Sparse partial least squares with group and subgroup structure

Integrative analysis of high dimensional omics datasets has been studied by many authors in recent years. By incorporating prior known relationships among the variables, these analyses have been successful in elucidating the relationships between different sets of omics data. In this article, our goal is to identify important relationships between genomic expression and cytokine data from a human immunodeficiency virus vaccine trial. We proposed a flexible partial least squares technique, which incorporates group and subgroup structure in the modelling process. Our new method accounts for both grouping of genetic markers (eg, gene sets) and temporal effects. The method generalises existing sparse modelling techniques in the partial least squares methodology and establishes theoretical connections to variable selection methods for supervised and unsupervised problems. Simulation studies are performed to investigate the performance of our methods over alternative sparse approaches. Our R package sgspls is available at https://github.com/matt‐sutton/sgspls .     

Application of structural equation modeling to health outcomes research

This article provides an overview of the basic underlying principles of structural equation modeling (SEM). SEM models have two basic elements: a measurement model and a structural model. The measurement model describes the associations between the indicators (observed measures) of the latent variables, whereas the structural model delineates the direct and indirect substantive effects among latent variables and between measured and latent variables. The application of SEM to health outcomes research is illustrated using two examples: (a) assessing the equivalence of the SF-36 and patient evaluations of care for English- and Spanish-language respondents and (b) evaluating a theoretical model of health in myocardial infarction patients. The results of SEM studies can contribute to better understanding of the validity of health outcome measures and of relationships between physiologic, clinical, and health outcome variables.     

Measuring asthma patient satisfaction in Sweden using partial least squares

Examines the relationships between different aspects involved in asthma treatment. Analyses each aspect's impact on overall patient satisfaction with asthma treatment. Also studies how outcome variables such as compliance with physician's recommendations, health-related quality of life and resource use are affected by the degree of patient satisfaction. The results refer to asthma patients as a group but not necessarily to each patient as an individual. The statistical technique applied for this analysis is partial least squares. Tests the suggested generic model on 599 respondents from a questionnaire survey. The structure of the suggested model is well supported by the data.     

脑卒中患者健康行为结构方程模型构建研究

目的 构建脑卒中患者健康行为结构方程模型,定量分析模型中影响因素对健康行为作用路径及强度。方法 以知信行模式为框架,构建健康行为结构方程模型,采用最大似然法对初始模型进行拟合,拟合度采用RMSEA、GFI、AGFI、IFI、TLI、CFI进行检验;采取路径系数分析健康知识、信念及行为之间作用路径。结果 整体适配度CMIN/DF<3,GFI、IFI、TLI、CFI均>0.90,AGFI>0.80,RMSEA<0.08;健康知识、信念、行为三个潜变量组合信度均>0.80,聚敛效度均>0.50;健康知识对健康信念直接效应0.45,健康信念对健康行为直接效应0.49,健康知识对健康行为直接效应0.31,健康知识对健康行为间接效应0.22,健康知识对健康行为总效应0.53。结论 模型具有较好整体适配度及内在结构适配度;健康知识及信念对健康行为有直接影响,健康知识对健康行为除直接作用外,还通过健康信念中介作用间接影响健康行为。     

On the use of structural equation modeling in health communication research

Structural equation modeling (SEM) is a multivariate technique suited for testing proposed relations between variables. In this article, the authors discuss the potential for SEM as a tool to advance health communication research both statistically and conceptually. Specifically, the authors discuss the advantages that latent variable modeling in SEM affords researchers by extracting measurement error. In addition, they argue that SEM is useful in understanding communication as a complex set of relations between variables. Moreover, the authors articulate the possibility for examining communication as an agent, mediator, and an outcome. Finally, they review the application of SEM to recursive models, interactions, and confirmatory factor analysis.     

结构方程模型在外来劳务工心理健康研究中的应用

目的 探讨结构方程模型在外来劳务工心理健康研究中的应用,为外来劳务工心理健康教育工作指明重点方向.方法 采取多阶段分层抽样的方法对深圳市宝安区1 716名外来劳务工进行症状自评量表(SCL-90)及一般情况的问卷调查,对调查数据进行结构方程模型拟合.结果 结构方程模型分析结果表明,在反映心理状况的9个因子中,标准化路径系数从大到小前5位的分别是焦虑、抑郁、人际关系敏感、精神病性及强迫.外来劳务工的生活经历、健康感知及家庭背景对心理状况的标准化路径系数分别为-0.165、0.506、-0.059,均具有统计学意义(均有P＜0.05).拟合模型的x2值与自由度(df)的比值为4.331,均方根的迈似误差(RMSEA)=0.044,拟合优度指数(GFI)=0.967,修正的拟合优度指数(AGFI)=0.954、比较拟合指数(CFI)=0.974、标准拟合指数( NFI)=0.967.结论 外来劳务工心理健康的结构方程模型拟合效果较好,外来劳务工的生活经历、健康感知及家庭背景是其心理健康的主要影响因素.     

Application of copulas to improve covariance estimation for partial least squares

Dimension reduction techniques, such as partial least squares, are useful for computing summary measures and examining relationships in complex settings. Partial least squares requires an estimate of the covariance matrix as a first step in the analysis, making this estimate critical to the results. In addition, the covariance matrix also forms the basis for other techniques in multivariate analysis, such as principal component analysis and independent component analysis. This paper has been motivated by an example from an imaging study in Alzheimer's disease where there is complete separation between Alzheimer's and control subjects for one of the imaging modalities. This separation occurs in one block of variables and does not occur with the second block of variables resulting in inaccurate estimates of the covariance. We propose the use of a copula to obtain estimates of the covariance in this setting, where one set of variables comes from a mixture distribution. Simulation studies show that the proposed estimator is an improvement over the standard estimators of covariance. We illustrate the methods from the motivating example from a study in the area of Alzheimer's disease. Copyright © 2012 John Wiley & Sons, Ltd.     

Analysis of additive risk model with high-dimensional covariates using partial least squares

In this paper, we construct a partial additive regression (PAR) model to predict the survival times of cancer patients based on microarray gene expression data with right censoring. The area under time-dependent receiver operating characteristic curve is used as a model evaluation criterion. We conduct a simulation study to compare the proposed method with other methods, i.e. partial Cox regression and supervised principal component analysis. Two data sets of breast cancer and diffuse large B-cell lymphoma are analyzed to illustrate our procedure. The outcome indicates great predictive performance on both dimension reduction and predictive performance of the proposed PAR model.     

Using hospitalization for ambulatory care sensitive conditions to measure access to primary health care: an application of spatial structural equation modeling

Background  

In data commonly used for health services research, a number of relevant variables are unobservable. These include population lifestyle and socio-economic status, physician practice behaviors, population tendency to use health care resources, and disease prevalence. These variables may be considered latent constructs of many observed variables. Using health care data from South Carolina, we show an application of spatial structural equation modeling to identify how these latent constructs are associated with access to primary health care, as measured by hospitalizations for ambulatory care sensitive conditions. We applied the confirmatory factor analysis approach, using the Bayesian paradigm, to identify the spatial distribution of these latent factors. We then applied cluster detection tools to identify counties that have a higher probability of hospitalization for each of the twelve adult ambulatory care sensitive conditions, using a multivariate approach that incorporated the correlation structure among the ambulatory care sensitive conditions into the model.     

Patient-reported outcomes helped predict survival in multiple myeloma using partial least squares analysis

OBJECTIVE: The prognostic value of Patient-Reported Outcomes (PRO) in predicting mortality during treatment of multiple myeloma (MM) patients was assessed using partial least square (PLS) regression, a statistical method that is well-adapted for highly correlated data. STUDY DESIGN AND SETTING: Four PRO measures, The European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30, the EORTC QLQ-MY24, the FACIT-Fatigue scale, and the FACT/GOG-Ntx scale, were administered during a trial designed to evaluate the efficacy and safety of bortezomib (VELCADE 1.3mg/m(2)) in MM patients (N=202). Clinical and PRO data were analyzed for predictive value by univariate and multivariate logistic regression methods and then by PLS regression. RESULTS: Fifteen baseline PRO parameters were significant in predicting mortality during treatment when univariate logistic regression was used. In contrast, only two variables were retained in the multivariate analysis, as correlated variables were excluded from the model. Using PLS regression, 14 of the 21 PRO predictors were significant in predicting mortality. Clinical and PRO data used together increased the predictive power of all models compared to clinical data alone. CONCLUSION: The prognostic value of PRO was established and was more informative using PLS regression. PLS regression may therefore be a valuable method for analyzing PRO data.     

An analysis of health related behavior using interpretive structural modeling

In order for people to know about and to adopt and maintain healthy living practices, 1) a theoretical overview regarding factors associated with health behavior, and 2) an understanding of the actual pattern of behavior in given situations are needed. While theoretical models are helpful in providing a perspective, these models are not practical enough for understanding actual patterns of behavior. In the present study, the ISM (Interpretive Structural Modeling) method was utilized to understand the actual pattern of health related behaviors. The ISM method is used in systems engineering for structurally modeling complex systems. In this study, the ISM method was applied to grasp the structure of coping behavior in the case of fever caused by a common cold under the following two conditions; i) a simplified situation using eight elementary behaviors, and ii) a more complicated situation using more than eight elementary behaviors. i) Subjects were 30 students of public health nursing. The sequence of eight elementary behaviors was determined by paired comparisons using the ISM matrix. The microcomputer made a network diagram of elementary behaviors. The 30 diagrams, none of which were the same, were classified into three types: 1) simple linear (7 subjects), 2) one junction (12 subjects), 3) two or more junctions (11 subjects). After the experiment, subjects were instructed to evaluate the validity of the ISM method. More than 80 percent of the subjects rated the ISM method as effective in increasing their cognition of the hierarchical structure of health related behaviors. ii) Two subjects (A and B) were instructed to come up with as many possible coping behaviors as they could imagine.(ABSTRACT TRUNCATED AT 250 WORDS)     

On the application of structural equation modeling for the construction of a health index

Objective  

The health of an individual is influenced by many factors. These could include factors that are related to the economy and the environment, as well as social and biological factors. Many studies have been carried out to study the effect of these factors on health, in terms of the individual factors or combined factors. The main purpose of this study was to demonstrate the value of structural equation modeling for the construction of an index to describe the health status of an individual.     

Violence,stigma and mental health among female sex workers in China: A structural equation modeling

Intimate partner violence is prevalent among female sex workers (FSWs) in China, and it is significantly associated with mental health problems among FSWs. However, limited studies have explored the mechanisms/process by which violence affects mental health. The purpose of this study was to explore the relationships among partner violence, internalized stigma, and mental health problems among FSWs. Data were collected using a self-administered cross-sectional survey administered to 1,022 FSWs in the Guangxi Zhuang Autonomous Region (Guangxi), China during 2008–2009. We used structural equation modeling to test the hypothesized relationships. Results indicated that violence perpetrated by either stable sexual partners or clients was directly and positively associated with mental health problems. Violence also had an indirect relation to mental health problems through stigma. Results highlight the need for interventions on counseling and care for FSWs who have experienced violence and for interventions to increase FSWs’ coping skills and empowerment strategies.     

What social workers in health care should know about lupus: a structural equation model

This article reports on findings from a cross-sectional study (N=378) of patients living with systemic lupus erythematosus (SLE). The purpose of this study was to identify and clarify the unique psychosocial challenges for those living with lupus. The specific analysis will help to develop a model to determine how different factors influence SLE patients' psychosocial needs. Key findings indicate that the highest general causes of depressive and anxious feelings were changes in appearance due to SLE and limitations in physical abilities due to SLE. The more chronic the symptoms, the more likely it was that feelings of depression would ensue. The more education subjects had, the less likely they were to report feeling depressed or anxious about their SLE challenges. Those with no health care insurance reported the highest levels of depressive and anxious feelings, those with Medicaid reported the second highest, and those with Medicare reported the least SLE-related depression and anxiety. The great majority of SLE patients on medications experienced a wide range of side effects, the most prominent being hair loss. These findings can inform policy and programs as well as clinical initiatives for those affected by SLE.     

健康危险因素与个体健康间的结构方程模型

目的探索健康危险因素与个体健康间的潜在关系,以及各潜变量对测量指标的影响。方法利用健康风险评估调查问卷调查西安市某三级甲等医院健康体检人群。构建健康危险因素与个体健康间的结构方程模型,解释不同因子间的相互作用及其对总体健康的效应权重。结果健康行为与健康意识之间呈显著正相关。健康行为对躯体健康的直接效用为0.85,健康意识不仅直接作用躯体健康,而且通过对心理健康的影响对躯体健康产生间接效应,总效应达0.78。结论健康行为、健康意识均是可控的健康危险因素,对躯体健康、心理健康产生一定影响。通过改变这些可控因素可以提高人们整体健康水平。     

结构方程模型GLS与WLS性能比较

目的通过构建结构方程模型的全模型,比较广义最小二乘法(GLS)和加权最小二乘法(WLS)在不同特征数据中的性能差异。方法建立包括12个外生显变量、3个外生潜变量和8个内生显变量、2个内生潜变量全模型的真模型和误设模型,运用SAS 9.1软件的IML模块生成模拟数据,通过CALIS过程进行模型拟合,采用两类错误频率对2种参数估计方法的性能进行评价。结果分布特征为多元正态分布、轻度偏态分布和重度偏态分布的数据,在采用相关系数矩阵和协方差矩阵时,GLS和WLS的两类错误频率均随相关系数或样本含量的增加而呈现下降趋势; GLS法表现为第一类错误频率较大而第二类错误频率较小,3种分布n>200即显变量个数的10倍以上时第二类错误频率<0.05,而第一类错误频率只有在n≥1 000即显变量个数的50倍及以上时才近似<0.05;WLS法第二类错误频率几乎均为0,但第一类错误频率较大,在数据特征条件相同时其相关系数矩阵的第一类错误频率小于协方差矩阵的第一类错误频率。结论GLS法与WLS法相比是比较稳健的结构方程模型参数估计方法。     

后新冠疫情时代对公众健康素养教育的思考——基于一项疫情期间的横断面调查和结构方程模型分析

目的 探讨公众对新冠肺炎疫情的常识性防疫能力、过度反应和预期的之间的关系,为防疫常态化阶段公众健康素养教育提供依据.方法 2020年3月1日至3月6日期间,采用便利抽样,通过网络问卷星问卷对553名中国居民进行调查.问卷分为三个维度:常识性防疫能力,过度反应和过高预期.采用logistic回归分别对三个维度的影响因素进...     

An exploration of socioeconomic variation in lifestyle factors and adiposity in the Ontario Food Survey through structural equation modeling

Title  

An exploration of socioeconomic variation in lifestyle factors and adiposity in the Ontario Food Survey through structural equation models.     

An exploration of socioeconomic variation in lifestyle factors and adiposity in the Ontario Food Survey through structural equation modeling

Background  

Multiple-risk-factor interventions offer a promising means for addressing the complex interactions between lifestyle behaviors, psychosocial factors, and the social environment. This report examines the long-term effects of a multiple-risk-factor intervention.