The term ‘insight’ in mental illness has evolved in its definition and conceptualization to favor multi-dimensional approach, but research on insight and its relationship with psychopathology has yielded inconsistent results. Some findings support...
The term ‘insight’ in mental illness has evolved in its definition and conceptualization to favor multi-dimensional approach, but research on insight and its relationship with psychopathology has yielded inconsistent results. Some findings support the direct causal link between the severity of psychopathology and insight, but others have failed to replicate such findings and instead have posited that insight should be considered as a separate domain of psychopathology. However, these studies mostly have not considered the complex inter-relationships among the symptom complexes and how they may affect insight through different stages of illness.
The overall aim of this thesis is therefore to present a series of studies to advance the conceptualization of insight as a symptom in schizophrenia that is 1) divisible from the traditionally defined domains of psychopathological symptoms and neurocognitive deficits, 2) a non-continuous phenomena, in which the relationships between insight and psychopathological and neurocognitive domains characteristically differ according to different stages of illness, and 3) reflected by the accuracy of the patient’s judgment of not only his/her symptoms but also other clinically related variables, such as subjective well-being and side effects, in relation to the evaluation of the clinicians.
Chapter 1 presents a brief overview of historical conceptualization and definition of insight, etiological models of lack of awareness of illness, and present status of insight research in schizophrenia.
In Chapter 2, it is hypothesized that different symptom domains interacted among each other in relation to insight and that there is a need to better refine such domains in order to construct conceptual model of insight. Hence, in Study 1, five symptom domains derived from factor analytic approach were applied to construct the causal model of insight and psychopathology, along with other feasible alternative models with structural equation modeling method (SEM). Such refinement was deemed necessary to clearly determine the independence between insight and psychopathology. As a result, the hypothesized causal model of insight, in which positive, negative, and autistic preoccupation (cognitive) symptoms were positioned as the primary predictors of insight with the activation factor mediating between positive and autistic preoccupation symptoms and insight, was found to satisfy all indices of goodness-of-fit. Hence, the activation symptoms played not only a mediating role for positive symptoms but also a moderating role for autistic preoccupation symptoms in predicting insight. In Study 2, a subgroup of same patients was followed longitudinally for longitudinal validation of the causal model of insight. Despite the temporal changes in the magnitude of associations among symptoms and their respective predictive strength, our causal model of insight was proven to reliably capture the dynamics between psychopathology and insight in schizophrenia patients at both stabilized (8-week) and chronic (1-year) stages. Also, the post hoc analysis of the causal relationship between depression and insight supported the notion that insight causes depression, rather than the opposite, partially discounting the psychological defense model of insight. In Study 3, the results obtained from the above studies were replicated with another group of patients with schizophrenia using self-reports rather than clinician-assessments of psychopathology. Hence, the causal model of insight constructed with this group also satisfied all indices of goodness-of-fit and self-reported symptoms. The clinician-rated psychopathology significantly correlated only among the patients with insight, and self-reported psychoticism and lack of insight were moderated by self-reported paranoid ideation and hostility. Lastly, in Study 4, differences between the uni-dimensional and multi-dimensional measure of insight in terms of their association with the psychopathology and validity of the model were examined. The results showed that the model constructed with the multi-dimensional measure of insight has relatively more percentage of variance, which can be explained by the contribution from the negative symptoms. The common finding from the above line of studies was that the relationship between insight and psychopathology qualitatively shifted according to the stage of illness and that the level of association between insight and psychopathology was at most moderate. In short, the notion that insight should be considered as a distinct primary symptom has been supported.
Included in Chapter 3 are studies that 1) expanded the present model to include neurocognitive functions by examining both the cross-sectional and longitudinal relationships between insight and neurocognitive variables, and 2) formulated how insight and other clinical variables predict subjective well-being in chronic patients of schizophrenia. Specifically, in Study 1, various domains of neurocognitive functioning, such as general intelligence, executive functioning, verbal and spatial memory, attention, and psychomotor speed, were examined in relation to insight. As a result, COWAT perseverative response, an executive function measure, was found to be the only neurocognitive functioning measure that predicts both cross-sectional level of and short-term longitudinal changes in insight. The cross-sectional model that included this variable as an intermediary variable between psychopathology and insight showed superior goodness-of-fit for both baseline and 8-week assessment. In Study 2, the possibility of expanding the model of insight in chronic patients of schizophrenia was examined by including primary predictors such as psychopathology and executive functioning, and intermediary variables such as insight and subjective side effects in predicting the quality of life. As a result, it was found that insight and executive functions mediate between psychopathology and side effects, and between psychopathology and QOL to determine the cumulative dysfunction in daily life for patients with chronic schizophrenia.
Chapter 4 presents two studies that demonstrate the feasibility of broadening the operational definition of insight and its conceptual paradigm. In other words, they deal with the conceptualization of insight that is based not only on the awareness and labeling of symptoms but also on the understanding of the one’s overall level of functioning and physical discomfort, including side effects of medication. Hence, in Study 1, the patients were stratified according to the level of correspondence between self-reported QOL and clinician-rated functioning based on the assumption that those with good insight would show higher levels of correspondence. As a result, those with higher levels of correspondence showed associations between patient- and clinician-rated psychopathology and with other measures that were stronger and more consistent than those of other discordant groups. In Study 2, drug-free patients mostly without insight were followed for 6-weeks after the patients were divided according to the presence of insight, measured for their heart rate variability, and assessed by the clinician for side effects. They also provided self-reports of side effects. As a result, those with insight showed higher and more consistent associations not only between the measures of side effects but also between self-reported side effects and HRV parameters. Such results confirmed our hypothesis that the conceptualization of insight should not be confined to the awareness of illness, but expanded to include the understanding of a broader range of consequences of illness.
Lastly, this thesis concludes with a brief overview of the findings, overall limitations of the studies, and implications for future studies.