Dataset Information

Caution against examining the role of reverse causality in Mendelian Randomization.

ABSTRACT: Recently, Mendelian Randomization (MR) has gained in popularity as a concept to assess the causal relationship between phenotypes in genetic association studies. An extension of standard MR methodology, the MR Steiger approach, has recently been developed to infer the causal direction between two phenotypes in prospective studies. Through simulation studies, we examined and quantified the ability of the MR Steiger approach to determine the causal direction between two phenotypes (i.e., effect direction). Through simulation studies, our results show that the MR Steiger approach may fail to correctly identify the direction of causality. This is true, especially in the presence of pleiotropy. We also applied the MR Steiger method to the COPDGene study, a case-control study of chronic obstructive pulmonary disease (COPD) in current and former smokers, to examine the role of smoking on lung function. We have created an R package on Github called reverseDirection which runs simulations for user-specified scenarios to examine when the MR Steiger approach can correctly determine the causal direction between two phenotypes in any user specified scenario. In summary, our results emphasize the importance of caution when the MR Steiger approach is used in to infer the direction of causality.

SUBMITTER: Lutz SM

PROVIDER: S-EPMC8222166 | biostudies-literature | 2021 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Caution against examining the role of reverse causality in Mendelian Randomization.

Lutz Sharon M SM Wu Ann Chen AC Hokanson John E JE Vansteelandt Stijn S Lange Christoph C

Genetic epidemiology 20210519 5

Recently, Mendelian Randomization (MR) has gained in popularity as a concept to assess the causal relationship between phenotypes in genetic association studies. An extension of standard MR methodology, the MR Steiger approach, has recently been developed to infer the causal direction between two phenotypes in prospective studies. Through simulation studies, we examined and quantified the ability of the MR Steiger approach to determine the causal direction between two phenotypes (i.e., effect di ...[more]

PMID: 34008876

Similar Datasets

Project description:Polymyositis is a prominent subgroup of idiopathic inflammatory myopathy, considered to have an autoimmune etiology. However, research exploring the condition between immunocytes and polymyositis remains limited, indicating the need for further investigation to unravel these intricate associations. We employed bidirectional Mendelian randomization (MR) analysis to ascertain causality between 731 immunocytes and polymyositis. We also compared the positive immunocytes with dermatomyositis. Our primary analytical method was inverse variance weighted, supplemented by 4 other MR techniques. Additionally, Cochran Q test was performed to assess heterogeneity, MR-Egger to appraise pleiotropy, and MR-PRESSO to identify and eliminate potential outliers. Furthermore, the leave-one-out test evaluated the impact of each instrumental variable (IV) on the causal effect. The inverse variance weighted results revealed that 10 immunocytes exert a protective effect against polymyositis (P < .05, OR < 1), while 16 immunocytes are connected with an elevated risk of the disease (P < .05, OR > 1). In reverse MR, polymyositis was found to decrease the levels of 2 immune cells (P < .05, OR < 1) and elevate the expression of 5 immune cell phenotypes (P < .05, OR > 1). A complex correlation was found between polymyositis and the immunocyte phenotypes CD8, CD33dim, HLA-DR, CD11b, and CD45. Additionally, it was discovered that 15 types of immune cells share a causal relationship between polymyositis and dermatomyositis. All analyses demonstrated no heterogeneity or horizontal pleiotropy (P > .05). Our study provides compelling evidence regarding the intricate causal relationships between immunocytes and polymyositis. Polymyositis and dermatomyositis share common immunocytes' regulatory mechanisms. CD8, CD33dim, HLA-DR, CD11b, and CD45 may represent potential immune cell markers for polymyositis. These findings hold implications for planning prognosis and therapeutic strategies for polymyositis, offering novel insights for drug development.

Project description:Background & aimsThe pathogenesis of hypertension involves a diverse range of genetic, environmental, hemodynamic, and more causative factors. Recent evidence points to an association between the gut microbiome and hypertension. Given that the microbiota is in part determined by host genetics, we used the two-sample Mendelian randomization (MR) analysis to address the bidirectional causal link between gut microbiota and hypertension.MethodsWe selected genetic variants (P < 1 × 10-5) for gut microbiota (n = 18,340) from the MiBioGen study. Genetic association estimates for hypertension were extracted from genome-wide association study (GWAS) summary statistics on 54,358 cases and 408,652 controls. Seven complementary MR methods were implemented, including the inverse-variance weighted (IVW) method, followed by sensitivity analyses to verify the robustness of the results. Reverse-direction MR analyses were further conducted to probe if there was a reverse causative relationship. Bidirectional MR analysis then examines a modulation of gut microbiota composition by hypertension.ResultsAt the genus level, our MR estimates from gut microbiome to hypertension showed that there were 5 protective factors Allisonella, Parabacteroide, Phascolarctobacterium, Senegalimassilia, and unknowngenus (id.1000000073), while 6 genera Clostridiuminnocuum, Eubacteriumcoprostanoligenes, Eubacteriumfissicatena, Anaerostipes, LachnospiraceaeFCS020, and unknowngenus (id.2041) are risk factors. The Alcaligenaceae and ClostridialesvadinBB60 were detrimental and beneficial at the family level, respectively. In contrast, the MR results of hypertension-gut flora showed hypertensive states can lead to an increased abundance of Eubacteriumxylanophilum, Eisenbergiella, and Lachnospiraceae and a lower abundance of Alistipes, Bilophila, Butyricimonas, and Phascolarctobacterium.ConclusionAltered gut microbiota is a causal factor in the development of hypertension, and hypertension causes imbalances in the intestinal flora. Substantial research is still needed to find the key gut flora and explore the specific mechanisms of their effects so that new biomarkers can be found for blood pressure control.

Project description:Background and purpose Observational studies have shown that sarcopenia and diabetic nephropathy (DN), are closely related; however, the causal relationship is unclear. This study aims to address this issue using a bidirectional Mendelian randomization (MR) study. Methodology We data from genome-wide association studies including appendicular lean mass (n = 244,730), grip strength (right: n = 461,089, left: n = 461026), walking speed (n = 459,915), and DN (3283 cases and 181,704 controls) to conduct a bidirectional MR study. First, we conducted a Forward MR analysis to evaluate the causality of sarcopenia on the risk of DN from the genetic perspective with appendicular lean mass, grip strength, and walking speed as exposure and DN as the outcome. Then, DN as the exposure, we performed a Reverse MR analysis to determine whether DN impacted the appendicular lean mass, grip strength, and walking speed of the appendices. Finally, a series of sensitivity studies, such as heterogeneity tests, pleiotropy evaluations, and Leave-one-out analyses, were conducted to assess the MR analysis’s accuracy further. Results According to a forward MR analysis, a genetically predicted decrease in appendicular lean mass is associated with an increased risk of developing DN risk (inverse variance weighting[IVW]: odd ratio [OR] = 0.863, 95% confidence interval [CI] 0.767-0.971; P = 0.014). According to reverse MR results, grip strength decreased as DN progressed (IVW: right β = 0.003, 95% CI: - 0.021 to - 0.009, P = 5.116e-06; left β = 0.003, 95% CI: - 0.024 to - 0.012, P = 7.035e-09). However, the results of the other MR analyses were not statistically different. Conclusion Notably, our findings suggest that the causal relationship between sarcopenia and DN cannot be generalized. According to analysis of the individual characteristic factors of sarcopenia, reducing in appendicular lean mass increases the risk of developing DN and DN is linked to reduced grip strength. But overall, there is no causal relationship between sarcopenia and DN, because the diagnosis of sarcopenia cannot be determined by one of these factors alone.

Dataset Information

Caution against examining the role of reverse causality in Mendelian Randomization.

Publications

Caution against examining the role of reverse causality in Mendelian Randomization.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets