Growth modeling with nonignorable dropout: alternative analyses of the STAR*D antidepressant trial

This article uses a general latent variable framework to study a series of models for nonignorable missingness due to dropout. Nonignorable missing data modeling acknowledges that missingness may depend not only on covariates and observed outcomes at previous time points as with the standard missing at random assumption, but also on latent variables such as values that would have been observed (missing outcomes), developmental trends (growth factors), and qualitatively different types of development (latent trajectory classes). These alternative predictors of missing data can be explored in a general latent variable framework with the Mplus program. A flexible new model uses an extended pattern-mixture approach where missingness is a function of latent dropout classes in combination with growth mixture modeling. A new selection model not only allows an influence of the outcomes on missingness but allows this influence to vary across classes. Model selection is discussed. The missing data models are applied to longitudinal data from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, the largest antidepressant clinical trial in the United States to date. Despite the importance of this trial, STAR*D growth model analyses using nonignorable missing data techniques have not been explored until now. The STAR*D data are shown to feature distinct trajectory classes, including a low class corresponding to substantial improvement in depression, a minority class with a U-shaped curve corresponding to transient improvement, and a high class corresponding to no improvement. The analyses provide a new way to assess drug efficiency in the presence of dropout.     

Intimate Partner Violence Within Gay Male Couples: Dimensionalizing Partner Violence Among Cuban Gay Men

Using a quantitative–qualitative mixed design, the current study investigated relationship violence within 35 gay male couples living in Santiago, Cuba. Informants narrated how violence was enacted within their relationship. Qualitative analyses revealed men’s construction of masculinity and the sequelae of economic hardships—which led to economic emasculinization—were primary contributors to relationship abuse. Gendered interpretations to the display of heightened masculinity contributed to intimate partner violence. Enacted interpersonal violence was not meaningfully associated with alcohol consumption or to personality measures. We suggest the economic constraints that are perceived by couples to precipitate relational stress may be similar across heterosexual and gay/lesbian couples, and that the relational power accorded to economic privilege cannot be aligned fully by gender.     

Activation and survival of immature neurons in the dentate gyrus with spatial memory is dependent on time of exposure to spatial learning and age of cells at examination

Neurogenesis continues to occur throughout life in the dentate gyrus of the hippocampus and may be related to hippocampus-dependent learning. We have recently reported that there is an enhancement of neurogenesis in the hippocampus only when BrdU is administered 6 days prior to starting spatial training but not when training started either 1 day or 11 days following BrdU administration. In that study, all rats were perfused on day 16 after BrdU injection in order to compare cells of the same age (i.e. 16 day old cells) and thus the survival time after learning was different between groups. This study was designed to address whether the amount of time that passed following training could also contribute to the effects of spatial learning on hippocampal neurogenesis and whether there was differential new neuron activation in response to spatial learning that depended on the age of new cells at the time of spatial learning. Here we tested whether a survival period of 5 days following spatial learning at either 1-5, 6-10 or 11-15 days following BrdU administration would alter cell survival and/or activation of new neurons. Our results indicate that 5 days after training in the Morris water task cell survival is unaltered by training on days 1-5, increased by training at days 6-10 and decreased when training occurs on days 11-15. Furthermore spatial learners trained on days 6-10 or 11-15 show greater activation of new neurons compared to cue-trained rats during a probe trial 5 days after training. In addition, rats trained on the spatial task on days 11-15 had a greater number of activated new neurons compared to rats trained on the spatial task on days 6-10. These results suggest there is a gradual removal of older BrdU-labeled new neurons following spatial learning perhaps due to a competitive interaction with a population of younger BrdU-labeled new neurons.     

Context fear learning specifically activates distinct populations of neurons in amygdala and hypothalamus

The identity and distribution of neurons that are involved in any learning or memory event is not known. In previous studies, we identified a discrete population of neurons in the lateral amygdala that show learning-specific activation of a c-fos-regulated transgene following context fear conditioning. Here, we have extended these studies to look throughout the amygdala for learning-specific activation. We identified two further neuronal populations, in the amygdalo-striatal transition area and medial amygdala, that show learning-specific activation. We also identified a population of hypothalamic neurons that show strong learning-specific activation. In addition, we asked whether these neurons are activated following recall of fear-conditioning memory. None of the populations of neurons we identified showed significant memory-recall-related activation. These findings suggest that a series of discrete populations of neurons are involved in fear learning in amygdala and hypothalamus. The lack of reactivation during memory recall suggests that these neurons either do not undergo substantial change following recall, or that c-fos is not involved in any such activation and change.     

Design of the Steps to Health Study of Physical Activity in Survivors of Endometrial Cancer: Testing a Social Cognitive Theory Model

Physical activity has been shown to benefit cancer survivors' physical functioning, emotional well-being, and symptoms. Physical activity may be of particular benefit to survivors of endometrial cancer because they are more likely to be obese and sedentary than the general population, as these are risk factors for the disease, and thus experience a number of related co-morbid health problems. However, there is little research systematically studying mechanisms of physical activity adherence in cancer survivor populations. This paper describes the design of the Steps to Health study, which applies a Social Cognitive Theory-based model of endometrial cancer survivors' adoption and maintenance of exercise in the context of an intervention to increase walking or other moderate intensity cardiovascular activity. In Steps to Health we will test the influence of self-efficacy and outcome expectations on adherence to exercise recommendations, as well as studying the determinants of self-efficacy. Endometrial cancer survivors who are at least 6 months post-treatment are provided with an intervention involving print materials and telephone counseling, and complete assessments of fitness, activity, self-efficacy and outcome expectations, and determinants of self-efficacy every two months for a six month period. In addition to testing an innovative model, the Steps to Health study employs multiple assessment methods, including ecological momentary assessment, implicit tests of cognitive variables, and ambulatory monitoring of physical activity. The study results can be used to develop more effective interventions for increasing physical activity in sedentary cancer survivors by taking into account the full complement of sources of self-efficacy information and outcome expectations.     

The prediction of intelligence in preschool children using alternative models to regression

Statistical prediction of an outcome variable using multiple independent variables is a common practice in the social and behavioral sciences. For example, neuropsychologists are sometimes called upon to provide predictions of preinjury cognitive functioning for individuals who have suffered a traumatic brain injury. Typically, these predictions are made using standard multiple linear regression models with several demographic variables (e.g., gender, ethnicity, education level) as predictors. Prior research has shown conflicting evidence regarding the ability of such models to provide accurate predictions of outcome variables such as full-scale intelligence (FSIQ) test scores. The present study had two goals: (1) to demonstrate the utility of a set of alternative prediction methods that have been applied extensively in the natural sciences and business but have not been frequently explored in the social sciences and (2) to develop models that can be used to predict premorbid cognitive functioning in preschool children. Predictions of Stanford–Binet 5 FSIQ scores for preschool-aged children is used to compare the performance of a multiple regression model with several of these alternative methods. Results demonstrate that classification and regression treesprovided more accurate predictions of FSIQ scores than does the more traditional regression approach. Implications of these results are discussed.     

Treatment Choice and Placebo Expectation Effects

Placebo expectations can have a strong influence on physical and mental health outcomes. In prior research examining placebo expectation effects, participants have been assigned placebo treatments rather than being actively involved in selecting their treatment. In modern health care, however, individuals are often involved in health care decisions. We review relevant literature regarding choice and placebo expectation effects and conclude that exercising choice over a treatment should strengthen placebo effects. Data relevant to this hypothesis is reviewed. Next, we provide a framework for understanding the key variables involved in mediating and moderating the influence of treatment choice on placebo expectation effects. Finally, theoretical and practical issues revolving around treatment choice in the context of placebo effects are raised.