Mixture-based Approach Research Articles

Associations between maternal exposure to multiple metals and metalloids and blood pressure in preschool children: A mixture-based approach

BackgroundExposure to metals during pregnancy can potentially influence blood pressure (BP) in children, but few studies have examined the mixed effects of prenatal metal exposure on childhood BP. We aimed to assess the individual and combined effects of prenatal metal and metalloid exposure on BP in preschool children. MethodsA total of 217 mother–child pairs were selected from the Zhuang Birth Cohort in Guangxi, China. The maternal plasma concentrations of 20 metals [e.g. lead (Pb), rubidium (Rb), cesium (Cs), and zinc (Zn)] in early pregnancy were measured by inductively coupled plasmamass spectrometry. Childhood BP was measured in August 2021. The effects of prenatal metal exposure on childhood BP were explored by generalized linear models, restricted cubic spline and Bayesian kernel machine regression (BKMR) models. ResultsIn total children, each unit increase in the log10-transformed maternal Rb concentration was associated with a 10.82-mmHg decrease (95% CI: −19.40, −2.24) in childhood diastolic BP (DBP), and each unit increase in the log10-transformed maternal Cs and Zn concentrations was associated with a 9.67-mmHg (95% CI: −16.72, −2.61) and 4.37-mmHg (95% CI: −8.68, −0.062) decrease in childhood pulse pressure (PP), respectively. The log10-transformed Rb and Cs concentrations were linearly related to DBP (P nonlinear=0.603) and PP (P nonlinear=0.962), respectively. Furthermore, an inverse association was observed between the log10-transformed Cs concentration and PP (β =-12.18; 95% CI: −22.82, −1.54) in girls, and between the log10-transformed Rb concentration and DBP (β =-12.54; 95% CI: −23.87, −1.21) in boys, while there was an increasing association between the log10-transformed Pb concentration and DBP there was an increasing in boys (β =6.06; 95% CI: 0.36, 11.77). Additionally, a U-shaped relationship was observed between the log10-transformed Pb concentration and SBP (P nonlinear=0.015) and DBP (P nonlinear=0.041) in boys. Although there was no statistically signiffcant difference, there was an inverse trend in the combined effect of maternal metal mixture exposure on childhood BP among both the total children and girls in BKMR. ConclusionsPrenatal exposure to both individual and mixtures of metals and metalloids influences BP in preschool children, potentially leading to nonlinear and sex-specific effects.

MixIR: Mixing Input and Representations for Contrastive Learning.

Recently, contrastive learning has shown significant progress in learning visual representations from unlabeled data. The core idea is training the backbone to be invariant to different augmentations of an instance. While most methods only maximize the feature similarity between two augmented data, we further generate more challenging training samples and force the model to keep predicting aggregated representation on these hard samples. In this article, we propose MixIR, a mixture-based approach upon the traditional Siamese network. On the one hand, we input two augmented images of an instance to the backbone and obtain the aggregated representation by performing an elementwise maximum of two features. On the other hand, we take the mixture of these augmented images as input and expect the model prediction to be close to the aggregated representation. In this way, the model could access more variant data samples of an instance and keep predicting invariant representations for them. Thus, the learned model is more discriminative compared with previous contrastive learning methods. Extensive experiments on large-scale datasets show that MixIR steadily improves the baseline and achieves competitive results with state-of-the-art methods. Our code is available at https://github.com/happytianhao/MixIR.

Association between multiple metals exposure and sleep disorders in a Chinese population: A mixture-based approach

ObjectivePrevious studies have suggested a possible association between metals and sleep disorders. This study aimed to explore the association between Zn, Cu, Se, Mg and Ca and sleep disorders in single and multi-metal co-exposure models. MethodsLogistic regression models, restricted cubic spline model (RCS), Quantile g computation (Q-gcomp), Weighted Quantile Sum (WQS), and Bayesian kernel machine regression (BKMR) models were used to investigate the association between metal levels and sleep disorders. ResultsLogistic regression showed that in the total population, the second, third, and fourth quartile Zn concentration exhibited a lower risk of sleep disorders compared with the first quartile, with odds ratios (ORs) of 0.783, 0.711, and 0.704, respectively. Compared with Zn/Cu and Zn/Se in the first quartile, the third and fourth quartiles showed a lower risk of sleep disorders. In the 30–59 years group, the risk of sleep disorders was 0.699 times greater for the fourth quartile Mg concentration than that for the first quartile. The risk of sleep disorders in Mg/Ca concentration in the third quartile was 0.737 times higher than in the first quartile. Q-gcomp, WQS, and BKMR model analysis showed the negative overall effect of mixtures of the five metals on sleep disorders, with Zn being the largest contributor. ConclusionOur study showed that plasma Zn, Mg, Zn/Cu, Zn/Se, and Mg/Ca reduced the risk of sleep disorders, and the combined effect of multiple metals was negatively associated with the risk of sleep disorders, with Zn being the largest contributor to this relationship.

Association between maternal exposure to per- and polyfluoroalkyl substances and serum markers of liver function during pregnancy in China: A mixture-based approach

Previous studies have shown that per- and polyfluoroalkyl substances (PFAS) may have hepatotoxic effects in animals. However, epidemiological evidence in humans, especially pregnant women, is limited. This study aimed to assess the association of single and multiple PFAS exposure with serum markers of liver function in pregnant women. A total of 420 pregnant women from the Guangxi Zhuang Birth Cohort were enrolled from June 2015 to April 2019. Nine PFAS were measured in the maternal serum in early pregnancy. Data for liver function biomarkers, namely, alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyltransferase (GGT), total bilirubin (TBIL), direct bilirubin (DBIL), and indirect bilirubin (IBIL), were obtained from medical records. In generalized linear model (GLM), there was a positive association of perfluorooctane sulfonate (PFOS) with ALT, perfluorodecanoic acid (PFDA) and perfluorobutanesulfonic acid (PFBS) with GGT, and perfluorohexane sulfonate (PFHxS) with TBIL and IBIL. In contrast, there was a negative association of perfluoroheptanoic acid (PFHpA) with TBIL. There were inverse U-shaped relationships of PFUnA with ALT and AST and PFDA with ALT by restricted cubic spline. The weighted quantile sum (WQS) regression model revealed the positive effects of the PFAS mixture on GGT, TBIL, DBIL, and IBIL. Bayesian kernel machine regression (BKMR) analysis confirmed that the PFAS mixture was positively associated with GGT, and PFBS was the main contributor. In addition, the BKMR model showed a positive association of individual PFBS with GGT, individual PFHxS with TBIL and IBIL, and a negative association of individual PFHpA with TBIL. Our findings provide evidence of an association between individual PFAS, PFAS mixture and maternal serum markers of liver function during pregnancy. Additionally, these findings also enhance concerns over PFAS exposure on maternal liver function and PFAS monitoring in pregnancy, reducing the effect of maternal liver dysfunction on maternal and infant health.

Perfluoroalkyl substance mixtures and cardio-metabolic outcomes in highly exposed male workers in the Veneto Region: A mixture-based approach.

BackgroundPerfluoroalkyl substances (PFAS) have been consistently associated with cardio-metabolic traits. Occupational exposures to multiple PFAS with health outcomes have been poorly investigated. The aim of the present study was to examine these associations among former workers involved in PFAS production. MethodsWe considered 232 male ex-employees who had worked in a factory (Trissino, Veneto Region, Italy), which produced PFAS and other chemicals during 1968–2018. Out of twelve serum PFAS, only four (PFOA, PFOS, PFHxS, and PFNA) were quantifiable in at least 50% of samples. Non-fasting serum total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured. The associations between serum PFAS mixture and considered outcomes were assessed through linear regression mixed models and Weighted Quantile Sum (WQS) regression, adjusting for potential confounders. ResultsPFOA was detected at the highest level, with a median concentration (in ng/mL) of 80.8 (min-max: 0.35–13,033), followed by PFOS (median: 8.55, min-max: 0.35–343), PFHxS (median: 6.8, min-max: 0.35–597) and PFNA (median: 0.8, min-max: 0.35–5). We observed that each A quartile increase in the WQS index was positively associated with the levels of TC (β: 8.41, 95% IC: 0.78–16.0), LDL-C (β: 8.02, 95% IC: 1–15.0) and SBP (β: 3.21, 95% IC: 0.82–5.60). No association of serum PFAS concentration on HDL cholesterol and DBP emerged. WQS analyses revealed a major contribution of PFNA and PFHxS for the cholesterol levels, although PFOA reported the highest concentration. PFOA and PFOS emerged as chemicals of concern regarding the association with SBP. ConclusionsThe results showed a clear association between serum PFAS levels and markers of cardiovascular risk and support the importance of clinical surveillance of cardiovascular risk factors in population with a high exposure to PFAS, especially in the occupational setting.

On discriminating between lognormal and Pareto tail: an unsupervised mixture-based approach

Many stochastic models in economics and finance are described by distributions with a lognormal body. Testing for a possible Pareto tail and estimating the parameters of the Pareto distribution in these models is an important topic. Although the problem has been extensively studied in the literature, most applications are characterized by some weaknesses. We propose a method that exploits all the available information by taking into account the data generating process of the whole population. After estimating a lognormal–Pareto mixture with a known threshold via the EM algorithm, we exploit this result to develop an unsupervised tail estimation approach based on the maximization of the profile likelihood function. Monte Carlo experiments and two empirical applications to the size of US metropolitan areas and of firms in an Italian district confirm that the proposed method works well and outperforms two commonly used techniques. Simulation results are available in an online supplementary appendix.

A robust and accurate finite element framework for cavitating flows with moving fluid-structure interfaces

In the current work, we present a variational mechanics framework for the coupled numerical prediction of cavitating turbulent flow and structural motion via a stabilized finite element formulation. To model the finite mass transfer rate in cavitation phenomena, we employ the homogenous mixture-based approach via phenomenological scalar transport differential equations given by the linear and nonlinear mass transfer functions. Stable linearizations of the finite mass transfer terms for the mass continuity equation and the reaction term of the scalar transport equations are derived for the robust and accurate implementation. The linearized matrices for the cavitation equation are imparted a positivity-preserving property to address numerical oscillations arising from high-density gradients typical of two-phase cavitating flows. The proposed formulation is strongly coupled in a partitioned manner with an incompressible 3D Navier-Stokes finite element solver, and the unsteady problem is advanced in time using a fully-implicit generalized-α time integration scheme. We first verify the implementation on the benchmark case of the Rayleigh bubble collapse. We demonstrate the accuracy and convergence of the cavitation solver by comparing the numerical solutions with the analytical solutions of the Rayleigh-Plesset equation for bubble dynamics. We find our solver to be robust for large time steps and the absence of spurious oscillations/spikes in the pressure field. The cavitating flow solver is coupled with a hybrid URANS-LES turbulence model. We validate the coupled solver for a very high Reynolds number turbulent cavitating flow over a NACA0012 hydrofoil section. Finally, the proposed method is applied in an Arbitrary Lagrangian-Eulerian framework to study turbulent cavitating flow over a pitching hydrofoil section and the characteristic features of cavitating flows such as re-entrant jet and periodic cavity shedding are discussed.

Perfluoroalkyl substance mixtures and cardio-metabolic outcomes in highly exposed male workers in the Veneto Region: a mixture-based approach

BACKGROUND AND AIM: Perfluoroalkyl substances (PFAS) have been consistently associated with cardio-metabolic traits, associations of combined exposures to multiple PFAS with health outcomes in highly exposed populations have been poorly investigated. The aim of the present study was to examine these associations in male workers. METHODS: We considered a population of 232 male employees who had worked in a factory located in Trissino (Veneto, Italy), which produced PFAS and other chemicals during 1968-2018. Twelve serum PFAS were measured by HPLC-MS during 2018-2020; four (PFOA, PFOS, PFHxS and PFNA) were quantifiable in at least 50% of samples. Non-fasting serum total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured. The associations between PFAS mixtures and each outcome were assessed through weighted quantile sum (WQS) regression model, taking into account potential confounders. RESULTS:Among the PFAS, PFOA was detected at the highest level, with a median concentration of 80.75 ng/mL and maximum levels up to 13,000 ng/mL. We observed that each quartile increase in the WQS index was associated with an increase in the levels of TC (β: 9.05, 95% IC: 2.04-16.05), LDL-C (β: 6.79, 95% IC: 0.51-13.07) and SBP (β: 2.64, 95% IC: 0.49-4.79). No effect of PFAS exposure on HDL cholesterol and DBP emerged. WQS analyses revealed a major contribution of PFNA and PFHxS for the cholesterol levels, although the main exposure was due to PFOA. PFOS and PFOA emerged as chemicals of concern regarding the association with SBP. CONCLUSIONS:We found strong positive association between a mixture of four PFAS and total cholesterol levels in particular. Within the mixture PFNA and PFHxS were the most influential. Further work will adjust for other possible confounding factors and apply other approaches to assess chemical mixtures. KEYWORDS: PFAS, Mixtures, Cardiovascular disease, Chemical exposures, Occupational exposures

Impact of menopause on the associations between high exposures to perfluoroalkyl substances and lipid profile: a mixture-based approach

BACKGROUND AND AIM: During their reproductive years, females show lower serum PFAS concentrations than males, most likely due to menstrual excretion, previous childbirth with transplacental transfer of PFAS and breastfeeding. Also, females in pre-menopause age have lower levels of low-density lipoprotein cholesterol (LDL-C) and higher levels of high-density lipoprotein cholesterol (HDL-C) compared to males. So far, little is known of the impact of menopause on dose-response relationships. Given the possible role of estrogens in regulating lipid metabolism, there can be an interaction and the aim of our study is to evaluate if the associations of PFAS and serum lipids changes with aging, comparing women in pre- and post-menopausal age groups. METHODS: A cross-sectional analysis was conducted among highly exposed women, 14,155 mostly pre-menopausal and 4,165 mostly post-menopausal women, based on a cut-off of 50 years. Associations between four PFAS (PFOA, PFOS, PFHxS and PFNA) and non-fasting serum total cholesterol (TC), HDL-C and LDL-C, were assessed through weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) models, adjusting for potential confounders. RESULTS:We observed that each PFAS in the WQS index was associated with an increase in the levels of TC in both groups, and pre-menopausal women showed a larger increase of TC per quartile increase of PFAS mixture (4.63, 95% CI 3.46-5.79) than post-menopausal women (β:1.84, 95% CI 0.30-3.38), mainly driven by the effect on LDL-C. WQS analyses revealed a major contribution of PFOS and PFNA to the association with cholesterol levels, although the larger contrast being in PFOA for both groups. CONCLUSIONS:Investigations of combined exposures to four PFAS showed a strong positive association with all three serum lipid parameters. The large differences between pre- and post-menopausal women may reflect complex relationships between PFAS, estrogens and cholesterol. Analysis on menopausal timing are ongoing. KEYWORDS: PFAS, Mixtures, Chemical exposure, Female

Urinary trace metals, maternal circulating angiogenic biomarkers, and preeclampsia: a single-contaminant and mixture-based approach

BackgroundExposures to toxic metals and deficiencies in essential metals disrupt placentation and may contribute to preeclampsia. However, effects of exposure to combinations of metals remain unknown.ObjectiveWe investigated the relationship between urinary trace metals, circulating angiogenic biomarkers, and preeclampsia using the LIFECODES birth cohort.MethodsUrine samples collected during pregnancy were analyzed for 17 trace metals and plasma samples were analyzed for soluble fms-like tyrosine-1 (sFlt-1) and placental growth factor (PlGF). Cox proportional hazard models were used to estimate the hazard ratios (HR) of preeclampsia associated with urinary trace metals. Linear regression models were used to estimate the relationship between urinary trace metals and angiogenic biomarkers. Principal components analysis (PCA) was used to identify groups of metals and interactions between principal components (PCs) loaded by toxic and essential metals were examined.ResultsIn single-contaminant models, several toxic and essential metals were associated with lower PlGF and higher sFlt-1/PlGF ratio. Detection of urinary chromium was associated with preeclampsia: HR (95% Confidence Interval [CI]) = 3.48 (1.02, 11.8) and an IQR-increase in urinary selenium was associated with reduced risk of preeclampsia (HR: 0.28, 95% CI: 0.08, 0.94). Using PCA, 3 PCs were identified, characterized by essential metals (PC1), toxic metals (PC2), and seafood-associated metals (PC3). PC1 and PC2 were associated with lower PlGF levels, but not preeclampsia risk in the overall cohort.ConclusionsTrace urinary metals may be associated with adverse profiles of angiogenic biomarkers and preeclampsia.

Modeling of Glass Transition Temperatures for Polymeric Coating Materials: Application of QSPR Mixture-based Approach.

Cross-linked block copolymers are structurally complex, and utilization of traditional methods of molecular representation in chemoinformatics is only of limited applicability. Therefore, we introduced new techniques of structural representation for block copolymers. We developed additive and combinatorial approaches that treat a copolymer as a mixture system. In this approach, DRAGON descriptors are concentration-weighted for all chemicals in the reaction mixture. As a proof of concept, we have studied glass transition temperatures of block copolymers of hydroxyalkyl- and dihydroxyalkyl carbamate terminated poly(dimethylsiloxane) oligomers with poly(-caprolactone) and developed four quantitative structure-property relationships (QSPR) models. The correlation coefficient (R2 ) for mentioned QSPR models ranges from 0.851 to 0.911 for the training set. In addition to the newly introduced technique we found that the octanol-water partition coefficient and 3D-MoRSE unweighted descriptors were the most important descriptors for the studied property. The results of the study demonstrated that all chemicals in reaction mixture influenced the glass transition temperatures.

Robust Student’s t Mixture Probability Hypothesis Density Filter for Multi-Target Tracking With Heavy-Tailed Noises

In order to improve filtering accuracy and restrain the degradation of filtering performance caused by the heavy-tailed process and measurement noises in multi-target tracking, this paper proposes a robust Student’s t mixture probability hypothesis density (PHD) filter. In the proposed method, a Student’s t mixture is implemented to the PHD filter, which recursively propagates the intensity as a mixture of Student’s t components in PHD filtering framework. Furthermore, with the advantage of a designed judging and re-weighting mechanism, an M-estimation-based dual-gating strategy is designed for the Student’s t mixture implementation to suppress the negative effect of the heavy-tailed noises. Our proposed approach not only utilizes the Student’s t distribution to match the real heavy-tailed non-Gaussian noise well but also enhances the robustness of the Student’s t mixture-based approach via the designed dual-gating strategy. The simulation results verify that the proposed algorithm can keep good filtering accuracy in the presence of the process and measurement outliers simultaneously.

A mixture-based approach to robust analysis of generalised linear models

ABSTRACTA method for robustness in linear models is to assume that there is a mixture of standard and outlier observations with a different error variance for each class. For generalised linear models (GLMs) the mixture model approach is more difficult as the error variance for many distributions has a fixed relationship to the mean. This model is extended to GLMs by changing the classes to one where the standard class is a standard GLM and the outlier class which is an overdispersed GLM achieved by including a random effect term in the linear predictor. The advantages of this method are it can be extended to any model with a linear predictor, and outlier observations can be easily identified. Using simulation the model is compared to an M-estimator, and found to have improved bias and coverage. The method is demonstrated on three examples.

Comparing consensus Monte Carlo strategies for distributed Bayesian computation

Consensus Monte Carlo is an algorithm for conducting Monte Carlo based Bayesian inference on large data sets distributed across many worker machines in a data center. The algorithm operates by running a separate Monte Carlo algorithm on each worker machine, which only sees a portion of the full data set. The worker-level posterior samples are then combined to form a Monte Carlo approximation to the full posterior distribution based on the complete data set. We compare several methods of carrying out the combination, including a new method based on approximating worker-level simulations using a mixture of multivariate Gaussian distributions. We find that resampling and kernel density based methods break down after 10 or sometimes fewer dimensions, while the new mixture-based approach works well, but the necessary mixture models take too long to fit.

Estimation of finite mixtures with symmetric components

It may sometimes be clear from background knowledge that a population under investigation proportionally consists of a known number of subpopulations, whose distributions belong to the same, yet unknown, family. While a parametric family is commonly used in practice, one can also consider some nonparametric families to avoid distributional misspecification. In this article, we propose a solution using a mixture-based nonparametric family for the component distribution in a finite mixture model as opposed to some recent research that utilizes a kernel-based approach. In particular, we present a semiparametric maximum likelihood estimation procedure for the model parameters and tackle the bandwidth parameter selection problem via some popular means for model selection. Empirical comparisons through simulation studies and three real data sets suggest that estimators based on our mixture-based approach are more efficient than those based on the kernel-based approach, in terms of both parameter estimation and overall density estimation.

Space–time latent component modeling of geo‐referenced health data

Latent structure models have been proposed in many applications. For space-time health data it is often important to be able to find the underlying trends in time, which are supported by subsets of small areas. Latent structure modeling is one such approach to this analysis. This paper presents a mixture-based approach that can be applied to component selection. The analysis of a Georgia ambulatory asthma county-level data set is presented and a simulation-based evaluation is made.

Divide and Conquer: A Mixture-Based Approach to Regional Adaptation for MCMC

The efficiency of Markov chain Monte Carlo (MCMC) algorithms can vary dramatically with the choice of simulation parameters. Adaptive MCMC (AMCMC) algorithms allow the automatic tuning of the parameters while the simulation is in progress. A multimodal target distribution may call for regional adaptation of Metropolis–Hastings samplers so that the proposal distribution varies across regions in the sample space. Establishing such a partition is not straightforward and, in many instances, the learning required for its specification takes place gradually, as the simulation proceeds. In the case in which the target distribution is approximated by a mixture of Gaussians, we propose an adaptation process for the partition. It involves fitting the mixture using the available samples via an online EM algorithm and, based on the current mixture parameters, constructing the regional adaptive algorithm with online recursion (RAPTOR). The method is compared with other regional AMCMC samplers and is tested on simulated as well as real data examples.Relevant theoretical proofs, code and datasets are posted as an online supplement.

Non-linear mixture models for cross-sectional financial log returns

High-frequency financial log returns are known to exhibit sharp departures from the usual normality assumption. This article carries out an in-depth evaluation of the effectiveness of multidimensional non-linear mixed models that arise from considering a scale mixture-based approach to modeling non-normal log returns over a cross-sectional period of time. Simulation-based first-order and second-order comparison measures that compare the non-linear mixed model with the linear model show strong preference in favour of the non-linear mixed models. The methodology is illustrated through a thorough analysis of quarterly intra-day log returns from CISCO, DELL, COKE, and S&P500 for the years 1998–2000.