Subgroup Memberships Research Articles

Profiles and Transitions of Non-Suicidal Self-Injury and Depressive Symptoms among Adolescent Boys and Girls: Predictive Role of Bullying Victimization.

Although non-suicidal self-injury (NSSI) and depressive symptoms often co-occur, as well as display distinct profiles in different genders, few studies examined the gender-specific profiles and transitions of NSSI and depressive symptoms among adolescents and the role of bullying victimization on their profiles and transitions. This study examined the profiles and transitions of NSSI and depressive symptoms for Chinese adolescent boys and girls separately, along with the predictive role of bullying victimization in subgroup memberships and transitions. A total of 3510 adolescents (52.9% boy, M age = 13.36, SD = 0.50) participated in two waves of the study over six months. The results indicated that four and three profiles identified for girls and boys separately. Adolescents in at-risk profiles showed varying degrees of transition. Adolescents with more bullying victimization were more likely to belong or transition to at-risk profiles for both genders, and bullying victimization exacerbated girls transitioning into co-occurring high profile but not in boys. The findings indicate distinct profile and transition patterns of NSSI and depressive symptoms, as well as different risk role of bullying victimization on their profiles and transitions in boys and girls, highlighting the importance of gender differences in understanding co-occurring and transitional nature of NSSI and depressive symptoms and the risk role of bullying victimization, informing effective strategies for prevention and intervention.

What Works for Whom? Evaluating Patterns and Mechanisms of Change Among Bullies, Victims, and Bully-Victims Participating in a School-Based Prevention Program

The effectiveness and program mechanisms of a whole-school anti-bullying program that builds on the socio-ecological framework were investigated by applying person-oriented methods of data analyses. A longitudinal cluster randomized control study was utilized comprising 1377 adolescents (48.5% girls, Mage = 11.7) who participated in a 1-year whole school anti-bullying program that was implemented in 13 schools and 665 adolescents (45.2% girls, Mage = 11.6) who attended 5 control schools. Student reports were collected at pre- and post-test within a 1-year interval. Applying latent profile analyses (LPA), students who were actively involved in bullying as perpetrators (e.g., bullies), who were the target of bullying (e.g., victims), who were both (e.g., bully-victims), and who were not involved either as perpetrators or victims (e.g., uninvolved) were identified. Latent transition analyses (LTA) investigated transition patterns between bully, victim, bully-victim, and uninvolved subgroups in the control and intervention group before and after the 1-year program implementation. Program effectiveness (e.g., a higher transition probability in the intervention compared to the control group) were found for victims and bully-victims. It was explored whether changes in program mechanisms were differently related to changes in these sub-group memberships in the intervention group. Victims transiting to the uninvolved sub-group increased in help-seeking and anti-bullying norms, while bully-victims transiting to the uninvolved sub-group increased help-seeking, perceived teacher intervention, and anti-bullying norms. It can be concluded that this program works differently for different sub-groups of students and that different mechanisms of change drive changes in different sub-groups of students.

Evaluating sensitivity to classification uncertainty in latent subgroup effect analyses

BackgroundIncreasing attention is being given to assessing treatment effect heterogeneity among individuals belonging to qualitatively different latent subgroups. Inference routinely proceeds by first partitioning the individuals into subgroups, then estimating the subgroup-specific average treatment effects. However, because the subgroups are only latently associated with the observed variables, the actual individual subgroup memberships are rarely known with certainty in practice and thus have to be imputed. Ignoring the uncertainty in the imputed memberships precludes misclassification errors, potentially leading to biased results and incorrect conclusions.MethodsWe propose a strategy for assessing the sensitivity of inference to classification uncertainty when using such classify-analyze approaches for subgroup effect analyses. We exploit each individual’s typically nonzero predictive or posterior subgroup membership probabilities to gauge the stability of the resultant subgroup-specific average causal effects estimates over different, carefully selected subsets of the individuals. Because the membership probabilities are subject to sampling variability, we propose Monte Carlo confidence intervals that explicitly acknowledge the imprecision in the estimated subgroup memberships via perturbations using a parametric bootstrap. The proposal is widely applicable and avoids stringent causal or structural assumptions that existing bias-adjustment or bias-correction methods rely on.ResultsUsing two different publicly available real-world datasets, we illustrate how the proposed strategy supplements existing latent subgroup effect analyses to shed light on the potential impact of classification uncertainty on inference. First, individuals are partitioned into latent subgroups based on their medical and health history. Then within each fixed latent subgroup, the average treatment effect is assessed using an augmented inverse propensity score weighted estimator. Finally, utilizing the proposed sensitivity analysis reveals different subgroup-specific effects that are mostly insensitive to potential misclassification.ConclusionsOur proposed sensitivity analysis is straightforward to implement, provides both graphical and numerical summaries, and readily permits assessing the sensitivity of any machine learning-based causal effect estimator to classification uncertainty. We recommend making such sensitivity analyses more routine in latent subgroup effect analyses.

SATs, sets and allegations of bias: The allocation of 11‐year‐old students to mathematics sets in some English schools in 2015. A response to Connolly et al., 2019

AbstractThis is a methodological critique of research by the Best Practice in Grouping Students (BPGS) project claiming teacher bias in allocating students to first‐year secondary school mathematics teaching sets (British Educational Research Journal, 45(4), 873–879). The research assumes that bias could be shown by non‐random relationships between ‘misallocations’ to sets and memberships of gender, ethnic and/or socio‐economic subgroups. This paper questions the authors’ prescriptions for correct set placements and demonstrates two ways in which non‐random relationships between subgroup memberships and (alleged) misplacements could be generated despite all students with similar scores within a school being given the same chance of a set position deemed correct by the researchers. It suggests that the BPGS project results are largely artefacts of their evaluation approach. No claim is made herein that teacher bias plays no part in set allocation, merely that, if it does, the approach by Connolly et al. could not evidence it. A lesson from the critique is that, before drawing inferences from data aggregated from diverse clusters, it is wise for researchers to investigate the way the data are generated at source and patterned by aggregation.

A fusion learning method to subgroup analysis of Alzheimer's disease

Uncovering the heterogeneity in the disease progression of Alzheimer's is a key factor to disease understanding and treatment development, so that interventions can be tailored to target the subgroups that will benefit most from the treatment, which is an important goal of precision medicine. However, in practice, one top methodological challenge hindering the heterogeneity investigation is that the true subgroup membership of each individual is often unknown. In this article, we aim to identify latent subgroups of individuals who share a common disorder progress over time, to predict latent subgroup memberships, and to estimate and infer the heterogeneous trajectories among the subgroups. To achieve these goals, we apply a concave fusion learning method to conduct subgroup analysis for longitudinal trajectories of the Alzheimer's disease data. The heterogeneous trajectories are represented by subject-specific unknown functions which are approximated by B-splines. The concave fusion method can simultaneously estimate the spline coefficients and merge them together for the subjects belonging to the same subgroup to automatically identify subgroups and recover the heterogeneous trajectories. The resulting estimator of the disease trajectory of each subgroup is supported by an asymptotic distribution. It provides a sound theoretical basis for further conducting statistical inference in subgroup analysis.

Synergistic relationships between tobacco smoking, tattooing, religiosity, and spirituality among Chinese Buddhist adolescents

Background Tobacco smoking and tattooing interact in multiple ways because they are risk factors to health and share social determinants and meanings (e.g. deviance). Existing research is limited and reliant on non-generalizable samples in Europe and the United States. Addiction studies on religious Chinese are scarce. In this study, we investigated the embeddedness of smoking in tattooing and tattoo-associated contextual factors in a representative sample of Chinese Buddhist adolescents. Method Latent class analyses based on smoking, tattooing, and tattoo norms were conducted on survey data from 1322 Chinese Dai students (aged 15–19 years, 41.2% females) in seven middle schools in Xishuangbanna, Yunnan Province, China. Subgroup membership was subsequently examined in terms of associations with religiosity, spirituality, and traditionalist tattoo esthetics. Results Three subgroups with different patterns of cooccurring smoking and tattooing were detected. Tattoo norms—especially peer norms—were found to be distinguishable indicators between subgroups. Traditionalist tattoo esthetics was associated with the lowest smoking-tattooing comorbidity subgroup. Religiosity and spirituality showed sporadic associations with subgroup memberships. Conclusions The results demonstrate the complexity of smoking-tattooing comorbidity and the importance of a culturally contextualized understanding of addiction. Greater smoking-tattooing comorbidity among adolescents with higher levels of religiosity and spirituality highlights the limitations of Eurocentric views of addiction. The role of tattoo esthetics suggests that body-related visual information can contribute to substance use. Further studies are needed on religious Chinese, a population overlooked in the literature.

Patterns of Paternal Involvement of Korean Fathers: A Person-Centered Approach

Given roles and expectations of father involvement in South Korea are in transition from traditional breadwinner to an involved caregiver to children, it is plausible that Korean fathers show diverse involvement behaviors in the contexts of work, family, and parenting. Using a person-centered approach, we explored if there were groupings of Korean fathers who could be identified from their involvement with their children. We also examined if those subgroup memberships were related to various factors in work, family, and parenting domains. With a sample of 212 married working fathers and the 12 items of involvement behaviors, we found four heterogeneous subgroups of people: low-involved, accessibility-focused, involved-but-less-accessible, and highly involved fathers. Significant differences among the four profiles were also found regarding various factors such as job stress, work and family conflict, work schedule, maternal employment, parenting satisfaction, and perceived level of involvement. Suggestions for future research, practitioners, and policymakers were discussed.

Stability and Changes in Traditional and Cyberbullying Perpetration and Victimization in Childhood: The Predictive Role of Depressive Symptoms.

Childhood bullying is a public health concern. The stability and changes in children's bullying profiles should provide useful information about specific risk groups and their developmental patterns. The current longitudinal study examined the stability and changes in bullying involvement in children along with the predictive roles of depressive symptoms in bullying subgroup memberships and transitions. A total of 4,321 Chinese children (55.1% male, Mage = 9.93) participated in self-report assessments at three time points with six-month intervals. Bullying subgroups were identified based on physical, relational, and cyberbullying using latent profile analysis. The transitions patterns were modeled using latent transition analysis. Depressive symptoms were added as a covariate in bullying subgroup memberships and transitions. Five bullying subgroups were found at Time 1 and Time 2: traditional bully-victims, traditional victims, (cyber) bully-victims, (cyber) victims, and noninvolved children, whereas four bullying subgroups were found at Time 3: traditional bully-victims, traditional victims, (cyber) bully-victims, and noninvolved children. Findings suggested that (a) childhood bullying involvement showed varying levels of stability, with noninvolved children displaying the most stability and cyberbullying-involved children displaying the least stability; (b) children reporting more severe depressive symptoms manifested a higher risk of bullying involvement; (c) initially noninvolved children showing depressive symptoms were at a higher risk for becoming traditional victims; and (d) children reporting depressive symptoms found it more difficult to escape from bullying than children without depressive symptoms. The findings offer support for both the cumulative continuity and interactional models of development. The findings also highlight the important the role of depressive symptoms in preventing and ameliorating children's bullying.

Using multiple imputation to classify potential outcomes subgroups.

With medical tests becoming increasingly available, concerns about over-testing, over-treatment and health care cost dramatically increase. Hence, it is important to understand the influence of testing on treatment selection in general practice. Most statistical methods focus on average effects of testing on treatment decisions. However, this may be ill-advised, particularly for patient subgroups that tend not to benefit from such tests. Furthermore, missing data are common, representing large and often unaddressed threats to the validity of most statistical methods. Finally, it is often desirable to conduct analyses that can be interpreted causally. Using the Rubin Causal Model framework, we propose to classify patients into four potential outcomes subgroups, defined by whether or not a patient's treatment selection is changed by the test result and by the direction of how the test result changes treatment selection. This subgroup classification naturally captures the differential influence of medical testing on treatment selections for different patients, which can suggest targets to improve the utilization of medical tests. We can then examine patient characteristics associated with patient potential outcomes subgroup memberships. We used multiple imputation methods to simultaneously impute the missing potential outcomes as well as regular missing values. This approach can also provide estimates of many traditional causal quantities of interest. We find that explicitly incorporating causal inference assumptions into the multiple imputation process can improve the precision for some causal estimates of interest. We also find that bias can occur when the potential outcomes conditional independence assumption is violated; sensitivity analyses are proposed to assess the impact of this violation. We applied the proposed methods to examine the influence of 21-gene assay, the most commonly used genomic test in the United States, on chemotherapy selection among breast cancer patients.

Electroretinography may reveal cognitive impairment among a cohort of subjects at risk of a major psychiatric disorder

IntroductionAlmost a third of the offspring of parents diagnosed with schizophrenia or bipolar disorder could develop a mental disorder or related symptoms. The objectives of this study were to test the existence of two distinct subgroups of youth at-risk, according to their retinal response to luminance measured with electroretinography (ERG), and to relate the resulting cluster memberships with the cognitive clusters previously reported. MethodologyA clustering analysis was performed with ERG measurements in 107 at-risk offspring. Each subgroup was compared to a healthy control group of 203 individuals. The ERG subgroup memberships were then associated with the cognitive clusters. ResultsA two-cluster solution was obtained: HR-Cluster1 (n=53) showed a control-like ERG profile and HR-Cluster2 (n=54) showed reduced rod amplitudes and prolonged cone latencies of the b-wave. Subjects in the HR-Cluster2 were 2.7 times more likely to belong to the most detrimental cognitive subgroup than subjects in the HR-Cluster1 (49% Vs 18%). ConclusionAt-risk offspring showed two distinct ERG profiles: a control-like and an altered profile. A higher risk of impaired cognitive function was observed in subjects with the altered ERG profile, suggesting the ERG as a potential biomarker of susceptibility to mental illness among youth at risk.

Typologies of Women With Orgasmic Difficulty and Their Relationship to Sexual Distress

BackgroundAbout 50% of women who report orgasmic difficulty (OD) during partnered sex are distressed by their condition, yet why some women are distressed and others are not is unclear. AimTo determine whether sexual distress is related to women's perceived causes of their OD during partnered sex. MethodsWe established homogenous subgroups of women based on their attributions for OD during partnered sex, and these groups were validated by comparing them on variables relevant to sexual response. We then predicted OD-related distress from subgroup memberships as well as from a number of sociodemographic, control, and empirically supported sexual response variables. Results3 distinct OD subgroups emerged: type 1—high psychological–high somatic reasons; type 2—partner-related reasons; and type 3—moderate psychological–low somatic reasons. These groups also differed on independent parameters related to sexual frequency and arousal. Subgroup membership, along with age, sexual relationship satisfaction, and frequency of partnered sex predicted sexual distress related to OD. Clinical ImplicationParticular perceptions regarding the causes for OD help predict women's sexual distress, and such factors might be considered in identifying sexual issues and managing them within the context of a sexual relationship. Strengths & LimitationsA large sample size drawn from a multinational population powered the study, while the cross-sectional nature of the sample could not rule out bidirectional associations between predictor covariates (including OD subgroup) and the outcome measure (sexual distress). ConclusionType 1 membership (high levels of psychological and somatic attributions) predicted greater levels of OD-related distress than type 2 (partner-related attributions) or type 3 (moderate psychological and low somatic attributions) membership, with type 1 women having a greater likelihood of internalizing (accepting responsibility/blame for) OD attributions.Hevesi K, Miklós E, Horváth Z, et al. Typologies of Women With Orgasmic Difficulty and Their Relationship to Sexual Distress. J Sex Med 2020;17:1144–1155.

Subgroup analysis based on structured mixed-effects models for longitudinal data

ABSTRACT In recent years, subgroup analysis has emerged as an important tool to identify unknown subgroup memberships. However, subgroup analysis is still under-studied for longitudinal data. In this paper, we propose a structured mixed-effects approach for longitudinal data to model subgroup distribution and identify subgroup membership simultaneously. In the proposed structured mixed-effects model, the heterogeneous treatment effect is modeled as a random effect from a two-component mixture model, while the membership of the mixture model is incorporated using a logistic model with respect to some covariates. One advantage of our approach is that we are able to derive the estimation of the treatment effects through an EM-type algorithm which keeps the subgroup membership unchanged over time. Our numerical studies and real data example demonstrate that the proposed model outperforms other competing methods.

RoBoT: a robust Bayesian hypothesis testing method for basket trials.

A basket trial in oncology encompasses multiple "baskets" that simultaneously assess one treatment in multiple cancer types or subtypes. It is well-recognized that hierarchical modeling methods, which adaptively borrow strength across baskets, can improve over simple pooling and stratification. We propose a novel Bayesian method, RoBoT (Robust Bayesian Hypothesis Testing), for the data analysis and decision-making in phase II basket trials. In contrast to most existing methods that use posterior credible intervals to determine the efficacy of the new treatment, RoBoT builds upon a formal Bayesian hypothesis testing framework that leads to interpretable and robust inference. Specifically, we assume that the baskets belong to several latent subgroups, and within each subgroup, the treatment has similar probabilities of being more efficacious than controls, historical, or concurrent. The number of latent subgroups and subgroup memberships are inferred by the data through a Dirichlet process mixture model. Such model specification helps avoid type I error inflation caused by excessive shrinkage under typical hierarchical models. The operating characteristics of RoBoT are assessed through computer simulations and are compared with existing methods. Finally, we apply RoBoT to data from two recent phase II basket trials of imatinib and vemurafenib, respectively.

Confucian Civilization and Identity Construction in Chosŏn Korea

기존의 연구들은 중종대의 다양한 정치적 논의의 의미를 ‘유교의 사회적 영향력’ 차원에서 주로 다루어 왔다. 이와는 달리 본 연구는 이러한 정치적 토론들을 사회적 개혁과 안정의 필요성을 자각하고 당면한 시대 문제를 해결하기 위해 당대 관료들이 행한 국가적 차원의 노력의 일환으로 이해, 접근한다. 이와 관련하여 본 연구는 당시 중앙관료들의 정책에 대해 살펴보고 이를 통해 소위 ‘유교화’를 통해 이루고자 한 그들의 정치적 의도와 목적은 무엇이었는지에 대해서 논의 한다. 본 논문은 또한 “민족적 의식 (national consciousness)의 기원은 인쇄 자본주의 (print capitalism)를 동반한 근대성 출현에 기인한다”는 Benedict Anderson의 주장을 일반화화여 모든 나라에 적용할 수 없음도 적시한다. 결론적으로 이 논문은 의례와 제도를 정비하고 체계화 하려던 조선 정부의 노력이 다양한 사회적 배경의 신민들로 하여금 그들이 강고히 지녀왔던 개별 정체성을 넘어선 확장된 ‘집단적 정체성’을 공유하게 하는데 기여하였다고 주장한다. 이는 한국 사회에 있어서 근대적 ‘민족’ 개념이 단순히 19세기 말 외부로부터 소개되어진 것이 아닌, 그 이전부터 축적되어온 역사적 경험의 산물이었을 가능성을 제시하는 것이다.This study argues that many of the political remarks and debates in the central government in the reign of King Chungjong, which in conventional views have been explained with excessive emphasis on the social influence of Confucianism, were related to government officials’ awareness of the need to reorganize and stabilize the social structure. In this regard, this study looks into the central government’s state politics and discusses the politicians’ various desires and purposes that may exist underneath the “Confucianization.” This paper also points out that Benedict Anderson’s assumption that developments associated with modernity such as print capitalism were crucial to the origin of national consciousness cannot be universally applied to all states. Finally, this paper concludes that with the systemization of rituals and institutions, the Chosŏn state could integrate various subgroups into a larger community whose members shared a “collective identity” which overcome individuals’ rigid sub-group memberships. This

Brain-derived neurotrophic factor modulates neuronal network structure and dynamics in striatal cultures: Implications for Huntington's disease

Almost a third of the offspring of parents diagnosed with schizophrenia or bipolar disorder could develop a mental disorder or related symptoms. The objectives of this study were to test the existence of two distinct subgroups of youth at-risk, according to their retinal response to luminance measured with electroretinography (ERG), and to relate the resulting cluster memberships with the cognitive clusters previously reported.A clustering analysis was performed with ERG measurements in 107 at-risk offspring. Each subgroup was compared to a healthy control group of 203 individuals. The ERG subgroup memberships were then associated with the cognitive clusters.A two-cluster solution was obtained: HR-Cluster1 (n=53) showed a control-like ERG profile and HR-Cluster2 (n=54) showed reduced rod amplitudes and prolonged cone latencies of the b-wave. Subjects in the HR-Cluster2 were 2.7 times more likely to belong to the most detrimental cognitive subgroup than subjects in the HR-Cluster1 (49% Vs 18%).At-risk offspring showed two distinct ERG profiles: a control-like and an altered profile. A higher risk of impaired cognitive function was observed in subjects with the altered ERG profile, suggesting the ERG as a potential biomarker of susceptibility to mental illness among youth at risk.

Mixture Modeling for Longitudinal Data

In this article, we propose an unbiased estimating equation approach for a two-component mixture model with correlated response data. We adapt the mixture-of-experts model and a generalized linear model for component distribution and mixing proportion, respectively. The new approach only requires marginal distributions of both component densities and latent variables. We use serial correlations from subjects’ subgroup memberships, which improves estimation efficiency and classification accuracy, and show that estimation consistency does not depend on the choice of the working correlation matrix. The proposed estimating equation is solved by an expectation-estimating-equation (EEE) algorithm. In the E-step of the EEE algorithm, we propose a joint imputation based on the conditional linear property for the multivariate Bernoulli distribution. In addition, we establish asymptotic properties for the proposed estimators and the convergence property using the EEE algorithm. Our method is compared to an existing competitive mixture model approach in both simulation studies and an election data application. Supplementary materials for this article are available online.

Relationship between single digit addition strategies and working memory reflects general reasoning sophistication

With age, children employ increasingly more sophisticated strategies to solve single-digit addition (SDA) number fact problems. Changes in SDA strategy-use are correlated with age-related changes in cognitive measures (e.g., working-memory capacity, WM); however, this correlation does not explain the basis of SDA strategy-use or change per se. Does strategy-use reflect differences in generally reasoning abilities or are they uniquely tied to SDA abilities? Answers to these questions would help clarify the nature of SDA abilities specifically and early math abilities more generally. To investigate these issues, we assessed 144 5- to 9-year-olds' SDA strategies, visuospatial working memory (VSWM), verbal WM, non-verbal IQ, basic RT, and visuospatial reasoning abilities. We focused on these abilities since recent research shows visuospatial processes support early math development. We used latent profile/class analysis to classify children's SDA strategies and visuospatial reasoning abilities, which yielded four ability subgroups in both cases, only partially related to age. Findings show SDA strategy subgroup and general reasoning subgroup memberships were related. More sophisticated subgroups had better WM, but did not differ in IQ and RT. Findings suggest SDA reasoning reflect general reasoning abilities. Implications for math skills acquisition and intervention are discussed.

Bayesian estimation of associations between identified longitudinal hormone subgroups and age at final menstrual period.

BackgroundAlthough follicle stimulating hormone (FSH) is known to be predictive of age at final menstrual period (FMP), previous methods use FSH levels measured at time points that are defined relative to the age at FMP, and hence are not useful for prospective prediction purposes in clinical settings where age at FMP is an unknown outcome. This study is aimed at assessing whether FSH trajectory feature subgroups identified relative to chronological age can be used to improve the prediction of age at FMP.MethodsWe develop a Bayesian model to identify latent subgroups in longitudinal FSH trajectories, and study the relationship between subgroup membership and age at FMP. Data for our study is taken from the Penn Ovarian Aging study, 1996–2010. The proposed model utilizes mixture modeling and nonparametric smoothing methods to capture hypothesized latent subgroup features of the FSH longitudinal trajectory; and simultaneously studies the prognostic value of these latent subgroup features to predict age at FMP.ResultsThe analysis identified two FSH trajectory subgroups that were significantly associated with FMP age: 1) early FSH class (15 %), which displayed initial increases in FSH shortly after age 40; and 2) late FSH class (85 %), which did not have a rise in FSH until after age 45. The use of FSH subgroup memberships, along with class-specific characteristics, i.e., level and rate of FSH change at class-specific pre-specified ages, improved prediction of FMP age by 20–22 % in comparison to the prediction based on previously identified risk factors (BMI, smoking and pre-menopausal levels of anti-mullerian hormone (AMH)).ConclusionsTo the best of our knowledge, this work is the first in the area to demonstrate the existence of subgroups in FSH trajectory patterns relative to chronological age and the fact that such a subgroup membership possesses prediction power for age at FMP. Earlier ages at FMP were found in a subgroup of women with rise in FSH levels commencing shortly after age 40, in comparison to women who did not exhibit an increase in FSH until after 45 years of age. Periodic evaluations of FSH in these age ranges are potentially useful for predicting age at FMP.Electronic supplementary materialThe online version of this article (doi:10.1186/s12874-015-0101-3) contains supplementary material, which is available to authorized users.

Hybrid Mixture Model for Subpopulation Identification

Personalized medicine aims to identify those patients who have good or poor prognosis for overall disease outcomes or therapeutic efficacy for a specific treatment. A well-established approach is to identify a set of biomarkers using statistical methods with a classification algorithm to identify patient subgroups for treatment selection. However, there are potential false positives and false negatives in classification resulting in incorrect patient treatment assignment. In this paper, we propose a hybrid mixture model taking uncertainty in class labels into consideration, where the class labels are modeled by a Bernoulli random variable. An EM algorithm was developed to estimate the model parameters, and a parametric bootstrap method was used to test the significance of the predictive variables that were associated with subgroup memberships. Simulation experiments showed that the proposed method averagely had higher accuracy in identifying the subpopulations than the Naive Bayes classifier and logistic regression. A breast cancer dataset was analyzed to illustrate the proposed hybrid mixture model.

Family conflict and somatic symptoms over 10 years: A growth mixture model analysis

ObjectiveWhile family conflict and somatic symptoms are mutually associated, few longitudinal studies have examined multiple (heterogeneous) trajectory subgroups for family conflict and somatic symptoms and their covariation over time. The aim of this study was to identify heterogeneous trajectory subgroups for family conflict and somatic symptoms and their joint trajectories. MethodsA representative sample of 424 community participants completed surveys at baseline and 1-, 4-, and 10-year follow-ups. Family conflict and somatic symptoms were assessed at each wave. Covariates (age, gender, marital status, education, and medical conditions) were assessed at baseline. Growth mixture modeling (GMM) was used to identify heterogeneous trajectory subgroups for family conflict and somatic symptoms. A parallel process GMM was used to examine joint trajectory subgroup membership between family conflict and somatic symptoms. ResultsThere were three trajectory subgroups for family conflict: stable low; worsening; and improving, and three somewhat similar trajectory subgroups for somatic symptoms: stable low; stable moderate; and improving. Family conflict and somatic symptom trajectory subgroup memberships were jointly associated. Individuals who had stable low family conflict were most likely to follow a stable low somatic symptom trajectory. Individuals who improved in family conflict were most likely to continue to have stable low somatic symptoms or improve in somatic symptoms. Moreover, individuals who had stable moderate somatic symptoms were most likely to show worsening family conflict. ConclusionThis study demonstrates heterogeneous family conflict and somatic symptom trajectories and indicates that these trajectories covary over time.