Mixture modeling methods for the assessment of normal and abnormal personality, part II: longitudinal models

Nuanced HEXACO: A Meta-Analysis of HEXACO Cross-Rater Agreement, Heritability, and Rank-Order Stability

Most Five-Factor Model (FFM) questionnaire items contain unique variance that is partly heritable, stable, and consensually observable, demonstrates consistent associations with age and sex, and predicts life outcomes beyond higher order factors. Extending these findings to the HEXACO model, we meta-analyzed single-item cross-rater agreement, heritability, and 2-year stability using samples from six countries. We analyzed raw item scores and their residual variance and adjusted the estimates for measurement unreliability. The median cross-rater agreement, heritability, and stability estimates were, respectively, .30, .30, and .57, for raw items and .10, .16, and .39, for item residuals. Adjusted for reliability, the respective medians were .46 and .25 for cross-rater agreement, .46 and .39 for heritability, and .87 and .94 for stability. These results are strikingly consistent with FFM-based findings, providing nondismissible evidence that single items index a partly unique level of the trait hierarchy-personality nuances-with trait properties comparable to those of higher-order traits.

Trajectories of resilience among young children involved with child protective services

Although child maltreatment is associated with short- and long-term maladaptive outcomes, some children are still able to display resilience. Currently, there is a limited understanding of how children's resilience changes over time after experiencing maltreatment, especially for young children. Therefore, the current study used a longitudinal, multidimensional approach to examine trajectories of resilience among very young children involved in child protective services and determine whether placement setting and caregiving behaviors are associated with resilience trajectories. This study used data from National Survey of Child and Adolescent Well-Being I and conducted repeated measures latent class analysis, focusing on children under 2 years old at baseline (n = 1,699). Results suggested that there were three trajectories of resilience: increasing resilience, decreasing resilience, and stable, low resilience. Caregiver cognitive stimulation was related to increasing trajectories of resilience compared to both decreasing and stable, low resilience. These findings illustrate the importance of caregiving behaviors for promoting resilience among a particularly vulnerable population.

Identifying differential trajectories and predictors for depressive symptoms in adolescents using latent class growth analysis: A population-based cohort study

INTRODUCTION

This study investigated the differential trajectories and relevant determinants of depressive symptoms in adolescents by following cohorts that included junior, senior, and vocational high school adolescents, over a 3-year period in Taiwan.

METHODS

Longitudinal data were obtained from 575 adolescents who participated in the Taiwan Adolescent to Adult Longitudinal Study. Data analysis included latent class growth with time-varying covariate, univariate, and multivariate analysis.

RESULTS

A three-class ("low but increasing trajectory," "moderate and stable trajectory," and "high but decreasing trajectory") model fit the data of the cohort. Our findings indicated that 29%, 38%, and 33% of the adolescents were in the low but increasing, moderate and stable, and high but decreasing trajectories, respectively. After confounders were controlled for, bullying experiences were identified as a risk factor for depressive symptoms. The protective factors against depressive symptoms included resilience and peer and social support.

CONCLUSIONS

The transitions between different educational stages critically influence the depressive symptoms of adolescents, and the adolescents follow different depressive trajectories, that have different etiology. Therefore, identifying adolescents at high risk for depression and designing student-centered intervention programs through individualized and multidimensional assessment of depressive symptoms are crucial for adolescents.

Age Profiles of Cognitive Decline and Dementia in Late Life in the Aging, Demographics, and Memory Study

OBJECTIVES

To better understand the temporal dynamics of progression from cognitive decline to onset of dementia in the dementia-free older population in the United States.

METHODS

We used longitudinal data from a diverse national population-based sample of older adults (N = 531) in the Aging, Demographics, and Memory Study from the Health and Retirement Study with repeated measures of cognitive function and dementia diagnosis during 12 years of follow-up from 1996 to 2009. We employed joint latent class mixed models to estimate the association between cognitive change and competing risks of dementia and nondementia death and identify heterogeneity in the age profiles of such association adjusting for baseline characteristics.

RESULTS

Our analyses found 3 latent classes with distinct age profiles of cognitive decline and associated risk of dementia and mortality: "Rapid Cognitive Decline" (19.6%), "Moderate Progression" (44.6%), and "Optimal Cognitive Aging" (35.8%). When simultaneously accounting for cognitive trajectories and time-to-dementia/death, we also found associations of baseline covariates with slope of cognitive decline (e.g., steeper decline among non-Hispanic Blacks and more educated) and risk of dementia (e.g., greater risk for females and apolipoprotein E-4 carriers, but no difference by education level) that differ substantially from those in separate longitudinal mixed models or survival models.

DISCUSSION

The differential age patterns of cognitive decline predicting dementia incidences identified in this study suggest variation in the course of cognitive aging in older adults that may inform future etiological and intervention studies.

An examination of the joint effects of adolescent interpersonal styles and parenting styles on substance use

The current study examined how parenting and adolescent interpersonal styles jointly influence youths' abilities to form close relationships - a central developmental milestone - yet avoid substance use, which predominantly occurs in the presence of peers. Nine annual waves from an adolescent sample (N = 387) were used to assess (a) combinations of interpersonal and parenting styles from early to middle adolescence using longitudinal latent profile analysis, (b) the validity of these profiles on indicators of adjustment, and (c) the relationships between the profiles and growth in substance use across adolescence as well as substance-related consequences in late adolescence. The results supported five distinct combinations of interpersonal and parenting styles, and validity analyses identified both risk and protective profiles. The protective profile submissive-communal interpersonal style + high-warmth-authoritative parenting style was associated with indicators of positive social adjustment (e.g., friendship quality, resistance to peer influence) as well as lower levels of substance use. Significant differences also emerged with respect to substance-related consequences. The findings of this study highlight how combinations of adolescent interpersonal style and parenting render adolescents more or less successful at navigating peer relationships while avoiding substance use behaviors.

Identifying mismatch and match between clinical needs and mental healthcare use trajectories in people with anxiety and depression: Results of a longitudinal study

BACKGROUND

Mismatch between need and mental healthcare (MHC) use (under-and overuse) has mainly been studied with cross-sectional designs, not accurately capturing patterns of persistence or change in clinical burden and MHC-use among persons with depressive and/or anxiety disorders.

AIMS

Determining and describing [mis]match of longitudinal trajectories of clinical burden and MHC-use.

METHODS

Six-year longitudinal burden and MHC-use data came from the Netherlands Study of Depression and Anxiety (n=2981). The sample was split into four subgroups: I) no clinical burden but constant MHC use, II) constant clinical burden but no MHC-use, III) changing clinical burden and MHC-use, and IV) healthy non-users. Within subgroups I)-III), specific clinical burden and MHC trajectories were identified (growth mixture modeling). The resulting classes' associations with predisposing, enabling, and need factors were investigated (regression analysis).

RESULTS

Subgroups I-III revealed different trajectories. I) increasing MHC without burden (4.1%). II) slightly increasing (1.9%), strongly increasing (2.4%), and decreasing (9.5%) burden without MHC. III) increasing (41.4%) or decreasing (19.4%) burden and concurrently increasing MHC use (first underuse, then matched care), thus revealing delayed MHC-use. Only having suicidal ideation (p<.001, Cohen's d= .6-1.5) was a significant determinant of being in latter classes compared to underusers (strongly increasing burden without MHC-use).

LIMITATIONS

More explanatory factors are needed to explain [mis]match.

CONCLUSION

Mismatch occurred as constant underuse or as delayed MHC-use in a high-income country (Netherlands). Additionally, no meaningful class revealed constantly matched care on average. Presence of suicidal ideation could influence the probability of symptomatic individuals receiving matched MHC or not.

Dynamic Mixture Modeling with dynr

Mixture modeling is commonly used to model sample heterogeneity by identifying unobserved classes of individuals with similar characteristics. Despite abundance of evidence in the literature suggesting that individuals are often characterized by different dynamic processes underlying their physiological, cognitive, psychological, and behavioral states, applications of dynamic mixture modeling are surprisingly lacking. We present here a proof-of-concept example of dynamic mixture modeling, where latent groups of individuals were identified based on different dynamic patterns in their time series. Our sample consists of 192 men who were in a heterosexual relationship. They were asked to complete a daily questionnaire involving emotions related to their relationship. Two latent groups were identified based on the strength of association between positive (e.g., loving) and negative (e.g., doubtful) affect. Men in the group characterized by a strong negative association (β=-.67) tended to be younger and had higher levels of anxiety toward their relationship than men in the other group, which was characterized by a weaker negative association (β=-.31). We illustrate the specification and estimation of dynamic mixture model using "dynr," an R package capable of handling a broad class of linear and nonlinear discrete- and continuous-time models with regime-switching properties.

Mind the gap: A review and recommendations for statistically evaluating Dual Systems models of adolescent risk behavior

According to Dual Systems models (Casey et al., 2008; Luna and Wright, 2016; Steinberg, 2008), a rapidly-developing socioemotional system and gradually-developing cognitive control system characterize adolescent brain development. The imbalance hypothesis forwarded by Dual Systems models posits that the magnitude of the imbalance between these two developing systems should predict the propensity for engaging in a variety of risk behaviors. The current integrative review argues that the excitement generated by the imbalance hypothesis and its implications for explaining adolescent risk behaviors has not been meet with equal efforts to rigorously test this hypothesis. The goal of the current review is to help guide the field to consider appropriate and rigorous methods of testing the imbalance hypothesis. First, we review the analytic approaches that have been used to test the imbalance hypothesis and outline statistical and conceptual limitations of these approaches. Next, we discuss the utility of two longitudinal analytic approaches (Latent Difference Scores and Growth Mixture Modeling) for testing the imbalance hypothesis. We utilize data from a large community adolescent sample to illustrate each approach and argue that Latent Difference Scores and Growth Mixture Modeling approaches enhance the specificity and precision with which the imbalance hypothesis is evaluated.

Fatigue after breast cancer treatment: Biobehavioral predictors of fatigue trajectories

OBJECTIVE

Fatigue is a common side effect of cancer treatment, but there is considerable variability in fatigue severity and persistence among survivors. This study aimed to characterize longitudinal trajectories of fatigue after breast cancer treatment and to identify predictors of varying fatigue trajectories.

METHODS

Women (N = 191) from the Mind-Body Study completed assessments after primary treatment for early stage breast cancer and at regular follow-ups that occurred up to 6 years after treatment (M = 4.3 years). Growth mixture models were used to characterize fatigue trajectories, and demographic, medical, and biobehavioral risk factors were examined as predictors of trajectory group.

RESULTS

Five trajectories were identified, characterized as High, Recovery, Late, Low, and Very Low fatigue. The High and Recovery groups (40% of sample) evidenced elevated fatigue at posttreatment that declined in Recovery but persisted in the High group. In bivariate analyses, trajectory groups differed significantly on depressive symptoms, sleep disturbance, childhood adversity, body mass index, and the inflammatory marker soluble TNF receptor type II, which were higher in the High and/or Recovery groups. In multivariate models, depressive symptoms and childhood adversity distinguished High and Recovery from other groups. Rates of chemotherapy were higher in the Recovery than in the High or Late group, whereas rates of endocrine therapy were higher in the High than in the Recovery group.

CONCLUSIONS

There are distinct longitudinal trajectories of fatigue after breast cancer treatment. Psychological factors are strongly associated with adverse fatigue trajectories, and together with treatment exposures may increase risk for cancer-related fatigue. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Promoting the development of resilient academic functioning in maltreated children

This study examined (a) the extent of heterogeneity in the patterns of developmental trajectories of language development and academic functioning in children who have experienced maltreatment, (b) how maltreatment type (i.e., neglect or physical abuse) and timing of abuse explained variation in developmental trajectories, and (c) the extent to which individual protective factors (i.e., preschool attendance, prosocial skills), relationship protective factors (i.e., parental warmth, absence of past-year depressive episode, cognitive/verbal responsiveness) and community protective factors (i.e., neighborhood safety) promoted the development of resilient language/academic functioning trajectories. Longitudinal data analyses were conducted using cohort sequential Growth Mixture Model (CS-GMM) with a United States national representative sample of children reported to Child Protective Services (n=1,776). Five distinct developmental trajectories from birth to age 10 were identified including two resilient groups. Children who were neglected during infancy/toddlerhood or physically abused during preschool age were more likely to be in the poorer language/academic functioning groups (decreasing/recovery/decreasing and high decreasing) than the resilient high stable group. Child prosocial skills, caregiver warmth, and caregiver cognitive stimulation significantly predicted membership in the two resilient academic functioning groups (low increasing and high stable), after controlling for demographics and child physical abuse and neglect. Results suggest that it is possible for a maltreated child to successfully achieve competent academic functioning, despite the early adversity, and identifies three possible avenues of intervention points. This study also makes a significant contribution to the field of child development research through the novel use of CS-GMM, which has implications for future longitudinal data collection methodology.

Identifying early pathways of risk and resilience: The codevelopment of internalizing and externalizing symptoms and the role of harsh parenting

Psychological disorders co-occur often in children, but little has been done to document the types of conjoint pathways internalizing and externalizing symptoms may take from the crucial early period of toddlerhood or how harsh parenting may overlap with early symptom codevelopment. To examine symptom codevelopment trajectories, we identified latent classes of individuals based on internalizing and externalizing symptoms across ages 3-9 and found three symptom codevelopment classes: normative symptoms (low), severe-decreasing symptoms (initially high but rapidly declining), and severe symptoms (high) trajectories. Next, joint models examined how parenting trajectories overlapped with internalizing and externalizing symptom trajectories. These trajectory classes demonstrated that, normatively, harsh parenting increased after toddlerhood, but the severe symptoms class was characterized by a higher level and a steeper increase in harsh parenting and the severe-decreasing class by high, stable harsh parenting. In addition, a transactional model examined the bidirectional relationships among internalizing and externalizing symptoms and harsh parenting because they may cascade over time in this early period. Harsh parenting uniquely contributed to externalizing symptoms, controlling for internalizing symptoms, but not vice versa. In addition, internalizing symptoms appeared to be a mechanism by which externalizing symptoms increase. Results highlight the importance of accounting for both internalizing and externalizing symptoms from an early age to understand risk for developing psychopathology and the role harsh parenting plays in influencing these trajectories.

Latent variable mixture modeling in psychiatric research--a review and application

Latent variable mixture modeling represents a flexible approach to investigating population heterogeneity by sorting cases into latent but non-arbitrary subgroups that are more homogeneous. The purpose of this selective review is to provide a non-technical introduction to mixture modeling in a cross-sectional context. Latent class analysis is used to classify individuals into homogeneous subgroups (latent classes). Factor mixture modeling represents a newer approach that represents a fusion of latent class analysis and factor analysis. Factor mixture models are adaptable to representing categorical and dimensional states of affairs. This article provides an overview of latent variable mixture models and illustrates the application of these methods by applying them to the study of the latent structure of psychotic experiences. The flexibility of latent variable mixture models makes them adaptable to the study of heterogeneity in complex psychiatric and psychological phenomena. They also allow researchers to address research questions that directly compare the viability of dimensional, categorical and hybrid conceptions of constructs.

Transition and protective agency of early childhood learning behaviors as portents of later school attendance and adjustment

This article reports on the study of differential change trajectories for early childhood learning behaviors as they relate to future classroom adjustment and school attendance. A large sample (N=2152) of Head Start children was followed through prekindergarten, kindergarten, and 1st grade. Classroom learning behaviors were assessed twice each year by teachers who observed gradual declines in Competence Motivation and Attentional Persistence as children transitioned through schooling. Cross-classified multilevel growth models revealed distinct transitional pathways for future adjustment versus maladjustment and sporadic versus chronic absenteeism. Generalized multilevel logistic modeling and receiver operating characteristic curve analyses showed that teachers' earliest assessments were substantially predictive of eventual good classroom adjustment and school attendance, with increasing accuracy for prediction of future sociobehavioral adjustment as time progressed.

Symptom-specific course trajectories and their determinants in primary care patients with Major Depressive Disorder: Evidence for two etiologically distinct prototypes

BACKGROUND

The course-heterogeneity of Major Depressive Disorder (MDD) hampers development of better prognostic models. Although latent class growth analyses (LCGA) have been used to explain course-heterogeneity, such analyses have failed to also account for symptom-heterogeneity of depressive symptoms. Therefore, the aim was to identify more specific data-driven subgroups based on patterns of course-trajectories on different depressive symptom domains.

METHODS

In primary care MDD patients (n=205), the presence of the MDD criterion symptoms was determined for each week during a year. Weekly 'mood/cognition' (MC) and 'somatic' (SOM) scores were computed and parallel processes-LCGA (PP-LCGA) was used to identify subgroups based on the course on these domains. The classes׳ associations with baseline predictors and 2-/3-year outcomes were investigated.

RESULTS

PP-LCGA identified four classes: quick recovery, persisting SOM, persisting MC, and persisting SOM+MC (chronic). Persisting SOM was specifically predicted by higher baseline somatic symptomatology and somatization, and was associated with more somatic depressive symptomatology at long-term follow-up. Persisting MC was specifically predicted by higher depressive severity, thinking insufficiencies, neuroticism, loneliness and lower self-esteem, and was associated with lower mental health related quality of life and more mood/cognitive depressive symptomatology at follow-up.

LIMITATIONS

The sample was small and contained only primary care MDD patients. The weekly depression assessments were collected retrospectively at 3-month intervals.

CONCLUSIONS

The results indicate that there are two specific prototypes of depression, characterized by either persisting MC or persisting SOM, which have different sets of associated prognostic factors and long-term outcomes, and could have different etiological mechanisms.

Factors of psychological distress: clinical value, measurement substance, and methodological artefacts

PURPOSE

Psychometric models and statistical techniques are cornerstones of research into latent structures of specific psychopathology and general mental health. We discuss "pivot points" for future research efforts from a psychometric epidemiology perspective, emphasising sampling and selection processes of both indicators that guide data collection as well as samples that are confronted with them.

METHOD

First, we discuss how a theoretical model of psychopathology determines which empirical indicators (questions, diagnoses, etc.) and modelling methods are appropriate to test its implications. Second, we deal with how different research designs introduce different (co-)variances between indicators, potentially leading to a different understanding of latent structures. Third, we discuss widening the range of statistical models available within the "psychometrics class": the inclusion of categorical approaches can help to enlighten the debate on the structure of psychopathology and agreement on a minimal set of models might lead to greater convergence between studies. Fourth, we deal with aspects of methodology that introduce spurious (co-)variance in latent structure analysis (response styles, clustered data) and differential item functioning to gather more detailed information and to guard against over-generalisation of results, which renders assessments unfair.

CONCLUSIONS

Building on established insights, future research efforts should be more explicit about their theoretical understanding of psychopathology and how the analysis of a given indicator-respondent set informs this theoretical model. A coherent treatment of theoretical assumptions, indicators, and samples holds the key to building a comprehensive account of the latent structures of different types of psychopathology and mental health in general.

Predicting response to intensive multimodal inpatient treatment: a comparison of single- and multiple-class growth modeling approaches

In a modest body of research, personality functioning assessed via performance-based instruments has been found to validly predict treatment outcome and, to some extent, differential response to treatment. However, state-of-the-science longitudinal and mixture modeling techniques, which are common in many areas of clinical psychology, have rarely been used. In this article, we compare multilevel growth curve modeling (MLM) and latent class growth modeling (LCGM) approaches with the same data set to illustrate the different research questions that can be addressed by each method. Global Assessment of Functioning (GAF) scores collected at 6 points during the course of a long-term multimodal inpatient treatment of 58 severely and persistently mentally ill adults were used to model the trajectory of treatment outcome. Pretreatment Rorschach-based markers of personality functioning and other markers of psychiatric severity were examined as covariates in each modeling approach. The results of both modeling approaches generally indicated that more psychologically impaired clients responded less favorably to treatment. The LCGM approach revealed 2 unique trajectories of improvement (a persistently low group and a higher starting, improving group). Personality functioning and baseline psychiatric variables significantly predicted group membership and the rate of change within the groups. A side-by-side examination of these 2 methods was found to be useful in predicting differential treatment response with personality functioning variables.