Search

We study the estimation property of the Elastic Net estimator in high-dimensional linear regression models where the number of parameters p is comparable to or larger than the sample size n. In such a situation, one often assumes sparsity of the true regression coefficient vector β ∗ ∈ R p , i.e., assuming that β ∗ belongs to an ℓ q -ball with radius R q , B q (R q) , for some q ∈ [ 0 , 1 ]. In this paper, we provide ℓ 2 -estimation error bounds for the Elastic Net and naive Elastic Net estimators under a unified framework for high-dimensional analysis of M-estimators proposed by Negahban et al. [A unified framework for high-dimensional analysis of M-estimators with decomposable regularizers. Adv Neural Inf Process Syst. 2009;22:1348–1356]. We show that for both cases of exact sparsity and weak sparsity, under the same conditions on the design matrix, the Elastic Net estimator achieves a slightly better error bound than the Lasso estimator by suitably choosing the tuning parameters. [ABSTRACT FROM AUTHOR]

With the advent of the big data era, the need to combine multiple individual data sets to draw causal effects arises naturally in many medical and biological applications. Especially each data set cannot measure enough confounders to infer the causal effect of an exposure on an outcome. In this article, we extend the method proposed by a previous study to causal data fusion of more than two data sets without external validation and to a more general (continuous or discrete) exposure and outcome. Theoretically, we obtain the condition for identifiability of exposure effects using multiple individual data sources for the continuous or discrete exposure and outcome. The simulation results show that our proposed causal data fusion method has unbiased causal effect estimate and higher precision than traditional regression, meta‐analysis and statistical matching methods. We further apply our method to study the causal effect of BMI on glucose level in individuals with diabetes by combining two data sets. Our method is essential for causal data fusion and provides important insights into the ongoing discourse on the empirical analysis of merging multiple individual data sources. [ABSTRACT FROM AUTHOR]

We analyze regional factors and spatial distribution of children's vision impairment in Shandong province, to explore the spatial changes brought by time and their influencing factors, so as to provide scientific basis for prevention of childhood vision impairment. This study covers five complete cross-sectional surveys from 2013 to 2017, involving about 29.24 million students. Spatial autocorrelation and hotspot analysis methods were used to analyze spatial features. The associated factors were analyzed by multinomial logistic regression. The vision impairment prevalence showed a trend of decreasing first and then increasing from 2013 to 2017, with slight changes. In terms of regional spatial differences, Weihai and Yantai have the highest VI rates in all years, and there was a large-scale spatial aggregation phenomenon. The southern low-value clusters, however, showed a trend of weakening from year to year. Further exploration revealed that economic factors and number of full-time teachers were verified as risk factors for regional vision impairment levels. The slight rebound of the prevalence of vision impairment and the high rate in the eastern and northern regions of Shandong province need more attention. It is suggested that relevant departments should focus on the influence of regional economic and educational factors when formulating relevant strategies. [ABSTRACT FROM AUTHOR]

As highlighted in Dawid (2000) and Pearl & Mackenzie (2018) , deducing the causes of given effects is a more challenging problem than evaluating the effects of causes in causal inference. Lu et al. (2023) proposed an approach for deducing causes of a single effect variable based on posterior causal effects. In many applications, there are multiple effect variables, and they can be used simultaneously to more accurately deduce the causes. To retrospectively deduce causes from multiple effects, we propose multivariate posterior total, intervention and direct causal effects conditional on the observed evidence. We describe the assumptions of no confounding and monotonicity, under which we prove identifiability of the multivariate posterior causal effects and provide their identification equations. The proposed approach can be applied for causal attributions, medical diagnosis, blame and responsibility in various studies with multiple effect or outcome variables. Two examples are used to illustrate the proposed approach. [ABSTRACT FROM AUTHOR]

Objective: High low-density-lipoprotein (LDL) cholesterol has been associated with an increased risk of coronary artery diseases (CAD) including acute myocardial infarction (AMI). However, whether lipids lowering drug treatment is causally associated with decreased risk of AMI remains largely unknown. We used Mendelian randomization (MR) to evaluate the influence of genetic variation affecting the function of lipid-lowering drug targets on AMI. Methods: Single-nucleotide polymorphisms (SNPs) associated with lipids as instruments were extracted from the Global Lipids Genetics Consortium (GLGC). The genome-wide association study (GWAS) data for AMI were obtained from UK Biobank. Two sample MR analysis was used to study the associations between high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides (TG) with AMI (n = 3,927). Genetic variants associated with LDL cholesterol at or near drug target gene were used to mimic drug effects on the AMI events in drug target MR. Results: Genetically predicted higher LDL-C (per one SD increase in LDL-C of 38.67 mg/dL, OR 1.006, 95% CI 1.004–1.007) and TG (per one SD increase in TG of 90.72 mg/dL, 1.004, 1.002–1.006) was associated with increased risk of AMI, but decreased risk for higher HDL-C (per one SD increase in HDL-C of 15.51 mg/dL, 0.997, 0.995–0.999) in univariable MR. Association remained significant for LDL-C, but attenuated toward the null for HDL-C and TG in multivariable MR. Genetically proxied lower LDL-C with genetic variants at or near the PCSK9 region (drug target of evolocumab) and NPC1L1 (drug target of ezetimibe) were associated with decreased risk of AMI (0.997, 0.994–0.999 and 0.986, 0.975–0.998, respectively), whereas genetic variants at HMGCR region (drug target of statin) showed marginal association with AMI (0.995, 0.990-1.000). After excluding drug target-related SNPs, LDL-C related SNPs outside the drug target region remained a causal effect on AMI (0.994, 0.993–0.996). Conclusions: The findings suggest that genetically predicted LDL-C may play a predominant role in the development of AMI. The drug MR results imply that ezetimibe and evolocumab may decrease the risk of AMI due to their LDL-C lowering effect, and there are other non-drug related lipid lowering pathways that may be causally linked to AMI. [ABSTRACT FROM AUTHOR]

Observational studies have suggested that the use of antihypertensive drugs was associated with the risk of frailty; however, these findings may be biased by confounding and reverse causality. This study aimed to explore the effect of genetically predicted lifelong lowering blood pressure (BP) through different antihypertensive medications on frailty. One‐sample Mendelian randomization (MR) and summary data‐based MR (SMR) were applied. We utilized two kinds of genetic instruments to proxy the antihypertensive medications, including genetic variants within or nearby drugs target genes associated with systolic/diastolic BP, and expression level of the corresponding gene. Among 298,618 UK Biobank participants, one‐sample MR analysis observed that genetically proxied BB use (relative risk ratios, 0.76; 95% CI, 0.65–0.90; p = 0.001) and CCB use (0.83; 0.72–0.95; p = 0.007), equivalent to a 10‐mm Hg reduction in systolic BP, was significantly associated with lower risk of pre‐frailty. In addition, although not statistically significant, the effect directions of systolic BP through ACEi variants (0.72; 0.39–1.33; p = 0.296) or thiazides variants (0.74; 0.53–1.03; p = 0.072) on pre‐frailty were also protective. Similar results were obtained in analyses for diastolic BP. SMR of expression in artery showed that decreased expression level of KCNH2, a target gene of BBs, was associated with lower frailty index (beta −0.02, p = 2.87 × 10−4). This MR analysis found evidence that the use of BBs and CCBs was potentially associated with reduced frailty risk in the general population, and identified KCNH2 as a promising target for further clinical trials to prevent manifestations of frailty. [ABSTRACT FROM AUTHOR]

Background To establish a polysocial risk score (PsRS) incorporating various social factors for capturing the dementia risk and investigate the benefits of favorable social conditions across different genetic backgrounds. Methods This prospective cohort study comprised 345 439 participants initially free of dementia from the UK Biobank. A total of 10 social factors were summed to create a PsRS. A polygenic risk score (PRS) was constructed based on genome-wide significant variants. Results During a median follow-up of 12.5 years, we documented 4 595 incident all-cause dementia events including 2 067 Alzheimer's disease (AD) events and 1 028 vascular dementia (VD) events. Each additional PsRS was associated with a 19% increased risk of all-cause dementia (hazard ratio [HR], 1.19; 95% confidence interval [CI], 1.17 to 1.21), a 13% increased risk of AD (1.13; 1.10 to 1.16), and a 24% increased risk of VD (1.24; 1.19 to 1.29). 29% (24% to 33%) of dementia cases, 22% (14% to 29%) of AD cases, and 39% (28% to 48%) of VD cases were associated with a disadvantageous social environment. In addition, among participants at a high genetic risk, the low social risk was linked to a lower incidence rate of all-cause dementia, AD, and VD compared to those who had a high social risk, with reductions of 67.8%, 64.5%, and 84.2%, respectively. Conclusions The PsRS could be effectively used in discriminating individuals at high risk of dementia. Around a quarter of dementia events could have a connection with a disadvantageous social environment, especially for those genetically susceptible to dementia. [ABSTRACT FROM AUTHOR]

Background: Proprotein convertase subtilisin/kexin type 9 (PCSK9) has been recognized as a novel lipid‐lowing target. Recent clinical studies suggested the value of inhibiting PCSK9 in decreasing the vulnerability of coronary plaques. However, the evidence of PCSK9‐regulated evolution of unstable carotid plaques is unclear, which has limited the use of PCSK9 inhibitor in carotid plaques. This study aimed to determine the effect and molecular mechanisms of PCSK9 on vulnerability of carotid plaques, to provide potential therapeutic targets for stabilizing carotid plaques. Methods: The expression of PCSK9 in stable and unstable carotid plaques were examined in tissue and plasma. Human aortic vascular smooth muscle cells (VSMCs) and carotid VSMCs were employed to transfect lentivirus for overexpression and knockdown of PCSK9, respectively. Morphological and functional changes of mitochondria were observed by live‐cell imaging. Cell apoptosis was evaluated by propidium iodide staining. RNA‐sequencing and biological examinations were performed to explore and validate the underlying mechanisms. Truncated plasmids were employed to identify the functional domain of PCSK9 in regulation of VSMCs' mitochondrial morphology, function and apoptosis. Results: Clinically, PCSK9 was closely related with vulnerability of human carotid plaques. Increased expression of PCSK9 in human VSMCs was accompanied by higher level of apoptosis. At subcellular level of VSMCs, the morphology of mitochondria was shifted toward the fission state, followed by mitochondrial dysfunction. Inhibition of p38 MAPK activation partially rescued the above morphological and behavioral changes caused by PCSK9. Furthermore, inhibiting of dynamin‐related protein 1 (DRP1) attenuated PCSK9‐related mitochondrial dysfunction and cell apoptosis. The 1‐149aa domain of PCSK9 protein was essential to achieve functional regulation to VSMCs. Conclusion: Our findings demonstrated that PCSK9 induced morphology‐related mitochondrial dysfunction and apoptosis of VSMCs, which may be related to increased vulnerability of carotid plaque. [ABSTRACT FROM AUTHOR]

Background Telomere length has been linked to various health outcomes. To comprehensively investigate the causal effects of telomere length throughout the human disease spectrum, we conducted a phenome-wide Mendelian randomization study (MR-PheWAS) and a systematic review of MR studies. Methods We conducted a PheWAS to screen for associations between telomere length and 1 035 phenotypes in the UK Biobank (n = 408 354). The exposure of interest was the genetic risk score (GRS) of telomere length. Observed associations passing multiple testing corrections were assessed for causality by 2-sample MR analysis. A systematic review of MR studies on telomere length was performed to harmonize the published evidence and complement our findings. Results Of the 1 035 phenotypes tested, PheWAS identified 29 and 78 associations of telomere length GRS at a Bonferroni- and false discovery rate-corrected threshold; 24 and 66 distinct health outcomes were causal in the following principal MR analysis. The replication MR using data from the FinnGen study provided evidence of causal effects of genetically instrumented telomere length on 28 out of 66 outcomes, including decreased risks of 5 diseases in respiratory diseases, digestive diseases, and myocardial infarction, and increased risks of 23 diseases, mainly comprised neoplasms, diseases of the genitourinary system, and essential hypertension. A systematic review of 53 MR studies found evidence to support 16 out of the 66 outcomes. Conclusions This large-scale MR-PheWAS identified a wide range of health outcomes that were possibly affected by telomere length, and suggested that susceptibility to telomere length may vary across disease categories. [ABSTRACT FROM AUTHOR]

Currently, most predictions of metabolic-associated fatty liver disease (MAFLD) in school-aged children utilize indicators that usually predict nonalcoholic fatty liver disease (NAFLD). The present study aimed to develop new predictive models and predictors for children with MAFLD, which could enhance the feasibility of MAFLD screening programs in the future. A total of 331 school-aged overweight/obese children were recruited from six primary schools in Ningbo city, China. Hepatic steatosis and fibrosis were detected with controlled attenuation parameter (CAP) and liver stiffness measurement (LSM), respectively. Machine learning methods were adapted to build a set of variables to predict MAFLD in children. Then, the area under the curve (AUC) of multiple models and indices was compared to predict pediatric MAFLD. Compared with non-MAFLD children, children with MAFLD had more obvious metabolic disturbances, as they had higher anthropometric indicators, alanine aminotransferase, fasting plasma glucose, and inflammation indicators (white blood cell count, hemoglobin, neutrophil count) (all P < 0.05). The optimal variables for all subjects selected by random forest (RF) were alanine aminotransferase, uric acid, insulin, and BMI. The logistic regression (LR) model performed best, with AUC values of 0.758 for males and 0.642 for females in predicting MAFLD. LnAI-BMI, LnAI, and LnAL-WHtR were approving indices for predicting pediatric MAFLD in all participants, boys and girls individually. Conclusions: This study developed LR models and sex-specific indices for predicting MAFLD in overweight/obese children that may be useful for widespread screening and identification of children at high risk of MAFLD for early treatment. What is Known: • Most of the indicators predicting pediatric MAFLD are derived from the predictive indicators for NAFLD, but the diagnostic criteria for MAFLD and NAFLD are not exactly the same. • The accuracy of predictors based on routine physical examination and blood biochemical indicators to diagnose MAFLD is limited. What is New: • This study developed indicators based on routine examination parameters that have approving performance for MAFLD, with AUC values exceeding 0.70. [ABSTRACT FROM AUTHOR]

Background: Major depressive disorder (MDD) is clinically documented to co-occur with multiple gastrointestinal disorders (GID), but the potential causal relationship between them remains unclear. We aimed to evaluate the potential causal relationship of MDD with 4 GID [gastroesophageal reflux disease (GERD), irritable bowel syndrome (IBS), peptic ulcer disease (PUD), and non-alcoholic fatty liver disease (NAFLD)] using a two-sample Mendelian randomization (MR) design. Methods: We obtained genome-wide association data for MDD from a meta-analysis (N = 480 359), and for GID from the UK Biobank (N ranges: 332 601–486 601) and FinnGen (N ranges: 187 028–218 792) among individuals of European ancestry. Our primary method was inverse-variance weighted (IVW) MR, with a series of sensitivity analyses to test the hypothesis of MR. Individual study estimates were pooled using fixed-effect meta-analysis. Results: Meta-analyses IVW MR found evidence that genetically predicted MDD may increase the risk of GERD, IBS, PUD and NAFLD. Additionally, reverse MR found evidence of genetically predicted GERD or IBS may increase the risk of MDD. Conclusions: Genetically predicted MDD may increase the risk of GERD, IBS, PUD and NAFLD. Genetically predicted GERD or IBS may increase the risk of MDD. The findings may help elucidate the mechanisms underlying the co-morbidity of MDD and GID. Focusing on GID symptoms in patients with MDD and emotional problems in patients with GID is important for the clinical management. [ABSTRACT FROM AUTHOR]

For the case with a single causal variable, Dawid et al. (2014) defined the probability of causation, and Pearl (2000) defined the probability of necessity to assess the causes of effects. For a case with multiple causes that could affect each other, this paper defines the posterior total and direct causal effects based on the evidence observed for post-treatment variables, which could be viewed as measurements of causes of effects. Posterior causal effects involve the probabilities of counterfactual variables. Thus, as with the probability of causation, the probability of necessity and direct causal effects, the identifiability of posterior total and direct causal effects requires more assumptions than the identifiability of traditional causal effects conditional on pre-treatment variables. We present assumptions required for the identifiability of posterior causal effects and provide identification equations. Further, when the causal relationships between multiple causes and an endpoint can be depicted by causal networks, we can simplify both the required assumptions and the identification equations of the posterior total and direct causal effects. Finally, using numerical examples, we compare the posterior total and direct causal effects with other measures for evaluating the causes of effects and the population attributable risks. [ABSTRACT FROM AUTHOR]

Background and Aims: Excessive cannabis use may lead to lower educational attainment. However, this association may be due to confounders and reverse causality. We tested the potential causal relationship between cannabis use disorder (CUD) or life‐time cannabis use (LCU) and educational attainment. Design: Bidirectional two‐sample Mendelian randomization (MR) study was conducted. Our primary method was inverse‐variance weighted (IVW) MR, with a series of sensitivity analyses. Multivariable MR (MVMR) was performed to estimate any direct effect independent of intelligence, smoking initiation or attention deficit hyperactivity disorder (ADHD). Setting and participants: European ancestry individuals. The sample sizes of the genome‐wide association study ranged from 55 374 to 632 802 participants. Measurements Genetic variants of CUD, LCU or educational attainment. Findings Using univariable MR, we found evidence of a potential causal effect of genetic liability to CUD on a lower educational attainment [MR, 95% confidence interval (CI)inverse variance weighted (IVW) = −1.2 month (−1.9 month, −0.5 month); P = 0.0008]. However, we found no evidence of an effect of genetic liability to LCU on educational attainment [MR, 95% CIIVW = 0.5 month (−1.5 month, 2.6 month), P = 0.6032]. Reverse direction analysis suggested that genetic liability to higher educational attainment had a potential causal effect on lower risk of CUD [odds ratio (OR), 95% CIIVW = 0.39 (0.29, 0.52), P = 1.69 × 10−10]. We also found evidence of potential causal effect from genetic liability to higher educational attainment to higher risk of LCU [OR, 95% CIIVW = 1.35 (1.11, 1.66), P = 0.0033]. Conclusions: Genetic liability to cannabis use disorder may lead to lower educational attainment. Genetic liability to higher educational attainment may also lead to higher life‐time cannabis use risk and lower cannabis use disorder risk. However, the bidirectional effect between cannabis use disorder and educational attainment may be due to shared risk factors (e.g. attention‐deficit hyperactivity disorder). [ABSTRACT FROM AUTHOR]

We examined the causal direction between gut microbiota-dependent metabolite trimethylamine N-oxide (TMAO) or its predecessors and cardiometabolic diseases, such as risk of type 2 diabetes mellitus (T2DM), coronary artery disease (CAD), myocardial infarction (MI), stroke, atrial fibrillation (AF), and chronic kidney disease (CKD). We used genetic variants as instruments to test the causal associations. Genetically predicted higher TMAO and carnitine were not associated with higher odds of T2DM, AF, CAD, MI, stroke, and CKD after Bonferroni correction (P ≤ 0.0005). However, we observed that genetically increased choline showed a suggestive association with higher risk of T2DM (odds ratio 1.84 [95% CI 1.00-3.42] per 10 units, P = 0.05). In contrast, genetically predicted higher betaine (0.68 [0.48-0.95] per 10 units, P = 0.023) was suggestively associated with a lower risk of T2DM. We observed a suggestive association of genetically increased choline with a lower level of body fat percentage (β ± SE -0.28 ± 0.11, P = 0.013) but a higher estimated glomerular filtration rate (0.10 ± 0.05, P = 0.034). We further found that T2DM (0.130 ± 0.036, P < 0.0001) and CKD (0.483 ± 0.168, P = 0.004) were causally associated with higher TMAO levels. Our Mendelian randomization findings support that T2DM and kidney disease increase TMAO levels and that observational evidence for cardiovascular diseases may be due to confounding or reverse causality. [ABSTRACT FROM AUTHOR]