Adverse Impact of Desulfovibrio spp. and Beneficial Role of Anaerostipes spp. on Renal Function: Insights from a Mendelian Randomization Analysis

Background: The microbiota composition is now considered as one of the main modifiable risk factors for health. No controlled study has been performed on the association between microbiota composition and renal function. We applied Mendelian randomization (MR) to estimate the casual impact of eight microbiota genera on renal function and the risk of chronic kidney disease (CKD). Methods: MR was implemented by using summary-level data from the largest-ever genome-wide association studies (GWAS) conducted on microbiota genera, CKD and renal function parameters. The inverse-variance weighted method (IVW), weighted median (WM)-based method, MR-Egger, MR-Robust Adjusted Profile Score (RAPS), MR-Pleiotropy RESidual Sum and Outlier (PRESSO) were applied. A sensitivity analysis was conducted using the leave-one-out method. Results: The Anaerostipes genus was associated with higher estimated glomerular filtration rate (eGFR) in the overall population (IVW: β = 0.003, p = 0.021) and non-diabetes mellitus (DM) subgroup (IVW: β = 0.003, p = 0.033), while it had a non-significant association with the risk of CKD and eGFR in DM patients. Subjects with higher abundance of Desulfovibrio spp. had a significantly lower level of eGFR (IVW: β = −0.001, p = 0.035); the same results were observed in non-DM (IVW: β = −0.001, p = 0.007) subjects. Acidaminococcus, Bacteroides, Bifidobacterium, Faecalibacterium, Lactobacillus and Megamonas had no significant association with eGFR in the overall population, DM and non-DM subgroups (IVW: p > 0.105 for all groups); they also presented no significant association with the risk of CKD (IVW: p > 0.201 for all groups). Analyses of MR-PRESSO did not highlight any outlier. The pleiotropy test, with very negligible intercept and insignificant p-value, also indicated no chance of pleiotropy for all estimations. The leave-one-out method demonstrated that the observed links were not driven by single single-nucleotide polymorphism. Conclusions: Our results suggest an adverse association of Desulfovibrio spp. and a beneficial association of Anaerostipes spp. with eGFR. Further studies using multiple robust instruments are needed to confirm these results.

Download Full-text

Longer sleep duration may negatively affect renal function

European Heart Journal ◽

10.1093/ehjci/ehaa946.2373 ◽

2020 ◽

Vol 41 (Supplement_2) ◽

Author(s):

M Mazidi ◽

N Shekoohi ◽

N Katsiki ◽

M Banach

Keyword(s):

Kidney Function ◽

Sleep Duration ◽

Total Population ◽

Association Studies ◽

P Value ◽

Genome Wide Association Studies ◽

Single Single Nucleotide Polymorphism ◽

Inverse Variance ◽

Leave One Out ◽

The Impact

Abstract Background Observational studies evaluating the link between sleep duration and kidney function reported controversial results. In the present study, Mendelian Randomization (MR) analysis was applied to obtain unconfounded estimates of the casual association of genetically determined sleep duration with estimated glomerular filtration rate (eGFR) and the risk of chronic kidney disease (CKD). Methods Data from the largest genome-wide association studies (GWAS) on self-reported and accelerometer derived sleep duration, eGFR and CKD were analysed in total, as well as separately in diabetic and non-diabetic individuals. Inverse variance weighted method (IVW), weighted median (WM)-based method, MR-Egger, as well as MR-Pleiotropy RESidual Sum and Outlier (PRESSO) were applied. To rule out the impact of single single-nucleotide polymorphism (SNP), the leave-one-out method was used. Results Overall, individuals with genetically longer self-reported sleep duration had a higher CKD risk (IVW: beta=0.358, p=0.047). Furthermore, in non-diabetics, longer self-reported sleep duration was negatively associated eGFR (IVW: beta=−0.024, p=0.020). Similarly, accelerometer derived sleep duration was negatively related to eGFR in the total population (IVW: beta=−0.019, p=0.047) and the non-diabetic individuals (IVW: beta=−0.025, p=0.014) (Table). No significant association was found between self-reported sleep duration and eGFR in the whole population (IVW: beta=−0.019, p=0.072) and T2DM patients (IVW: beta=0.028, p=0.484). None of the estimated associations was subjected to a significant level of heterogeneity. Furthermore, MR-PRESSO analysis did not show any chance of outliers for all estimates. The pleiotropy test, with very negligible intercept and insignificant p value. The results of the MR-RAPS were identical with the IVW estimates, highlighting again no possibility of pleiotropy. The leave-one-out method demonstrated that the links were not driven by single SNPs. Conclusions For the first time, the present study shed a light on the potential harmful effects of longer sleep duration (measured both objectively and subjectively) on kidney function. This finding was observed in the total population and in non-diabetic individuals, but not in those with diabetes. Further research is needed to elucidate the links between sleep duration, eGFR and CKD. Funding Acknowledgement Type of funding source: None

Download Full-text

Mendelian randomization studies of brain MRI yield insights into the pathogenesis of neuropsychiatric disorders

BMC Genomics ◽

10.1186/s12864-021-07661-8 ◽

2021 ◽

Vol 22 (S3) ◽

Author(s):

Weichen Song ◽

Wei Qian ◽

Weidi Wang ◽

Shunying Yu ◽

Guan Ning Lin

Keyword(s):

Mendelian Randomization ◽

Diffusion Tensor ◽

Association Studies ◽

Brain Mri ◽

Null Distribution ◽

Brain Structures ◽

Neuropsychiatric Disorders ◽

The Body ◽

P Value ◽

Genome Wide Association Studies

Abstract Background Observational studies have identified various associations between neuroimaging alterations and neuropsychiatric disorders. However, whether such associations could truly reflect causal relations remains still unknown. Results Here, we leveraged genome-wide association studies (GWAS) summary statistics for (1) 11 psychiatric disorders (sample sizes varied from n = 9,725 to 1,331,010); (2) 110 diffusion tensor imaging (DTI) measurement (sample size n = 17,706); (3) 101 region-of-interest (ROI) volumes, and investigate the causal relationship between brain structures and neuropsychiatric disorders by two-sample Mendelian randomization. Among all DTI-Disorder combinations, we observed a significant causal association between the superior longitudinal fasciculus (SLF) and the risk of Anorexia nervosa (AN) (Odds Ratio [OR] = 0.62, 95 % confidence interval: 0.50 ~ 0.76, P = 6.4 × 10− 6). Similar significant associations were also observed between the body of the corpus callosum (fractional anisotropy) and Alzheimer’s disease (OR = 1.07, 95 % CI: 1.03 ~ 1.11, P = 4.1 × 10− 5). By combining all observations, we found that the overall p-value for DTI − Disorder associations was significantly elevated compared to the null distribution (Kolmogorov-Smirnov P = 0.009, inflation factor λ = 1.37), especially for DTI − Bipolar disorder (BP) (λ = 2.64) and DTI − AN (λ = 1.82). In contrast, for ROI-Disorder combinations, we only found a significant association between the brain region of pars triangularis and Schizophrenia (OR = 0.48, 95 % CI: 0.34 ~ 0.69, P = 5.9 × 10− 5) and no overall p-value elevation for ROI-Disorder analysis compared to the null expectation. Conclusions As a whole, we show that SLF degeneration may be a risk factor for AN, while DTI variations could be causally related to some neuropsychiatric disorders, such as BP and AN. Also, the white matter structure might have a larger impact on neuropsychiatric disorders than subregion volumes.

Download Full-text

Coffee Types and Cardiac Metabolic Risks: Large-scale Cross-phenotype Association Study and Mendelian Randomization Analysis

10.21203/rs.3.rs-776922/v1 ◽

2021 ◽

Author(s):

Xinpei Wang ◽

Jinzhu Jia ◽

Tao Huang

Keyword(s):

Antigen Processing ◽

Large Scale ◽

Mendelian Randomization ◽

Association Studies ◽

P Value ◽

Separate Analysis ◽

Genome Wide Association Studies ◽

Coffee Intake ◽

Different Types ◽

Ground Coffee

Abstract Purpose: To explore whether coffee intake is associated with cardiac metabolic risks from a genetic perspective, and whether this association remains the same among different types of coffee consumers.Methods: We utilised the summary-level results of 28 genome-wide association studies (total sample size: ~5,000,000). First, we used linkage disequilibrium score regression and cross-phenotypic association analysis to estimate the genetic correlation and identify shared genes between coffee intake and various cardiac metabolic risks. Second, we used Mendelian randomization (MR) analysis to test whether there was a significant genetically predicted causal association between coffee intake and cardiac metabolic risks. For all the analyses above, we also conducted a separate analysis for different types of coffee consumers, in addition to total coffee intake.Results: Genetically, coffee intake and choice for decaffeinated/instant coffee had significant positive correlation with body mass index (BMI) and some other cardiac metabolic risks, while choice for ground coffee was significantly negatively associated with these risks. Between these genetically related phenotypes, there were 1708 genomic shared regions, of which 139 loci were novel. Enrichment analysis showed that these shared genes were significantly enriched in antigen processing related biological processes. MR analysis indicated that higher genetically proxied coffee intake may increase BMI (b: 0.35, p-value: 1.80ⅹ10-05), while genetically proxied choice for ground coffee can reduce BMI (b: -0.08, p-value: 6.50ⅹ10-05), and the risk of T2D (T2D: b: -0.2, p-value: 4.70ⅹ10-10; T2D adjusted for BMI: b: -0.11, p-value: 4.60ⅹ10-05).Conclusions: Compared with other types of coffee, ground coffee has a significant negative genetic and genetically predicated causal relationship with cardiac metabolic risks. And this association is likely to be mediated by immunity. The effect of different coffee types on cardiac metabolic risks is not equal, researchers on coffee should pay more attention to distinguishing between coffee types.

Download Full-text

Serum anti-inflammatory and inflammatory markers have no causal impact on telomere length: a Mendelian randomization study

Archives of Medical Science ◽

10.5114/aoms/119965 ◽

2021 ◽

Vol 17 (3) ◽

pp. 739-751

Author(s):

Mohsen Mazidi ◽

Niloofar Shekoohi ◽

Niki Katsiki ◽

Michal Rakowski ◽

Dimitri Mikhailidis ◽

...

Keyword(s):

Telomere Length ◽

Inflammatory Markers ◽

Mendelian Randomization ◽

Causal Effect ◽

Association Studies ◽

Telomere Maintenance ◽

Genome Wide Association Studies ◽

Biological Index ◽

Anti Inflammatory ◽

Leave One Out

IntroductionThe relationship between inflammatory and anti-inflammato�ry markers and telomere length (TL), a biological index of aging, is still poor�ly understood. By applying a 2-sample Mendelian randomization (MR), we investigated the causal associations between adiponectin, bilirubin, C-reac�tive protein (CRP), leptin, and serum uric acid (SUA) with TL.Material and methodsMR was implemented by using summary-level data from the largest ever genome-wide association studies (GWAS) conducted on our interested exposure and TL. Inverse variance weighted method (IVW), weighted median (WM)-based method, MR-Egger, MR-Robust Adjusted Pro�file Score (RAPS), and MR-Pleiotropy RESidual Sum and Outlier (PRESSO) were applied. Sensitivity analysis was conducted using the leave-one-out method.ResultsWith regard to adiponectin, CRP, leptin, and SUA levels, we found no effect on TL for all 4 types of tests (all p > 0.108). Results of the MR-Egger (p = 0.892) and IVW (p = 0.124) showed that bilirubin had no effect on telomere maintenance, whereas the results of the WM (p = 0.030) and RAPS (p = 0.022) were negative, with higher bilirubin concentrations linked to shorter TL. There was a low likelihood of heterogeneity for all the estima�tions, except for bilirubin (IVW p = 0.026, MR Egger p = 0.018). MR-PRESSO highlighted no outlier. For all the estimations, we observed negligible inter�cepts that were indicative of low likelihood of the pleiotropy (all p > 0.161). The results of leave-one-out method demonstrated that the links are not driven because of single nucleotide polymorphisms (SNPs).ConclusionsOur results highlight that neither the anti-inflammatory nor pro-inflammatory markers tested have any significant causal effect on TL. The casual role of bilirubin on TL still needs to be investigated.

Download Full-text

The putative causal effect of type 2 diabetes in risk of cataract: a Mendelian randomization study in East Asian

10.1101/2021.02.08.430342 ◽

2021 ◽

Author(s):

Haoyang Zhang ◽

Xuehao Xiu ◽

Angli Xue ◽

Yuedong Yang ◽

Yuanhao Yang ◽

...

Keyword(s):

Type 2 Diabetes ◽

Genetic Correlation ◽

Mendelian Randomization ◽

Causal Effect ◽

Association Studies ◽

Population Based ◽

P Value ◽

Genome Wide Association Studies ◽

Genome Wide

AbstractBackgroundThe epidemiological association between type 2 diabetes and cataract has been well-established. However, it remains unclear whether the two diseases share a genetic basis, and if so, whether this reflects a causal relationship.MethodsWe utilized East Asian population-based genome-wide association studies (GWAS) summary statistics of type 2 diabetes (Ncase=36,614, Ncontrol=155,150) and cataract (Ncase=24,622, Ncontrol=187,831) to comprehensively investigate the shared genetics between the two diseases. We performed 1. linkage disequilibrium score regression (LDSC) and heritability estimation from summary statistics (ρ-HESS) to estimate the genetic correlation and local genetic correlation between type 2 diabetes and cataract; 2. multiple Mendelian randomization (MR) analyses to infer the putative causality between type 2 diabetes and cataract; and 3. Summary-data-based Mendelian randomization (SMR) to identify candidate risk genes underling the causality.ResultsWe observed a strong genetic correlation (rg=0.58; p-value=5.60×10−6) between type 2 diabetes and cataract. Both ρ-HESS and multiple MR methods consistently showed a putative causal effect of type 2 diabetes on cataract, with estimated liability-scale MR odds ratios (ORs) at around 1.10 (95% confidence interval [CI] ranging from 1.06 to 1.17). In contrast, no evidence supports a causal effect of cataract on type 2 diabetes. SMR analysis identified two novel genes MIR4453HG (βSMR=−0.34, p-value=6.41×10−8) and KCNK17 (βSMR=−0.07, p-value=2.49×10−10), whose expression levels were likely involved in the putative causality of type 2 diabetes on cataract.ConclusionsOur results provided robust evidence supporting a causal effect of type 2 diabetes on the risk of cataract in East Asians, and posed new paths on guiding prevention and early-stage diagnosis of cataract in type 2 diabetes patients.Key MessagesWe utilized genome-wide association studies of type 2 diabetes and cataract in a large Japanese population-based cohort and find a strong genetic overlap underlying the two diseases.We performed multiple Mendelian randomization models and consistently disclosed a putative causal effect of type 2 diabetes on the development of cataract.We revealed two candidate genes MIR4453HG and KCNK17 whose expression levelss are likely relevant to the causality between type 2 diabetes and cataract.Our study provided theoretical fundament at the genetic level for improving early diagnosis, prevention and treatment of cataract in type 2 diabetes patients in clinical practice

Download Full-text

What multiple Mendelian randomization approaches reveal about obesity and gout

10.1101/2021.03.26.21254420 ◽

2021 ◽

Author(s):

Charleen D. Adams ◽

Brian Boutwell

Keyword(s):

Observational Studies ◽

Mendelian Randomization ◽

Association Studies ◽

Hdl Cholesterol ◽

P Value ◽

Genome Wide Association Studies ◽

Cholesterol Levels ◽

Genome Wide ◽

The Impact ◽

Apparent Ability

Background/Objectives: Gout is a painful arthritic disease. A robust canon of observational literature suggests strong relationships between obesity, high urate levels, and gout. But findings from observational studies can be fraught with confounding and reverse causation. They can conflict with findings from Mendelian randomization (MR), designed to tackle these biases. We aimed to determine whether the relationships between obesity, higher urate levels, and gout were causal using multiple MR approaches, including an investigation of how other closely related traits, LDL, HDL cholesterol, and triglyceride levels fit into the picture. Subjects/Methods: Summary results from genome-wide association studies of the five above-mentioned traits were extracted and used to perform two-sample (univariable, multivariable, and two-step) MR and MR mediation analysis. Results Obesity increased urate (beta=0.127; 95% CI=0.098, 0.157; P-value=1.2E-17) and triglyceride levels (beta=0.082; 95% CI=0.065, 0.099; P-value=1.2E-21) and decreased HDL cholesterol levels (beta=-0.083; 95% CI=-0.101, -0.065; P-value=2.5E-19). Higher triglyceride levels increased urate levels (beta=0.198; 95% CI=0.146, 0.251; P-value=8.9E-14) and higher HDL levels decreased them (beta=-0.109; 95% CI=-0.148, -0.071; P-value=2.7E-08). Higher urate levels (OR=1.030; 95% CI=1.028, 1.032; P-value=1.1E-130) and obesity caused gout (OR=1.003; 95% CI=1.001, 1.004; P-value=1.3E-04). The mediation MR of obesity on gout with urate levels as a mediator revealed, however, that essentially all of the effect of obesity on gout is mediated through urate. The impact of obesity on LDL cholesterol was null (beta=-0.011; 95% CI=-0.030, 0.008; P-value=2.6E-01), thus it was not included in the multivariable MR. The multivariable MR of obesity, HDL cholesterol, and triglycerides on urate levels revealed that obesity has an effect on urate levels even when accounting for HDL cholesterol and triglyceride levels. Conclusions: Obesity impacts gout indirectly by influencing urate levels and possibly other traits, such as triglycerides, that increase urate levels. Obesity's impact on urate is exacerbated by its apparent ability to decrease HDL cholesterol.

Download Full-text

Abstract 15527: Association Between Adrenergic Receptor Modulation and the Risk of Heart Failure: A Two-sample Mendelian Randomization Study

Circulation ◽

10.1161/circ.142.suppl_3.15527 ◽

2020 ◽

Vol 142 (Suppl_3) ◽

Author(s):

Claire Baudier ◽

Françoise Fougerousse ◽

Amand F Schmidt ◽

Folkert W Asselbergs ◽

Mickael Guedj ◽

...

Keyword(s):

Heart Failure ◽

Mendelian Randomization ◽

Association Studies ◽

P Value ◽

Sample Population ◽

Genome Wide Association Studies ◽

Receptor Modulation ◽

Genome Wide ◽

Genes Encoding ◽

The Impact

Introduction: The impact of the sympathetic nervous system (SNS) modulation on the risk of heart failure (HF) outside of ß1 receptor blockade remains controversial. Methods: We performed a two-sample Mendelian randomization (MR) study using common independent genetic variants located in the cis region of genes encoding the 9 SNS receptors (α1 A, B, D, α2 A, B, C and ß 1, 2 and 3) that were associated at genome-wide significance (P-value ≤ 5х10 –8 ) with blood pressure (BP) and/or heart rate (HR) in published genome-wide association studies (GWAS) available for BP and HR. Variants were filtered out by Linkage Disequilibrium clumping (LD R 2 > 0.1) and based on their minor allele frequency (MAF < 0.01). The effects of selected variants on the genetic risk of HF were extracted from a GWAS of HF from the HERMES consortium, based on a non-overlapping sample population. MR estimates were obtained using the Wald estimator for a single variant or the inverse variance weighted method for multiple variants. Results: 542,362 controls and 40,805 HF cases were evaluated. Independent variants in genes encoding 4 SNS receptors associated with BP or HR were identified as follows: α1A (diastolic BP), α2B (diastolic BP and HR), ß1 and ß2 (diastolic and systolic BP). MR analysis of α1A and ß1, weighted by their effects on diastolic BP, estimated an association with a higher risk of HF, while α2B variants were associated with a lower risk. We found no evidence for an effect of ß2. A similar relationship with systolic BP was found for ß1 and ß2. HR increasing effect of α2B variants was associated with a decreased odd of HF. Conclusions: Mindful of pleiotropic effects, these findings are consistent with the known benefits of ß1 blockade in HF and support a similar role for α1A blockade; conversely, they suggest a detrimental lowering effect of BP and HR through α2B modulation that deserves further studies. No evidence for a role of ß2 in HF was found.

Download Full-text

The Parkinson’s Disease Mendelian Randomization Research Portal

10.1101/604033 ◽

2019 ◽

Cited By ~ 2

Author(s):

Alastair J Noyce ◽

Sara Bandres-Ciga ◽

Jonggeol Kim ◽

Karl Heilbron ◽

Demis Kia ◽

...

Keyword(s):

Parkinson’S Disease ◽

Parkinson's Disease ◽

Mendelian Randomization ◽

Causal Effect ◽

Association Studies ◽

Sensitivity Analyses ◽

P Value ◽

Phenotypic Traits ◽

Genome Wide Association Studies ◽

Summary Statistics

ABSTRACTBackgroundMendelian randomization (MR) is a method for exploring observational associations to find evidence of causality.ObjectiveTo apply MR between multiple risk factors/phenotypic traits (exposures) and Parkinson’s disease (PD) in a large, unbiased manner, and to create a public resource for research.MethodsWe used two-sample MR in which the summary statistics relating to SNPs from genome wide association studies (GWASes) of 5,839 exposures curated on MR Base were used to assess causal relationships with PD. We selected the highest quality exposure GWASes for this report (n=401). For the disease outcome, summary statistics from the largest published PD GWAS were used. For each exposure, the causal effect on PD was assessed using the inverse variance weighted (IVW) method, followed by a range of sensitivity analyses. We used a false discovery rate (FDR) corrected p-value of <0.05 from the IVW analysis to prioritize traits of interest.ResultsWe observed evidence for causal associations between twelve exposures and risk of PD. Of these, nine were causal effects related to increasing adiposity and decreasing risk of PD. The remaining top exposures that affected PD risk were tea drinking, time spent watching television and forced vital capacity, but the latter two appeared to be biased by violations of underlying MR assumptions.DiscussionWe present a new platform which offers MR analyses for a total of 5,839 GWASes versus the largest PD GWASes available (https://pdgenetics.shinyapps.io/pdgenetics/). Alongside, we report further evidence to support a causal role for adiposity on lowering the risk of PD.

Download Full-text

Genetically Higher Level of Mannose Has No Impact on Cardiometabolic Risk Factors: Insight from Mendelian Randomization

Nutrients ◽

10.3390/nu13082563 ◽

2021 ◽

Vol 13 (8) ◽

pp. 2563

Author(s):

Mohsen Mazidi ◽

Abbas Dehghan ◽

Maciej Banach ◽

Keyword(s):

Risk Factors ◽

Cardiometabolic Risk ◽

Mendelian Randomization ◽

Association Studies ◽

Fasting Blood Glucose ◽

Density Lipoprotein ◽

Cardiometabolic Risk Factors ◽

Significant Heterogeneity ◽

Genome Wide Association Studies ◽

Leave One Out

Background: There is a handful of controversial data from observational studies on the serum levels of mannose and risks of coronary artery disease (CAD) and other cardiometabolic risk factors. We applied Mendelian Randomization (MR) analysis to obtain estimates of the causal effect of serum mannose on the risk of CAD and on cardiometabolic risk factors. Methods: Two-sample MR was implemented by using summary-level data from the largest genome-wide association studies (GWAS) conducted on serum mannose and CAD and cardiometabolic risk factors. The inverse variance weighted method (IVW) was used to estimate the effects, and a sensitivity analysis including the weighted median (WM)-based method, MR-Egger, MR-Pleiotropy RESidual Sum and Outlier (PRESSO) were applied. Radial MR Methods was applied to remove outliers subject to pleiotropic bias. We further conducted a leave-one-out analysis. Results: Mannose had no significant effect on CAD (IVW: odds ratio: 0.96 (95% Confidence Interval (95%CI): 0.71−1.30)), total cholesterol (TC) (IVW: 95%CI: 0.60−1.08), low density lipoprotein (LDL) (IVW: 95%CI = 0.68−1.15), high density lipoprotein (HDL) (IVW: 95%CI = 0.85−1.20), triglycerides (TG) (IVW: 95%CI = 0.38−1.08), waist circumference (WC) (IVW: 95%CI = 0.94−1.37), body mass index (BMI) (IVW: 95%CI = 0.93−1.29) and fasting blood glucose (FBG) (IVW: 95%CI = 0.92−1.33), with no heterogeneity for CAD, HDL, WC and BMI (all p > 0.092), while a significant heterogeneity was observed for TC (IVW: Q = 44.503), LDL (IVW: Q = 33.450), TG (IVW: Q = 159.645) and FBG (IVW: Q = 0. 32.132). An analysis of MR-PRESSO and radial plots did not highlight any outliers. The results of the leave-one-out method demonstrated that the links were not driven by a single instrument. Conclusions: We did not find any effect of mannose on adiposity, glucose, TC, LDL, TG and CAD.

Download Full-text

A comprehensive evaluation of methods for Mendelian randomization using realistic simulations and an analysis of 38 biomarkers for risk of type 2 diabetes

International Journal of Epidemiology ◽

10.1093/ije/dyaa262 ◽

2021 ◽

Author(s):

Guanghao Qi ◽

Nilanjan Chatterjee

Keyword(s):

Type 2 Diabetes ◽

Mendelian Randomization ◽

Association Studies ◽

Real Data ◽

Causal Effects ◽

Type I ◽

Genome Wide Association Studies ◽

Simulation Studies ◽

Sample Sizes

Abstract Background Previous studies have often evaluated methods for Mendelian randomization (MR) analysis based on simulations that do not adequately reflect the data-generating mechanisms in genome-wide association studies (GWAS) and there are often discrepancies in the performance of MR methods in simulations and real data sets. Methods We use a simulation framework that generates data on full GWAS for two traits under a realistic model for effect-size distribution coherent with the heritability, co-heritability and polygenicity typically observed for complex traits. We further use recent data generated from GWAS of 38 biomarkers in the UK Biobank and performed down sampling to investigate trends in estimates of causal effects of these biomarkers on the risk of type 2 diabetes (T2D). Results Simulation studies show that weighted mode and MRMix are the only two methods that maintain the correct type I error rate in a diverse set of scenarios. Between the two methods, MRMix tends to be more powerful for larger GWAS whereas the opposite is true for smaller sample sizes. Among the other methods, random-effect IVW (inverse-variance weighted method), MR-Robust and MR-RAPS (robust adjust profile score) tend to perform best in maintaining a low mean-squared error when the InSIDE assumption is satisfied, but can produce large bias when InSIDE is violated. In real-data analysis, some biomarkers showed major heterogeneity in estimates of their causal effects on the risk of T2D across the different methods and estimates from many methods trended in one direction with increasing sample size with patterns similar to those observed in simulation studies. Conclusion The relative performance of different MR methods depends heavily on the sample sizes of the underlying GWAS, the proportion of valid instruments and the validity of the InSIDE assumption. Down-sampling analysis can be used in large GWAS for the possible detection of bias in the MR methods.

Download Full-text