Evaluation of supplemental samples in longitudinal research with non-normal missing data

Adolescent Girls' Musculoskeletal Pain is More Affected by Insomnia Than Boys, and Through Different Psychological Pathways

Prior research has established that insomnia is predictive of pain in adolescents and that psychological mechanisms have a crucial role in this relationship. Adolescent girls report more insomnia and pain than boys, yet little is known of gender differences in how insomnia influences pain. This study assessed gender differences in levels and trajectories of insomnia and pain during adolescence, and whether rumination and negative mood mediated the effect of insomnia on pain. Longitudinal survey data measured on 5 annual occasions (Nbaseline = 2,767) were analyzed in a multigroup longitudinal serial mediation model. A final model was generated with insomnia as the predictor, rumination and depressed mood as mediators, pain as the outcome, and gender as the grouping variable. The results showed that insomnia predicted pain in adolescents, with an effect 3.5 times larger in girls than boys. Depressed mood was the main mediator in boys. In girls, rumination was the only significant mediator. There were significant gender differences in the effects of insomnia on rumination and pain, and in the effects of rumination on depressed mood and pain, with stronger effects in girls. These results highlight that girls and boys should be considered separately when studying the relationship between insomnia and pain. PERSPECTIVE: Levels of insomnia and pain are progressively higher in adolescent girls than boys, across adolescence. The predictive strength of insomnia symptoms for future pain is 3.5 times greater in girls, with distinct gender-specific underlying pathways: rumination partially mediates this effect in girls, while depressed mood does so in boys.

Anxiety sensitivity and COVID-19 mental health, fatigue, and well-being: a longitudinal examination among adults from the United States during March-October 2020

There is widespread empirical evidence that the COVID-19 pandemic contributed to elevated risk of mental and physical health symptoms and decreased quality of life. The present investigation sought to examine if individual differences in anxiety sensitivity was associated with mental health, psychosomatic, and well-being among a sample of US adults during a 6-month period early in the COVID-19 pandemic. Employing longitudinal research methodology, we tested the hypothesis that the anxiety sensitivity global factor would be related to increased risk of anxiety, depression, fatigue, and lower well-being. Secondary analyses evaluated the lower order anxiety sensitivity factors for the same criterion variables. The sample consisted of 778 participants with an average age of 37.96 (SD = 11.81; range 18-73). Results indicated that, as hypothesized, anxiety sensitivity was associated with increased risk for more severe anxiety, depression, fatigue, and lesser well-being; the observed effects of anxiety sensitivity were relatively robust and evident in adjusted models that controlled for numerous theoretically and clinically relevant factors (e.g. perceived health status). Overall, these results suggest that pandemic functioning could likely be improved via interventions that target elevated anxiety sensitivity as a vulnerability factor for a broad range of aversive psychosomatic symptoms and personal well-being.

Bias Correction for Replacement Samples in Longitudinal Research

Missing data are commonly encountered problem in longitudinal research. One way researchers handle missing data is through the use of supplemental samples (i.e., the addition of new participants to the original sample after missing data appear at the second or later measurement occasion). Two types of supplemental sample approaches are commonly used: a refreshment approach (additional participants are randomly selected from the population of interest) and a replacement approach (additional participants are selected based on auxiliary variables that explain missingness in the original data). Past research demonstrates that using a replacement approach produces biased parameter estimates because the addition of the replacement sample results in an unrepresentative sample of the population. However, replacement samples have been used in previous studies and the estimation bias has not been corrected. Thus, for this study, we propose and evaluate four ways to correct the bias introduced by replacement samples: a parametric bootstrapping replacement sample correction, a non-parametric bootstrapping replacement sample correction, a primary inverse probability reweighting correction, and a likelihood-based inverse probability reweighting correction. We evaluate their performance using a simulation study and an empirical study.

Applying planned missingness designs to longitudinal panel studies in developmental science: An overview

Longitudinal panel studies are widely used in developmental science to address important research questions on human development across the lifespan. These studies, however, are often challenging to implement. They can be costly, time-consuming, and vulnerable to test-retest effects or high attrition over time. Planned missingness designs (PMDs), in which partial data are intentionally collected from all or some of the participants, are viable solutions to these challenges. This article provides an overview of several PMDs with potential utilities in longitudinal studies, including the multi-form designs, multi-method designs, varying lag designs, accelerated longitudinal designs, and efficient designs for analysis of change. For each of the designs, the basic rationale, design considerations, data analysis, advantages, and limitations are discussed. The article is concluded with some general recommendations to developmental researchers and promising directions for future research.

Longitudinal relations of self-criticism with disordered eating behaviors and nonsuicidal self-injury

OBJECTIVE

Self-criticism has been proposed as a transdiagnostic predictor of disordered eating and nonsuicidal self-injury (NSSI). First, this study explored cross-sectional associations of multiple disordered eating behaviors, NSSI, and self-criticism. Second, it tested longitudinal relations of self-criticism with disordered eating and NSSI, adjusting for baseline levels of both behaviors.

METHODS

In Sub-study 1, undergraduates (N = 251, 79.5% female, M_age = 19.1 years) completed self-report measures of disordered eating, NSSI, and self-criticism at baseline and after 8 weeks. In Sub-study 2, community-based young adults with histories of disordered eating, NSSI, or both (N = 517, 88.8% female, M_age = 24.7 years) completed measures of disordered eating, NSSI, and self-criticism at baseline and after 4 weeks. All measures were completed online.

RESULTS

In Sub-study 1, both disordered eating and NSSI showed significant cross-sectional associations with self-criticism, and self-criticism was significantly related to binge eating, fasting, and NSSI at follow-up. In Sub-study 2, both behaviors again showed significant cross-sectional associations with self-criticism. Self-criticism showed significant longitudinal relations with fasting, purging, and excessive exercise. Longitudinal relations of self-criticism with NSSI varied across disordered eating behaviors.

DISCUSSION

NSSI showed cross-sectional associations with a range of disordered eating behaviors. Self-criticism reflects a common correlate of both disordered eating and NSSI. Evidence supported transdiagnostic longitudinal impact of self-criticism across multiple forms of disordered eating but provided more limited support for impacts on NSSI.

Evaluating Supplemental Samples in Longitudinal Research: Replacement and Refreshment Approaches

Despite the wide application of longitudinal studies, they are often plagued by missing data and attrition. The majority of methodological approaches focus on participant retention or modern missing data analysis procedures. This paper, however, takes a new approach by examining how researchers may supplement the sample with additional participants. First, refreshment samples use the same selection criteria as the initial study. Second, replacement samples identify auxiliary variables that may help explain patterns of missingness and select new participants based on those characteristics. A simulation study compares these two strategies for a linear growth model with five measurement occasions. Overall, the results suggest that refreshment samples lead to less relative bias, greater relative efficiency, and more acceptable coverage rates than replacement samples or not supplementing the missing participants in any way. Refreshment samples also have high statistical power. The comparative strengths of the refreshment approach are further illustrated through a real data example. These findings have implications for assessing change over time when researching at-risk samples with high levels of permanent attrition.