Quantifying psychological resilience and elucidating its mechanisms using multivariate modelling

Abstract

It is estimated that approximately 30% of individuals worldwide are affected by mental health problems during their lifetime. Currently, Major Depressive Disorder (MDD) is one of the most prevalent psychiatric disorders and a leading cause of non-lethal disability worldwide. However, despite exposure to known risk factors for MDD, human responses to it vary widely. Whilst some individuals develop MDD, others develop only mild and transient symptoms or no depressive symptomology at all. This ability to 'bounce back' from or 'escape‘ the development of psychiatric illness is referred to as psychological resilience (Chapter 1). Scientific and clinical interest in resilience has grown exponentially over recent decades, but wide discrepancies are still found in both its definition and measurement. As such, resilience is rarely measured directly, but inferred from the measurement of two specific points of convergence; adversity (its antecedents) and positive adaptation (its consequences). Whilst the study of adversity and positive adaptation has informed our knowledge of resilience it often fails to consider other putative risk factors for MDD (such as genetics), or potential protective factors that may foster resilience despite risk. More recently, examining protective factors have become a focus of research in relation to resilience. This research suggests that numerous protective factors coalesce to contribute to resilient outcomes which give rise to a dynamic resilience process that varies contextually and temporally. Although investigating resilience may be expected to reveal similar findings to studying MDD itself, it does represent a new facet to scientific and clinical research. Specifically, resilience focuses on intervention long before the development of MDD when effects on subsequent suffering may be ameliorated. For this reason, it is imperative to address the concept of resilience, concentrating on the core components of adversity, positive adaptation and protective factors, to move beyond description towards an understanding of individual differences in resilience (Chapter 2). In this thesis, three studies will be presented which aim to examine psychological resilience from multiple perspectives to further delineate the concept. In Chapter 3, the associations and interactions between neuroticism and general intelligence (g) on MDD, and psychological distress were examined in GS:SFHS (Generation Scotland: Scottish Family Health Study) to investigate whether g mitigates the detrimental effects of neuroticism on mental health, as such an association has previously been identified for physical health and mortality. A larger replication was also performed in UK Biobank using a self-reported measure of depression. Across two large samples it was found that intelligence provides protection against psychological distress and self-reported depression in individuals high in neuroticism, but intelligence confers no such protection against clinical MDD in those high in neuroticism. In Chapter 4, a new dataset is presented which was designed to investigate psychological resilience and mental health. Specifically, the STRADL (Stratifying Resilience and Depression Longitudinally) dataset aimed to re-contact existing GS:SFHS participants to obtain repeat measures of MDD and psychological distress in addition to obtaining data on resilience, coping style and adverse life experiences. This dataset has the potential to identify mechanisms and pathways to resilience but also elucidate causal mechanisms and pathways of depression sub-types. Chapter 5 investigated whether neuroticism and resilience are downstream mediators of genetic risk for depression, and whether they contribute independently to such risk. Specifically, the moderating and mediating relationships between polygenic risk scores (PRS) for depression, neuroticism, resilience, and both clinical and self-reported MDD were examined in STRADL. Regression analyses indicated that neuroticism and PRS for depression independently associated with increased risk for both clinical and self-reported MDD, whereas resilience associated with reduced risk. Structural equation modelling suggested that polygenic risk for depression associates with vulnerability for both clinical and self-reported MDD through two partially independent mediating mechanisms in which neuroticism increases vulnerability and resilience reduces it. In Chapter 6, the proportion of phenotypic variance that is attributable to genetic and shared-familial environment was estimated for resilience and three main coping styles; task-, emotion-, and avoidance-oriented coping. Bivariate analyses were conducted to estimate the genetic correlations between these traits and neuroticism. Our results indicate that common genetics affect both resilience and coping style. However, in addition, early shared-environmental effects from the nuclear family influence resilience whereas recent shared-environment effects from a spouse influence coping style. Furthermore, strong genetic overlap between resilience, emotion-oriented coping, and neuroticism suggests a relationship whereby genetic factors that increase negative emotionality lead to decreased resilience. These studies highlight the necessity for complementary multivariate techniques in resilience research to elucidate tractable methodologies to potentially identify mechanisms and modifiable risk factors to protect against psychiatric illness (Chapter 7).