Individuals with bipolar disorder have a lower quality of life than people without the disorder, even during euthymia (Michalak et al., 2010). Several parameters have been associated with quality of life in bipolar disorder: disturbances in sleep and biological rhythms (Slyepchenko et al., 2019), self-stigmatization and residual depressive signs (Post et al., 2018), comorbid anxiety disorder (Yapici Eser et al., 2018), and hypomanic residual signs (Yoldi-Negrete et al., 2019). To date, few studies have investigated the links between cognition and quality of life in bipolar disorders and their results are heterogeneous: some studies report no link (Dias et al., 2008; Faurholt-Jepsen et al., 2020) while other studies find that better neurocognition is associated with a better quality of life transversally (Cotrena et al., 2016) and 6 months later (Mackala et al., 2014). In fact, a recent study shows that it is cognitive reserve rather than the current cognitive performance that is associated with quality of life (Cotrena et al., 2020). In this project, we propose to significantly advance the question of the links between neurocognition and quality of life by analyzing data from the 12 bipolar expert centers of the FondaMental foundation. Cognition will be measured by a complete neuropsychological battery and quality of life by the EQ-5D. Three measurement times over two years of follow-up will be included in the analyses. Intermediate variables such as residual depression, anxiety-status, functioning, metacognition will be integrated to better understand the causal pathway linking neurocognition and quality of life. Cross-sectional and longitudinal SEM (using cross-lagged panel modeling) will be used with neurocognition as the latent variable. If you are interested in this project, please contact Paul Roux.

Our initial work on the cross-sectional database of people followed at the Versailles Expert Center for Bipolar Disorder suggested a detrimental impact of hyperactivation of maladaptive early schema on functioning independently from the neuropsychological performance. We have the project to analyze longitudinal data of the impact of these schemas on functioning and quality of life one year later (N=200) using multiple linear regressions. Different covariates will be entered into the model such as residual depression, neurocognition, trait anxiety, and metacognition. If you are interested in this project, please contact Paul Roux.

Schizophrenia is characterized by difficulties in specific metacognition manifested by a mismatch between the judgment on performance at a specific cognitive task and the performance itself. Patients may show overconfidence in errors and a lack of confidence in correct answers. They may interrupt the accumulation of evidence prematurely when evidence is needed to respond to a cognitive task, or they may not use the evidence sufficiently to make judgments. This overconfidence bias may play a crucial role in delusions (Moritz and Woodward, 2006; Bacon and Izaute, 2009). The field of metacognition in social cognition has been little explored to date in people with schizophrenia. Studies report overconfidence in errors of perception of mental states on the Read the Mind in the Eye Test (Köther et al., 2012) and of facial emotions (Moritz et al., 2012). However, more recent studies do not confirm these results, reporting instead a tendency for people with schizophrenia to be underconfident in their emotional perception (Cornacchio et al., 2017; Jones et al., 2019) with no difference in adjustment between performance and confidence between patients and controls (Pinkham et al., 2018). This heterogeneity seems to be related to methodological issues of first-order performance control, as overconfidence in emotional perception seems to be closely associated with first-order social and non-social cognition performance (Perez et al., 2020). The determinants of this social metacognition are unknown in schizophrenia. In this project, we propose to search for clinical features (positive symptoms, disorganization, depression, mania, amount of antipsychotic treatment, impulsivity, clinical insight, psychosocial functioning, and quality of life) that modulate the link (statistically significant interaction) between objective and subjective mental performance. In the task used in this study, we have presented 10 intentional attribution comic strips, and for each of them to ask the participant first to choose which picture ends the story and then to rate how confident he or she is on a 4-level Likert scale: not at all confident, not confident, almost confident, totally confident. We included 143 patients with two measurement times one year apart. For each participant, we will calculate a logistic regression coefficient of first-order performance on second-order performance at inclusion and assess whether this measure of metacognitive sensitivity predicts functioning and quality of life one year later. If you are interested in this project, please contact Paul Roux.

Several previous studies have explored the impact of neurocognition on quality of life in schizophrenia. However, the results are contradictory, as several authors have found a positive association between cognitive performance and quality of life (Herman, 2004; Alptekin et al., 2005; Savilla et al., 2008; Woon et al., 2010; Kim et al., 2021), while others have reported a negative association (Narvaez et al., 2008; Meesters et al., 2013; DeRosse et al., 2018; Kurtz et al., 2019). Finally, other studies report no significant association (Brissos et al., 2008; Chino et al., 2009). In this context, we propose to conduct a systematic review of the literature using the PRISMA method to identify moderators who explain the heterogeneity of the links between neurocognition and quality of life in schizophrenia. In this project, we will then conduct a meta-analysis to measure the size of this association and reduce heterogeneity by grouping studies according to the moderators identified above (type of quality of life measure? type of neuropsychological measures? duration of disorder evolution? control for confounding factors in the association such as depression, insight or self-stigmatization?) and specifying the effect sizes in these subgroups. We will also conduct a multivariate analysis by meta-analytic SEM (metaSEM) to estimate a multiple regression model with the different neuropsychological domains as predictors and quality of life as a dependent variable. The metaSEM will evaluate the unique and combined predictive capacity of quality of life through the different neuropsychological dimensions. The unique predictive capacity of a domain measures the impact of that domain by controlling for the influence of other domains. If you are interested in this project, please contact Paul Roux.

This internship will explore the following question: Do the difficulties of individuals with schizophrenia in representing the emotions, intentions, beliefs and desires of others lead to an increased risk of heteroaggression? In non-clinical samples, theory of mind has been associated with aggression in children (Austin et al 2017) and fear recognition is impaired in violent adult offenders proportionally to their level of psychopathy (Jusyte et al., 2016). If the clinical determinants of heteroagressivity in schizophrenia have been extensively studied, much less is known about their cognitive underpinnings. A metanalysis suggests that neurocognitive is slightly more impaired in violent patients compared to non-violent ones (Reinharth et al., 2014). Such a metanalysis is lacking for social cognition, despite a growing body of evidence for an association between violence in schizophrenia and ToM (Abu-Akel et al., 2004; Altamura et al. 2015; Engelstad et al. 2019; Johansson et al., 2020), emotion recognition (Weiss et al., 2006; Brugman et al., 2016; Bulgari et al. 2019), intentional bias (Waldheter et al, 2005 ) and social reasoning (O’Reilly et al., 2016). If you are interested in this project, please contact Paul Roux.

Cognitive impairments are hallmarks of schizophrenia and to a lesser extent of bipolar disorders. However, the functional impact of psychotropic medication, and particularly antipsychotics, is poorly understood. For instance, a recent study has reported that anticholinergic load had a negative impact on episodic memory, well-being, recovery, and medication adherence in individuals with psychosis (Verdoux et al., 2020). The aims of this internship are:

1°) Improving the measure of anticholinergic load by taking into account medication dosage and validating the psychometric properties of this measure.

2°) Measuring the impact of anticholinergic and antidopaminergic loads on cognition, functioning, quality of life, and relapses in bipolar and schizophernia spectrum disorders in different databases.

3°) Elaborating recommendations about anticholinergic load for clinicians

4°) Developping online tools to improve the prescription of psychotropics that would be integrated to drug prescription software

For this internship, the student will have access to several databases:

• clinical cohorts from the 12 bipolar expert centers and the 10 schizophrenia expert centers of the FondaMental foundation.

• electronic prescriptions for inward patients from several Departments of Psychiatry

• if results are concluding, access to the prescriptions of CNAM french national database for outward patients will be asked

If you are interested in this project, please contact Paul Roux or Eric Brunet-Gouet.