Interpreting findings from Mendelian randomization using the MR-Egger method

Serum magnesium levels and risk of coronary artery disease: Mendelian randomisation study

BACKGROUND

Observational studies have shown that serum magnesium levels are inversely associated with risk of cardiovascular disease, but whether this association is causal is unknown. We conducted a Mendelian randomisation study to investigate whether serum magnesium levels may be causally associated with coronary artery disease (CAD).

METHODS

This Mendelian randomisation analysis is based on summary-level data from the CARDIoGRAMplusC4D consortium's 1000 Genomes-based genome-wide association meta-analysis of 48 studies with a total of 60,801 CAD cases and 123,504 non-cases. Six single-nucleotide polymorphisms associated with serum magnesium levels at genome-wide significance were used as instrumental variables.

RESULTS

A genetic predisposition to higher serum magnesium levels was inversely associated with CAD. In conventional Mendelian randomisation analysis, the odds ratio of CAD was 0.88 (95% confidence interval [CI] 0.78 to 0.99; P = 0.03) per 0.1-mmol/L (about 1 standard deviation) increase in genetically predicted serum magnesium levels. Results were consistent in sensitivity analyses using the weighted median and heterogeneity-penalised model averaging methods, with odds ratios of 0.84 (95% CI 0.72 to 0.98; P = 0.03) and 0.83 (95% CI 0.71 to 0.96; P = 0.02), respectively.

CONCLUSIONS

This study based on genetics provides evidence that serum magnesium levels are inversely associated with risk of CAD. Randomised controlled trials elucidating whether magnesium supplementation lowers the risk of CAD, preferably in a setting at higher risk of hypomagnesaemia, are warranted.

MR-PheWAS: exploring the causal effect of SUA level on multiple disease outcomes by using genetic instruments in UK Biobank

Objectives

We aimed to investigate the role of serum uric acid (SUA) level in a broad spectrum of disease outcomes using data for 120 091 individuals from UK Biobank.

Methods

We performed a phenome-wide association study (PheWAS) to identify disease outcomes associated with SUA genetic risk loci. We then implemented conventional Mendelianrandomisation (MR) analysis to investigate the causal relevance between SUA level and disease outcomes identified from PheWAS. We next applied MR Egger analysis to detect and account for potential pleiotropy, which conventional MR analysis might mistake for causality, and used the HEIDI (heterogeneity in dependent instruments) test to remove cross-phenotype associations that were likely due to genetic linkage.

Results

Our PheWAS identified 25 disease groups/outcomes associated with SUA genetic risk loci after multiple testing correction (P<8.57e-05). Our conventional MR analysis implicated a causal role of SUA level in three disease groups: inflammatory polyarthropathies (OR=1.22, 95% CI 1.11 to 1.34), hypertensive disease (OR=1.08, 95% CI 1.03 to 1.14) and disorders of metabolism (OR=1.07, 95% CI 1.01 to 1.14); and four disease outcomes: gout (OR=4.88, 95% CI 3.91 to 6.09), essential hypertension (OR=1.08, 95% CI 1.03 to 1.14), myocardial infarction (OR=1.16, 95% CI 1.03 to 1.30) and coeliac disease (OR=1.41, 95% CI 1.05 to 1.89). After balancing pleiotropic effects in MR Egger analysis, only gout and its encompassing disease group of inflammatory polyarthropathies were considered to be causally associated with SUA level. Our analysis highlighted a locus (ATXN2/S2HB3) that may influence SUA level and multiple cardiovascular and autoimmune diseases via pleiotropy.

Conclusions

Elevated SUA level is convincing to cause gout and inflammatory polyarthropathies, and might act as a marker for the wider range of diseases with which it associates. Our findings support further investigation on the clinical relevance of SUA level with cardiovascular, metabolic, autoimmune and respiratory diseases.

Coagulation Factors and the Risk of Ischemic Heart Disease: A Mendelian Randomization Study

BACKGROUND

Coagulation plays a role in ischemic heart disease (IHD). However, which coagulation factors are targets of intervention is unclear. We assessed how genetically predicted vWF (von Willebrand factor), ETP (endogenous thrombin potential), FVIII (factor VIII), d-dimer, tPA (tissue-type plasminogen activator), and PAI (plasminogen activator inhibitor)-1 affected IHD. We similarly estimated effects on lipids to determine whether any associations were independent of lipids.

METHODS AND RESULTS

Separate sample instrumental variable analysis with genetic instruments, that is, Mendelian randomization, was used to obtain unconfounded estimates of effects on IHD using extensively genotyped studies of coronary artery disease/myocardial infarction, CARDIoGRAMplusC4D Metabochip (64 374 cases, 130 681 controls) and CARDIoGRAMplusC4D 1000 Genomes (60 801 cases, 123 504 controls), and on lipids using the Global Lipids Genetics Consortium Results (n=196 475). Genetically predicted ETP was positively associated with IHD (odds ratio, 1.05 per log-transformed SD; 95% confidence interval, 1.03-1.07) based on 15 single-nucleotide polymorphisms, as were vWF (odds ratio, 1.05 per SD; 95% confidence interval, 1.02-1.08) and FVIII (odds ratio, 1.06 per SD; 95% confidence interval, 1.03-1.09) based on 16 and 6 single-nucleotide polymorphisms, respectively, but the latter associations were null after considering pleiotropy. vWF and FVIII were associated with higher LDL (low-density lipoprotein) cholesterol, but not after considering pleiotropy. Genetically predicted d-dimer, tPA, and PAI-1 were not clearly associated with IHD or lipids based on 3, 3, and 5 single-nucleotide polymorphisms, respectively.

CONCLUSIONS

ETP may affect IHD. Assessing the role of its drivers in more precisely phenotyped studies of IHD could be worthwhile.

Modifiable pathways in Alzheimer's disease: Mendelian randomisation analysis

OBJECTIVE

To determine which potentially modifiable risk factors, including socioeconomic, lifestyle/dietary, cardiometabolic, and inflammatory factors, are associated with Alzheimer's disease.

DESIGN

Mendelian randomisation study using genetic variants associated with the modifiable risk factors as instrumental variables.

SETTING

International Genomics of Alzheimer's Project.

PARTICIPANTS

17 008 cases of Alzheimer's disease and 37 154 controls.

MAIN OUTCOME MEASURES

Odds ratio of Alzheimer's per genetically predicted increase in each modifiable risk factor estimated with Mendelian randomisation analysis.

RESULTS

This study included analyses of 24 potentially modifiable risk factors. A Bonferroni corrected threshold of P=0.002 was considered to be significant, and P<0.05 was considered suggestive of evidence for a potential association. Genetically predicted educational attainment was significantly associated with Alzheimer's. The odds ratios were 0.89 (95% confidence interval 0.84 to 0.93; P=2.4×10-6) per year of education completed and 0.74 (0.63 to 0.86; P=8.0×10-5) per unit increase in log odds of having completed college/university. The correlated trait intelligence had a suggestive association with Alzheimer's (per genetically predicted 1 SD higher intelligence: 0.73, 0.57 to 0.93; P=0.01). There was suggestive evidence for potential associations between genetically predicted higher quantity of smoking (per 10 cigarettes a day: 0.69, 0.49 to 0.99; P=0.04) and 25-hydroxyvitamin D concentrations (per 20% higher levels: 0.92, 0.85 to 0.98; P=0.01) and lower odds of Alzheimer's and between higher coffee consumption (per one cup a day: 1.26, 1.05 to 1.51; P=0.01) and higher odds of Alzheimer's. Genetically predicted alcohol consumption, serum folate, serum vitamin B12, homocysteine, cardiometabolic factors, and C reactive protein were not associated with Alzheimer's disease.

CONCLUSION

These results provide support that higher educational attainment is associated with a reduced risk of Alzheimer's disease.

A Mendelian Randomization Study of Metabolite Profiles, Fasting Glucose, and Type 2 Diabetes

Mendelian randomization (MR) provides us the opportunity to investigate the causal paths of metabolites in type 2 diabetes and glucose homeostasis. We developed and tested an MR approach based on genetic risk scoring for plasma metabolite levels, utilizing a pathway-based sensitivity analysis to control for nonspecific effects. We focused on 124 circulating metabolites that correlate with fasting glucose in the Erasmus Rucphen Family (ERF) study (n = 2,564) and tested the possible causal effect of each metabolite with glucose and type 2 diabetes and vice versa. We detected 14 paths with potential causal effects by MR, following pathway-based sensitivity analysis. Our results suggest that elevated plasma triglycerides might be partially responsible for increased glucose levels and type 2 diabetes risk, which is consistent with previous reports. Additionally, elevated HDL components, i.e., small HDL triglycerides, might have a causal role of elevating glucose levels. In contrast, large (L) and extra large (XL) HDL lipid components, i.e., XL-HDL cholesterol, XL-HDL-free cholesterol, XL-HDL phospholipids, L-HDL cholesterol, and L-HDL-free cholesterol, as well as HDL cholesterol seem to be protective against increasing fasting glucose but not against type 2 diabetes. Finally, we demonstrate that genetic predisposition to type 2 diabetes associates with increased levels of alanine and decreased levels of phosphatidylcholine alkyl-acyl C42:5 and phosphatidylcholine alkyl-acyl C44:4. Our MR results provide novel insight into promising causal paths to and from glucose and type 2 diabetes and underline the value of additional information from high-resolution metabolomics over classic biochemistry.