PLmixed: An R Package for Generalized Linear Mixed Models With Factor Structures.

Pharmacokinetic-Pharmacodynamic Modeling and Simulation

Background: Advanced breast cancer (ABC) is affecting substantial number of Chinese women in Hong Kong. Understanding their unmet supportive care needs (SCNs) is important for health care system to precisely allocate resources to areas demanded for help the most and for patients to achieve better quality of life.
\n
\nObjectives: (1) To validate Chinese version of Supportive Care Needs Survey Questionnaire (SCNS-SF34), (2) to address the prevalence of unmet SCNs at baseline (newly diagnosis of ABC) and explore relevant factors associating with baseline unmet SCNs, (3) to identify the trajectories of unmet SCNs from baseline, 6-week, 12-week, 18-week to one-year post-diagnosis of ABC among Hong Kong Chinese women and identify the predictors related to individual resources.
\n
\nMethods: The study consisted of two phases. In Phase I, breast cancer (BC) patients were recruited from six public hospitals and the SCNS-SF34 (which covers five domains of needs) was administered concurrently with measures of psychological distress CHQ-12 (Chinese Health Questionnaire-12), HADS (Hospital Anxiety and Depression Scale), symptom distress (MSAS-SF), and patient satisfaction (ChPSQ-9) to explore factor structure by using Exploratory Factor Analysis (EFA) and to examine internal consistency, and convergent, divergent and discriminant validities of the identified factor structure. In phase II, women newly diagnosed with ABC were recruited and followed up to assess their unmet SCN trajectories one year after diagnosis. Prevalence of initial baseline SCNs and associated demographic, medical and psychological factors were identified. Linear Mixed Modeling (LMM) was performed to assess trajectories for each SCNs domain. Hypothesized variables reflecting personal and social resources (optimism, trait hope, social support, psychological distress and patient satisfaction) were examined for association with the changes of unmet SCNs after adjusting for demographic and medical characteristics.
\n
\nResults: Instead of five-factor structure identified in the original SCNS-SF34, a four-factor structure with 33 items was identified, comprising: 1) Health system, information and patient support needs (HSIPS), 2) Psychological needs (PSYC), 3) Physical and daily living needs (PDL) and 4) Sexuality needs (SEX). The SCNS-SF33-C demonstrated moderate-to-good internal consistency (Cronbach’s alphas=0.75-0.92) across all domains. Acceptable convergent and divergent validity were demonstrated. Discriminant validity was demonstrated in the SCNS-SF33-C’s ability to differentiate between clinically distinct patient groups (ABC vs. localized BC and active treatment vs. no active treatment). Of the top 15 unmet SCNs, all belonged to the HSIPS domain. There were significant linear declines in unmet HSIPS and PSYC needs over the year after diagnosis, but not in PDL and SEX. After adjusting for demographic and medical factors, LMM identified symptom distress, patient satisfaction and patient satisfaction x time are predictors of HSIP. Total symptom distress, optimism, anxiety and anxiety x time predicted PSYC. Total symptom distress was predictor of PDL. Anxiety was predictor of SEX.
\n
\nConclusions: The SCNS-SF33-C has a suitable factor structure and psychometric properties for the use in assessing unmet psychosocial SCN among Chinese women with BC. Generally, unmet HSIP and PSYC tended to decline, while levels of unmet PDL and SEX tended to persist over time. Specific individual resources predicted the future change of unmet SCNs.

Privileged by rapid increase in available epigenomic data, epigenome‐wide association studies (EWAS) are to make a profound contribution to understand the molecular mechanism of DNA methylation in cognitive aging. Current statistical methods used in EWAS are dominated by models based on multiple assumptions, for example, linear relationship between molecular profiles and phenotype, normal distribution for the methylation data and phenotype. In this study, we applied an assumption‐free method, the generalized correlation coefficient (GCC), and compare it to linear models, namely the linear mixed model and kinship model. We use DNA methylation associated with a cognitive score in 400 and 206 twins as discovery and replication samples respectively. DNA methylation associated with cognitive function using GCC, linear mixed model, and kinship model, identified 65 CpGs (p < 1e‐04) from discovery sample displaying both nonlinear and linear correlations. Replication analysis successfully replicated 9 of these top CpGs. When combining results of GCC and linear models to cover diverse patterns of relationships, we identified genes like KLHDC4, PAPSS2, and MRPS18B as well as pathways including focal adhesion, axon guidance, and some neurological signaling. Genomic region‐based analysis found 15 methylated regions harboring 11 genes, with three verified in gene expression analysis, also the 11 genes were related to top functional clusters including neurohypophyseal hormone and maternal aggressive behaviors. The GCC approach detects valuable methylation sites missed by traditional linear models. A combination of methylation markers from GCC and linear models enriched biological pathways sensible in neurological function that could implicate cognitive performance and cognitive aging.

A simulation study was conducted to assess Type I error and statistical power of linear mixed models, generalized linear mixed models, and linear quantile mixed models for the analysis of longitudinal cytometry by time-of-flight data. Findings indicate that, while generalized linear mixed models have superior statistical power, they also suffer from inflated Type I error rates. Linear mixed models can have substantially lower statistical power than the other methods at larger effect sizes. While linear quantile mixed models have slightly lower statistical power at smaller effect sizes, they have intermediate statistical power at larger effect sizes and good Type I error control. Taken altogether, these results generally recommend the use of linear quantile mixed models for longitudinal datasets such as for cytometry by time-of-flight data, although linear mixed models may be slightly more useful at small effect sizes and sample sizes.

In the previous chapters, we discussed autoregressive linear mixed effects models. In this section, we discuss the relationships between the autoregressive linear mixed effects models and nonlinear mixed effects models, growth curves, and differential equations. The autoregressive model shows a profile approaching an asymptote, where the change is proportional to the distance remaining to the asymptote. Autoregressive models in discrete time correspond to monomolecular curves in continuous time. Autoregressive linear mixed effects models correspond to monomolecular curves with random effects in the baseline and asymptote, and special error terms. The autoregressive coefficient is a nonlinear parameter, but all random effects parameters in the model are linear. Therefore, autoregressive linear mixed effects models are nonlinear mixed effects models without nonlinear random effects and have a closed form of likelihood. When there are time-dependent covariates, autoregressive linear mixed effects models are represented by a differential equation and random effects. The monomolecular curve is one of the popular growth curves. We introduce other growth curves, such as the logistic curves and von Bertalanffy curves, and generalizations of growth curves. Re-parameterization is often performed in nonlinear models, and various representations of re-parameterization in monomolecular and other curves are provided herein.

Root-cause identification for quality-related problems is a key issue in quality and productivity improvement for a manufacturing process. Unfortunately, root-cause identification is also a very challenging engineering problem, particularly for a multistage manufacturing process. In this paper, root-cause identification is formulated as a problem of estimation and hypothesis testing of a general linear mixed model. First, a linear mixed fault-quality model is built to describe the cause-effect relationship between the process faults and product quality. Then, the estimation algorithms developed for a general linear mixed model are adapted to estimate the process mean and variance. Finally, a hypothesis testing method is developed to determine if process faults exist in terms of statistical significance. A detailed experimental study illustrated the effectiveness of the proposed methodology.<br><br>Note to Practitioners-Economic globalization brings intense competition among manufacturing enterprises. The key to succeed in this competitive climate is to rapidly respond to fast-changing market demands with high-quality and competitively priced products. To achieve this, we need to quickly identify root causes of quality-related problems in a complicated manufacturing system. However, the current widely adopted quality-control techniques focus more on monitoring than on root-cause identification. These techniques can efficiently detect the changes in the process but the root cause identification is often left to the plant engineers or operators. In this paper, a systematic estimation and testing method is proposed to identify the variational root causes in multistage manufacturing processes. First, a linear model is built based on the design information to describe the cause-effect relationship between the process faults and product quality. Then, an algorithm is developed to estimate the mean and variance of the process faults from the quality measurements of products. Finally, a statistical testing method is developed to determine if process faults (i.e. root causes) exist in terms of statistical significance. A detailed experimental study illustrates the effectiveness of this method. The method presented in this paper is a new quality-control technique and can be used for quality improvement of multistage manufacturing processes.

A central point of disagreement, in certain long-standing discussions about a particular whaling dataset in the Scientific Committee of the International Whaling Commission, has directly involved model selection issues for linear mixed effect models. The biological question under discussion is associated with a clearly defined parameter of primary interest, a focus parameter, which makes model selection with the Focused Information Criterion (FIC) more appropriate than other selection methods. Since the existing FIC methodology has not covered the case of linear mixed effects models, this article sets up the required framework and develops the necessary formulae for the relevant FIC. Our new criterion requires the asymptotic distribution of estimators derived for a given candidate linear mixed model but with behaviour examined under a wider linear mixed model. These results, needed here to build our FIC, also have independent interest.

In this exploratory study, a theoretical model proposed by Sriraman (2005) consisting of five theoretical principles for optimizing creativity in a K–12 setting was investigated empirically. This was accomplished in two steps. In the first study, the five principles were operationalized by generating a questionnaire consisting of 45 items intended to capture the dimension of each principle. An exploratory maximum‐likelihood factor analysis indicated a relatively robust five factor structure that corresponded with the theoretical model. In the second study, the five factor model was validated using a confirmatory factor analysis. The model was then investigated using a two‐level linear mixed model with a random intercept. The results revealed that motivation and mathematical achievement were significant predictors of mathematical creativity.

Utilizes the 'lme4' and 'optimx' packages (previously the optim() function from 'stats') to estimate (generalized) linear mixed models (GLMM) with factor structures using a profile likelihood approach, as outlined in Jeon and Rabe-Hesketh (2012) &lt;<a href="https://doi.org/10.3102%2F1076998611417628" target="_top">doi:10.3102/1076998611417628</a>&gt; and Rockwood and Jeon (2019) &lt;<a href="https://doi.org/10.1080%2F00273171.2018.1516541" target="_top">doi:10.1080/00273171.2018.1516541</a>&gt;. Factor analysis and item response models can be extended to allow for an arbitrary number of nested and crossed random effects, making it useful for multilevel and cross-classified models.

We study model selection for clustered data, when the focus is on cluster specific inference. Such data are often modelled using random effects, and conditional Akaike information was proposed in Vaida & Blanchard (2005) and used to derive an information criterion under linear mixed models. Here we extend the approach to generalized linear and proportional hazards mixed models. Outside the normal linear mixed models, exact calculations are not available and we resort to asymptotic approximations. In the presence of nuisance parameters, a profile conditional Akaike information is proposed. Bootstrap methods are considered for their potential advantage in finite samples. Simulations show that the performance of the bootstrap and the analytic criteria are comparable, with bootstrap demonstrating some advantages for larger cluster sizes. The proposed criteria are applied to two cancer datasets to select models when the cluster-specific inference is of interest.

Using long-term maternal pedigree data, microsatellite analysis, and behavioral tests, we examined whether personality differences in wild Japanese macaques (Macaca fuscata) are associated with additive genetic effects, maternal influences, or belonging to a particular social group. Behaviors elicited by novel-object tests were defined by a component related to caution around novel-objects (Ob-PC1) and behaviors elicited by novel food-tests were defined by correlated components related to consummatory responses (Fo-PC1) and caution around novel foods (Fo-PC2). The repeatability of Ob-PC1 was modest and not significant; the repeatabilities of Fo-PC1 and Fo-PC2 were moderate and significant. Linear mixed effects models found that sex, age, sex × age, provisioning, trial number, date, time of day, season, and distance to the closest monkey were not related to personality. Linear mixed effects models of females older than 2 years found that high rank was associated with greater caution around novel objects. Linear models were used to determine whether sex, age, group membership, maternal kinship, or relatedness had independent effects on the personality similarity of dyads. These analyses found that pairs of macaques that lived in the same group were less similar in their caution around novel objects, more closely related pairs of macaques were more similar in their tendency to eat novel food, and that pairs of macaques in the same group were more similar in how cautious they were around novel foods. Together, these findings suggest that personality in this population of wild monkeys was driven by rank, genetic effects, and group effects, the latter possibly including the need to exploit different niches in the environment.

The dissertation proposes contributions to longitudinal data analysis and moderated regression analysis and is structured into three parts. The first part develops a decorated parallel coordinate plot for longitudinal categorical data, featuring a jitter mechanism revealing the diversity of observed longitudinal patterns, allowing the tracking of each individual pattern and filter options for highlighting typical patterns. The second and the third parts develop semi-parametric methods for moderated regression that combine linear regression models with tree-based algorithms. Specifically, the method of the second part allows fitting tree-structured varying coefficients in multivariate generalized linear mixed models for longitudinal data, and the method of the third part implements coefficient-wise tree-structured varying coefficients in generalized linear models. For general use, the methods are implemented in the freely available packages TraMineR and vcrpart for the statistical software environment R.

Earlier studies to assess the effects of risk factors on child nutritional status in Bangladesh have used conventional regression models that are inadequate to capture a complete scenario of effects. Therefore, this study aimed to evaluate the heterogeneous effects of factors at different points of conditional height-for-age Z-score (HAZ) distribution accounting for cluster-level variation using linear quantile mixed model (LQMM) and to compare them with a linear mixed model (LMM). In addition, an unconditional quantile model (UQM) was used to measure the effect of factors on the unconditional (marginal) HAZ distribution. A total of 6340 children aged 0–59 months extracted from the 2014 Bangladesh Demographic and Health Survey. Different factors – maternal characteristics (age, occupation, nutritional status, parity, birth interval), parental education, child age, breastfeeding status, and morbidity had significant heterogeneous effects on HAZ distribution. For example, secondary or higher educated parents had substantial differential impacts on the lower tail and upper tail of the child HAZ distribution, which was masked by LMM estimate. Moreover, significant cluster-level variations found across all quantiles of child HAZ. During intervention design, heterogeneous effects of factors and cluster variation ought to consider addressing the undernutrition problem in Bangladesh.

In this paper, we discuss two different data transformation techniques for dealing with censored data: Kaplan-Meier weights and the k-nearest neighbor imputation method. The main objective of this paper is to find penalized spline estimates for the components of a linear mixed effect model with right-censored data. In the context of a mixed model setting, the estimation procedure is performed based on the modified or transformed dataset obtained via these transformation techniques. In order to compare the outcomes from a linear mixed model using these two approaches, a Monte Carlo simulation and two real data examples are presented. According to our results, the k-nearest neighbor imputation is very successful in dealing with censored observations.

Excessive Erythrocytosis and Chronic Mountain Sickness in the Highest City in the World: A Longitudinal Study