A longitudinal study of emotion predictors and outcomes of disordered eating trajectory classes across adolescence

2017-02-22T01:44:28Z (GMT) by Mitchell, Sarah Anne
Adolescence is a vulnerable period for the emergence of disordered eating (DE) which is associated with a range of negative physical and psychological outcomes. A large body of research has examined prevalence rates of DE in adolescents, but fewer studies have examined differential developmental trajectories of DE symptoms and predictors of trajectories across this period. Emerging evidence suggests that emotion may be a particularly important construct in differentiating alternate trajectories of DE across adolescence. Previous research has demonstrated that abnormalities in emotion processes are associated with the maintenance of DE. Further, in clinical eating disorder (ED) samples, research suggests that abnormalities in emotion processes are associated with structural abnormalities in brain regions implicated in emotion. There is however limited research examining the role of structural abnormalities in brain regions implicated in emotion and how they relate to different DE trajectories across adolescence. Consequently, this thesis aimed to characterise discrete trajectory classes of DE across adolescence using Growth Mixture Modeling and to examine key emotion predictors and outcomes of membership to different patterns of DE across adolescence. To examine these aims, the thesis comprised 3 studies. Study 1 was a narrative review synthesising the literature examining the role of emotion in the development and maintenance of ED’s and DE. Study 2 examined different patterns of change in DE across adolescence using the statistical technique Growth Mixture Modeling (GMM). In Study 3, the identified trajectories were then used to examine how structural abnormalities in early adolescence in brain regions implicated in emotion differentiated those who exhibited trajectories characterised by higher levels of DE compared to those with lower levels of DE across adolescence. The narrative review (Study 1) synthesised the literature examining a range of emotion processes in individuals with ED’s and DE (e.g. including emotion perception, self-reported emotion experience, emotion expression, physiological reactions) and adopting various methodologies to assess emotion (e.g. self-report, observational methods, physiological techniques, neuroimaging and genetics). In regards to predisposing factors, studies indicated that the cognitive emotion regulation strategy rumination predicted the onset of DE. When reviewing the role of abnormalities in emotion processes in perpetuating ED’s and DE several factors were identified including; higher levels of both inwards and outwards anger expression, higher levels of negative self-reported emotions and maladaptive cognitive emotion regulation strategies. These relationships are integrated into a model discussed in Chapter 2. The sample assessed for Study 2 and Study 3 comprised individuals participating in the Adolescent Development Study, a longitudinal study conducted in Melbourne, Australia. Participants included 212 individuals (105 males) from 4 time points (Time 1 Mage = 12.43, Time 2 Mage = 14.99, Time 3 Mage = 16.62 years, Time 4 Mage = 18.86). Participants underwent magnetic resonance imaging scans using a 3-Tesla GE scanner at Time 1 and completed the Eating Disorders Examination Questionnaire- Fourth Edition (EDE-Q; which was used to estimate trajectories) at Time 2, 3 and 4. Emotion and psychosocial outcomes were assessed at Time 4 using a semi-structured clinical interview and self-report measures. Using factor scores based on confirmatory factor analysis of subscales and behavioural frequency items from the EDE-Q, GMM indicated that a 4-Class model (high decreasing, moderate increasing, low stable, and low decreasing) provided the best fit of the DE data. When examining outcomes of trajectory classes, results revealed that the high decreasing and moderate increasing trajectories were associated with poorer emotion and psychosocial outcomes (e.g. higher lifetime prevalence of mood and anxiety disorders, higher levels of anxiety and depression symptoms poorer sleep quality, higher levels of rumination). In Study 3, the identified trajectories were used to examine how structural abnormalities in brain regions implicated in emotion in early adolescence in predict different patterns of DE across adolescence. Results indicated that smaller right vetnrolateral prefrontal cortex volumes predicted membership to the high decreasing trajectory relative to the low stable and low decreasing trajectories, and smaller right nucleus accumbens volumes predicted membership to the high decreasing trajectory relative to the moderate increasing trajectory. Further, smaller right hippocampal volumes predicted membership to the moderate increasing trajectory compared to the high decreasing and low decreasing trajectories and smaller left caudal anterior cingluate cortex volumes and larger right medial medial orbitofrontal cortex volumes predicted membership to the moderate increasing trajectory relative to the low stable trajectory. A general discussion integrates the major findings of the thesis and discusses the results in the context of implications for prevention and intervention efforts. In particular, it is highlighted that individuals experiencing high and moderate levels of DE at age 12 are an at-risk group and would likely benefit from intervention. The potential benefit of addressing abnormalities in emotion processes for the prevention and intervention of DE in adolescents is also discussed. It is recommended that future research extend current findings by examining whether functional abnormalities in brain regions implicated in emotion are also related to DE trajectories. Additional research is also needed examining how the identified structural abnormalities interact with disorder specific risk factors to predict DE trajectories. An increased understanding of different developmental patterns of change in DE across adolescence and predictors of alternate trajectories will inform prevention and intervention efforts to decrease the negative outcomes of these potentially chronic symptoms.