether angiotensin II receptor blockers and angiotensin converting enzyme inhibitors were associated with risk of mortality or inpatient hospitalization for patients with dementia compared to other antihypertensive medications. We extracted a clinical cohort of 6,290 patients with dementia from the United Kingdom General Practice Research Database, prescribed antihypertensive medication at diagnosis of dementia with around 10 years follow-up. Using survival analysis we estimated associations of exposure to antihypertensive medication with subsequent hospitalization and mortality risk, stratified by dementia type (Alzheimer's disease, vascular and other dementias). Angiotensin converting enzyme inhibitors (but not angiotensin II receptor blockers) were associated with an increased risk of mortality in patients with Alzheimer's disease (adjusted hazard ratio: 1.19; 95% CI 1.07, 1.33, p = 0.002), but no convincing evidence of increased hospitalization. Angiotensin II receptor blockers were inversely associated with hospitalization for any form of dementia, but after adjustment for covariates, these associations became consistent with chance. Further evidence is required to either support or refute the observation that exposure to angiotensin converting enzyme inhibitors in patients with dementia is associated with increased mortality.