Robust multivariable Mendelian randomization based on constrained maximum likelihood.

Mendelian randomization (MR) is a powerful tool for causal inference with observational genome-wide association study (GWAS) summary data. Compared to the more commonly used univariable MR (UVMR), multivariable MR (MVMR) not only is more robust to the notorious problem of genetic (horizontal) pleiotropy but also estimates the direct effect of each exposure on the outcome after accounting for possible mediating effects of other exposures. Despite promising applications, there is a lack of studies on MVMR's theoretical properties and robustness in applications. In this work, we propose an efficient and robust MVMR method based on constrained maximum likelihood (cML), called MVMR-cML, with strong theoretical support. Extensive simulations demonstrate that MVMR-cML performs better than other existing MVMR methods while possessing the above two advantages over its univariable counterpart. An application to several large-scale GWAS summary datasets to infer causal relationships between eight cardiometabolic risk factors and coronary artery disease (CAD) highlights the usefulness and some advantages of the proposed method. For example, after accounting for possible pleiotropic and mediating effects, triglyceride (TG), low-density lipoprotein cholesterol (LDL), and systolic blood pressure (SBP) had direct effects on CAD; in contrast, the effects of high-density lipoprotein cholesterol (HDL), diastolic blood pressure (DBP), and body height diminished after accounting for other risk factors.

Associations of hearing loss and structural changes in specific cortical regions: a Mendelian randomization study.

INTRODUCTION: Previous studies have suggested a correlation between hearing loss (HL) and cortical alterations, but the specific brain regions that may be affected are unknown. METHODS: Genome-wide association study (GWAS) data for 3 subtypes of HL phenotypes, sensorineural hearing loss (SNHL), conductive hearing loss, and mixed hearing loss, were selected as exposures, and GWAS data for brain structure-related traits were selected as outcomes. The inverse variance weighted method was used as the main estimation method. RESULTS: Negative associations were identified between genetically predicted SNHL and brain morphometric indicators (cortical surface area, cortical thickness, or volume of subcortical structures) in specific brain regions, including the bankssts (ß = -0.006 mm, P = 0.016), entorhinal cortex (ß = -4.856 mm2, P = 0.029), and hippocampus (ß = -24.819 cm3, P = 0.045), as well as in brain regions functionally associated with visual perception, including the pericalcarine (ß = -10.009 cm3, P = 0.013). CONCLUSION: Adaptive changes and functional remodeling of brain structures occur in patients with genetically predicted HL. Brain regions functionally associated with auditory perception, visual perception, and memory function are the main brain regions vulnerable in HL.

The causal relationship between physical activity, sedentary behavior and brain cortical structure: a Mendelian randomization study.

Physical activity and sedentary behavior, both distinct lifestyle behaviors associated with brain health, have an unclear potential relationship with brain cortical structure. This study aimed to determine the causal link between physical activity, sedentary behavior, and brain cortical structure (cortical surface area and thickness) through Mendelian randomization analysis. The inverse-variance weighted method was primarily utilized, accompanied by sensitivity analyses, to confirm the results' robustness and accuracy. Analysis revealed nominally significant findings, indicating a potential positive influence of physical activity on cortical thickness in the bankssts (ß = 0.002 mm, P = 0.043) and the fusiform (ß = 0.002 mm, P = 0.018), and a potential negative association of sedentary behavior with cortical surface area in the caudal middle frontal (ß = -34.181 mm2, P = 0.038) and the pars opercularis (ß = -33.069 mm2, P = 0.002), alongside a nominally positive correlation with the cortical surface area of the inferior parietal (ß = 58.332 mm2, P = 0.035). Additionally, a nominally significant negative correlation was observed between sedentary behavior and cortical thickness in the paracentral (ß = -0.014 mm, P = 0.042). These findings offer insights into how lifestyle behaviors may influence brain cortical structures, advancing our understanding of their interaction with brain health.

Causal effects of osteoporosis on structural changes in specific brain regions: a Mendelian randomization study.

Observational studies have reported that osteoporosis is associated with cortical changes in the brain. However, the inherent limitations of observational studies pose challenges in eliminating confounding factors and establishing causal relationships. And previous observational studies have not reported changes in specific brain regions. By employing Mendelian randomization, we have been able to infer a causal relationship between osteoporosis and a reduction in the surficial area (SA) of the brain cortical. This effect is partially mediated by vascular calcification. We found that osteoporosis significantly decreased the SA of global brain cortical (ß = -1587.62 mm2, 95%CI: -2645.94 mm2 to -529.32 mm2, P = 0.003) as well as the paracentral gyrus without global weighted (ß = - 19.42 mm2, 95%CI: -28.90 mm2 to -9.95 mm2, P = 5.85 × 10-5). Furthermore, we estimated that 42.25% and 47.21% of the aforementioned effects are mediated through vascular calcification, respectively. Osteoporosis leads to a reduction in the SA of the brain cortical, suggesting the presence of the bone-brain axis. Vascular calcification plays a role in mediating this process to a certain extent. These findings establish a theoretical foundation for further investigations into the intricate interplay between bone, blood vessels, and the brain.

Investigation of causal relationships between cortical structure and osteoporosis using two-sample Mendelian randomization.

The mutual interaction between bone characteristics and brain had been reported previously, yet whether the cortical structure has any relevance to osteoporosis is questionable. Therefore, we applied a two-sample bidirectional Mendelian randomization analysis to investigate this relationship. We utilized the bone mineral density measurements of femoral neck (n = 32,735) and lumbar spine (n = 28,498) and data on osteoporosis (7300 cases and 358,014 controls). The global surficial area and thickness and 34 specific functional regions of 51,665 patients were screened by magnetic resonance imaging. For the primary estimate, we utilized the inverse-variance weighted method. The Mendelian randomization-Egger intercept test, MR-PRESSO, Cochran's Q test, and "leave-one-out" sensitivity analysis were conducted to assess heterogeneity and pleiotropy. We observed suggestive associations between decreased thickness in the precentral region (OR = 0.034, P = 0.003) and increased chance of having osteoporosis. The results also revealed suggestive causality of decreased bone mineral density in femoral neck to declined total cortical surface area (ß = 1400.230 mm2, P = 0.003), as well as the vulnerability to osteoporosis and reduced thickness in the Parstriangularis region (ß = -0.006 mm, P = 0.002). Our study supports that the brain and skeleton exhibit bidirectional crosstalk, indicating the presence of a mutual brain-bone interaction.

Causal effect of severe and non-severe malaria on dyslipidemia in African Ancestry individuals: A Mendelian randomization study.

BACKGROUND: Dyslipidemia is becoming prevalent in Africa, where malaria is endemic. Observational studies have documented the long-term protective effect of malaria on dyslipidemia; however, these study designs are prone to confounding. Therefore, we used Mendelian randomization (MR, a method robust to confounders and reverse causation) to determine the causal effect of severe malaria (SM) and the recurrence of non-severe malaria (RNM) on lipid traits. METHOD: We performed two-sample MR using genome wide association study (GWAS) summary statistics for recurrent non-severe malaria (RNM) from a Benin cohort (N = 775) and severe malaria from the MalariaGEN dataset (N = 17,000) and lipid traits from summary-level data of a meta-analyzed African lipid GWAS (MALG, N = 24,215) from the African Partnership for Chronic Disease Research (APCDR) (N = 13,612) and the Africa Wits-IN-DEPTH partnership for genomics studies (AWI-Gen) dataset (N = 10,603). RESULT: No evidence of significant causal association was obtained between RNM and high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol and triglycerides. However, a notable association emerged between severe malarial anaemia (SMA) which is a subtype of severe malaria and reduced HDL-C levels, suggesting a potential subtype-specific effect. Nonetheless, we strongly believe that the small sample size likely affects our estimates, warranting cautious interpretation of these results. CONCLUSION: Our findings challenge the hypothesis of a broad causal relationship between malaria (both severe and recurrent non-severe forms) and dyslipidemia. The isolated association with SMA highlights an intriguing area for future research. However, we believe that conducting larger studies to investigate the connection between malaria and dyslipidemia in Africa will enhance our ability to better address the burden posed by both diseases.

Immune-mediated inflammatory diseases and periodontal disease: a bidirectional two-sample mendelian randomization study.

BACKGROUND: Previous observational studies have shown a bidirectional association between immune-mediated inflammatory disorders (IMID) and periodontal disease. However, evidence regarding the causal role of IMID and periodontal disease is still lacking. Therefore, we conducted a bidirectional two-sample Mendelian randomization (MR) study to uncover the potential genetic causal effects between IMID and periodontal disease. METHODS: Bidirectional two-sample MR analysis was employed. Data for ten IMIDs were sourced from genome-wide association studies (GWAS) conducted by the FinnGen Consortium (range from 1023 to 36321 cases) and UK Biobank (UKB) (range from 150 to 17574 cases). Furthermore, GWAS data for periodontal disease were obtained from the FinnGen Consortium (87497 cases), UKB (458 cases), and Gene Lifestyle Interactions in Dental Endpoints (GLIDE) consortium (17,353 periodontitis cases). Subsequently, the causal relationships were analyzed by random effects inverse variance weighting, weighted median, and MR-Egger. Sensitivity analyses were performed using the Cochrane Q test, funnel plot, and Mr-Egger intercept test to ensure robustness. Eventually, replication analysis and meta-analysis across different databases were carried out. RESULTS: Systemic lupus erythematosus (SLE) [IVW: OR = 1.079 (95% CI: 1.032-1.128) and P < 0.001], Sjogren syndrome [IVW: OR = 1.082 (95% CI: 1.012-1.157) and P = 0.022] and hypothyroidism [IVW: OR = 1.52 (95% CI: 1.13-2.04) and P = 0.005] may increase the risk of periodontal disease. In addition, periodontal disease may reduce the risk of SLE [IVW: OR = 0.8079 (95% CI: 0.6764-0.9650) and P = 0.019] and hyperthyroidism [IVW: OR = 5.59*10-9 (95% CI: 1.43*10-15-2.18*10-2) and P = 0.014]. Meta-analysis indicated a causal correlation between SLE and an increased risk of periodontal disease: [OR = 1.08 (95% CI: 1.03-1.13), P = 0.0009]. No significant evidence suggests bilateral causal relationships between other IMIDs and periodontal disease. No significant estimation of heterogeneity or pleiotropy is detected. CONCLUSIONS: Our study has confirmed a genetic causal relationship between IMIDs and periodontal disease, thereby unveiling novel potential mechanisms underlying IMIDs and periodontal disease. This discovery is promising in fostering interdisciplinary collaboration between clinicians and stomatologists to facilitate appropriate and precise screening, prevention, and early treatment of IMIDs and periodontal disease.

Genetic evidence for the causal association between type 1 diabetes and the risk of polycystic ovary syndrome.

BACKGROUND: Accumulating observational studies have identified associations between type 1 diabetes (T1D) and polycystic ovary syndrome (PCOS). Still, the evidence about the causal effect of this association is uncertain. METHODS: We performed a two-sample Mendelian randomization (MR) analysis to test for the causal association between T1D and PCOS using data from a large-scale biopsy-confirmed genome-wide association study (GWAS) in European ancestries. We innovatively divided T1D into nine subgroups to be analyzed separately, including: type1 diabetes wide definition, type1 diabetes early onset, type 1 diabetes with coma, type 1 diabetes with ketoacidosis, type 1 diabetes with neurological complications, type 1 diabetes with ophthalmic complications, type 1 diabetes with peripheral circulatory complications, type 1 diabetes with renal complications, and type 1 diabetes with other specified/multiple/unspecified complications. GWAS data for PCOS were obtained from a large-scale GWAS (10,074 cases and 103,164 controls) for primary analysis and the IEU consortium for replication and meta-analysis. Sensitivity analyses were conducted to evaluate heterogeneity and pleiotropy. RESULTS: Following rigorous instrument selection steps, the number of SNPs finally used for T1D nine subgroups varying from 6 to 36 was retained in MR estimation. However, we did not observe evidence of causal association between type 1 diabetes nine subgroups and PCOS using the IVW analysis, MR-Egger regression, and weighted median approaches, and all P values were > 0.05 with ORs near 1. Subsequent replicates and meta-analyses also yielded consistent results. A number of sensitivity analyses also did not reveal heterogeneity and pleiotropy, including Cochran's Q test, MR-Egger intercept test, MR-PRESSO global test, leave-one-out analysis, and funnel plot analysis. CONCLUSION: This is the first MR study to investigate the causal relationship between type 1 diabetes and PCOS. Our findings failed to find substantial causal effect of type 1 diabetes on risk of PCOS. Further randomized controlled studies and MR studies are necessary.

The societal impact of individual placement and support implementation on employment outcomes for young adults receiving temporary health-related welfare benefits: a difference-in-differences study.

BACKGROUND: Individual placement and support (IPS) is an evidence-based practice that helps individuals with mental illness gain and retain employment. IPS was implemented for young adults at a municipality level through a cross-sectoral collaboration between specialist mental healthcare, primary mental healthcare, and the government funded employment service (NAV). We investigated whether IPS implementation had a causal effect on employment outcomes for all young adults in receipt of a temporary health-related rehabilitation (work assessment allowance, WAA) welfare benefit, measured at the societal level compared to municipalities that did not implement IPS. METHOD: We used a difference in differences design to estimate the effects of IPS implementation on the outcome of workdays per year using longitudinal registry data. We estimate the average effect of being exposed to IPS implementation during four-years of implementation compared to ten control municipalities without IPS for all WAA recipients. RESULTS: We found a significant, positive, causal effect on societal level employment outcomes of 5.6 (p = 0.001, 95% CI 2.7-8.4) increased workdays per year per individual, equivalent to 12.7 years of increased work in the municipality where IPS was implemented compared to municipalities without IPS. Three years after initial exposure to IPS implementation individuals worked, on average, 10.5 more days per year equating to 23.8 years of increased work. CONCLUSIONS: Implementing IPS as a cross sectoral collaboration at a municipality level has a significant, positive, causal, societal impact on employment outcomes for all young adults in receipt of a temporary health-related rehabilitation welfare benefit.

Causal relationship between female reproductive factors, sex hormones and uterine leiomyoma: a Mendelian randomization study.

RESEARCH QUESTION: Are the observed associations between female reproductive factors and sex hormones with the risk of uterine leiomyoma truly causal associations? DESIGN: The putative causal relationships between female reproductive factors and sex hormones with uterine leiomyoma were investigated using two-sample Mendelian randomization. Statistics on exposure-associated genetic variants were obtained from genome-wide association studies (GWAS). The uterine leiomyoma GWAS from the FinnGen and FibroGENE consortia were used as outcome data for discovery and replication analyses, respectively. Results were pooled by meta-analysis. Sensitivity analyses ensured robustness of the Mendelian randomization analysis. RESULTS: When FinnGen GWAS were used as outcome data, a causal relationship was found between age at menarche (OR 0.84, P < 0.0001), age at menopause (OR 1.08, P < 0.0001), number of live births (OR 0.25, P < 0.001) and total testosterone levels (OR 0.90, P < 0.001) with the risk of uterine leiomyoma. When FibroGENE GWAS were used as outcome data, Mendelian randomization results for age at menopause, the number of live births and total testosterone levels were replicated. In the meta-analysis, a later age at menopause (OR 1.08, P < 0.0001) was associated with an increased risk of uterine leiomyoma. A higher number of live births (OR 0.25, P < 0.0001) and higher total testosterone levels (OR 0.90, P < 0.0001) were associated with a decreased risk of uterine leiomyoma. CONCLUSIONS: A causal relationship between later age at menopause, lower number of live births and lower total testosterone levels with increased risk of uterine leiomyoma was found.

Doubly robust estimation and sensitivity analysis for marginal structural quantile models.

The marginal structure quantile model (MSQM) provides a unique lens to understand the causal effect of a time-varying treatment on the full distribution of potential outcomes. Under the semiparametric framework, we derive the efficiency influence function for the MSQM, from which a new doubly robust estimator is proposed for point estimation and inference. We show that the doubly robust estimator is consistent if either of the models associated with treatment assignment or the potential outcome distributions is correctly specified, and is semiparametric efficient if both models are correct. To implement the doubly robust MSQM estimator, we propose to solve a smoothed estimating equation to facilitate efficient computation of the point and variance estimates. In addition, we develop a confounding function approach to investigate the sensitivity of several MSQM estimators when the sequential ignorability assumption is violated. Extensive simulations are conducted to examine the finite-sample performance characteristics of the proposed methods. We apply the proposed methods to the Yale New Haven Health System Electronic Health Record data to study the effect of antihypertensive medications to patients with severe hypertension and assess the robustness of the findings to unmeasured baseline and time-varying confounding.

Exploring causal effects of gut microbiota and metabolites on body fat percentage using two-sample Mendelian randomization.

AIM: The relationship between the gut microbiota, metabolites and body fat percentage (BFP) remains unexplored. We systematically assessed the causal relationships between gut microbiota, metabolites and BFP using Mendelian randomization analysis. MATERIALS AND METHODS: Single nucleotide polymorphisms associated with gut microbiota, blood metabolites and BFP were screened via a genome-wide association study enrolling individuals of European descent. Summary data from genome-wide association studies were extracted from the MiBioGen consortium and the UK Biobank. The inverse variance-weighted model was the primary method used to estimate these causal relationships. Sensitivity analyses were performed using pleiotropy, Mendelian randomization-Egger regression, heterogeneity tests and leave-one-out tests. RESULTS: In the aspect of phyla, classes, orders, families and genera, we observed that o_Bifidobacteriales [ß = -0.05; 95% confidence interval (CI): -0.07 to -0.03; false discovery rate (FDR) = 2.76 × 10-3], f_Bifidobacteriaceae (ß = -0.05; 95% CI: -0.07 to -0.07; FDR = 2.76 × 10-3), p_Actinobacteria (ß = -0.06; 95% CI: -0.09 to -0.03; FDR = 6.36 × 10-3), c_Actinobacteria (ß = -0.05; 95% CI: -0.08 to -0.02; FDR = 1.06 × 10-2), g_Bifidobacterium (ß = -0.05; 95% CI: -0.07 to -0.02; FDR = 1.85 × 10-2), g_Ruminiclostridium9 (ß = -0.03; 95% CI: -0.06 to -0.01; FDR = 4.81 × 10-2) were negatively associated with BFP. G_Olsenella (ß = 0.02; 95% CI: 0.01-0.03; FDR = 2.16 × 10-2) was positively associated with BFP. Among the gut microbiotas, f_Bifidobacteriales, o_Bifidobacteriales, c_Actinobacteria and p_Actinobacteria were shown to be significantly associated with BFP in the validated dataset. In the aspect of metabolites, we only observed that valine (ß = 0.77; 95% CI: 0.5-1.04; FDR = 8.65 × 10-6) was associated with BFP. CONCLUSIONS: Multiple gut microbiota and metabolites were strongly associated with an increased BFP. Further studies are required to elucidate the mechanisms underlying this putative causality. In addition, BFP, a key indicator of obesity, suggests that obesity-related interventions can be developed from gut microbiota and metabolite perspectives.

Sensitivity analysis for principal ignorability violation in estimating complier and noncomplier average causal effects.

An important strategy for identifying principal causal effects (popular estimands in settings with noncompliance) is to invoke the principal ignorability (PI) assumption. As PI is untestable, it is important to gauge how sensitive effect estimates are to its violation. We focus on this task for the common one-sided noncompliance setting where there are two principal strata, compliers and noncompliers. Under PI, compliers and noncompliers share the same outcome-mean-given-covariates function under the control condition. For sensitivity analysis, we allow this function to differ between compliers and noncompliers in several ways, indexed by an odds ratio, a generalized odds ratio, a mean ratio, or a standardized mean difference sensitivity parameter. We tailor sensitivity analysis techniques (with any sensitivity parameter choice) to several types of PI-based main analysis methods, including outcome regression, influence function (IF) based and weighting methods. We discuss range selection for the sensitivity parameter. We illustrate the sensitivity analyses with several outcome types from the JOBS II study. This application estimates nuisance functions parametrically - for simplicity and accessibility. In addition, we establish rate conditions on nonparametric nuisance estimation for IF-based estimators to be asymptotically normal - with a view to inform nonparametric inference.

Causal relationship between glycemic traits and bone mineral density in different age groups and skeletal sites: a Mendelian randomization analysis.

INTRODUCTION: Previous research has confirmed that patients with type 2 diabetes mellitus tend to have higher bone mineral density (BMD), but it is unknown whether this pattern holds true for individuals without diabetes. This Mendelian randomization (MR) study aims to investigate the potential causal relationship between various glycemic trait (including fasting glucose, fasting insulin, 2-h postprandial glucose, and glycated hemoglobin) and BMD in non-diabetic individuals. The investigation focuses on different age groups (15-30, 30-45, 45-60, and 60 + years) and various skeletal sites (forearm, lumbar spine, and hip). MATERIALS AND METHODS: We utilized genome-wide association study data from large population-based cohorts to identify robust instrumental variables for each glycemic traits parameter. Our primary analysis employed the inverse-variance weighted method, with sensitivity analyses conducted using MR-Egger, weighted median, MR-PRESSO, and multivariable MR methods to assess the robustness and potential horizontal pleiotropy of the study results. RESULTS: Fasting insulin showed a negative modulating relationship on both lumbar spine and forearm. However, these associations were only nominally significant. No significant causal association was observed between blood glucose traits and BMD across the different age groups. The direction of fasting insulin's causal effects on BMD showed inconsistency between genders, with potentially decreased BMD in women with high fasting insulin levels and an increasing trend in BMD in men. CONCLUSIONS: In the non-diabetic population, currently available evidence does not support a causal relationship between glycemic traits and BMD. However, further investigation is warranted considering the observed gender differences.

Causal effects of dietary habits on COVID-19 susceptibility, hospitalisation and severity: a comprehensive Mendelian randomisation study.

This study aimed to investigate the causal effect of dietary habits on COVID-19 susceptibility, hospitalisation and severity. We used data from a large-scale diet dataset and the COVID-19 Host Genetics Initiative to estimate causal relationships using Mendelian randomisation. The inverse variance weighted (IVW) method was used as the main analysis. For COVID-19 susceptibility, IVW estimates indicated that milk (OR: 0·82; 95 % CI (0·68, 0·98); P = 0·032), unsalted peanut (OR: 0·53; 95 % CI (0·35, 0·82); P = 0·004), beef (OR: 0·59; 95 % CI (0·41, 0·84); P = 0·004), pork (OR: 0·63; 95 % CI (0·42, 0·93); P = 0·022) and processed meat (OR: 0·76; 95 % CI (0·63, 0·92); P = 0·005) were causally associated with reduced COVID-19 susceptibility, while coffee (OR: 1·23; 95 % CI (1·04, 1·45); P = 0·017) and tea (OR: 1·17; 95 % CI (1·05, 1·31); P = 0·006) were causally associated with increased risk. For COVID-19 hospitalisation, beef (OR: 0·51; 95 % CI (0·26, 0·98); P = 0·042) showed negative correlations, while tea (OR: 1·54; 95 % CI (1·16, 2·04); P = 0·003), dried fruit (OR: 2·08; 95 % CI (1·37, 3·15); P = 0·001) and red wine (OR: 2·35; 95 % CI (1·29, 4·27); P = 0·005) showed positive correlations. For COVID-19 severity, coffee (OR: 2·16; 95 % CI (1·25, 3·76); P = 0·006), dried fruit (OR: 1·98; 95 % CI (1·16, 3·37); P = 0·012) and red wine (OR: 2·84; 95 % CI (1·21, 6·68); P = 0·017) showed an increased risk. These findings were confirmed to be robust through sensitivity analyses. Our findings established a causal relationship between dietary habits and COVID-19 susceptibility, hospitalisation and severity.

Lack of causal association between epilepsy and dementia: A Mendelian randomization analysis.

OBJECTIVE: Epidemiological studies have reported an association between epilepsy and dementia. However, the causal relationship between epilepsy and the risk of dementia is not clear. We aimed to inspect the causal effect of epilepsy on memory loss and dementia. METHODS: We analyzed summary data of epilepsy, memory loss, and dementia from the genome-wide association study (GWAS) using the two-sample Mendelian randomization (MR) method. We used the estimated odds ratio of memory loss and dementia associated with each of the genetically defined traits to infer evidence for a causal relationship with the following exposures: all epilepsy, focal epilepsy (including focal epilepsy with hippocampal sclerosis, lesion-negative focal epilepsy, and focal epilepsy with other lesions), and genetic generalized epilepsy (including childhood absence epilepsy, generalized tonic-clonic seizures alone, Juvenile absence epilepsy, and Juvenile myoclonic epilepsy). RESULTS: According to the result of MR using the inverse variance weighted method (IVW), we found that genetically predicted epilepsy did not causally increase the risk of memory loss and dementia (p > 0.05). Results of the MR-Egger and weighted median method were consistent with the IVW method. CONCLUSIONS: No evidence has been found to support the notion that epilepsy can result in memory loss and dementia. The associations observed in epidemiological studies could be attributed, in part, to confounding or nongenetic determinants.

Effect of cytomegalovirus infection on post-transplant hospitalization days among children undergoing allogeneic hematopoietic cell transplantation: A marginal structural model approach.

BACKGROUND: Cytomegalovirus (CMV) commonly reactivates after allogeneic hematopoietic cell transplant (HCT), potentially leading to CMV disease and significant morbidity and mortality. To reduce morbidity and mortality, many centers conduct weekly CMV blood polymerase chain reaction (PCR) surveillance testing with subsequent initiation of antiviral therapy upon CMV DNAemia detection. However, the impact of CMV DNAemia on subsequent hospitalization risk has not been assessed using models accounting for the time-varying nature of the exposure, outcome, and confounders. METHODS: All allogeneic HCTs at the Children's Hospital of Philadelphia from January 2004-April 2017 were considered for inclusion. Patients were monitored with CMV surveillance via PCR testing for up to 105 days after HCT receipt. We estimated the association between CMV DNAemia and rate of hospitalization using marginal structural models (MSM). RESULTS: There were 343 allogeneic HCT episodes in 330 with CMV surveillance; median age was 9.0 (range: 0.1-26.2) and 46.5% were female. And 24.1% of HCT patients had at least one positive CMV blood PCR during the follow-up period. Median time to CMV DNAemia detection was 19 days (range: 4-97). The MSM estimated the incidence rate ratios for an association of CMV DNAemia with hospitalization to be 1.24, (95% confidence interval: 1.04-1.47). CONCLUSIONS: CMV DNAemia was associated with an increased hospitalization in the post-HCT period. The MSM accounted for time-varying nature of the outcome, exposure and confounders. The findings support prevention of CMV DNAemia in this population. We recommend further investigation into the effectiveness and safety of prophylaxis versus pre-emptive CMV prevention approaches.

Obesity causally influencing brain cortical structure: a Mendelian randomization study.

Obesity may lead to cognitive impairment and psychiatric disorders, which are associated with alterations in the brain cortical structure. However, the exact causality remains inconclusive. We aimed to conduct two-sample Mendelian randomization (MR) analysis to identify the causal associations of obesity [body mass index (BMI), waist-hip ratio (WHR), and waist-hip ratio adjusted for BMI ((WHRadjBMI)) and brain cortical structure (cortical thickness and cortical surface area). Inverse-variance weighted (IVW) method was used as the main analysis, whereas a series of sensitivity analyses were employed to assess heterogeneity and pleiotropy. The main MR results showed that higher BMI significantly increased the cortical surface area of the transverse temporal (ß = 5.13 mm2, 95% confidence interval [CI]: 2.55-7.71, P = 9.9 × 10-5); higher WHR significantly decreased cortical surface area of the inferior temporal (ß = -38.60, 95% CI: -56.67- -20.54, P = 1.2 × 10-5), but significantly increased cortical surface area of the isthmus cingulate (ß = 14.25, 95% CI: 6.97-21.54, P = 1.2 × 10-4). No significant evidence of pleiotropy was found in the MR analyses. This study supports that obesity has a causal effect on the brain cortical structure. Further studies are warranted to understand the clinical outcomes caused by these effects.

No evidence for a genetic causal effect of breast cancer on venous thromboembolism: a mendelian randomization study.

Active cancer is known to contribute to venous thromboembolism (VTE), but the cause-and-effect association of breast cancer on VTE is not yet clear. In order to investigate the possible causal relationships, we used a Mendelian randomization analysis. Data for generically predicted breast cancer were identified based on the BCAC consortium. A meta-analysis of genome-wide association study (GWAS) comprising 1,500,861 participants for VTE as well as data from the FinnGen study for VTE, DVT and PE was used for the causal-effect estimation. Our primary method was inverse-variance weighted (IVW), and our supplementary methods included weighted median and MR-Egger. We also carried out sensitivity analysis for the study. No evidence of causal-effect was detected of overall breast cancer on VTE in both the GWAS meta-analysis (OR=1.01, 95%CI:0.98-1.04, p = 0.495) and the FinnGen consortium (OR=1.00,95%CI:0.96-1.04, p = 0.945). In addition, the presence of ER-positive or ER-negative disease did not significantly influence the incidence of VTE and its subtypes. In conclusion, no genetic cause-and-effect of breast cancer on VTE risk was detected in the large MR analysis.

Causal effect of atrial fibrillation on pulmonary embolism: a mendelian randomization study.

Atrial fibrillation (AF) can increase thrombosis, especially arterial thrombosis, and some studies show that AF patients have a higher risk of developing pulmonary embolism (PE). The objective of our study is to investigate whether there is a direct causal effect of AF on PE. A two-sample Mendelian randomization (MR) approach was utilized to determine whether there is a causal relationship between AF and PE. European population-based consortia provided statistical data on the associations between Single Nucleotide Polymorphisms (SNPs) and relevant traits. The AF dataset was obtained from genome-wide association studies (GWAS) comprising 60,620 cases and 970,216 controls, while a GWAS of 1846 cases and 461,164 controls identified genetic variations associated with PE. Estimation of the causal effect was mainly performed using the random effects inverse-variance weighted method (IVW). Additionally, other tests such as MR-Egger intercept, MR-PRESSO, Cochran's Q test, "Leave-one-out," and funnel plots were conducted to assess the extent of pleiotropy and heterogeneity. Using 70 SNPs, there was no evidence to suggest an association between genetically predicted AF and risk of PE with multiplicative random-effects IVW MR analysis (odds ratio = 1.0003, 95% confidence interval: 0.9998-1.0008, P = 0.20). A null association was also observed in other methods. MR-Egger regression and MR-PRESSO respectively showed no evidence of directional (intercept, - 2.25; P = 0.94) and horizontal(P-value in the global heterogeneity test = 0.99) pleiotropic effect across the genetic variants. No substantial evidence was found to support the causal role of AF in the development of PE.