The relationship between cognitive deficits and impaired short-term functional outcome in clinical high-risk for psychosis participants: A machine learning and modelling approach

Timeframe for Conversion to Psychosis From Individuals at Clinical High-Risk: A Quantile Regression

BACKGROUND AND HYPOTHESIS

The time taken for an individual who is at the clinical high-risk (CHR) stage to transition to full-blown psychosis may vary from months to years. This temporal aspect, known as the timeframe for conversion to psychosis (TCP), is a crucial but relatively underexplored dimension of psychosis development.

STUDY DESIGN

The sample consisted of 145 individuals with CHR who completed a 5-year follow-up with a confirmed transition to psychosis within this period. Clinical variables along with functional variables such as the Global Assessment of Function (GAF) score at baseline (GAF baseline) and GAF-drop from the highest score in the past year. The TCP was defined as the duration from CHR identification to psychosis conversion. Participants were categorized into 3 groups based on TCP: "short" (≤6 months, ≤33.3%), "median" (7-17 months, 33.3%-66.6%), and "long" (≥18 months, ≥66.6%). The quantile regression analysis was applied.

STUDY RESULTS

The overall sample had a median TCP of 11 months. Significant differences among the three TCP groups were observed, particularly in GAF-drop (χ2 = 8.806, P = .012), disorganized symptoms (χ2 = 7.071, P = .029), and general symptoms (χ2 = 6.586, P = .037). Greater disorganized symptoms (odds ratio [OR] = 0.824, P = .009) and GAF-drop (OR = 0.867, P = .011) were significantly associated with a shorter TCP, whereas greater general symptoms (OR = 1.198, P = .012) predicted a longer TCP. Quantile regression analysis demonstrated a positive association between TCP and GAF baseline above the 0.7 quantile and a negative association between TCP rank and GAF drop below the 0.5 quantile.

CONCLUSIONS

This study underscores the pivotal role of functional characteristics in shaping TCP among individuals with CHR, emphasizing the necessity for a comprehensive consideration of temporal aspects in early prevention efforts.

Characteristics and clinical correlates of risk symptoms in individuals at clinical high-risk for psychosis: A systematic review and meta-analysis

Emerging evidence suggests that the duration of risk symptoms (DUR) may have an impact on clinical outcomes in clinical high-risk for psychosis (CHRP) participants. To explore this hypothesis, we performed a meta-analysis on studies that examined DUR in CHR-P individuals in relation to their clinical outcomes. This review was conducted in accordance with the PRISMA guidelines and the protocol was registered with PROSPERO on 16th April 2021 (ID no. CRD42021249443). Literature searches were conducted using PsycINFO and Web of Science in March and November 2021, for studies reporting on DUR in CHR-P populations, in relation to transition to psychosis or symptomatic, functional, or cognitive outcomes. The primary outcome was transition to psychosis, while the secondary outcomes were remission from CHR-P status and functioning at baseline. Thirteen independent studies relating to 2506 CHR-P individuals were included in the meta-analysis. The mean age was 19.88 years (SD = 1.61) and 1194 individuals (47.65 %) were females. The mean length of DUR was 23.61 months (SD = 13.18). There was no meta-analytic effect of DUR on transition to psychosis at 12-month follow-up (OR = 1.000, 95%CI = 0.999-1.000, k = 8, p = .98), while DUR was related to remission (Hedge's g = 0.236, 95%CI = 0.014-0.458, k = 4, p = .037). DUR was not related to baseline GAF scores (beta = -0.004, 95%CI = -0.025-0.017, k = 3, p = .71). The current findings suggest that DUR is not associated with transition to psychosis at 12 months, but may impact remission. However, the database was small and further research in this area is required.

'Feeling' or 'sensing' the future? Testing for anomalous cognitions in clinical versus healthy populations

In the study and treatment of psychosis, emotional intelligence (EI) and thinking styles are important patient characteristics for successful outcomes in clinical intervention. Anticipation of unpredictable stimuli (AUS) may be understood as an anomalous perception and anomalous cognition in which an individual supposedly senses and recognizes future stimuli in an unexpected way, also referred to as “hunches or premonitions.” This examined the roles of EI and thinking styles in AUSs in convenience samples of healthy participants (n = 237) versus patients diagnosed with psychosis (n = 118). We adjusted several quadratic and exponential regression models according to the obtained functions. Group means were also compared to examine differences in EI scores for participants with psychosis compared to healthy participants. In the healthy group, EI predicted AUSs with a weight between 42% and 58%. Thinking styles were not correlated with AUSs. However, EI was not correlated with AUSs in the clinical group. Patients with psychosis tended to score higher on AUSs and lower on EI and thinking styles compared to participants in the healthy group. We discuss EI as a variable that can contextualize some anomalous perceptions which are otherwise difficult to classify or measure within the classic psychosis continuum model.

Predictive models for social functioning in healthy young adults: A machine learning study integrating neuroanatomical, cognitive, and behavioral data

Poor social functioning is an emerging public health problem associated with physical and mental health consequences. Developing prognostic tools is critical to identify individuals at risk for poor social functioning and guide interventions. We aimed to inform prediction models of social functioning by evaluating models relying on bio-behavioral data using machine learning. With data from the Human Connectome Project Healthy Young Adult sample (age 22-35, N = 1,101), we built Support Vector Regression models to estimate social functioning from variable sets of brain morphology to behavior with increasing complexity: 1) brain-only model, 2) brain-cognition model, 3) cognition-behavioral model, and 4) combined brain-cognition-behavioral model. Predictive accuracy of each model was assessed and the importance of individual variables for model performance was determined. The combined and cognition-behavioral models significantly predicted social functioning, whereas the brain-only and brain-cognition models did not. Negative affect, psychological wellbeing, extraversion, withdrawal, and cortical thickness of the rostral middle-frontal and superior-temporal regions were the most important predictors in the combined model. Results demonstrate that social functioning can be accurately predicted using machine learning methods. Behavioral markers may be more significant predictors of social functioning than brain measures for healthy young adults and may represent important leverage points for preventative intervention.

Characterising cognitive heterogeneity in individuals at clinical high-risk for psychosis: a cluster analysis with clinical and functional outcome prediction

Schizophrenia is characterised by cognitive impairments that are already present during early stages, including in the clinical high-risk for psychosis (CHR-P) state and first-episode psychosis (FEP). Moreover, data suggest the presence of distinct cognitive subtypes during early-stage psychosis, with evidence for spared vs. impaired cognitive profiles that may be differentially associated with symptomatic and functional outcomes. Using cluster analysis, we sought to determine whether cognitive subgroups were associated with clinical and functional outcomes in CHR-P individuals. Data were available for 146 CHR-P participants of whom 122 completed a 6- and/or 12-month follow-up; 15 FEP participants; 47 participants not fulfilling CHR-P criteria (CHR-Ns); and 53 healthy controls (HCs). We performed hierarchical cluster analysis on principal components derived from neurocognitive and social cognitive measures. Within the CHR-P group, clusters were compared on clinical and functional variables and examined for associations with global functioning, persistent attenuated psychotic symptoms and transition to psychosis. Two discrete cognitive subgroups emerged across all participants: 45.9% of CHR-P individuals were cognitively impaired compared to 93.3% of FEP, 29.8% of CHR-N and 30.2% of HC participants. Cognitively impaired CHR-P participants also had significantly poorer functioning at baseline and follow-up than their cognitively spared counterparts. Specifically, cluster membership predicted functional but not clinical outcome. Our findings support the existence of distinct cognitive subgroups in CHR-P individuals that are associated with functional outcomes, with implications for early intervention and the understanding of underlying developmental processes.

Using combined environmental-clinical classification models to predict role functioning outcome in clinical high-risk states for psychosis and recent-onset depression

BACKGROUND

Clinical high-risk states for psychosis (CHR) are associated with functional impairments and depressive disorders. A previous PRONIA study predicted social functioning in CHR and recent-onset depression (ROD) based on structural magnetic resonance imaging (sMRI) and clinical data. However, the combination of these domains did not lead to accurate role functioning prediction, calling for the investigation of additional risk dimensions. Role functioning may be more strongly associated with environmental adverse events than social functioning.

AIMS

We aimed to predict role functioning in CHR, ROD and transdiagnostically, by adding environmental adverse events-related variables to clinical and sMRI data domains within the PRONIA sample.

METHOD

Baseline clinical, environmental and sMRI data collected in 92 CHR and 95 ROD samples were trained to predict lower versus higher follow-up role functioning, using support vector classification and mixed k-fold/leave-site-out cross-validation. We built separate predictions for each domain, created multimodal predictions and validated them in independent cohorts (74 CHR, 66 ROD).

RESULTS

Models combining clinical and environmental data predicted role outcome in discovery and replication samples of CHR (balanced accuracies: 65.4% and 67.7%, respectively), ROD (balanced accuracies: 58.9% and 62.5%, respectively), and transdiagnostically (balanced accuracies: 62.4% and 68.2%, respectively). The most reliable environmental features for role outcome prediction were adult environmental adjustment, childhood trauma in CHR and childhood environmental adjustment in ROD.

CONCLUSIONS

Findings support the hypothesis that environmental variables inform role outcome prediction, highlight the existence of both transdiagnostic and syndrome-specific predictive environmental adverse events, and emphasise the importance of implementing real-world models by measuring multiple risk dimensions.