Response to Hartwig and Davies

Bowden, Jack; Burgess, Stephen; Smith, George Davey

doi:10.1093/ije/dyw252

Cited by 129 publications

(30 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performance of sisVIVE in scenarios where InSIDE holds and there is no indirect pleiotropy is more encouraging, but more generally, we would recommend that MR studies should be based on a sensitivity analysis involving robust MR‐Egger and “valid IV” methods to capture whether differences between the estimates point to the presence of pleiotropic SNPs. We would also concur with others who have suggest using unweighted polygenic risk scores or a two‐sample strategy or both . For a one‐sample strategy, it appears that imprecision in the estimated weights of an IPRS, while improving efficiency, can lead to substantial bias.…”

Section: Discussionsupporting

confidence: 85%

“…We would also concur with others who have suggest using unweighted polygenic risk scores or a two-sample strategy or both. [31][32][33] For a one-sample strategy, it appears that imprecision in the estimated weights of an IPRS, while improving efficiency, can lead to substantial bias. While SPRSs perform well here, it is important to note that, in further simulations, we found (results not shown) that severe bias was introduced if the effect-allele coding of the SNPs led tô(or̃) being positive when true j would have led us to code it the other way; such "flip flopping" is possible, 34 even in the absence of population stratification, but remains a potential source of bias for SPRS despite its being discounted elsewhere (eg, see page 1883 in the work of Burgess et al).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Assessing the robustness of sisVIVE in a Mendelian randomization study to estimate the causal effect of body mass index on income using multiple SNPs from understanding society

Bao

Clarke

Smart

2018

Statistics in Medicine

View full text Add to dashboard Cite

The “some invalid, some valid instrumental variable estimator” (sisVIVE) is a lasso‐based method for instrumental variables (IVs) regression of outcome on an exposure. In principle, sisVIVE is robust to some of the IVs in the analysis being invalid, in the sense of being related to the outcome variable through pathways not mediated by the exposure. In this paper, we consider the application of sisVIVE to a Mendelian randomization study in which multiple genetic variants are used as IVs to estimate the causal effect of body mass index on personal income in the presence of unobserved confounding. In addition to analyzing data from the large‐scale longitudinal household survey Understanding Society , we conduct a simulation study to (a) assess the performance of sisVIVE in relation to that of competing robust methods like “MR‐Egger” and “MR‐Median” and (b) identify scenarios under which its absolute performance is poor. We find that sisVIVE outperforms alternative robust methods, in terms of mean‐square error, across a wide range of scenarios, but that its performance is poor in absolute terms when the presence of indirect pleiotropy leads to failure of the “InSIDE” condition, which is not explicitly required for identification. We argue that this is because the consistency criterion for sisVIVE does not identify the true causal effect when InSIDE fails.

show abstract

Section: Discussionsupporting

confidence: 85%

Section: Discussionmentioning

confidence: 99%

Assessing the robustness of sisVIVE in a Mendelian randomization study to estimate the causal effect of body mass index on income using multiple SNPs from understanding society

Bao

Clarke

Smart

2018

Statistics in Medicine

View full text Add to dashboard Cite

show abstract

“…We investigated this using MR-Egger, weighted median, and weighted mode estimators in summary data analyses. (43)(44)(45) We included the inverse variance weighted estimates for comparison. These estimates use the educational attainment GWAS discovery sample coefficients.…”

Section: Sensitivity Analysesmentioning

confidence: 99%

“…The estimated heterogeneity in the estimated effect of education across SNPs. Estimated using the I-squared statistic (43). …”

mentioning

confidence: 99%

The effect of education on adult mortality, health, and income: triangulating across genetic and policy reforms

Davies

Dickson

Smith

et al. 2018

Preprint

Self Cite

View full text Add to dashboard Cite

1AbstractOn average, educated people are healthier, wealthier and have higher life expectancy than those with less education. Numerous studies have attempted to determine whether these differences are caused by education, or are merely correlated with it and are ultimately caused by another factor. Previous studies have used a range of natural experiments to provide causal evidence. Here we exploit two natural experiments, perturbation of germline genetic variation associated with education which occurs at conception, known as Mendelian randomization, and a policy reform, the raising of the school leaving age in the UK in 1972. Previous studies have suggested that the differences in outcomes associated with education may be due to confounding. However, the two independent sources of variation we exploit largely imply consistent causal effects of education on outcomes much later in life.

show abstract

“…MR-Egger, weighted median and weighted mode estimates were compared to those obtained from the inverse-variance weighted (IVW) for two-sample MR 66 . For these analyses, the first-stage estimates (coefficients of the association between each SNP and BMI) were obtained from an independent external source, as to not induce weak instrument bias 69,70 , and the second-stage estimates (natural logarithm of the HR for each mortality outcome with each SNP) were obtained directly from UK Biobank.…”

Section: Sensitivity Analysesmentioning

confidence: 99%

Body mass index and mortality in UK Biobank: revised estimates using Mendelian randomization

Wade¹,

Carslake²,

Sattar

et al. 2018

Preprint

View full text Add to dashboard Cite

ObjectiveObtain estimates of the causal relationship between different levels of body mass index (BMI) and mortality.MethodsMendelian randomization (MR) was conducted using genotypic variation reliably associated with BMI to test the causal effect of increasing BMI on all-cause and cause-specific mortality in participants of White British ancestry in UK Biobank.ResultsMR analyses supported existing evidence for a causal association between higher levels of BMI and greater risk of all-cause mortality (hazard ratio (HR) per 1kg/m2: 1.02; 95% CI: 0.97,1.06) and mortality from cardiovascular diseases (HR: 1.12; 95% CI: 1.02, 1.23), specifically coronary heart disease (HR: 1.19; 95% CI: 1.05, 1.35) and those other than stroke/aortic aneurysm (HR: 1.13; 95% CI: 0.93, 1.38), stomach cancer (HR: 1.30; 95% CI: 0.91, 1.86) and oesophageal cancer (HR: 1.08; 95% CI: 0.84, 1.38), and with decreased risk of lung cancer mortality (HR: 0.97; 95% CI: 0.84, 1.11). Sex-stratified analyses supported a causal role of higher BMI in increasing the risk of mortality from bladder cancer in males and other causes in females, but in decreasing the risk of respiratory disease mortality in males. The characteristic J-shaped observational association between BMI and mortality was visible with MR analyses but with a smaller value of BMI at which mortality risk was lowest and apparently flatter over a larger range of BMI.ConclusionResults support a causal role of higher BMI in increasing the risk of all-cause mortality and mortality from other causes. However, studies with greater numbers of deaths are needed to confirm the current findings.

show abstract

Response to Hartwig and Davies

Cited by 129 publications

References 13 publications

Assessing the robustness of sisVIVE in a Mendelian randomization study to estimate the causal effect of body mass index on income using multiple SNPs from understanding society

Assessing the robustness of sisVIVE in a Mendelian randomization study to estimate the causal effect of body mass index on income using multiple SNPs from understanding society

The effect of education on adult mortality, health, and income: triangulating across genetic and policy reforms

Body mass index and mortality in UK Biobank: revised estimates using Mendelian randomization

Contact Info

Product

Resources

About