Psychiatric Symptom Dimensions Are Associated With Dissociable Shifts in Metacognition but Not Task Performance

Abstract

Background: Distortions in metacognition-the ability to reflect on and control other cognitive processes-are thought to be characteristic of poor mental health. However, it remains unknown whether such shifts in self-evaluation are due to specific alterations in metacognition and/or a downstream consequence of changes in decision-making processes.

Methods: Using perceptual decision making as a model system, we employed a computational psychiatry approach to relate parameters governing both decision formation and metacognitive evaluation to self-reported transdiagnostic symptom dimensions in a large general population sample (N = 995).

Results: Variability in psychopathology was unrelated to either speed or accuracy of decision formation. In contrast, leveraging a dimensional approach, we revealed independent relationships between psychopathology and metacognition: a symptom dimension related to anxiety and depression was associated with lower confidence and heightened metacognitive efficiency, whereas a dimension characterizing compulsive behavior and intrusive thoughts was associated with higher confidence and lower metacognitive efficiency. Furthermore, we obtained a robust double dissociation-whereas psychiatric symptoms predicted changes in metacognition but not decision performance, age predicted changes in decision performance but not metacognition.

Conclusions: Our findings indicate a specific and pervasive link between metacognition and mental health. Our study bridges a gap between an emerging neuroscience of decision making and an understanding of metacognitive alterations in psychopathology.

Keywords: Cognitive neuroscience; Computational psychiatry; Confidence; Decision making; Metacognition; Psychopathology.

Figure 1

Decision-making task and behavior in experiment 1. (A) Perceptual decision-making task. Subjects were asked to judge which box contained the higher number of dots and to provide a confidence rating in each decision. Choice and confidence responses were unspeeded. (B, C) Behavioral data and drift-diffusion model fits. As difference in dots became greater, accuracy increased (B), and response times decreased (C). These features of the data were well captured by the drift-diffusion model. Error bars reflect SEM. (D) Average confidence rating distributions for correct and incorrect trials. Subjects gave higher confidence ratings for correct (green) than incorrect (red) trials. Shaded areas denote SEM; vertical lines denote the average confidence level for each response class. (E, F) Distributions of mean choice accuracy (E) and confidence level (F) across subjects (n = 498).

Figure 2

Association between decision (left) and metacognitive (right) variables with self-reported psychopathology in experiment 1 (n = 498). Each psychiatric symptom was examined in a separate regression, additionally controlling for the influence of age, gender, and IQ. The y axis indicates the change in each dependent variable for each change of 1 SD of symptom scores. Anxiety and depression symptoms were related to lower confidence level in the absence of a change in decision accuracy. Error bars denote SE. *p < .05, **p < .01, corrected for multiple comparisons over the number of dependent variables tested. See also Supplemental Figure S7A. OCD, obsessive-compulsive disorder.

Figure 3

Three latent factors (dimensions) explained the shared variance between all questionnaire items. (A) Correlation matrix of 209 questionnaire items showing significant correlations between the answers to questionnaire items across subjects. The color scale indicates the correlation coefficient. (B) Eigenvalues from the factor analysis revealing a three-factor solution that best accounted for our data. We labeled these factors anxious-depression, compulsive behavior and intrusive thought, and social withdrawal, according to the strongest individual item loadings. The inset corresponds to a zoom on the first few factors. (C) Item loadings onto each factor, color-coded by questionnaire. See also Supplemental Figures S1 and S2. OCD, obsessive-compulsive disorder.

Figure 4

Factor analysis on the correlation matrix of 209 questionnaire items revealed a three-factor solution comprising anxious-depression, compulsive behavior and intrusive thought, and social withdrawal dimensions. Entry of these factors into a multiple regression model predicting decision formation and metacognition revealed bidirectional effects of anxious-depression and compulsive behavior and intrusive thought factors on confidence level, despite no relationships with performance. ^op < .05 uncorrected, ***p < .001 corrected for multiple comparisons over the number of dependent variables tested. See also Supplemental Figures S7C and S8. NS, not significant.

Figure 5

Model comparison. Taking into account both goodness of fit and parsimony, model comparison provided strong evidence for including psychiatric factors in addition to age and IQ for explaining confidence level. Age/IQ model: Variable ∼ age + IQ + gender. Age/IQ + psychiatric factors model: Variable ∼ anxious-depression factor score + compulsive behavior and intrusive thought factor score + social withdrawal factor score + age + IQ + gender. See also Supplemental Figure S6C. BIC, Bayesian information criterion.